Dork Polls

[hurleyfor3]

Want to start a dicsussion on the difference between what the computer polls think and how this differs from the public (pollsters, commentators and us) perception of us.

We are not only #2 Pomeroy, but also we have maintained that ranking through the Georgetown and NC State losses. The distance between us and #3 (Bucky) is far greater than that between any other two teams.

We are also #2 Sagarin, although the Cuse is quite close to us and Kansas is way north of everyone. Our "predictor" subrating is #1, but year after year we tend to do better there.

So who thinks we're really the second-best team out there? On the one hand I think after multiple early-round exits more of us have become more sceptical of mid-February rankings than ever. I became jaded to this phenomenon earlier than most, so I'm on board here. But I stand by my earlier implications that no one is really that good or that consistent, and that Kansas isn't even that dominant. We could see a lot of early round losses by 1-2-3 seeds.

The dork polls really like the ACC. The ACC is either #1 or #2 sagarin (behind the Televen by one measure) and #1 Pomeroy. I think the middle/bottom of the ACC, eliminating ncsu and possibly unc, is pretty strong in fact.

So what are the computers thinking that the human pollsters (and we) disagree with? Is it just recollections of past tournaments? Again, I don't feel like we should be #2 either, but right now what's a better idea and why?

[dukebluelemur]

I think the formula poles overvalue (to many human rankings) the value of all those decent-but-not-great ACC teams. This punishes the BE and its top teams for playing the utter crap bottom teams, while it helps Duke.

[Kedsy]

The public doesn't see any significant difference between playing Charlotte and playing Stetson. The computers see a huge difference. IMO, this explains much of the discrepancy.

The computers can't see a player actually play, so they rely on their creators being able to find the proper statistics that tell them as much as seeing the games tell us. In many ways the computer way is more accurate, because they can't be fooled by flash and high-flying entertainment. In other ways the computers miss things because there are no statistics to measure certain aspects of the game.

As far as why the pollsters, media, and public, don't completely agree with the computers, it is human nature to discount things we don't entirely understand, especially when our eyes tell us something different.

As far as why Duke doesn't get the same respect from humans as from computers, you have to remember that according to the computers we've been similarly dominant almost every season. For example, according to Sagarin, we were #4 in 2005, #1 in 2006, #9 in 2008, and #6 in 2009. And the funny part is we may really have earned those numbers in each of those years. Because here's the last word on computer ratings: they're not so good at predicting the outcome of a single game. As long as NCAAT performance is the primary factor in the perception of who is the best, the computer ranking is not the bottom line.

[tbyers11]

I think as knowledgeable Duke fans that (mostly :D) populate this board, we are undervaluing, or not putting a lot of stock in, the computer polls (Pomeroy, Sagarin and the like) because of our recent poor showings in the tournament. I agree that I am not sold that we are the 2nd best team in the country, but I am rather optimistic about our chances due to improved interior defense and will be really psyched if our interior offense continues to improve.

On a national poll/commentator level, I think that Duke is not given the perception of a #2 or even #6 (AP) team partly because of 1) recent tourney performance, 2) laying an egg in our highest profile non-con game @ Georgetown and 3) the lack of ranked teams in the ACC this season (the Duke only has a gaudy record b/c the ACC sucks camp).

UNC's awful season (that is fun to type) has really lowered national perceptions of the strength of the ACC. While the ACC is not as top heavy as the Big East (having 4 teams in basically the top 10 all season long has helped the Big East rep), I think that the ACC depth is better than the Big East. The top 4 Big East teams as a whole may be stronger than the top 3 ACC teams, but I think teams 4-8 in the ACC are much better than 5-10 in the Big East.


Right on topic with this thread, Vegas Watch has written a really good piece with a simulated 2010 tourney (http://vegaswatch.net/2010/02/simulating-2010-tournament.html?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+VegasWatch+%28Vegas+Watch%29&utm_content=Google+Reader) (h/t Rush the Court (http://rushthecourt.net/)), based on their modified version of the Pomeroy rankings which factor in recent point spreads and performance relative to the point spreads. The simulation results with Duke as the second most likely to win at 16.8% aren't that different from Pomeroy. They do bump up Purdue and Kentucky a couple of notches and take Wisconsin and Syracuse down a couple.

The interesting part to me is comparing the sim results to current Vegas odds to win the tourney and figuring the return on investment (ROI) a bet would get. Duke has the second highest ROI at 170% of the 15 teams with greater than 1% chance to win. Only Missouri has a higher ROI and they only have a 1.6% chance to win. The only other teams with a positive ROI in this scenario are Purdue, WVa and Wisconsin. The most negative ROI's belong Kentucky, Syracuse, Villanova and Michigan State.

Some interesting food for thought in the numbers vs perception debate on how good Duke (and other top teams) are this year.

[superdave]

Right on topic with this thread, Vegas Watch has written a really good piece with a simulated 2010 tourney (http://vegaswatch.net/2010/02/simulating-2010-tournament.html?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+VegasWatch+%28Vegas+Watch%29&utm_content=Google+Reader) (h/t Rush the Court (http://rushthecourt.net/)), based on their modified version of the Pomeroy rankings which factor in recent point spreads and performance relative to the point spreads. The simulation results with Duke as the second most likely to win at 16.8% aren't that different from Pomeroy. They do bump up Purdue and Kentucky a couple of notches and take Wisconsin and Syracuse down a couple.


This poll last year had Duke with better than 7% chance of winning. Nova was not even on the list, and we all know how poorly that turned out for us.

[Troublemaker]

As I've written before, it would be nice if the computer guys (Pom, Sagarin) would do home / road splits for their ratings. I think Duke would rank a little bit lower in "away from home" ratings, but I do like this Duke team a lot relative to the competition. The field is just weak.

Anyway, I wanted to post the current national championship odds from a well-regarded online book. Just to add to the discussion. Sorry, the formatting won't be great and you can ignore the first column. For those unfamiliar with odds, "Duke +1615" means you bet $100 to win $1615 on Duke.

2451 Kansas +185
2452 Kentucky +525
2453 Villanova +1215
2454 Texas +1215
2455 Syracuse +815
2456 Ohio State +1615
2457 Duke +1615
2458 Purdue +1615
2459 Michigan State +2550
2460 West Virginia +3050
2461 Georgetown +3550
2462 Kansas State +3550
2463 Connecticut +4050
2464 Tennessee +6050
2465 Pittsburgh +6050
2466 Washington +6050
2467 Xavier +6050
2468 Illinois +6050
2469 Wisconsin +6050
2470 Baylor +6550
2471 California +6550
2472 Louisville +8050
2473 Notre Dame +8050
2474 Gonzaga +6050
2475 North Carolina +8550
2476 Mississippi State +8550
2477 Memphis +8550
2478 Georgia Tech +8550
2479 Clemson +8550
2480 Oklahoma State +8550

[loldevilz]

I think one of the issues is that Duke fans don't put a lot of stock in home wins. If kentucky or kansas played us at cameron they would have a very difficult time pulling out a W. Unfortunately tournament games are played at neutral sites and not at home.

[jipops]

As I've written before, it would be nice if the computer guys (Pom, Sagarin) would do home / road splits for their ratings. I think Duke would rank a little bit lower in "away from home" ratings, but I do like this Duke team a lot relative to the competition. The field is just weak.

Anyway, I wanted to post the current national championship odds from a well-regarded online book. Just to add to the discussion. Sorry, the formatting won't be great and you can ignore the first column. For those unfamiliar with odds, "Duke +1615" means you bet $100 to win $1615 on Duke.

2451 Kansas +185
2452 Kentucky +525
2453 Villanova +1215
2454 Texas +1215
2455 Syracuse +815
2456 Ohio State +1615
2457 Duke +1615
2458 Purdue +1615
2459 Michigan State +2550
2460 West Virginia +3050
2461 Georgetown +3550
2462 Kansas State +3550
2463 Connecticut +4050
2464 Tennessee +6050
2465 Pittsburgh +6050
2466 Washington +6050
2467 Xavier +6050
2468 Illinois +6050
2469 Wisconsin +6050
2470 Baylor +6550
2471 California +6550
2472 Louisville +8050
2473 Notre Dame +8050
2474 Gonzaga +6050
2475 North Carolina +8550
2476 Mississippi State +8550
2477 Memphis +8550
2478 Georgia Tech +8550
2479 Clemson +8550
2480 Oklahoma State +8550

How is North Carolina still in that list?

[MADevil30]

This may be a little tangential, but one of the things that I think drops us down in the human polls compared to the computer polls is the thinking that "Duke plays poorly on the road and wins a lot at home, but none of the NCAA tournament games will be played in cameron." A bunch of ESPN guys and others have written this and it makes so sense to me. No NCAA tournament games will be played in any other team's home gym either! This argument completely ignores the fact that we are 4-0 on neutral courts (you know, the type all the NCAA games WILL be played on), including a win over a very legit Gonzaga team. I can't imagine that our neutral site record wont look even better after the ACC tournament. So count me amongst those who don't buy that our "road woes" (aka only two losses) will hamper our chances at the end of the year.

[hc5duke]

How is North Carolina still in that list?

same reason people could buy gm stock the day before it filed for bankruptcy?

[RoyalBlue08]

I have two points:

1.) That Kansas is a sucker bet...wow! better than a 2-1 favorite to win the thing, really?

2.)I think Duke (and the ACC as a whole for that matter) do a very good job with pre conference scheduling by avoiding those schools at the very bottom of DI that will drag all your numbers down. But do I think that if Duke wins a bunch of games against mediocre schools early it makes them better than teams that win games against really terrible teams. Not really. I think it is much more informative to look at how you have done in your own conference and against other Big 6 and mid major schools. In other words, I understand while the computers have us at #2, but I don't honestly think we are the second best team in the country. Just my opinion obviously.

[GODUKEGO]

Sagarin ratings
http://www.usatoday.com/sports/sagarin/bkt0910.htm

Latest RPI
http://www.realtimerpi.com/rpi_Men.html

Latest AP
http://nmsn.foxsports.com/cbk/polls

[Olympic Fan]

As far as why Duke doesn't get the same respect from humans as from computers, you have to remember that according to the computers we've been similarly dominant almost every season. For example, according to Sagarin, we were #4 in 2005, #1 in 2006, #9 in 2008, and #6 in 2009.

Actually, your example indicates that Duke gets almost exactly the same respect from humans as from the computers.

In 2005, when Duke finished #4 in Sagarin, Duke finished #3 in the final AP poll

In 2006, when Duke finished #1 in Sagarin, Duke finished #1 in the final AP poll

In 2008, when Duke finished #9 in Sagarin, Duke finished #9 in the final AP poll

In 2006, when Duke finished #6 in Sagarin, Duke finished #6 in the final AP poll

So it seems that before THIS season, Sagarin (and RPI and Pomeroy) have pretty much agreed with the human voters. I agree that you can't use the "dork" polls to predict one-game results -- such as in the NCAA Tournament. But over the long term, the computers pretty much agree with the human voters -- at least as far as Duke is concerned.

[Dr. Rosenrosen]

I see the computer rankings as more reflective of a team's net potential. That is, given past performance and a variety of statistics, computer rankings say this year we have great potential to go all the way.

Whereas the traditional polls are more refelctive of writers' and coaches' perceptions of each team at the moment. Why else would a team move up and down from week to week? If you really believe Nova is #3 in the country and nothing drastic has changed about the team, why all of a sudden should they be bumped down 3 or 4 spots in just one week, only to earn their way back up as teams above them lose. It's a very short term focus. Which may in fact be more predictive of how the teams might play "in the moment" during tourney time.

Which is really to say, I have a hard time reconciling the computer rankings. They're nice to look at but... And honestly I have a hard time with the regular polls, too.

In the end, isn't it awesome that it all gets settled on the court!

[Troublemaker]

How is North Carolina still in that list?

Believe it or not, I think people are still betting them (and hoping UNC can sneak into the tournament). It makes sense for books to offer odds on as many teams as possible as long as the odds are in their favor.


I have two points:

1.) That Kansas is a sucker bet...wow! better than a 2-1 favorite to win the thing, really?


They're all bad bets with negative expected value, imo. The books wouldn't be doing their jobs correctly otherwise. But you're right, the Kansas odds seem especially egregious, because the Jayhawks are probably the most heavily bet on team this season and those +185 odds really punish the bettor. I mean, I think Kansas is probably the best team, but they'd have to win the championship 35% of the time just for you to break even at +185. Kansas isn't that good, imo.

[Kedsy]

In 2008, when Duke finished #9 in Sagarin, Duke finished #9 in the final AP poll

Actually, I committed a typo. We were #7 in Sagarin in 2008, not #9. But that shouldn't detract from your point.

[91_92_01_10_15]

I've been thinking all season that it would be interesting to see how Pomeroy's predictions for individual games compare with Vegas odds. From my own observations, probably 25 games or so, I'd estimate that Pomeroy has been right against the spread 75% of the time.

Don't get me wrong. I'm not really a bettor or anything like that. I'm more interested in using any data on this as a measure of the accuracy of his predictions, and the implication that may have as to the accuracy of his ranking system in general.

It seems obvious to me that someone must have done some research on this because of the obvious potential financial implications; however, maybe they just aren't sharing it. ;)

[Kedsy]

I've been thinking all season that it would be interesting to see how Pomeroy's predictions for individual games compare with Vegas odds. From my own observations, probably 25 games or so, I'd estimate that Pomeroy has been right against the spread 75% of the time.

Don't get me wrong. I'm not really a bettor or anything like that. I'm more interested in using any data on this as a measure of the accuracy of his predictions, and the implication that may have as to the accuracy of his ranking system in general.

It seems obvious to me that someone must have done some research on this because of the obvious potential financial implications; however, maybe they just aren't sharing it. ;)

I haven't studied it for Pomeroy, but years ago I had my own computer system that I used to place bets with real bookies (only a few bets in Vegas, mostly just my corner bookie). The answer is you can make a little money if you play smart, but it's nowhere close to 75% because the real spread is almost never what you see in the newspapers. In most of the games where there are large disparities between Pomeroy and the spread (let's say 2 points or more -- you don't have much of an advantage with a one point difference), either the spread moves quickly away from the early spread (which was probably close to what you see in the paper) or the disparity is caused by some piece of information (e.g., an injury) that the bookies factor in but the computer generally doesn't.

Put another way, if Pomeroy says team A will win by 5 but the line you see is only 2, when you call your bookie to place a bet the most likely results are either (a) he won't let you lay 2 but instead will insist on 4 (give or take a half point); or (b) will let you lay 2 but only because team A's sixth man sprained his ankle in practice the day before.

[Trooper]

I agree that the fact that our home court advantage is greater than most any school, so our numbers are skewed vs. some other teams. This gives us a bit of a bump in the computers, but I don't think that's the major reason we are always high in these ratings. We just happen to be in a year where our home/road splits are unusually large (we're normally great on the road too), so naturally our minds are floating to that issue.

Another factor that undoubtedly plays a role is Coach K's intelligent scheduling, which eliminates what I'd call the "computer bottom feeders". People have correctly brought that up, but I still don't think this is the primary difference.

I think the biggest issue we see is due to how amazing of a motivator Coach K is. Most years, Duke is one of the 10-15 best teams in the nation (often much better). And every year I see nearly all of those other teams have 6-8 games where they simply don't come to play. They don't necessarily lose those games, but they may beat a far inferior team by 2-3 points instead of 15, and at times they will get upset. For Duke, that happens 2-3 times a year (NC State & G'Town this season). Simply put, Coach K gets his guys to play their hardest in nearly every game...even against teams that the players probably expect to beat. That's an amazing ability he has as a coach. It has the net effect of increasing our adjusted scoring margin relative to others, which is a huge driver of every computer ranking (especially pomeroy and sagarin predictor where we always fair the best).

The only downside of this is that we lose this advantage come tourney time, because at that point every team is playing as if it's their last game, because, well, it could be.

If we were to play Kansas on a neutral court tomorrow, I'd say they'd be favored in that game by maybe 4.5pts or so (at least in my mind). However, if you told me both teams were playing the 42nd ranked team in the country this week on neutral courts, I'd probably wager that Duke would win by more points. Not to knock Kansas or Bill Self in particular, but you get my point...

[Jderf]

Not perfectly relevant, but I always wonder what these rankings would look like if you cut out the last 1:30 or so of every game, where the play usually breaks down (whether a tight game or a blowout) and the game/stats become a crapshoot. Take our G'town game this year. We were much further down than the final score would suggest, only because we cranked up our score when G'town wasn't really trying anymore. The computer rankings weigh these points equally, which I feel they shouldn't.

Although I guess you could argue that those endgame situations are just as much a part of the sport as the rest and need to be calculated equally. Still though, I'm not convinced.

[MChambers]

There are several reasons that I am a little bit skeptical about the dork polls. (This pains me, because I prefer cold, hard numbers to subjective opinions.)

In many recent years, Duke has been highly ranked in those polls, only to have a disappointing end to the season. How to explain this?

First, as many of us have noted, Coach K tends to keep our best players in one-sided games longer than some coaches would do. That undoubtedly leads to bigger scoring margins. (On the other hand, Coach K also goes to the delay game earlier than some teams, which can reduce scoring margins.)

Second, Duke players always play hard. In contrast to some other teams (I’ll let you, Ol’ Roy, and Calhoun guess who), there is never any question about effort. That also leads to bigger scoring margins.

Other teams, especially veteran, talented teams, may loaf through parts of the season. I had my hopes raised last year during the season, when UNC's defensive efficiency rating was not very good. Unfortunately, at the end of the season, that team improved its defense markedly. I wonder if Duke’s admirable effort means that it is harder for the team to raise its level of play at year end, at least as far as a computer can tell.

It would be interesting to see how the 1991, 1992, and 2001 teams looked in the dork polls at this point in the season.

[GoingFor#5]

I am just going to let things play out. I wouldn't be surprised either way. From what I am seeing, we are looking great against all-comers lately, but the Georgetown game still sticks in the back of my head. Was it just a bad game or are we going to get owned again by a stellar guard in March? Who knows....

[RoyalBlue08]

I think the biggest issue we see is due to how amazing of a motivator Coach K is. Most years, Duke is one of the 10-15 best teams in the nation (often much better). And every year I see nearly all of those other teams have 6-8 games where they simply don't come to play. They don't necessarily lose those games, but they may beat a far inferior team by 2-3 points instead of 15, and at times they will get upset. For Duke, that happens 2-3 times a year (NC State & G'Town this season). Simply put, Coach K gets his guys to play their hardest in nearly every game...even against teams that the players probably expect to beat. That's an amazing ability he has as a coach. It has the net effect of increasing our adjusted scoring margin relative to others, which is a huge driver of every computer ranking (especially pomeroy and sagarin predictor where we always fair the best).

The only downside of this is that we lose this advantage come tourney time, because at that point every team is playing as if it's their last game, because, well, it could be.

If we were to play Kansas on a neutral court tomorrow, I'd say they'd be favored in that game by maybe 4.5pts or so (at least in my mind). However, if you told me both teams were playing the 42nd ranked team in the country this week on neutral courts, I'd probably wager that Duke would win by more points. Not to knock Kansas or Bill Self in particular, but you get my point...

These are some really nice points. I especially like the point that Duke gives it's best effort more often than other schools, and it is an advantage that we lose come tournament time. I definitely think this is true. I know I mentioned this in another thread somewhere, but this is why I have no desire to see a team like Kentucky come tournament time. you could make an argument that we have had just as productive a season as them to this point, given our SOS, but if they ever played as hard as our guys play, they would be scary.

[MChambers]

It's a system, or at least two systems, created by some math professors, some of whom are at Georgia Tech. Claims to be a better predictor of NCAA tourney performance than Sagarin, RPI, or the NCAA committee. (Hasn't been compared to Pomeroy.)

http://www2.isye.gatech.edu/people/faculty/Joel_Sokol/lrmc/lrmc.sort0.html

[crimsonandblue]

I think the biggest issue we see is due to how amazing of a motivator Coach K is. Most years, Duke is one of the 10-15 best teams in the nation (often much better). And every year I see nearly all of those other teams have 6-8 games where they simply don't come to play. They don't necessarily lose those games, but they may beat a far inferior team by 2-3 points instead of 15, and at times they will get upset. For Duke, that happens 2-3 times a year (NC State & G'Town this season). Simply put, Coach K gets his guys to play their hardest in nearly every game...even against teams that the players probably expect to beat. That's an amazing ability he has as a coach. It has the net effect of increasing our adjusted scoring margin relative to others, which is a huge driver of every computer ranking (especially pomeroy and sagarin predictor where we always fair the best).

The only downside of this is that we lose this advantage come tourney time, because at that point every team is playing as if it's their last game, because, well, it could be.

If we were to play Kansas on a neutral court tomorrow, I'd say they'd be favored in that game by maybe 4.5pts or so (at least in my mind). However, if you told me both teams were playing the 42nd ranked team in the country this week on neutral courts, I'd probably wager that Duke would win by more points. Not to knock Kansas or Bill Self in particular, but you get my point...

I think I agree with this almost completely, but will also add a point or two. The "almost" is the notion of this being a "knock" on Kansas or Self (or any other coaches/teams) at all. It may be that it's a conscious decision. It may be that some coaches, like jockeys, feel they can only whip their kids into a frenzy so often before it's counterproductive.

I guess one question is whether Duke's recent NCAA tournament struggles are the result of teams finally matching Duke's motivation level or is it Duke being worn down over the course of the year from running through walls to get that extra two wins that a comparable Kansas team wouldn't, or both? I don't know.

One other factor is just style of play. Duke's personnel and style of play lend themselves to throttling lesser teams. The overplays, the lack of turnovers, the three ball, and getting to the line (I'm talking traditionally here). You compare that to a Kansas approach that is much more Big Ten in its manner (less extended man and more emphasis on working the ball inside to bigs - lending itself to turnovers) and style of play combined with a less frenzied manner day-in-and-day-out. Duke's just built to kill lesser teams and they're more motivated to do it. Come NCAA tourney time, things do change.

[Chitowndevil]

One thing to mention is that both Pomeroy and Sagarin's ratings depend on margin of victory (more precisely, the Sagarin combines two components, one that does and one that doesn't). Sagarin himself has acknowledged this as a concern:


This is the most obvious problem with the system - there is no cap on margin of victory. It’s not that I’m particularly comfortable with it, but I’ve looked at quite a few ways to limit the impact of MOV, and I haven’t found one that I like, yet.

As others have mentioned, Duke has a significant home court advantage. Every year that translates to a few lopsided home wins over some pretty good teams, such as this year's home wins over Maryland, Georgia Tech, Wake, and Clemson. These carry a lot of weight in predictive ratings.

Duke is currently rated #5 is the non-predictive component of Sagarin's poll, so the discrepancy isn't that huge. I went back a few seasons and compared Duke's final Sagarin ratings in the two components, Elo Chess (which is NOT based on scoring margin), and Predictor (which is):

2008-09: #7 overall, #8 Elo Chess, #6 Predictor
2007-08: #7 overall, #13 Elo Chess, #6 Predictor
2006-07: #24 overall, #32 Elo Chess, #12 Predictor
2005-06: #1 overall, #3 Elo Chess, #1 Predictor
2004-05: #4 overall, #4 Elo Chess, #3 Predictor
2003-04: #1 overall, #4 Elo Chess, #1 Predictor

Of course this doesn't mean home court advantage explains the discrepancy, that's just my favorite hypothesis. Glancing at the data, there's a similar (though I'd argue weaker) pattern for Carolina, so the ACC being consistently strong in the computer ratings probably has something to do with it, too.

[Kedsy]

Duke is currently rated #5 is the non-predictive component of Sagarin's poll, so the discrepancy isn't that huge. I went back a few seasons and compared Duke's final Sagarin ratings in the two components, Elo Chess (which is NOT based on scoring margin), and Predictor (which is):

2008-09: #7 overall, #8 Elo Chess, #6 Predictor
2007-08: #7 overall, #13 Elo Chess, #6 Predictor
2006-07: #24 overall, #32 Elo Chess, #12 Predictor
2005-06: #1 overall, #3 Elo Chess, #1 Predictor
2004-05: #4 overall, #4 Elo Chess, #3 Predictor
2003-04: #1 overall, #4 Elo Chess, #1 Predictor

Of course this doesn't mean home court advantage explains the discrepancy, that's just my favorite hypothesis. Glancing at the data, there's a similar (though I'd argue weaker) pattern for Carolina, so the ACC being consistently strong in the computer ratings probably has something to do with it, too.

If I'm not mistaken, his "Elo Chess" system is similar to the RPI, which I don't think is a particularly good rating system (despite its annual love for Duke). Although it's obviously not perfect, the "Predictor" system is a much better indicator of a team's strength.

One problem we are all having in this discussion is we seem to be accepting as an unshakable premise that the computer rankings have been wrong because we've been ranked really high but haven't played to that ranking in the NCAAT. In my opinion, the premise we should be accepting is these computer rankings don't guarantee a winner, in which case they may have been spot on every year in their evaluation of Duke as a top team.

Pomeroy has a "consistency" ranking, although I'm not sure exactly how he arrives at it (Duke currently ranks 284th in the country in "consistency"; last year we were 287th). He doesn't seem to do a lot with it, but it seems to me this idea is on the right track. How good a team is on any given day should be described as a range, not as a point. When two teams play, the amount of uncertainty regarding who will win can be described by the overlap in their ranges. After that, a lot of randomish/unpredictable factors determine where each team falls in its own range. If your team is inconsistent (like Duke, according to Pomeroy) the range is bigger, the randomish factors are more important, and the results have a decent chance of being different than the single-point rating would suggest.

[MChambers]

Pomeroy has a "consistency" ranking, although I'm not sure exactly how he arrives at it (Duke currently ranks 284th in the country in "consistency"; last year we were 287th). He doesn't seem to do a lot with it, but it seems to me this idea is on the right track. How good a team is on any given day should be described as a range, not as a point. When two teams play, the amount of uncertainty regarding who will win can be described by the overlap in their ranges. After that, a lot of randomish/unpredictable factors determine where each team falls in its own range. If your team is inconsistent (like Duke, according to Pomeroy) the range is bigger, the randomish factors are more important, and the results have a decent chance of being different than the single-point rating would suggest.

Explanation from http://kenpom.com/blog/index.php/weblog/ratings_explanation/

"Consistency is basically the standard deviation of scoring difference by game for a team. Again, it’s not included in the ratings calculation. It can be an aid in determining which teams are overrated by my system. Highly rated teams that are inconsistent tend to look beatable more often. As of this writing, Georgia is ranked 329 in consistency and Oklahoma is at 334. They’ve played their best games against poor teams, and their worst against good ones.

Ideally, I’d synthesize the consistency and rating into one number, but I haven’t found a way I’m comfortable with. So right now, I’m throwing this system out there with all its warts for everyone to see. The warts tend to decrease as more games are played, but at least I’ve made you aware of them and where they can pop up."

I suspect the NCSU game has a lot to do with Duke's consistency rating.

[Chitowndevil]

If I'm not mistaken, his "Elo Chess" system is similar to the RPI, which I don't think is a particularly good rating system (despite its annual love for Duke). Although it's obviously not perfect, the "Predictor" system is a much better indicator of a team's strength.

One problem we are all having in this discussion is we seem to be accepting as an unshakable premise that the computer rankings have been wrong because we've been ranked really high but haven't played to that ranking in the NCAAT. In my opinion, the premise we should be accepting is these computer rankings don't guarantee a winner, in which case they may have been spot on every year in their evaluation of Duke as a top team.

Pomeroy has a "consistency" ranking, although I'm not sure exactly how he arrives at it (Duke currently ranks 284th in the country in "consistency"; last year we were 287th). He doesn't seem to do a lot with it, but it seems to me this idea is on the right track. How good a team is on any given day should be described as a range, not as a point. When two teams play, the amount of uncertainty regarding who will win can be described by the overlap in their ranges. After that, a lot of randomish/unpredictable factors determine where each team falls in its own range. If your team is inconsistent (like Duke, according to Pomeroy) the range is bigger, the randomish factors are more important, and the results have a decent chance of being different than the single-point rating would suggest.

Interesting points. I'm not a fan of the RPI, either, but I do think you have to limit the impact of victory margin somehow.

I actually disagree somewhat with Sagarin's premise that a low consistency rating implies a team is "beatable". Consistency is based on the variability in a team's victory margin. A team that wins by an average of 30 points, winning by 10 in half its games and 50 in the other half, would look "inconsistent" by this metric. But final score is highly dependent on the sequence of scores throughout the game. Just because our hypothetical team won by "only" 10 points does not mean it was in danger of losing - it could have been up by 30 with 8 minutes to go and chosen to run the clock out.

I actually think Duke's low consistency rating may support my hypothesis that home court advantage may be responsible for Duke's high ratings in the computer ratings relative to their performance in March. Those same large home victories I mention add substantial variability to Duke's scoring margin. Notice that of the current top 10 Pomeroy teams, only Kansas and Kentucky are in the top 50 in consistency. I think this has a lot to do with good-but-not-great teams often winning by large margins at home. Note that all of these ratings DO account for home/road, but not for the fact certain teams (in particlar, often good teams) have a much bigger discrepancy between expected scoring margin in home vs. road games. Of course consistency should pick up some of this variation, so maybe it's a useful ingredient in predicting tournament outcomes after all.

[Kedsy]

I actually think Duke's low consistency rating may support my hypothesis that home court advantage may be responsible for Duke's high ratings in the computer ratings relative to their performance in March. Those same large home victories I mention add substantial variability to Duke's scoring margin. Notice that of the current top 10 Pomeroy teams, only Kansas and Kentucky are in the top 50 in consistency. I think this has a lot to do with good-but-not-great teams often winning by large margins at home. Note that all of these ratings DO account for home/road, but not for the fact certain teams (in particlar, often good teams) have a much bigger discrepancy between expected scoring margin in home vs. road games. Of course consistency should pick up some of this variation, so maybe it's a useful ingredient in predicting tournament outcomes after all.

Actually, Kansas and Kentucky are top 50 in Pomeroy's "luck" rating. Of Pomeroy's top 10, only Purdue is in the top 50 in "consistency." Kansas is ranked 240, not too far ahead of Duke.

Putting that aside, I agree that Duke having more of a home team advantage than the system expects home teams to have could be a possible partial explanation for our high computer ranking. Or it could be an indicator of how good we can be when our players feel comfortable and confident.

[uncwdevil]

After this weekend's games, Duke is #1 in Pomeroy by the slimmest of margins.

[Neals384]

Well, the human polls seem to value losses more than the computer rankings. The two teams with 1 loss are #1 and #2.

The computer rankings value Duke's strength of schedule and winning margin enough to make up for having 4 losses...

[uncwdevil]

Well, the human polls seem to value losses more than the computer rankings. The two teams with 1 loss are #1 and #2.

The computer rankings value Duke's strength of schedule and winning margin enough to make up for having 4 losses...

Yes, the computer rankings seek to determine how good a team truly is, while human polls simply move teams down an arbitrary number of spots for each loss and let them slowly climb up the rankings as they win, even if they are barely beating weak teams. Also, the starting ranks for teams are determined before anybody has even played a game yet.

I would rather be #1 in the computer rankings and #6 in the human rankings than vice versa, especially at this point in the season.

[gotham devil]

After this weekend's games, Duke is #1 in Pomeroy by the slimmest of margins.

http://twitter.com/kenpomeroy

I can't win. RT @tallguyDuke: man, even @kenpomeroy hates Duke
about 6 hours ago via TweetDeck


I can only hope! RT @BeyndArcMMiller: This keeps up, Kansas is gonna make Duke's reign atop @kenpomeroy ratings real short.
about 7 hours ago via TweetDeck

[patentgeek]

Currently, Duke is ranked No. 1 in the Pomeroy ratings, largely on the strength of its offense being ranked as the most efficient offense in the country (and by a rather wide margin as these things go). Of course, this is a great showing. Duke's defense is also obviously very good (No. 12 currently). What I find somewhat odd, though, as I read various posts on this board and listen to commentators, is that 95 percent of the angst/worry/criticism/bashing of the 2009-10 Blue Devils is focused on the offense (and very little criticism is leveled at the defense, other than saying "Mason misses assignments" or the like).

We all know the offensive shortcomings people point to (too reliant on Earth Wind & Scheyer, too reliant on the 3-pointer, no inside game, no breakdown point guard, etc.). Yet somehow, with all of these issues, Duke is still, by a great deal, the most efficient offense in the country. What this means, at least in my interpretation, is that all of these other teams that people are so impressed with/scared of playing in March (e.g., Kansas, Kentucky, Syracuse, Villanova) must also have some offensive issues of their own - and apparently bigger issues than Duke has, or else you'd thank that some of these teams would be ahead of Duke in Off. Eff.

How is this the case (or, put differently, why are so many people still down on the Duke offense)? I think it may have to do with the fact that the things Duke does well (spread the court, hit the open man, make the extra pass, shoot the 3 with multiple players, avoid turnovers, hit the offensive boards, make free throws) are not the kind of thing that sticks in your mind - at least not in the same way that a quick slash to the basket or a resounding dunk does. But clearly Duke is quietly and very effectively putting points on the board, even though the team has the perceived weaknesses people talk about.

So, take heart the next time you see John Wall break someone's ankles on a drive, or Cole Aldrich posterize a defender with a dunk - those points, while maybe visually stunning, are worth no more than two free throws from Scheyer or a simple Zoubek put-back. Duke has figured out a formula that works for its personnel, and it works really well.

Does this mean another national championship, or even another Final Four appearance, is in the cards? Maybe, maybe not - lots of factors, including luck, are involved. But if the team isn't able to accomplish either of these feats, I don't think it will be because Duke didn't put together a really effective offensive team (and scheme). Sure, it would be great to have a stud post player who's automatic, or a blur of a point guard, but what we do have is still apparently pretty strong.

[MarkD83]

fans and experts underestimate the value of rebounding, especially offensive rebounding. I only do this analysis after games and don't keep records, but take a look at Duke's shooting percentage when you subtract offensive reobounds from shots taken. The shooting looks really good.

In addition, when good shooters miss and then are given the ball back a few seconds later they usually make the adjustment. It is sort of like warm-ups, (oops I missed give me the ball and let me try again...swish.)

This does not look pretty but it is very effective.

[DukieInBrasil]

great point ie show points vs. actual points, they're all worth the same. There are 3 things which Duke is doing pretty well that help our off. eff. remain at a high level. Our 3 pt shooting is still quite good, our 4 most prolific 3pt shooters are shooting slightly under 40% as a group, that's like shooting almost 60% from 2pt land. Our FT shooting is excellent, top 5 nationally I believe. Scoring points with the clock stopped is always a plus. We are also quite good as an offensive rebounding team, especially recently. This helps erase some of our poor FG shooting performances.

[RockyMtDevil]

This team has been through more than most over the last 3-4 years. While they haven't broken out in the NCAA Tourny, we have guys that are seasoned in this arena. The Texas game last year was a maturation game, though we got skilled against a very, very hot Nova team.

I can see us making a series threat at the Final Four, simply based on the mental and physical toughness of this group. They've seen failure, they know they aren't the "Duke of old" and I think they are relishing in that role. There is something to be said of a team of veterans who've been kicked around a bit and are now bloody tired of it, especially if two of them are some of the best shooters, play makers in the league.

I don't believe anyone of us saw the Zoubek outbreak that has been thrown against the last three oponents. he has truly had three games for the ages, if he continues this smart, crafty play, we may just keep dancing with our alarmingly un-athletic dancers....I'm all for it!

[Verga3]

Thank you, Pomeroy, but I don't want to be ranked #1.....yet.

[D.C. Devil]

I won't speak for others, but I'm worried because Duke's failed three of its four hardest tests (@Wisco, @GU, @ GT, winning only @ Clem) and laid a stinker of a bad loss @ NCSU. So there is a feeling that the team doesn't have the high end to run with the big boys AND that it's prone to the upset. I'll feel a lot better if the guys come through at the Comcast Center to make it two of five.

That said, I agree that the team has shown real heart this year (Miami and VT being the two most recent examples, but I also count wins @ UNC and @ Clemson in this category). The team has shown that it can win even when it's not firing on all cylinders.

My hope is that the steady play down low will continue this streak. If we see improved play from MP2 and 'dre, though, Duke is a favorite in every game (in my mind at least).

[kexman]

I won't speak for others, but I'm worried because Duke's failed three of its four hardest tests (@Wisco, @GU, @ GT, winning only @ Clem) and laid a stinker of a bad loss @ NCSU.

This is an aside, but the line above triggered the thought. We almost always show up to play. We seldom lose to below average teams and seldom are non competitive with good teams. Last year the clemson game was a blowout and I remembered thinking that I hadn't really felt like that since the UCLA game in the mid 90's. We may not always play our very best, but we seldom don't compete. That is a tribute to the coaches and players that they get up for every game.

I think this team can legitimately reach the elite 8 and than it is a coin toss in the next few games. I think Kansas may be a step better and Kentucky's talent is balanced by their youth.

[COYS]

I won't speak for others, but I'm worried because Duke's failed three of its four hardest tests (@Wisco, @GU, @ GT, winning only @ Clem) and laid a stinker of a bad loss @ NCSU. So there is a feeling that the team doesn't have the high end to run with the big boys AND that it's prone to the upset. I'll feel a lot better if the guys come through at the Comcast Center to make it two of five.

That said, I agree that the team has shown real heart this year (Miami and VT being the two most recent examples, but I also count wins @ UNC and @ Clemson in this category). The team has shown that it can win even when it's not firing on all cylinders.

My hope is that the steady play down low will continue this streak. If we see improved play from MP2 and 'dre, though, Duke is a favorite in every game (in my mind at least).

I would add the Gonzaga game as a tough test . . . it was just so easy it didn't seem like a real test.

[cptnflash]

People are down on our offense for three reasons:

1) Most people who follow college basketball don't yet understand tempo-free statistics like offensive efficiency (although this is changing rapidly... God bless the internet!). TV personalities reinforce the ignorance by continuing to cite meaningless stats like points per game (unadjusted for quality of competition and hugely impacted by pace) as a measure of offensive capability.

2) Duke is playing at a slower pace this year (odd given their incredible efficiency margin but perhaps understandable given the number of minutes being logged by the Big 3).

3) While our average offensive (and defensive) efficiency is excellect, our consistency is not (#274 in the country according to Pomeroy). This is probably due in large part to our heavy reliance on 3-pointers, although I'm speculating there. In any case, people remember the few games where we've laid an egg offensively and it gives their perception a downward bias.

The last point is significant, however, for a single elimination tournament where all it takes is one off night from the 3 point line to end the season.

[theAlaskanBear]

I would add the Gonzaga game as a tough test . . . it was just so easy it didn't seem like a real test.

I was going to say UConn at MSG. That was a tough game, and we beat the will out of them.

The bad losses I think about are Georgetown and NC State, and both of those seemed to clearly be aberrations.

The losses at Wisconsin and at Gtech are tough losses, but I am not ashamed of them by any means.

[COYS]

I was going to say UConn at MSG. That was a tough game, and we beat the will out of them.

The bad losses I think about are Georgetown and NC State, and both of those seemed to clearly be aberrations.

The losses at Wisconsin and at Gtech are tough losses, but I am not ashamed of them by any means.

Wow, I forgot UCONN, too. I guess the luster from that win faded as UCONN struggled through most of the season. Maybe their wins this month will give that victory a little more significance, again.

I agree with you about GaTech and Wiscy. Neither loss is a loss to be ashamed about. Losing to G-Town at their place is not too bad either. I think it is more about the way we lost. Somehow, however, we let NC State post the highest offensive efficiency rating against us of any team by far. Not that we should be ashamed of that, but it is definitely something to scratch your head over.

[gumbomoop]

We all know the offensive shortcomings people point to (too reliant on Earth Wind & Scheyer, too reliant on the 3-pointer, no inside game, no breakdown point guard, etc.). Yet somehow, with all of these issues, Duke is still, by a great deal, the most efficient offense in the country. What this means, at least in my interpretation, is that all of these other teams that people are so impressed with/scared of playing in March (e.g., Kansas, Kentucky, Syracuse, Villanova) must also have some offensive issues of their own - and apparently bigger issues than Duke has, or else you'd thank that some of these teams would be ahead of Duke in Off. Eff.

How is this the case (or, put differently, why are so many people still down on the Duke offense)? I think it may have to do with the fact that the things Duke does well (spread the court, hit the open man, make the extra pass, shoot the 3 with multiple players, avoid turnovers, hit the offensive boards, make free throws) are not the kind of thing that sticks in your mind - at least not in the same way that a quick slash to the basket or a resounding dunk does. But clearly Duke is quietly and very effectively putting points on the board, even though the team has the perceived weaknesses people talk about. Duke has figured out a formula that works for its personnel, and it works really well.

Several good points in this analysis:
- somehow effective
- not flashy, but efficient
- quietly effective
- figured out formula that matches personnel, and works really well

I've said a time or 2 on threads long forgotten that the emotional/psychological/spiritual [ugh??] attribute I think most important in an individual and in a team is [I]relentlessness.

KS has had it from day 1.

JS, too, but his "soft" look throws us off, until we gradually become aware that he never stops moving, thinking, calculating [e.g., those occasions when, after a Duke basket he turns to run back to D, then stops to intercept inbound pass], planning. He never takes play off, physically or..... inellectually [gimme another word for this, but if you cannot, I'm sticking with "JS is intellectually relentless on the court"].

NS? Well, I wondered whether he had it, as he didn't show it much his first 2 years, but he sure looks relentless these days. I think he'd like to smile a lot, but mostly he just scowls, and often shoots daggers.

LT always demonstrated energy, but his play until this year was often so undisciplined that his energy was mostly - repeat as constructive criticism: mostly - wasted and disruptive to his own team. This year is different, not perfect, but more often relentless [leading toward efficiency] rather than chaotic [leading to ..... arrggghhhhhh!]

Z - I missed, where others saw, his potential. I thought last spring that he should redshirt this season, to make up for lost time due to injuries. I did not expect to see anything like what he's done the last few games.

IMO, what patentgeek says in the quoted post re "effective and efficient" is very closely tied to this team's individual and collective intensity, usually discussed as "toughness." In VaT postgame, K referred to Wojo's halftime reminder to the team that their "identity" was based on defense and rebounding. It strikes me that both of these are most effective when undertaken relentlessly, taking no plays off.

This is hardly a group of "fluid" athletes. NS, yes, MP2, someday, DD, if handle develops. But the core 5 are collectively relentless.

And btw, KS is pissed off pretty much all the time.

[Kedsy]

In any case, people remember the few games where we've laid an egg offensively and it gives their perception a downward bias.

The last point is significant, however, for a single elimination tournament where all it takes is one off night from the 3 point line to end the season.

In games where we've "laid an egg," we did so defensively, not offensively. We've had plenty of off nights from the 3-point line this year, and we've won most (if not all) of those games. The fact is, in our four losses this year we've shot an aggregate 41% from three, which is not bad at all. You are talking about Duke teams from the past, not this year's model.

[pfrduke]

3) While our average offensive (and defensive) efficiency is excellect, our consistency is not (#274 in the country according to Pomeroy). This is probably due in large part to our heavy reliance on 3-pointers, although I'm speculating there. In any case, people remember the few games where we've laid an egg offensively and it gives their perception a downward bias.

The last point is significant, however, for a single elimination tournament where all it takes is one off night from the 3 point line to end the season.

I'm going to quote myself here:


Percentage of Duke shots that are 3s: 32.8%
National rank: 152nd
3 point attempts on the season: 541
3 point attempts per game: 20.03
# of games we've taken 30 3s: 1


Also, we've had 14 games this season where we're 33.3% or less from 3 this season. We're 12-2 in those games, with a collective offensive rating of 1.084 points per possession. Our offense is better when we hit threes, as a general matter, but any offense is. This season, our offense is still good when we don't.

[patentgeek]

In games where we've "laid an egg," we did so defensively, not offensively. We've had plenty of off nights from the 3-point line this year, and we've won most (if not all) of those games. The fact is, in our four losses this year we've shot an aggregate 41% from three, which is not bad at all. You are talking about Duke teams from the past, not this year's model.

I agree - certainly the Georgetown and NCState games were defensive disasters, and Wisconsin wasn't good either (73 points in 63 possessions - 1.159 points/possession - is way above what Duke would typically allow). It appears that as long as the defense stays solid, Duke will be in just about every game, and if Duke shoots well, can beat anybody.

(This may help to explain why Lance continues to get heavy minutes despite some offensive shortcomings, and why the Plumlees and Dakwins sit more than some observers would prefer.)

Also, one of my points from my earlier post was that other teams have their own offensive problems too. We watch a lot of Duke basketball, so we see the warts along with the good stuff, and may not readily recognize some of the flaws in these other teams that appear to be offensive juggernauts.

[Kedsy]

The fact is, in our four losses this year we've shot an aggregate 41% from three...

I want to apologize for getting the stats wrong. The real facts are, in our four losses we shot an aggregate of 31% (our opponents shot 41%). However, we did shoot our threes pretty well in 2 of the 4 losses, and really the only horrendous shooting performance was against State (6 for 28). I stand by my original point that poor defense and not poor 3-point shooting is what can lead this team to defeat.

[JasonEvans]

(Raises hand and stands)

"ummm, I'm Jason E."

"Hello, Jason E."

"Yeah, ummm, I'm Jason E and I am addicted to Ken Pomeroy's statistics."
---------------------------------------------------------------------------

So, going into tonight's game against so-so Tulsa (Pomeroy rates them the #69 team in the country with a .8148 rating), Duke had the following ratings--


Pyth AdjO/Rnk AdjD/Rnk
.9821 123.5/1 87.2/12

Pomeroy projected Duke to beat Tulsa 80-61. The final margin of 70-52 was similar but we were probably less efficient on offense than he expected and more efficient on D. Frankly, our O rating has seemed a tad high to me for a while and I like the "win with defense" mantra so I am cool if we can improve our defensive efficiency.

Coming into the night, Duke was .0023 ahead of Kansas for the #1 ranking in the nation (.9821 - .9798).

I'll be interested in seeing how these stats change. I am not sure if any previous Duke opponent played Thursday night so any change would probably only come from the result of this game, not from other results changing Pom's view of old opponents.

--Jason "by the way, Pom gives Duke only a 59% chance of winning at Maryland but he thinks the Terps game at Va Tech is a tossup" Evans

[kestrel]

For the record, as of this morning Duke edged up to .9827 in today's KenPom, taking a slight nudge offensively (123.0/1) but making up for it on D (86.6/6). Overall, that increases the lead over Kansas to .003 (.9827-.9797).

Can't say I'm unhappy with this direction. I realize that stats are more reflective than predictive, but if we've been trending better and better, then that bodes well for the next month or so.

[bluesin]

Well we went from the #12 adjusted defense to the #6 adjusted defense after the Tulsa game, which makes me happy. I think that just about covers our huge drops after the GT and NCSU games, which did not make me nearly as happy :p.

Actually, I am also addicted to the Kenpom numbers, and I've been keeping an eye on a couple to see how we progress over the season, especially the FTA/FGA number on offense (and to a limited extent defense, though I wasn't ever really concerned with that number). I was worried at the beginning of the year that we were in the low 190s/180s in that particular stat (and briefly in the 200s), but since our switch to an even more "motion" motion offense and the second half of our ACC schedule that number has improved quite a bit (currently 158) and even better it seems to be improving at a fairly consistent rate. I'm hoping that trend continues.

The only area for concern for me at the moment from a statistical point of view is our current 2pt shooting slump. The fact that we haven't really been efficient from 2pt in the last few games has born out in the numbers, as our point distribution has gone from being pretty well balanced in the 3 areas (3's, 2's and FT's) to being more 3pt heavy.

I'm not too discouraged by the change in the numbers, as I know we've been shooting the same amount as we have earlier from both 2 and 3, we just aren't hitting from 2 as much. What I am worried about is more that the team begins to trend toward what is working better and gets away from the 2pt basket attempts, though thankfully I haven't seen that happening yet, and I find it unlikely it will (I have to have something to worry irrationally about right?).

Anyone else have anything they're looking at in the numbers, or any thoughts on what I mentioned I'd be glad to hear it, that's not everything I've found kenpom to offer up this season but I thought going deeper in depth might reveal my dorkitude more than necessary. :D

[uncwdevil]

With Brian's increased minutes lately, he now officially qualifies for Pomeroy's individual stats leaders. His offensive rebouding percentage of 23.0 is leading the NATION as he has a slight lead over Demarcus Cousins (a player who is just about as big as Zoubek and a lot more athletic).

http://kenpom.com/leaders.php?c=ORPct

When Zoubek is in the game, he rebounds over 1/5th of our missed shots. Pretty impressive.

[MChambers]

Anyone else have anything they're looking at in the numbers, or any thoughts on what I mentioned I'd be glad to hear it, that's not everything I've found kenpom to offer up this season but I thought going deeper in depth might reveal my dorkitude more than necessary. :D

I'm a dork, too. I think Duke is improving at the turnover differential lately, turning the ball over less but also forcing a higher % of turnovers.

We're also putting the opposition on the line a little less, except for Miami, which may be attributable to Zoubs's newfound ability to avoid silly fouls.

[bird]

In games where we've "laid an egg," we did so defensively, not offensively. We've had plenty of off nights from the 3-point line this year, and we've won most (if not all) of those games. The fact is, in our four losses this year we've shot an aggregate 41% from three, which is not bad at all. You are talking about Duke teams from the past, not this year's model.

I've enjoyed looking at individual game statistics available by clicking "GamePlan" on the Duke page in KenPom's website. Indeed, these numbers show poor defensive efficiency in three of the four losses: Wisc (1.15); NC State (1.27) and Georgetown (1.21). Ga. Tech was a mediocre but not poor 1.05; our offensive efficiency in that game was the third lowest of the year (.99). Stats say that Ga. Tech was primarily an offensive problem.

Duke's offensive efficiency numbers have held up despite poor shooting: last night we were at respectable 1.08 offensive efficiency despite 32.8 percent shooting. Offensive rebounding, drawing and converting free throws, decent three-point productivity and lack of turnovers are bouying the numbers.

To quote K, if our shooting percentages go up to normal levels, we will be on the "yellow brick road" indeed. Our defensive efficiency is holding up here late in the season (no signs of the defensively-led collapses of recent seasons); and the other pieces of the puzzle are performing at a high level and if you add good shooting our offensive could be become, ahem --

A MONSTER.

That's that the geek stats are shouting right now.

[cptnflash]

To quote K, if our shooting percentages go up to normal levels, we will be on the "yellow brick road" indeed.

What if our shooting percentage continues to go down? I'm generally skeptical of the whole "heavy minutes = tired legs = poor shooting" argument we hear on TV all the time, but it's worth noting that three of our four worst shooting performances (as measured by Pomeroy's effective FG%) have come since Feb 10th (against UNC, VT, and last night). And eFG% has an 0.82 correlation with offensive efficiency, significant at 99%. Scary. I hope I'm wrong to be worried about this. Will need to watch closely for signs of improvement.

[Chitowndevil]

Many here probably already know this, but I was asked by a friend recently how Duke can shoot such a low percentage (particularly from 2) in many games and still be #1 in Pomeroy's Offensive Efficiency ratings. I thought I'd post my answer here too.

Pomeroy's offensive and defensive efficiency are rough estimates of points scored and allowed, respectively, per 100 possessions. From Pomeroy's stats page (http://kenpom.com/blog/index.php/weblog/stats_explained/),


We can estimate possessions very well from box score stats by using this formula.

FGA-OR+TO+0.475xFTA

Note the role played by offensive rebounds and turnovers. Duke is only 97th in the country in effective FG% (field goal percentage treating made 3s as 1.5 baskets). However, they are 15th in the country in Turnover %, turning the ball over on only 16.5% of their possessions. They also have 3 players with double digit OR% (percentage of missed shots rebounded), including Zoubek at 23%, and not including Lance Thomas's 9.7%.

Here's what this does. Ignore 3's for the moment and suppose a team is shooting 50% from the field. But suppose they also turn the ball over on 25% of their possessions (that is, one turnover per four field goal attempts). All else equal, that team is scoring the same points per possession as a team that shoots 40% and NEVER turns it over.

Now suppose that team shooting 50% and gets an offensive rebound on one third of its missed shots. All else equal, that team is scoring the same points per possession as a team that shoots 75% from the field and NEVER gets an offensive rebound.

If you want a 'real' comparison, look at Duke vs. Syracuse. The 'Cuse are shooting 58% from inside the arc, tops in the country. Duke is shooting 46.9%, 202nd in the country. However, Duke's offensive efficiency is higher. Part of that is because Duke is better from 3 point range. It is also (moreso) because Syracuse is turning the ball over on 21.8% of its possessions, and has only one player in double digit OR% (Rick Jackson at 13.5%, followed by Onuaku at 9.9%).

[jimsumner]

Note that Duke had a whopping five turnovers last night against Tulsa. That kind of ball possession is exceptional.

Not impressed by the competition? Duke had 7 turnovers against Maryland, a team that presses from the time they get off the bus.

Duke isn't always this good with the ball, of course. But Duke has about 100 fewer turnovers than the competition. That bodes well for March.

[patentgeek]

The low turnover percentage Jim notes is particularly impressive in view of the fact that no less than six of Duke's ACC opponents (VT, GT, FSU, Clemson, Md and Wake) as well as Wisconsin are ranked in the top 30 in defensive efficiency. In other words, these low turnover numbers are coming against what we should probably assume are tough defenses.

[bird]

Having been watching the statistics this season, primarily Kenpom, I have been moving toward the conclusion that Duke is better than what our deceiving eyes tell us. One thread suggested that Duke is a "weird" looking team. I have suggested elsewhere that the lack of an obvious NBA body (such as a Lawal or Favors) creates an instinctive feeling that a lid is on the team's potential. Now, I will acknowledge that in prior years when we flamed out the computers loved Duke in a similar fashion, but we have not seen any late season fade yet as measured by efficiency numbers. Shooting percentage yes, efficiency no. I have duplicated a late season scatter plot similar to what the Chronical did, and replicated its then-finding of a lack of fade in ACC play. And with Tulsa the lack of fade is further confirmed.

I am very much an optimist. Despite having personally attended the Villanova game last year at an all-in total cost of $2,000.

[JasonEvans]

The low turnover percentage Jim notes is particularly impressive in view of the fact that no less than six of Duke's ACC opponents (VT, GT, FSU, Clemson, Md and Wake) as well as Wisconsin are ranked in the top 30 in defensive efficiency. In other words, these low turnover numbers are coming against what we should probably assume are tough defenses.

Worth noting that Duke overall SOS according to KenPom is #3.

But really worth noting that KenPom says our opponents are the toughest in the nation at defense. Our opponent's D rank is 94.5, #1.

So, as amazing as it is to ponder, Duke has the nation's #1 most efficient offernse despite playing against the #1 most efficient defense.

Folks, that is stunning.

--Jason "Sagarin also rates us with the #3 SOS while the RPI says our SOS is #4" Evans

[pfrduke]

Worth noting that Duke overall SOS according to KenPom is #3.

But really worth noting that KenPom says our opponents are the toughest in the nation at defense. Our opponent's D rank is 94.5, #1.

So, as amazing as it is to ponder, Duke has the nation's #1 most efficient offernse despite playing against the #1 most efficient defense.

Folks, that is stunning.

--Jason "Sagarin also rates us with the #3 SOS while the RPI says our SOS is #4" Evans

Well, we actually have the #1 offense, according to the Pomeroy numbers, because we've played the toughest defensive schedule. In raw terms, our efficiency is 5th in the nation, behind Notre Dame, Kansas, Utah State, and St. Mary's. When you adjust for strength of schedule, 5th best offense against toughest set of defenses = most efficient offense, all else equal.

[patentgeek]

Worth noting that Duke overall SOS according to KenPom is #3.

But really worth noting that KenPom says our opponents are the toughest in the nation at defense. Our opponent's D rank is 94.5, #1.

So, as amazing as it is to ponder, Duke has the nation's #1 most efficient offernse despite playing against the #1 most efficient defense.

Folks, that is stunning.

--Jason "Sagarin also rates us with the #3 SOS while the RPI says our SOS is #4" Evans

I think these efficiency numbers may be related, i.e., that one of the reasons Duke's offensive efficiency is so high is Duke's performance against really good defenses. In other words, with exactly the same points/game and pace numbers, had Duke been playing against worse defenses, Duke's offensive efficiency would be lower. Don't get me wrong - having the No. 1 offense is great - but it's less surprising because we've been playing against strong defenses.

[JasonEvans]

I think these efficiency numbers may be related, i.e., that one of the reasons Duke's offensive efficiency is so high is Duke's performance against really good defenses. In other words, with exactly the same points/game and pace numbers, had Duke been playing against worse defenses, Duke's offensive efficiency would be lower. Don't get me wrong - having the No. 1 offense is great - but it's less surprising because we've been playing against strong defenses.

And I feel like a total fool for not recognizing that. Duuh! You and pfr are much bigger dorks than me!

;)

--Jason "Can my dorkness be revoked?" Evans

[-jk]

--Jason "Can my dorkness be revoked?" Evans

To quote one of our usernames: Uh, No.

-jk

[ArtVandelay]

A couple of things about some of the stuff said in this thread:

1) SOS - It's true that part of the reason for our strong SOS is the lack of games against really bad non-conference teams, but we've also played a lot of teams that KenPom thinks are very good. What I find REALLY interesting is how much more KenPom likes the ACC as compared to the national polls:

Maryland - 13
Florida St. - 15
Clemson - 17
Ga Tech - 20

Granted, we've played none of the top 10, but when you throw in G'Town (16), we've played a quarter of the top 20. Not too shabby.

2) Unless I'm wrong, KenPom doesn't take into account margin of victory (maybe Sagarin does), so blowouts don't have much to do with it. It's just an aggregate of all offensive and defensive possessions on the season.

3) Does anyone know if it's possible to locate the pre-tournament KenPom rankings from years past? One problem I have with looking at the historical data is that the previous year's rankings are going to be skewed a bit in favor of teams that succeeded in the tournament, which limits their predictive value.

[patentgeek]

And I feel like a total fool for not recognizing that. Duuh! You and pfr are much bigger dorks than me!

;)

--Jason "Can my dorkness be revoked?" Evans

I can't speak for pfr (and wouldn't want to defame him), but I'm sure that anyone I know personally would say I'm a much bigger dork than you.

[InSpades]

A couple of things about some of the stuff said in this thread:

1) SOS - It's true that part of the reason for our strong SOS is the lack of games against really bad non-conference teams, but we've also played a lot of teams that KenPom thinks are very good. What I find REALLY interesting is how much more KenPom likes the ACC as compared to the national polls:

Maryland - 13
Florida St. - 15
Clemson - 17
Ga Tech - 20

Granted, we've played none of the top 10, but when you throw in G'Town (16), we've played a quarter of the top 20. Not too shabby.

2) Unless I'm wrong, KenPom doesn't take into account margin of victory (maybe Sagarin does), so blowouts don't have much to do with it. It's just an aggregate of all offensive and defensive possessions on the season.

3) Does anyone know if it's possible to locate the pre-tournament KenPom rankings from years past? One problem I have with looking at the historical data is that the previous year's rankings are going to be skewed a bit in favor of teams that succeeded in the tournament, which limits their predictive value.

On point #2... I think KenPom does take into account margin of victory. I admittedly haven't figured out exactly how he does his rankings but... if you click the read this (http://kenpom.com/blog/index.php/weblog/ratings_explanation/) link from the ratings site it talks about not capping margin of victory (so I assume this means it counts).

On point #1... I'm not sure but I would guess that KenPom rankings give the ACC a lot of credit for UNC's demise. I mean UNC performed well against other top teams (Texas, Ohio St., etc.) and then bombed in the conference. The ACC teams beating UNC must get credit for that, no? In reality I think it's just UNC falling apart but the computer doesn't see that.

[uncwdevil]

On point #2... I think KenPom does take into account margin of victory. I admittedly haven't figured out exactly how he does his rankings but... if you click the read this (http://kenpom.com/blog/index.php/weblog/ratings_explanation/) link from the ratings site it talks about not capping margin of victory (so I assume this means it counts).

On point #1... I'm not sure but I would guess that KenPom rankings give the ACC a lot of credit for UNC's demise. I mean UNC performed well against other top teams (Texas, Ohio St., etc.) and then bombed in the conference. The ACC teams beating UNC must get credit for that, no? In reality I think it's just UNC falling apart but the computer doesn't see that.

Actually, my understanding is that KenPom doesn't consider margin of victory per se, or even wins and losses for that matter. His ratings are solely based on per-possession performance. However, you could say that margin-of-victory is baked-in to his ratings if you consider a team's per-possession stats are going to be more impressive in a blow-out than a close game. He does look at record at the back end when calculating the "luck" stat. Based on his method, "luckier" teams have won more games than their per-possession stats indicate that they should have.

[ArtVandelay]

On point #2... I think KenPom does take into account margin of victory. I admittedly haven't figured out exactly how he does his rankings but... if you click the read this (http://kenpom.com/blog/index.php/weblog/ratings_explanation/) link from the ratings site it talks about not capping margin of victory (so I assume this means it counts).

On point #1... I'm not sure but I would guess that KenPom rankings give the ACC a lot of credit for UNC's demise. I mean UNC performed well against other top teams (Texas, Ohio St., etc.) and then bombed in the conference. The ACC teams beating UNC must get credit for that, no? In reality I think it's just UNC falling apart but the computer doesn't see that.

Interesting - I'll have to look into point #2 more. My understanding was always that the only way lopsided wins factor in is that in blowouts you're much more likely to have a lot of productive offensive possessions, thereby skewing the stats.

As for Point#1 - KenPom's rankings don't give you any "credit" per se based on who you beat. It's completely neutral as to the identity, as it were, of your opponents, obviously. It just looks at the quality of your competition overall to create the adjusted offensive and defensive efficiencies. So the only way the ACC gets "credit" for beating Carolina is if Carolina is well-ranked in the KenPom rankings.

Which brings me to Carolina. I think KenPom's rankings are revealing in some ways. For one, UNC has been one of the unluckiest teams in the country. That is, they have badly under-performed their pythagorean record. Their overall ranking is only 61, which is not very good, but note that their record is worse than a lot of the teams that are ranked near them. I think it's fair to say that UNC "should" have a slightly better record than what they do have, but they still are not "good." Also, the tempo-free stats definitely prove what your eyes show you - they are simply awful when it comes to taking care of the ball.

[MChambers]

I guess Duke must've emphasized defense in practice yesterday, because they've climbed from 6th to 4th in defense in Pomeroy, without playing a game. I guess that means that on average the teams we've played had good offensive days yesterday, or that some of the teams ahead of us had bad defensive games (Kansas, for example).

[Jumbo]

Duke remains first in offensive efficiency according to kenpom, and is now third in defensive efficiency. Remarkable.

[loldevilz]

I guess Duke must've emphasized defense in practice yesterday, because they've climbed from 6th to 4th in defense in Pomeroy, without playing a game. I guess that means that on average the teams we've played had good offensive days yesterday, or that some of the teams ahead of us had bad defensive games (Kansas, for example).

I believe that before we were tied for fifth and only .4 behind #4. Duke went up .8 (I think) holding another opponent to their season low in scoring. I still astonishes me that we are first in scoring considering that we hardly ever hit 80 points in games and have hit 70 only in half of the last 8 or so. Clearly K's changes to the offense are working very effectively.

[tele]

Clearly K's changes to the offense are working very effectively.

Not to mention the defensive changes as well. Less pressure on the ball and overplaying the passing lanes and more contesting the three and stopping dribble penetration. Have to give credit to Zoubek and Thomas for being mobile enough on D to play this way, and also versatile enough to switch and/or show out on the perimeter. (also has a remarkable tendency to keep your primary scorers out of foul trouble, not a bad thing just by itself).

The offensive number must put a good deal of weight on the low turnover rate and to a lesser degree, offensive rebounds.

Works for me. Go Duke!

[gumbomoop]

Duke remains first in offensive efficiency according to kenpom, and is now third in defensive efficiency. Remarkable.

Remarkable ain't the half of it. Not that I fully [or even half-ly] understand all the dorkness, but all the DBR poster-dorks have sure got my attention with all this KenPom stuff. It's way more interesting than the regular polls.

The following, from KenPom his very self also got my attention early this morn: "Just so everyone can prepare - barring a hellacious draw, Duke will be the favorite to win the NCAA's when we do log5 on BP in two weeks."

In my own glass-9/10ths-full loony optimism, I'm of course willing to accept KenPom at his numbers. It's charming that KenPom feels compelled to "prepare" folks for declaring Duke the favorite; presumably his numbers catch hell from ubiquitous Duke-haters. [Not to mention from understandably skeptical, because previously burned, DBR posters.]

Back to Jumbo's point above, referring to KenPom's efficiency rankings: as a dork-wannabe who must for now still rely principally on the eye test, I can see, literally, our defensive efficiency, but I have to look a lot more closely to see the O-eff. I credit numerous DBR posters for their good advice on what to look for.

Maybe our guys still sometimes look inconsistent on O because...... what?..... I want every single play to work? I gotta learn more dork.

[91_92_01_10_15]

The following, from KenPom his very self also got my attention early this morn: "Just so everyone can prepare - barring a hellacious draw, Duke will be the favorite to win the NCAA's when we do log5 on BP in two weeks."


Anyone know how accurate Pomeroy's system has been in predicting past Final Fours/champs?

[ice-9]

What's interesting to me is how Coach K is careful to call this team "good but not great." That's the theme of his assessment of this team for the entire season. Why is that? Is it because he truly believes that this team isn't there yet, or is it a psychological tactic?

All the numbers seem to indicate Duke is a top team, a national title contender, but Coach K's words seem to caution against that.

I vaguely remember him being a lot more positive and effusive in praise in the strength of previous teams.

[gumbomoop]

What's interesting to me is how Coach K is careful to call this team "good but not great." That's the theme of his assessment of this team for the entire season. Why is that? Is it because he truly believes that this team isn't there yet, or is it a psychological tactic?

All the numbers seem to indicate Duke is a top team, a national title contender, but Coach K's words seem to caution against that.

I'm pretty sure K thinks his guys are near-top and contenders. He's said publicly, however, what the general consensus is, namely, that there are no great teams this year, just a lot of very good ones.

For comparison, think UNLV '90, UNLV '91 [a great team that lost to a pretty good one], Duke '92 [a great team that won a great game against a decent team], others. Or think how many potentially great teams we might see next year if "everybody came back": UK, Duke, UNC [well, everybody but Roy.....]

[kong123]

Duke is only 1-2 against teams in the top 25. While they have a strong overall SOS, they have only played 3 top 25 teams and only 1 of which has been in the top 10 at any point during the year. While all of these statistics that you are throwing out are interesting, do not forget that Duke has not played anyone that has a chance to go to the elite 8 other than maybe Georgetown, and that depends on which GU team shows up.

[Duvall]

Duke is only 1-2 against teams in the top 25.

Top 25 of what? No matter the answer, I'm pretty sure that statement will be wrong by noon.

[davekay1971]

Duke is only 1-2 against teams in the top 25. While they have a strong overall SOS, they have only played 3 top 25 teams and only 1 of which has been in the top 10 at any point during the year. While all of these statistics that you are throwing out are interesting, do not forget that Duke has not played anyone that has a chance to go to the elite 8 other than maybe Georgetown, and that depends on which GU team shows up.

Yeah, but we beat Carolina, and they're number 4 in the nation.

Oh, wait, that's right...that was preseason number 4. Then some games were actually played and Carolina turned out to suck. You're right kong, at least one of our 25 wins was against a team with absolutely no chance to get to the elite 8...because to get to the elite 8 you actually have to get in the NCAA tournament. Better luck on making the tournament next year. But, hey, congrats on getting that fourth ACC win. Quite an accomplishment...you guys are down to, what, 2 McDonald's All-Americans per ACC win now, right?

[kong123]

Top 25 of what? No matter the answer, I'm pretty sure that statement will be wrong by noon.

Glad you are counting on an unranked team to crack the top 25. Still no teams in the top 10. In fact, At this moment, Duke has played the least amount of games against top 25 teams. You have to go all the way down to Northern Iowa to find a team that has played less top 25 teams. I know, I know, wait till noon!

[airowe]

Duke is only 1-2 against teams in the top 25. While they have a strong overall SOS, they have only played 3 top 25 teams and only 1 of which has been in the top 10 at any point during the year. While all of these statistics that you are throwing out are interesting, do not forget that Duke has not played anyone that has a chance to go to the elite 8 other than maybe Georgetown, and that depends on which GU team shows up.

And yet Duke has more wins against the RPI Top 50 than any team in the nation. As for teams that have a chance to make it to the Elite 8, can you please show us your criteria for this?

When you're done, please show us carolina's record against the RPI Top 100.

[Duvall]

Glad you are counting on an unranked team to crack the top 25.

Well, I think they are better than most of the teams ranked last week. So there's that.

And why are you talking about the human polls in a thread about the computer ratings?

[kong123]

Congrats, Duke has the same chance of winning the NC as does UNC this year.... NONE

oh yeah, we've won two NC's in the last five years. Enjoy our short downfall, we know how you have suffered through our recent success.

[gumbomoop]

Duke is only 1-2 against teams in the top 25. While they have a strong overall SOS, they have only played 3 top 25 teams and only 1 of which has been in the top 10 at any point during the year. While all of these statistics that you are throwing out are interesting, do not forget that Duke has not played anyone that has a chance to go to the elite 8 other than maybe Georgetown, and that depends on which GU team shows up.

Fair enough, but when you say "you are throwing out," you mean, "you [we] are citing from KenPom." And when you say "interesting," I'm guessing you mean "not really."

You are fairly raising the "eye test," I assume. And honestly, there's been a pretty heated debate [that I describe as "friendly fire"] on DBR as this season has progressed about our NCAAT hopes. But the eye test, it turns out, requires looking carefully, especially because this team seems to play much more efficiently than its sometimes poor FG% would suggest.

[kong123]

Well, I think they are better than most of the teams ranked last week. So there's that.

And why are you talking about the human polls in a thread about the computer ratings?

I am talking about RPI and that is a component of how the RPI is formed

[uncwdevil]

I am talking about RPI and that is a component of how the RPI is formed

Uh, no it's not. Maybe you are thinking about the BCS?

[davekay1971]

Uh, no it's not. Maybe you are thinking about the BCS?

C'mon, don't bother the guy with facts and accuracy, or any desire for evidence. Kong's proclamations, alone, are gospel. Of all Duke's opponents to date, only Georgetown as any chance of making the Elite 8. And Duke has NO chance of winning the NC. Kong said it. Bank, baby.

[gumbomoop]

Congrats, Duke has the same chance of winning the NC as does UNC this year.... NONE

oh yeah, we've won two NC's in the last five years. Enjoy our short downfall, we know how you have suffered through our recent success.

I hope you will at least consider doing 2 things:

1. Email KenPom to inform him he's an idiot; i.e., there's no way KU and 'Cuse should be rated that high, so his numbers must be cooked.

2. If you want to engage in a serious debate, upgrade your tone ["Congrats...... NONE....... oh yeah"] from that of the typical I[diot]C-poster. I[diot]C is pathetic.

[kong123]

I hope you will at least consider doing 2 things:

1. Email KenPom to inform him he's an idiot; i.e., there's no way KU and 'Cuse should be rated that high, so his numbers must be cooked.

2. If you want to engage in a serious debate, upgrade your tone ["Congrats...... NONE....... oh yeah"] from that of the typical I[diot]C-poster. I[diot]C is pathetic.


Go and read some of the posts under all the "Pile on Roy" or "UNC is Sucks" thread, the only difference is that I am a UNC fan.

[davekay1971]

I hope you will at least consider doing 2 things:

1. Email KenPom to inform him he's an idiot; i.e., there's no way KU and 'Cuse should be rated that high, so his numbers must be cooked.

2. If you want to engage in a serious debate, upgrade your tone ["Congrats...... NONE....... oh yeah"] from that of the typical I[diot]C-poster. I[diot]C is pathetic.

I love your first idea.

As for point number 2, I must at this point confess that I probably egged the Konger into that kind of post. Mea culpa for my part in momentarily dragging DBR down to IC's level. Of course, when a Carolina fan comes over on a Duke board for the sole purpose, with each and every post he/she makes, of taking shots at Duke, I get a little feisty.

[gumbomoop]

Go and read some of the posts under all the "Pile on Roy" or "UNC is Sucks" thread, the only difference is that I am a UNC fan.

I have myself penned several of those posts, usually with one of 2 objectives:

1. to suggest, seriously, that Roy "stop digging" himself deeper and deeper into a hole regarding his sanity, his relationship to his players, his 3d-person-self-obsession; he's come too close to some uncomfortable edge this year.

2. to make a joke, with a twinkle in the eye. Said attempts at humor, however frail, do not generally begin with "Oh yeah."

As I read your posts in this thread, they are neither serious analysis nor leavened by any hint of whimsy. You seem upset that Duke is doing well, that Carolina isn't. The mere thought that "Duke isn't dead" is driving you nuts. Just as - again, fair's fair - the mere thought that your team might somehow gain momentum from the Wake game, well, that gives me the willies.

[BlueintheFace]

oh yeah, we've won two NC's in the last five years. Enjoy our short downfall, we know how you have suffered through our recent success.

Easy Tiger


Rational Sounding but Trolling - Snarky and semi-insulting comments mainly to get under the skin of posters.

[JasonEvans]

Kong has been given a timeout for what was clearly an attempt to troll. I had considered deleting all his posts and the various responses, but ya'll did a nice job of staying on the high road and refuting what he had to say so I left them up.

I would hope we can now return this conversation to the highly sophisticated computer analysis that continues to say that Duke is the #1 team in the land and the most likely to win the national title. Woo hoo!!

-Jason "#1 in the Sagarin predictor model... that matters!" Evans

[davekay1971]

I would hope we can now return this conversation to the highly sophisticated computer analysis that continues to say that Duke is the #1 team in the land and the most likely to win the national title. Woo hoo!!

-Jason "#1 in the Sagarin predictor model... that matters!" Evans

Comparing the "Eyeball Test" to the computers...

One of the frequent comments regarding Duke is that we just don't pass the "Eyeball Test" as a NC contender. Games like the G'town game and NCSU game tend to stick in the mind. Too many games with poor shooting. Too much reliance on the Big 3, too little offensive contribution from the inside, too many instances of quick guards able to get penetration, etc. So, basically, sophisticated, objective computer analysis says one thing. Highly subjective observational analysis based on experience watching hoops says something completely different to many.

My question...which of the other top contenders does pass the eyeball test? Kenpom #3 Syracuse does pass my eyeball test. They have a number of great pieces, can produce nicely on offense, and are playing Boeheim's zone defense about as well as it can be played. Kenpom #2 Kansas also passes the eyeball test...just really good players, good experience, relatively consistent, strong on both sides of the ball.

Kentucky seems to pass the eyeball test of a lot of experts, but they don't score great on Kenpom, and they don't pass my amatuer eyeball test - they have incredible athletes, but look like a team that could win the title or be taken out in the second round. Not what you want from your NC favorite.

Kenpom's teams 4-9 don't pass my eyeball test as being real NC contenders. Purdue, of course, took a hit with Hummel's injury, but the others just don't seem to have "it". I won't go into detail on each because this post is already too long. But I'll just post the simple question...in looking at teams other than Duke, outside of Syracuse and Kansas, do any of Kenpom's top 10 pass the eyeball as a better NC contender than Duke?

[roywhite]

Short answer to the "eyeball test"---no, other than Kansas and Syracuse, don't see other teams that appear to be playing better overall than the Blue Devils. One possible disclaimer---the ACC just isn't that good? and the Georgetowns and Villanovas of the Big East are better teams that face tougher opponents and pick up losses occasionally?

One key factor in Duke's success IMO is the perimeter defense, specifically against 3-point shots. Opponents are shooting less than 27% from 3-point, a Duke school record so far for lowest percentage allowed. At the same time, even with some cold games, Duke is shooting approx. 39% from 3-point. Huge difference...is that apparent when watching the games or too subtle for the eyeball test?

[gumbomoop]

...... what he had to say....

I would hope we can now return this conversation to the highly sophisticated computer analysis that continues to say that Duke is the #1 team in the land and the most likely to win the national title. Woo hoo!!

-Jason "#1 in the Sagarin predictor model... that matters!" Evans

I got 2 dogs in this hunt, so:

1. Wait, what the hell are you talking about - "what he had to say"? Are you nuts?

Sorry. The idea of re-perusing several posts to find out what "was said" has temporarily affected my composure.

2. Back on point: dork-land. Even I cannot quite believe/trust KenPom's prediction. But unlike our recent visitor, and ubiquitous Duke-skeptics the world 'round, I do have to admit the obvious: his numbers for several teams that sure look eye-tested are real high, and so "make sense" even outside dork-land. And his numbers seem to account for our bewildering winning ways on bad-shooting nights. Or more accurately, his numbers seem to take into account - indeed, measure - those underappreciated factors that overbalance the shooting deficiencies. Our guys apparently have a striking number of efficiencies to counter our one [or is it more?] deficiency.

[COYS]

One key factor in Duke's success IMO is the perimeter defense, specifically against 3-point shots. Opponents are shooting less than 27% from 3-point, a Duke school record so far for lowest percentage allowed. At the same time, even with some cold games, Duke is shooting approx. 39% from 3-point. Huge difference...is that apparent when watching the games or too subtle for the eyeball test?

I think this is an important stat that is overlooked by many in the national media. Our defense is configured very differently than in previous seasons and we do a great job taking away the oppositions' ability to score in transition or hit the three, which are the two worst things a defense can give up. Transition buckets are generally high percentage shots, but we do a good job protecting the basketball and limiting these opportunities. The three point shot has been called the great equalizer for a reason (Long Island game last year, anyone?), yet we've all but managed to eliminate it as a threat against us most games this season. This means anyone who wants to beat Duke has to beat us with their half court offense without the three point shot. With our size and rebounding ability, I'll take that matchup against most teams. As long as we continue to take care of the ball on offense, our defense matches up really well with just about anyone.

[MChambers]

Short answer to the "eyeball test"---no, other than Kansas and Syracuse, don't see other teams that appear to be playing better overall than the Blue Devils. One possible disclaimer---the ACC just isn't that good? and the Georgetowns and Villanovas of the Big East are better teams that face tougher opponents and pick up losses occasionally?

One key factor in Duke's success IMO is the perimeter defense, specifically against 3-point shots. Opponents are shooting less than 27% from 3-point, a Duke school record so far for lowest percentage allowed. At the same time, even with some cold games, Duke is shooting approx. 39% from 3-point. Huge difference...is that apparent when watching the games or too subtle for the eyeball test?

That's the best three point shooting defense in the country, in terms of percentage. Not something I expected, given Duke's personnel.

[roywhite]

That's the best three point shooting defense in the country, in terms of percentage. Not something I expected, given Duke's personnel.

Had to laugh the other day when one of the TarHeel players (Drew?) was explaining their success against Wake Forest and said something about how the coaches "had really emphasized getting up on the shooters"... Well, duh...

Among other things, I think our perimeter defense numbers are testimony to what good, underrated defenders Jon Scheyer and Kyle Singler are (and that's not to undermine Nolan Smith...just that he probably is recognized more generally as a good on-the-ball defender).

[jv001]

Had to laugh the other day when one of the TarHeel players (Drew?) was explaining their success against Wake Forest and said something about how the coaches "had really emphasized getting up on the shooters"... Well, duh...

Among other things, I think our perimeter defense numbers are testimony to what good, underrated defenders Jon Scheyer and Kyle Singler are (and that's not to undermine Nolan Smith...just that he probably is recognized more generally as a good on-the-ball defender).

And having a better inside game with better inside players helps in that regard also. Our guards do not have to help as much in defending the paint because we have bodies and fouls that we can throw at our opponents this year. This allows our perimeter to stay home more. Haven't had that luxury in a long time. Go Duke!

[gumbomoop]

One key factor in Duke's success IMO is the perimeter defense, specifically against 3-point shots. Opponents are shooting less than 27% from 3-point, a Duke school record so far for lowest percentage allowed. At the same time, even with some cold games, Duke is shooting approx. 39% from 3-point. Huge difference...is that apparent when watching the games or too subtle for the eyeball test?

Excellent response to roywhite's final question here by COYS at post #100.

I'll add: It's too subtle even for most who watch lots of games, even DBR posters who watch every second of Duke games, as evidenced by the fact that when we hear the numbers, we [most of us, me] say, "Wow," which means, "I didn't have any idea."

It's been really interesting reading so many threads whose underlying theme is, "Are we for real?" I'm a near-total optimist, yet until other posters added lots of informed detail to my half-understanding of our weird strengths, I couldn't figure us out.

So now I'm thinking "outlier theory" : The [I]only 2 reasons we could possibly lose again this year are (1) an outlier performance by Duke on D [= outlier performance by opponent on O??], or (2) ACC is outlier-mediocre, and we really haven't been tested, and we're in for an unpleasant awakening, a la West Va/Nova.

[NSDukeFan]

Had to laugh the other day when one of the TarHeel players (Drew?) was explaining their success against Wake Forest and said something about how the coaches "had really emphasized getting up on the shooters"... Well, duh...

Among other things, I think our perimeter defense numbers are testimony to what good, underrated defenders Jon Scheyer and Kyle Singler are (and that's not to undermine Nolan Smith...just that he probably is recognized more generally as a good on-the-ball defender).

I agree with what you and COYS said in your last post(s). I agree that Jon and Kyle don't get the credit they deserve as smart 40 minute solid defenders. I wouldn't call them stoppers individually, but add in some help defense in a good defensive scheme and they do a very good job of containing whoever they are guarding. I believe at the beginning of the year, the concern with Kyle was defending his position without concern about offense. My impression has been that Kyle has been very solid defending perimeter players in our defensive scheme.
All of this is too boring to pass the eyeball test of big dunks, huge blocks, loads of steals and ankle breaking crossovers, however.

[Kedsy]

My question...which of the other top contenders does pass the eyeball test? Kenpom #3 Syracuse does pass my eyeball test. They have a number of great pieces, can produce nicely on offense, and are playing Boeheim's zone defense about as well as it can be played. Kenpom #2 Kansas also passes the eyeball test...just really good players, good experience, relatively consistent, strong on both sides of the ball.

I'm not entirely sold on Syracuse. Defensively, yes, they pass whatever test you want. But on offense they seem to turn the ball over a lot and they don't shoot free throws very well, and those two issues could be important in a high-pressure tournament game.

Of course, that's only my eye-test assessment. According to Pomeroy their offense is better than their defense (9th in adjusted offense and 10th in adjusted defense).

I've seen Kansas in person and they passed my eye test with flying colors.

[roywhite]

So now I'm thinking "outlier theory" : The [I]only 2 reasons we could possibly lose again this year are (1) an outlier performance by Duke on D [= outlier performance by opponent on O??], or (2) ACC is outlier-mediocre, and we really haven't been tested, and we're in for an unpleasant awakening, a la West Va/Nova.

That's what I'm thinking, yet dare not say it. :)

Glad you did.

[MChambers]

IAnd his numbers seem to account for our bewildering winning ways on bad-shooting nights. Or more accurately, his numbers seem to take into account - indeed, measure - those underappreciated factors that overbalance the shooting deficiencies. Our guys apparently have a striking number of efficiencies to counter our one [or is it more?] deficiency.

KenPom's numbers are just based on points per possession or, more accurately, points per 100 possessions. His more detailed statistics do give you a sense, however, of how Duke is managing to be so efficient on offense: low turnover rate, fantastic offensive rebounding, and very good three point shooting.

[Relics]

...or (2) ACC is outlier-mediocre, and we really haven't been tested, and we're in for an unpleasant awakening, a la West Va/Nova.

Here are our last 10 opponents ranked from best 3-ball offense to worst, then in the second column what we held them to, then the differential (in %):

GTOWN 39.5 46.2 +6.7
UMD 39.3 15.4 -23.9
GT 36.6 25.0 -11.6
Miami 36.6 33.3 -3.3
Tulsa 35.8 10.0 -25.8
UVA 35.3 16.7 -18.6
UNC 34.5 26.3 -8.2
FSU 33.6 30.8 -2.8
BC 33.5 14.3 -19.2
VT 30.4 13.3 -17.1

IMO, it looks like we aren't stopping reasonably better shooting teams any better or worse than marginal or poor shooting teams. (And look at that Maryland win!) The Georgetown loss, of course, being the exception, which fits in with your OOC haven't-been-tested litmus test theory to an extent.

Wednesday's game for peace-of-mind gold?

[ArtVandelay]

Comparing the "Eyeball Test" to the computers...

One of the frequent comments regarding Duke is that we just don't pass the "Eyeball Test" as a NC contender. Games like the G'town game and NCSU game tend to stick in the mind. Too many games with poor shooting. Too much reliance on the Big 3, too little offensive contribution from the inside, too many instances of quick guards able to get penetration, etc. So, basically, sophisticated, objective computer analysis says one thing. Highly subjective observational analysis based on experience watching hoops says something completely different to many.

I am really glad that we've been able to have an extended discussion of Pomeory's rating system here. I am far from an expert on tempo-free stats, but I do think some of us are (understandably) missing a few things:

1) The whole point of using a purely objective rating system is to REMOVE the "eyeball test" from the equation. The premise is that our "eyeballs" can be deceiving and almost always tend to bring a whole lot of subjective and biased elements into our analysis. KenPom's system applies a steely-eyed detachment from all of those factors and cares only about the numbers.

2) Going strictly by the eyeball test, I would note that Kansas (probably considered the #1 team by the majority of the eyeballs in the country) got beat down pretty handily by Okla St this weekend, a team that is probably not as good as Georgetown. This, to me, just proves the fallacy of the eyeball test. If you judge a team by one game, you don't get the whole story. Nobody can say they've watched every possession of every game by every team, but KenPom's system has.

3) KenPom's ratings have historically been quite good at picking the top performers in the tournament, but it's somewhat hard to tell exactly because his historical data includes the tournament games (as far as I know). This means that Final Four teams are naturally going to look higher in his end-of-year rankings than before the tournament.

4) Don't over-emphasize what Duke being #1 means. Sure, log5 will predict Duke to win it all on a sheer probability basis. But the difference in the odds for Duke as opposed to other teams at the top is so small that in a one-and-done situation, those predictions don't mean much. Pomeroy doesn't pretend like he can predict the future.

[davekay1971]

I am really glad that we've been able to have an extended discussion of Pomeory's rating system here. I am far from an expert on tempo-free stats, but I do think some of us are (understandably) missing a few things:

1) The whole point of using a purely objective rating system is to REMOVE the "eyeball test" from the equation. The premise is that our "eyeballs" can be deceiving and almost always tend to bring a whole lot of subjective and biased elements into our analysis. KenPom's system applies a steely-eyed detachment from all of those factors and cares only about the numbers.

2) Going strictly by the eyeball test, I would note that Kansas (probably considered the #1 team by the majority of the eyeballs in the country) got beat down pretty handily by Okla St this weekend, a team that is probably not as good as Georgetown. This, to me, just proves the fallacy of the eyeball test. If you judge a team by one game, you don't get the whole story. Nobody can say they've watched every possession of every game by every team, but KenPom's system has.

3) KenPom's ratings have historically been quite good at picking the top performers in the tournament, but it's somewhat hard to tell exactly because his historical data includes the tournament games (as far as I know). This means that Final Four teams are naturally going to look higher in his end-of-year rankings than before the tournament.

4) Don't over-emphasize what Duke being #1 means. Sure, log5 will predict Duke to win it all on a sheer probability basis. But the difference in the odds for Duke as opposed to other teams at the top is so small that in a one-and-done situation, those predictions don't mean much. Pomeroy doesn't pretend like he can predict the future.

Wow, thanks for all that. I really had no basic understanding of the point of a computer-based system like Pomeroy's, what all this probability stuff meant, or why it might be superior to an eyeball test. I can actually feel my knuckles dragging less markedly on the floor as I walk after reading your post.

[gumbomoop]

I am really glad that we've been able to have an extended discussion of Pomeory's rating system here.

4) Don't over-emphasize what Duke being #1 means. Sure, log5 will predict Duke to win it all on a sheer probability basis. But the difference in the odds for Duke as opposed to other teams at the top is so small that in a one-and-done situation, those predictions don't mean much. Pomeroy doesn't pretend like he can predict the future.

Ditto for me on the value of hearing from posters more attuned to KenPom, what his system does [and doesn't], and why it's a useful corrective to simple eye-test.

For me, it's been serendipitous, coming roughly simultaneously with my latching onto, and foisting on this board, the "this team is weird" idea. Weirdly assembled and weirdly winning, I thought. KenPom, and especially fellow posters' translations thereof, do seem to be able to explain the weirdness, which probably really means to demystify it, and to show its complexities. Turns out, if I understand what others have explained, that weirdness is a misleading, if innocent and convenient, handle to summarize several well-coached and relentlessly well-executed efficiencies. Yes? If you don't do it relentlessly ["Duke plays every play"], it isn't likely to be so efficient. Yes?

(4) KenPom's prediction - Ok, important point that he's not exactly predicting Duke as NC. Still, that he would actually use the word "predict" is intriguing; moreover, that Duke may be near-top in his pre-NCAAT rankings means that this Duke bunch is better, over the season so far, than the simple eye-comparison with the "wow" teams [KU, UK, Cuse, Nova, a couple others, intermittently] would suggest. Yes?

[Dukeface88]

I am really glad that we've been able to have an extended discussion of Pomeory's rating system here. I am far from an expert on tempo-free stats, but I do think some of us are (understandably) missing a few things:

1) The whole point of using a purely objective rating system is to REMOVE the "eyeball test" from the equation. The premise is that our "eyeballs" can be deceiving and almost always tend to bring a whole lot of subjective and biased elements into our analysis. KenPom's system applies a steely-eyed detachment from all of those factors and cares only about the numbers.

2) Going strictly by the eyeball test, I would note that Kansas (probably considered the #1 team by the majority of the eyeballs in the country) got beat down pretty handily by Okla St this weekend, a team that is probably not as good as Georgetown. This, to me, just proves the fallacy of the eyeball test. If you judge a team by one game, you don't get the whole story. Nobody can say they've watched every possession of every game by every team, but KenPom's system has.

3) KenPom's ratings have historically been quite good at picking the top performers in the tournament, but it's somewhat hard to tell exactly because his historical data includes the tournament games (as far as I know). This means that Final Four teams are naturally going to look higher in his end-of-year rankings than before the tournament.

4) Don't over-emphasize what Duke being #1 means. Sure, log5 will predict Duke to win it all on a sheer probability basis. But the difference in the odds for Duke as opposed to other teams at the top is so small that in a one-and-done situation, those predictions don't mean much. Pomeroy doesn't pretend like he can predict the future.

On the other hand, my understanding is that the computers also have no way of knowing about personnel changes (read, injuries). That may cause the rankings to be off for specific teams as players leave and return. I would guess that Ohio State is currently underrated and Purdue overrated by the dork polls.

[roywhite]

Well, all this stuff can't be totally esoteric. Digger Phelps has jumped on the Duke bandwagon.

[gumbomoop]

Well, all this stuff can't be totally esoteric. Digger Phelps has jumped on the Duke bandwagon.

I'd like to start the rumor that Digger studies KenPom. [Shoot, I'd like to start the rumor that I study KenPom].

But I can already hear the wiseacres: "Dude, Ol' Roy the First's horse is more likely to be a KenPom aficionado than Digger."

So I won't start any rumors in this post.

[ArtVandelay]

Ditto for me on the value of hearing from posters more attuned to KenPom, what his system does [and doesn't], and why it's a useful corrective to simple eye-test.

For me, it's been serendipitous, coming roughly simultaneously with my latching onto, and foisting on this board, the "this team is weird" idea. Weirdly assembled and weirdly winning, I thought. KenPom, and especially fellow posters' translations thereof, do seem to be able to explain the weirdness, which probably really means to demystify it, and to show its complexities. Turns out, if I understand what others have explained, that weirdness is a misleading, if innocent and convenient, handle to summarize several well-coached and relentlessly well-executed efficiencies. Yes? If you don't do it relentlessly ["Duke plays every play"], it isn't likely to be so efficient. Yes?

(4) KenPom's prediction - Ok, important point that he's not exactly predicting Duke as NC. Still, that he would actually use the word "predict" is intriguing; moreover, that Duke may be near-top in his pre-NCAAT rankings means that this Duke bunch is better, over the season so far, than the simple eye-comparison with the "wow" teams [KU, UK, Cuse, Nova, a couple others, intermittently] would suggest. Yes?

As to both of your points, I would say that yes, the KenPom numbers reveal that Duke is, according to his system anyway, a better team than your eye-test (or the national polls, to an extent) gives them credit for. But note that Duke is ranked #5 in the country, so it's not like we're talking about big differences here.

As to why this is, the main point is that we as fans often tend to get fixated on the wrong things. E. just one g., rebound margin (cited, as it happens, on the front page wrap-up of the UVA game). This is a pretty meaningless statistic because it is not rate-based; that is, it mostly measures how many opportunities you got to rebound, not how often you capitalized on your opportunities. But you hear guys on TV talk about rebound margin all the time like it's important. A much more egregious example is Len Elmore's "index" that he uses to measure players. This is like war-crime level stupid. He arbitrarily decided to weight total points, rebounds, assists, etc. at certain levels with no real justification for it and then comes out with a number that proclaims to determine the best player in the ACC.

KenPom has attempted to use math to try to isolate the factors that are most predictive of success and measure the teams using those factors.

[ArtVandelay]

On the other hand, my understanding is that the computers also have no way of knowing about personnel changes (read, injuries). That may cause the rankings to be off for specific teams as players leave and return. I would guess that Ohio State is currently underrated and Purdue overrated by the dork polls.

Well, this is definitely true, as far as I understand it. His system just looks at possessions, regardless of who was injured, tired, in foul trouble, etc. on any given night. This is reflective of one problem with his system: college basketball seasons are pretty short so there are definitely some small sample-size concerns. So anyway yes, I would agree that your analysis of Purdue and Ohio St. is probably correct.

[91_92_01_10_15]

3) KenPom's ratings have historically been quite good at picking the top performers in the tournament, but it's somewhat hard to tell exactly because his historical data includes the tournament games (as far as I know). This means that Final Four teams are naturally going to look higher in his end-of-year rankings than before the tournament.


The above is definitely an issue, but I put these numbers together nonetheless. The first column is the Pomeroy ranking of the champ, and the second column contains the Pomeroy rankings of the other three Final Four participants, in order by their final rank.


Year ChampRank OtherFinal4Rank
2009 1 8,14, 3
2008 1 2,4,3
2007 2 4,6,5
2006 1 3,23,10
2005 1 2,7,5
2004 2 7,1,3


After watching closely as Duke slipped in the Pomeroy rankings towards the end of last season, I am encouraged that we have actually improved over the last month.

[uncwdevil]

I was just looking at whether we could possibly move up to #1 in KenPom's defensive rating in order to achieve the dorky trifecta (offense, defense, overall)...

We are currently #3 in defense, FSU is #2 which didn't surprise me too much as I know they have a lot of long athletes and play tough. Southern Cal at #1 caught me a little off guard though. They apparently play great defense (even on an adjusted basis) but they are a .500 team in a bad Pac-10. Without looking it up I cannot name their head coach or a single one of their players.

FSU and USC are both pretty far ahead of Duke, so I'm not sure catching them will be possible over the next couple of weeks.

[Kedsy]

The above is definitely an issue, but I put these numbers together nonetheless. The first column is the Pomeroy ranking of the champ, and the second column contains the Pomeroy rankings of the other three Final Four participants, in order by their final rank.


Year ChampRank OtherFinal4Rank
2009 1 8,14, 3
2008 1 2,4,3
2007 2 4,6,5
2006 1 3,23,10
2005 1 2,7,5
2004 2 7,1,3


After watching closely as Duke slipped in the Pomeroy rankings towards the end of last season, I am encouraged that we have actually improved over the last month.


Unfortunately, these numbers are not so helpful in evaluating how good Pomeroy is. I don't have pre-tournament numbers for the other years, but last year going into the tourney Pomeroy ranked UNC #2, UConn #3, Michigan State #13, and Villanova #19.

Of Pomeroy's pre-tournament top 8, one went down in the first round (#8 West Virginia), three went down in the Sweet 16 (#1 Memphis, #5 Gonzaga, and #7 Duke), two made the Elite 8 (#4 Louisville and #6 Pitt), and two made the Final Four (#2 UNC and #3 UConn).

By contrast, all four Final Four teams were in the pre-tournament AP top 10 (UNC #2, UConn #5, Michigan State #8, Villanova #10). Of the pre-tourney AP's top 8, two lost in the Sweet 16 (#3 Memphis and #6 Duke), three made the Elite Eight (#1 Louisville, #4 Pitt, #7 Oklahoma), and three made the Final Four.

So, last year at least, the AP poll (a source one might guess to be a very poor predictor) was significantly better at picking the tournament winners than Pomeroy was. Not saying Pomeroy doesn't mean anything. Just sayin'.

[JasonEvans]

So, last year at least, the AP poll (a source one might guess to be a very poor predictor) was significantly better at picking the tournament winners than Pomeroy was. Not saying Pomeroy doesn't mean anything. Just sayin'.

Luckily, Duke is a Final Four team in the AP poll, the coach's poll, KenPom, Sagarin, the RPI, and darn near any other measure of excellence.

--Jason "I like our chances ;) " Evans

[Acymetric]

I was just looking at whether we could possibly move up to #1 in KenPom's defensive rating in order to achieve the dorky trifecta (offense, defense, overall)...

We are currently #3 in defense, FSU is #2 which didn't surprise me too much as I know they have a lot of long athletes and play tough. Southern Cal at #1 caught me a little off guard though. They apparently play great defense (even on an adjusted basis) but they are a .500 team in a bad Pac-10. Without looking it up I cannot name their head coach or a single one of their players.

FSU and USC are both pretty far ahead of Duke, so I'm not sure catching them will be possible over the next couple of weeks.

I really can't understand how a .500 team can be tops in defensive efficiency, are they like dead last in offense?

[loldevilz]

I really can't understand how a .500 team can be tops in defensive efficiency, are they like dead last in offense?

265th

[MChambers]

With last night's suffocating defense, Duke is now back up to #3 in defensive efficiency. Also, last night's offense was pretty good too (too bad the second half wasn't like the first in that regard), so that Duke is still #1 in offensive efficiency and a clear, but not overwhelming, #1 overall.

[gw67]

At this time of year, I like to look at the weighted rackup put together by Massey. It includes the polls and various rating systems. It is updated every Monday and includes about 50 different rankings. The Devils were 3rd last Monday. I presume that they will be 3rd or 4th tomorrow.

http://www.masseyratings.com/cb/compare.htm

gw67

[Exiled_Devil]

There are several reasons that I am a little bit skeptical about the dork polls. (This pains me, because I prefer cold, hard numbers to subjective opinions.)

In many recent years, Duke has been highly ranked in those polls, only to have a disappointing end to the season. How to explain this?



The simple answer here is that the Dork polls do one of two things - examine past success and/or predict probability of future success.

No game is certain. Sometimes with the rankings we as fans forget that, but the reality is that games are games because they are uncertain. Otherwise they would movies of the week.

[CDu]

The simple answer here is that the Dork polls do one of two things - examine past success and/or predict probability of future success.

No game is certain. Sometimes with the rankings we as fans forget that, but the reality is that games are games because they are uncertain. Otherwise they would movies of the week.

Exactly. Which is also a reason why you don't take one-game samples as definitive evidence of who is better (for example, us losing badly to G'Town). Any single game is subject to such variation that it is inappropriate to use as a measuring stick. That's why the committee seeds based on the entire body of work, rather than simply head-to-head (though head-to-head can be used as a tiebreaker).

The tournament is lots of fun, but it's only loosely indicative of whom the best teams are. The "dork polls" are designed only to identify who played the best in the past and can be used to predict the most likely outcome in the tournament. But even the most likely outcome is still HIGHLY unlikely to occur, because there are just so many different potential outcomes that can occur in a 6-round, single-elimination tournament.

[MChambers]

The simple answer here is that the Dork polls do one of two things - examine past success and/or predict probability of future success.

No game is certain. Sometimes with the rankings we as fans forget that, but the reality is that games are games because they are uncertain. Otherwise they would movies of the week.

I understand that the statistical analyses are not perfect predictors. I was asking for a less simple explanation of why Duke in recent years had looked great through January (and sometimes later) and then struggled. I was wondering if it was related to our style of play, personnel, or coaching. (I think part of it is that Coach K demands full effort, all the time, which may make Duke look better on a relative basis than it really is.) Luckily, Duke seems to be improving this year, so maybe this year will be different.

[Lulu]

Not perfectly relevant, but I always wonder what these rankings would look like if you cut out the last 1:30 or so of every game, where the play usually breaks down (whether a tight game or a blowout) and the game/stats become a crapshoot. Take our G'town game this year. We were much further down than the final score would suggest, only because we cranked up our score when G'town wasn't really trying anymore. The computer rankings weigh these points equally, which I feel they shouldn't.

Although I guess you could argue that those endgame situations are just as much a part of the sport as the rest and need to be calculated equally. Still though, I'm not convinced.

I almost completely agree, except for the reverse case where a team (Kentucky?) doesn't even try until the last few minutes because they do what they have to do to not fall behind the whole game... Actually, it's the whole "playing down to the competition" factor that computers also cannot discern from true effort in a do-or-die situation.

[CDu]

I understand that the statistical analyses are not perfect predictors. I was asking for a less simple explanation of why Duke in recent years had looked great through January (and sometimes later) and then struggled. I was wondering if it was related to our style of play, personnel, or coaching. (I think part of it is that Coach K demands full effort, all the time, which may make Duke look better on a relative basis than it really is.) Luckily, Duke seems to be improving this year, so maybe this year will be different.

That certainly may have something to do with it. We're usually better prepared early in the season than most teams, which boosts our "dork poll" rankings (which take into account full-season success). So our ratings probably slightly overrate us a bit in March, as other teams are maybe playing better but playing catchup at that point.

I think part of it too is that we've caught some bad matchups (Villanova, MSU, LSU). And then, the rest of it is just catching the wrong team on the wrong day, and not having our best day (Indiana, VCU).

But this year, as you note, I think we're playing better in February/March than earlier in the season. So hopefully that'll mean good things for the tourneys.

[MChambers]

That certainly may have something to do with it. We're usually better prepared early in the season than most teams, which boosts our "dork poll" rankings (which take into account full-season success). So our ratings probably slightly overrate us a bit in March, as other teams are maybe playing better but playing catchup at that point.

I think part of it too is that we've caught some bad matchups (Villanova, MSU, LSU). And then, the rest of it is just catching the wrong team on the wrong day, and not having our best day (Indiana, VCU).

But this year, as you note, I think we're playing better in February/March than earlier in the season. So hopefully that'll mean good things for the tourneys.

Some of it is bad matchups, no doubt, especially last year, but in several recent years Duke's efficiency ratings, especially on defense, has dropped at the end of the season.

I've wondered if the pressure defense just wasn't as effective against better competition, especially quick guards with good handles. Of course, the ACC this year doesn't have particularly strong backcourts, except for Maryland (and perhaps VT). And we're playing "compact" pressure defense, so maybe that is a better fit.

I am looking forward to seeing some extended pressure defense next year, however. A topic for April, I guess.

[Clipsfan]

I really can't understand how a .500 team can be tops in defensive efficiency, are they like dead last in offense?

I've seen them play in person (against UCLA) and it was just ugly. I think the finals stats were 7-8 assists for each team with 17-20 turnovers each. Just an ugly game.

[NSDukeFan]

I just looked at Pomeroy's rankings today and found it interesting that after Friday's upsets, the last ranked team in the ACC is ranked #63 in the country. That is really impressive conference depth to not have a single team ranked lower than that. Of course we all know who is ranked #63. I'll give you a hint, they are the lowest ranked team in the ACC to not win a game in the tournament.

[InSpades]

Virginia is actually #75, but your point remains.

Looking at Miami's results is actually kind of fascinating. They played an *awful* non-conference schedule, but they didn't lose a game. They beat South Carolina and Minnesota. Which isn't exactly impressive, but when you consider that Miami finished last in the ACC it becomes pretty respectable. How many other last place teams can say they have 5 wins over NCAA teams (and possibly 6 if Minnesota can win today). I'm guessing exactly none.

[NSDukeFan]

Virginia is actually #75, but your point remains.

Looking at Miami's results is actually kind of fascinating. They played an *awful* non-conference schedule, but they didn't lose a game. They beat South Carolina and Minnesota. Which isn't exactly impressive, but when you consider that Miami finished last in the ACC it becomes pretty respectable. How many other last place teams can say they have 5 wins over NCAA teams (and possibly 6 if Minnesota can win today). I'm guessing exactly none.

Thanks for the correction, though it is too bad that Carolina is not the lowest ranked team in the conference though.

[91_92_01_10_15]

but still #1:

http://kenpom.com/rate.php

[tbyers11]

Thanks largely to Texas Tech destroying Seton Hall and UNC squeaking by William and Mary, Duke has fallen just behind Kansas in Pomeroy's rankings (http://kenpom.com/rate.php). Man, those Tar Heels never do us any favors do they? Although pretty much quitting after getting slapped around for the first ten minutes in Cameron was nice :D.

In other news, to all those people who say that Duke only looks good on computer thanks to the home court advantage of Cameron. Basketball Prospectus examines home road efficiency splits (http://basketballprospectus.com/article.php?articleid=1003) (intra-conference only) for the 11 conferences that he follows weekly in his Tuesday truths column (6 BCS plus A10, CUSA, Missouri Valley, Mountain West and West Coast).

Duke does have a much better home than road performance (0.23 points per possession vs 0.05), but overall we are tied for 9th out of the 44 teams in these conferences that made the tourney. Our 0.18 difference is identical to Ohio St and Baylor and less than the 0.20 difference of Maryland. How come no one claims that Baylor, Ohio St or Maryland only got their high seeds because of home court advantage? :)

[Kedsy]

Duke does have a much better home than road performance (0.23 points per possession vs 0.05), but overall we are tied for 9th out of the 44 teams in these conferences that made the tourney. Our 0.18 difference is identical to Ohio St and Baylor and less than the 0.20 difference of Maryland. How come no one claims that Baylor, Ohio St or Maryland only got their high seeds because of home court advantage? :)

That's good to hear about Baylor. They may have the crowd in Houston (assuming they make it) but it won't be a home game.

[MChambers]

There is a third dork poll, behind Sagarin and Pomeroy, that is worth considering:

http://www.bizjournals.com/atlanta/stories/2010/03/15/daily37.html

Has us playing Kansas in the title game. Losing, unfortunately, but five more wins would be nice.

[91_92_01_10_15]

back to #1:

http://kenpom.com/rate.php

[94duke]

back to #1:

http://kenpom.com/rate.php

I'm assuming that if we win tonight, our rating will go down like Kansas.
Agree/Disagree?

[91_92_01_10_15]

I'm assuming that if we win tonight, our rating will go down like Kansas.
Agree/Disagree?

I guess that would depend on how we win.

[MChambers]

I guess that would depend on how we win.

If we beat them like they're UNC, our rating goes up.

[InSpades]

If we beat them like they're UNC, our rating goes up.

So we have to break their will within the 1st ten minutes and have them not really try for the rest of the game? That seems really really unlikely. Guess we will have to settle for 2nd in this dork poll :).