Duke is Still Duke (so far)

[BlueDevilBrowns]

I know the board seems kinda down this week considering the egg Duke laid on Saturday, but here's some perspective(from a statistical standpoint) that should make all of us feel MUCH better about how the season has gone and where it could STILL go.

This is from Peter Tiernan from cbssports.com and bracketscience.com

http://www.cbssports.com/collegebask...rolific-enough

The highlights:

The last 12 champions have possessed 8 key stats. 7 teams so far this year match the profile of a champion using these stats - Duke, Indiana, Michigan, Florida, Syracuse, Kansas, and Louisville.

Only 5 teams meet KenPom statistical model of the past 9 champions dating back to 2004 - Duke, Syracuse, Indiana, Florida, and Pittsburgh(although they don't meet the above-mentioned model).

[cptnflash]

The problem is that more than half of our overall statistical profile was accumulated when we had Ryan. If you only looked at the games we've played since he got injured, we wouldn't be anywhere near the level of past champions.

The correct title for this thread should be "Duke was still Duke when Ryan Kelly was healthy."

[BlueDevilBrowns]

The problem is that more than half of our overall statistical profile was accumulated when we had Ryan. If you only looked at the games we've played since he got injured, we wouldn't be anywhere near the level of past champions.

The correct title for this thread should be "Duke was still Duke when Ryan Kelly was healthy."

It's true we're not the same team without Kelly but from everything we've been told by media, Coach K, and Kelly himself is that he IS returning at some point this year. Therefore, the statistical profile should still hold(perhaps improve) once he returns.

My point with the title of the thread was to provide some optimism that was much needed on the board this week. Basically, to take a step back, breathe, and chill, realizing the season is long and we've accomplished alot already and their's plenty of time and opportunity to accomplish more. The title is a take-off on Pete Gillen's classic line "Duke is Duke".

Of course, we could nose-dive like last year but over the course of the last 3 decades or so, we usually end up winning either the ACC Regular Season, ACC Tournament, or at least reach the second weekend of the NCAAT.

All impressive accomplishments.

[NashvilleDevil]

I know the board seems kinda down this week considering the egg Duke laid on Saturday

They did not lay an egg Sat. They lost a 2 point game to a team that had a week to prepare and nothing to lose. My definition of laying an egg is the Miami game.

[throatybeard]

They did not lay an egg Sat. They lost a 2 point game to a team that had a week to prepare and nothing to lose. My definition of laying an egg is the Miami game.

Exactly. It was a one-possession game where Cook's last 3 hit the back of the rim. If that goes in, we win.

[BlueDevilBrowns]

They did not lay an egg Sat. They lost a 2 point game to a team that had a week to prepare and nothing to lose. My definition of laying an egg is the Miami game.

Look, the point I was originally trying to make was to be positive about the season up to now and Duke's chances going forward using statistical analysis as a basis.

The Maryland loss did touch a nerve with a lot of people on the board so I was just trying to bring some good news.

Whether you want to call the game an "egg" or not, I don't care.

[OldSchool]

I'll tell you DID lay an egg: Maryland against Boston College tonight.

After their loss tonight, Lunardi declared that the Terps washed away whatever good they did for their tourney chances in beating Duke.

At BC, Mason carried us with 19 points and hit the game-winning free throws.

At BC, Alex "Lottery Pick" Len had 4 points and fouled out.

[mo.st.dukie]

They did not lay an egg Sat. They lost a 2 point game to a team that had a week to prepare and nothing to lose. My definition of laying an egg is the Miami game.

Yes, and in both the NCSU and Maryland losses Duke fought hard and played well, the other team just played a little bit better in one of their biggest games of the year on their own floor. The Miami loss was depressing but the other two were not surprising in the least. I think what may have some people really down about the Maryland loss isn't that game itself but the fact that we didn't blow out BC and UNC and instead needed late surges (and some luck) to win those games. If we had beaten BC by double digits and destroyed UNC 82-50 style then perhaps people wouldn't be as down on the team about a 2 point loss.

[throatybeard]

Note that Florida just lost by a possession at Mizzou, their fifth loss.

22-3 with no losses OOC looks pretty sweet.

[darthur]

The problem is that more than half of our overall statistical profile was accumulated when we had Ryan. If you only looked at the games we've played since he got injured, we wouldn't be anywhere near the level of past champions.

The correct title for this thread should be "Duke was still Duke when Ryan Kelly was healthy."

I fail to see how that's a problem. If you attribute our struggles to Kelly's injury, then you should be pretty optimistic about our chances once he gets back.

[uh_no]

I fail to see how that's a problem. If you attribute our struggles to Kelly's injury, then you should be pretty optimistic about our chances once he gets back.

i think the implication might be that a) despite the optimistic views, he might not get back or end up reinjured, or b) when he does get back, he is either not up to the quality of play that he gave a few months ago, or the team does not play as well with him as they did a few months ago.

I hope the team pops right back...but we'll see

either way I'm hoping for a return on sunday :P

[75Crazie]

i think the implication might be that a) despite the optimistic views, he might not get back or end up reinjured, or b) when he does get back, he is either not up to the quality of play that he gave a few months ago, or the team does not play as well with him as they did a few months ago.

After all, it's not as if we've never seen that happen before ... say two years ago or so.

[wallyman]

Note that Florida just lost by a possession at Mizzou, their fifth loss.

22-3 with no losses OOC looks pretty sweet.

Florida is 21-4, with a good chance of winning out from here. Toughest game left is at cratering Kentucky, and SEC tournament is not murderer's row...

[CDu]

The problem is that more than half of our overall statistical profile was accumulated when we had Ryan. If you only looked at the games we've played since he got injured, we wouldn't be anywhere near the level of past champions.

The correct title for this thread should be "Duke was still Duke when Ryan Kelly was healthy."

Actually, we'd still qualify on all but one of the metrics, even without Kelly. The only number we fall short of is the scoring margin > 7. Our numbers are 76.4 ppg scored vs 72.2 ppg allowed (4.2 ppg margin).

Though obviously we're much more likely to win it with a healthy Kelly.

[Dev11]

Florida is 21-4, with a good chance of winning out from here. Toughest game left is at cratering Kentucky, and SEC tournament is not murderer's row...

I believe 22-3 referred to Duke and our chances of getting a #1.

[Mtn.Devil.91.92.01.10.15]

Interesting... even with the conference being in another down year, Duke has the #1 SOS.

I would have thought that one of the Big Ten schools would have taken that over by now.

Go Duke!

[Dr. Rosenrosen]

I know the board seems kinda down this week considering the egg Duke laid on Saturday, but here's some perspective(from a statistical standpoint) that should make all of us feel MUCH better about how the season has gone and where it could STILL go.

This is from Peter Tiernan from cbssports.com and bracketscience.com

http://www.cbssports.com/collegebask...rolific-enough

The highlights:

The last 12 champions have possessed 8 key stats. 7 teams so far this year match the profile of a champion using these stats - Duke, Indiana, Michigan, Florida, Syracuse, Kansas, and Louisville.

Only 5 teams meet KenPom statistical model of the past 9 champions dating back to 2004 - Duke, Syracuse, Indiana, Florida, and Pittsburgh(although they don't meet the above-mentioned model).

I like the optimism. Nothing wrong with finding positive things to focus on given how easy it is to focus on the negative - which many on this board are guilty of. I still cannot fathom how some can be so down about a 22-3 year. We're clearly not what we were with Ryan but that has a very good chance of turning back in our favor.

So how about some good vibes around here already?

Let's not try to break down every positive post and statistic with the goal of caveating and debunking something really should make us all feel good and appreciate what we are witnessing... which is another VERY good year... at least by 99.9999999999% of other peoples'/programs' standards!

[CDu]

Interesting... even with the conference being in another down year, Duke has the #1 SOS.

I would have thought that one of the Big Ten schools would have taken that over by now.

Go Duke!

Depends on whose formula you are viewing. Pomeroy has us at #5 in SOS, behind (you guessed it) four big-10 schools (Minnesota, MSU, Illinois, and Wisconsin).

[ArtVandelay]

I know the board seems kinda down this week considering the egg Duke laid on Saturday, but here's some perspective(from a statistical standpoint) that should make all of us feel MUCH better about how the season has gone and where it could STILL go.

This is from Peter Tiernan from cbssports.com and bracketscience.com

http://www.cbssports.com/collegebask...rolific-enough

The highlights:

The last 12 champions have possessed 8 key stats. 7 teams so far this year match the profile of a champion using these stats - Duke, Indiana, Michigan, Florida, Syracuse, Kansas, and Louisville.

Only 5 teams meet KenPom statistical model of the past 9 champions dating back to 2004 - Duke, Syracuse, Indiana, Florida, and Pittsburgh(although they don't meet the above-mentioned model).

I would caution against reading too much into these types of attempts to predict future winners based on characteristics of past winners. In particular, the "8 key stats" they mention seem to be a bit arbitrary. Trying to find a group of stats that just so happen to fit the last 7 winners is not exactly a rigorous statistical method for predicting the next winner. It feels like an example of an overfit model, as Nate Silver would say, or the Kate Upton effect, as I would say. But I digress.

Also, I am a bit skeptical of the KenPom numbers. I've been keeping his pre-tourney rankings for the past several years (but not back to 2004), and UConn was ranked 21/32 in AdjO and AdjD efficiency before the 2011 tournament. UNC was ranked 32nd in AdjD. These don't line up with what the author of this article said. Can anyone correct me if I'm off base on this? I have Duke at 2/5 in 2010, and UK was 2/6 last year. But if you take just those four winners, it would be tough to draw any firm conclusions about the relationship between pre-tourney KenPom rankings and tournament success. I'd be interested to see how far back anyone has kept this data.

Anyway, I'm not saying that these aren't interesting data points, but you have to be skeptical of anyone's efforts to accurately predict NCAA tournament results using these types of statistical comparisons to past winners (particularly when the stats seem cherry-picked). The Ken Pom numbers can be used to create predictions (which Ken Pom himself does), but they're just that -- predictions. They can be, and often are, wrong.

[Reilly]

Depends on whose formula you are viewing. Pomeroy has us at #5 in SOS ...

sports-reference.com has us at #3 in SOS ... so, 1, 3, 5 ... still tough ... 7 of the top 9 in SOS at s-r are Big 10 schools:

http://www.sports-reference.com/cbb/...standings.html