MBB: Duke 61, Butler 59 - 2010 NCAA Champs!!!

[pfrduke]

OK then, what does make someone the best team then? Your position assumes there is some metric, other than performance when it counts most, that makes sense. I contend that is a fallacy. The only thing that matters is the result. The point of a team isn't to impress. It's to win.

Do you follow this argument all the way down? I.e., that Butler was the 2nd best team this season. West Virginia and Michigan State the 3rd/4th best? Kansas somehwere between 17th-32nd best?

[BattierBattalion]

OK then, what does make someone the best team then? Your position assumes there is some metric, other than performance when it counts most, that makes sense. I contend that is a fallacy. The only thing that matters is the result. The point of a team isn't to impress. It's to win.

And with that, I'm leaving this topic. Posting too often about one thing can get you into trouble around here.

It's all good, man. Don't feel ostracised. We just all want some polite discourse and disagree with some of your points.

I don't think there's an easy answer to who the best team in a given season is. I think it's clear that using only the NCAAT is not the best indicator. Look at this way, the NCAAT is a hugely stochastic affair. Let's assume that the "best" team going into the tournament has 99% chance of winning the 1st game, followed by 95% for game 2, 90% for game 3, and 85% for the last few games. Keep in mind that these are very generous probabilities. No team has even been close (based on predictors).

.99*.95*.90*.85*.85*.85 = 0.52

So this superteam, which no team in NCAAB history has been close to, still only has a 52% chance of winning the whole thing. That's quite far from a sure thing.

This should go to show you just how difficult it is to win the NCAAT, even if you have the best team in the land. We should be thankful for years like 2010 where Duke had a mixture of toughness, will, and fortune to win the whole thing. At the same time, we should appreciate years such as 1986, 1994, 1999, and 2004 where we were capable and deserving of a title but had a few plays go the other way.

[94duke]

Since I'm re-watching tournament games tonight...

Duke Won!!!
Nat'l Champs!!!

That is the topic of this thread, right?

[crimsondevil]

.99*.95*.90*.85*.85*.85 = 0.52

So this superteam, which no team in NCAAB history has been close to, still only has a 52% chance of winning the whole thing. That's quite far from a sure thing.

I like this way of thinking about it. Maybe adjust the #s a bit:
.999 (0 #16 wins in what, 104 and counting?) * .97 (if superteam is really all that) * .9 *.85 *.85 *.8 (make up for the extra in the first two rounds) = 0.50, so a nice, even money bet.

[CDu]

I like this way of thinking about it. Maybe adjust the #s a bit:
.999 (0 #16 wins in what, 104 and counting?) * .97 (if superteam is really all that) * .9 *.85 *.85 *.8 (make up for the extra in the first two rounds) = 0.50, so a nice, even money bet.

Yeah, and that would be an unbelievable team. If you take it a step further, assume this:
100% chance of winning the opening round (since no #16 has ever won - this is too high, but I'm simplifying)
98% chance of winning round 2 (probably too high in reality)
90% chance of winning Sweet 16 game (again, probably too high)
80% chance of winning Elite 8 game (ditto)
70% chance of winning Final 4 game
70% chance of winning championship game.

100%*98%*90%*80%*70%*70% = 34.6%. So even assuming that the team is so unbelievably good as to be a 70% chance of beating their opponent in the title game, they'd still only be about a 35% chance at winning the championship at the tournament's onset.

In reality, the very best of teams have no more than about a 33% chance of winning the title. The tournament doesn't determine the best team - it determines the team that had the best results in that tournament. To just say "the team that wins is proven to be the best team" is an overly simplistic and incorrect way to view the tournament.

[pfrduke]

Yeah, and that would be an unbelievable team. If you take it a step further, assume this:
100% chance of winning the opening round (since no #16 has ever won - this is too high, but I'm simplifying)
98% chance of winning round 2 (probably too high in reality)
90% chance of winning Sweet 16 game (again, probably too high)
80% chance of winning Elite 8 game (ditto)
70% chance of winning Final 4 game
70% chance of winning championship game.

100%*98%*90%*80%*70%*70% = 34.6%. So even assuming that the team is so unbelievably good as to be a 70% chance of beating their opponent in the title game, they'd still only be about a 35% chance at winning the championship at the tournament's onset.

In reality, the very best of teams have no more than about a 33% chance of winning the title. The tournament doesn't determine the best team - it determines the team that had the best results in that tournament. To just say "the team that wins is proven to be the best team" is an overly simplistic and incorrect way to view the tournament.

Thanks to Ken Pomeroy, we can actually run this for Duke this season (although, I have no way of assessing how accurate the win probability numbers he uses are):

1. Arkansas-Pine Bluff: 98.6%
2. California: 78.5%
3. Purdue: 78.6%
4. Baylor: 71.2%
5. West Virginia: 71.9%
6. Butler: 76.2%

All that together totals 23.7%.

[MarkD83]

In reality, the very best of teams have no more than about a 33% chance of winning the title.

No matter how you run the statistics it is difficult to win the tournament. That is why we all need to temper our expectations for 2010-11. No matter how great a team looks on paper playing your best in the last 6 games of the year is all that matters to win an NCAA championship. In 2009-10 Duke played their best basketball of the year in the last 10 games of the year. This was due to hard work and great coaching so credit all goes to the players and coaches.

[moonpie23]

what were uconn's chances in 99?

[CDu]

No matter how you run the statistics it is difficult to win the tournament. That is why we all need to temper our expectations for 2010-11. No matter how great a team looks on paper playing your best in the last 6 games of the year is all that matters to win an NCAA championship. In 2009-10 Duke played their best basketball of the year in the last 10 games of the year. This was due to hard work and great coaching so credit all goes to the players and coaches.

I absolutely agree. I think a distinction needs to be made in this discussion. When I say that the NCAA champion doesn't have to be the best team, I don't mean that in any way to denigrate the championship. It's a terrific accomplishment and the teams that win it are very deserving of praise. Just because it doesn't prove who the best team is doesn't mean it's not an amazing accomplishment to win it. They are two entirely separate issues.

Villanova in 1985 certainly deserves a ton of credit for the way they played in that tournament. They earned that championship, and it is deserving of great honor. But that doesn't mean they were the best team that season. It just means they had the best performance in that tournament. But they are absolutely deserving of praise for what they accomplished.

[Duvall]

what were uconn's chances in 99?

Well, they were the #2 team in the country, so - pretty good?

Let's go back to CDu's scenario above. Even if we assume a 30% chance of beating this hypothetical great team in the final, we still get 100%*98%*90%*80%*70%*30% = 14.8%. Hardly world-shocking territory.

[BoozerWasFouled]

I absolutely agree. I think a distinction needs to be made in this discussion. When I say that the NCAA champion doesn't have to be the best team, I don't mean that in any way to denigrate the championship. It's a terrific accomplishment and the teams that win it are very deserving of praise. Just because it doesn't prove who the best team is doesn't mean it's not an amazing accomplishment to win it. They are two entirely separate issues.

Villanova in 1985 certainly deserves a ton of credit for the way they played in that tournament. They earned that championship, and it is deserving of great honor. But that doesn't mean they were the best team that season. It just means they had the best performance in that tournament. But they are absolutely deserving of praise for what they accomplished.

This entire dead-end debate is a result of the philosophically flawed assumption that "the best team" is an empirical attribute that could be measured if only the tournament had the right set-up.

In reality, "best team" is an abstract quality, not a measurable attribute. You can't empirically measure "best team" the way you can measure "tallest person" or "heaviest weight." "Best team" is a concept used to generalize across specific performances, but it's only the performances themselves that are measurable, not the "team" itself.

[gumbomoop]

This entire dead-end debate is a result of the philosophically flawed assumption that "the best team" is an empirical attribute that could be measured if only the tournament had the right set-up.

In reality, "best team" is an abstract quality, not a measurable attribute. You can't empirically measure "best team" the way you can measure "tallest person" or "heaviest weight." "Best team" is a concept used to generalize across specific performances, but it's only the performances themselves that are measurable, not the "team" itself.

I have participated in this debate, but in a few earlier posts found it necessary to clarify exactly what my own limited view [in terms of the several different sub-debates] was: namely, that in '10, the "actual best team" did in fact win the NC. [Not that the best team always, or maybe even usually, wins....]

In your view, does or does not use of KemPom as "evidence" for Duke's arguable pre-NCAAT "bestness" constitute "measurable performance"? If yes, is it or is it not fair to label, roughly/colloquially, as "best" a team whose performance is measurably best?

[Kedsy]

I have participated in this debate, but in a few earlier posts found it necessary to clarify exactly what my own limited view [in terms of the several different sub-debates] was: namely, that in '10, the "actual best team" did in fact win the NC. [Not that the best team always, or maybe even usually, wins....]

In your view, does or does not use of KemPom as "evidence" for Duke's arguable pre-NCAAT "bestness" constitute "measurable performance"? If yes, is it or is it not fair to label, roughly/colloquially, as "best" a team whose performance is measurably best?

Personally, I agree with you that KenPom's ratings are good evidence in an argument about who is the best team. I don't think it is conclusive evidence. There are other legitimate computer rankings, and there are legitimate non-computer evaluations that can also be used as evidence.

I also think it's possible that Duke was, in fact, the best team in 2010. Possible. But I actually believe the best team was Kansas.

This entire dead-end debate is a result of the philosophically flawed assumption that "the best team" is an empirical attribute that could be measured if only the tournament had the right set-up.

In reality, "best team" is an abstract quality, not a measurable attribute. You can't empirically measure "best team" the way you can measure "tallest person" or "heaviest weight." "Best team" is a concept used to generalize across specific performances, but it's only the performances themselves that are measurable, not the "team" itself.

I disagree. There is a best team. We just may not have a definitive way to measure which team it is. But just because we can't measure it very well doesn't mean one team isn't best.

Playing a one-and-done tournament is certainly not a very good way to measure which team is best (and I don't believe it was intended for that purpose) but it's a very exciting way to determine a champion. They're two different things, that's all.

[COYS]

Personally, I agree with you that KenPom's ratings are good evidence in an argument about who is the best team. I don't think it is conclusive evidence. There are other legitimate computer rankings, and there are legitimate non-computer evaluations that can also be used as evidence.

I also think it's possible that Duke was, in fact, the best team in 2010. Possible. But I actually believe the best team was Kansas.



I disagree. There is a best team. We just may not have a definitive way to measure which team it is. But just because we can't measure it very well doesn't mean one team isn't best.

I think any declaration of "best team" must also incorporate a time period. Players go in and out of form, face injuries, go into shooting slumps, etc. Kenpom would even agree with this as Kansas was the "best team" based on Pomeroy's stats for much of the first part of the season only to be surpassed by Duke as the season went on. Kansas may indeed have been the "best team" for the longest period of time during last season, but by the final third of the season they appeared to be passed by Duke. Does this mean that Duke really was the best team all along or simply that the best team was different depending on when in the season we decide to look?

[CDu]

I have participated in this debate, but in a few earlier posts found it necessary to clarify exactly what my own limited view [in terms of the several different sub-debates] was: namely, that in '10, the "actual best team" did in fact win the NC. [Not that the best team always, or maybe even usually, wins....]

In your view, does or does not use of KemPom as "evidence" for Duke's arguable pre-NCAAT "bestness" constitute "measurable performance"? If yes, is it or is it not fair to label, roughly/colloquially, as "best" a team whose performance is measurably best?

I have to disagree with BoozerWasFouled's claim that "the best team" isn't measurable. I'd say it's absolutely measurable. The problem is that it is very difficult to measure. And I'd argue that none of the measures in existence are perfect (including Pomeroy, though his method and some others are very good).

And I don't necessarily have a disagreement with the idea that Duke this year was the best team. I think an argument can be made that we were the best. My only contention is with the idea that us winning the championship proves that we're the best team. I think a strong argument could be made for Kansas as well, though they hurt themselves with a lackluster tourney effort. My statement that the tournament doesn't decide the best team has nothing to do with the outcome of this year's tournament and whether or not Duke was actually the best. That's an entirely separate debate for me.

My point is simply that the best team and the NCAA champion do not have to be one and the same. They can be one and the same, and often they are one and the same. But winning the tournament does not in itself determine the best team. It just determines the champion.

[94duke]

Duke Won!!!
Nat'l Champs!!!

I re-watched the WVa game last night.
I'm looking forward to re-watching the Butler game.
I haven't actually watched the game on TV yet, since I was in Indy for the Final Four!
I suspect I will be extremely tense even though I know the outcome!

[gumbomoop]

I have to disagree with BoozerWasFouled's claim that "the best team" isn't measurable. I'd say it's absolutely measurable. The problem is that it is very difficult to measure. And I'd argue that none of the measures in existence are perfect (including Pomeroy, though his method and some others are very good).

And I don't necessarily have a disagreement with the idea that Duke this year was the best team. I think an argument can be made that we were the best. My only contention is with the idea that us winning the championship proves that we're the best team. I think a strong argument could be made for Kansas as well, though they hurt themselves with a lackluster tourney effort. My statement that the tournament doesn't decide the best team has nothing to do with the outcome of this year's tournament and whether or not Duke was actually the best. That's an entirely separate debate for me.

My point is simply that the best team and the NCAA champion do not have to be one and the same. They can be one and the same, and often they are one and the same. But winning the tournament does not in itself determine the best team. It just determines the champion.

Agree with pretty much every word here. As one who is statistically-challenged, I'm just out of my league on the issue of measurable or not. But my intuition is that if measurable, you're surely right that it's very difficult to measure. Also agree with you and Kedsy that KenPom, however sophisticated, wouldn't be the sole measure. Goodness knows, I'm not even opposed to the eye test as one measure. [But there is the eye-of-the-beholder problem, I suppose....]

I also agree with you on the specific point that in any particular season the best team and the NC are not necessarily the same team. My own role in this sub-set of this thread [which sub-set, I can understand, irritates those posters who'd prefer to keep this thread absolutely focused on the thrill of the NC], goes back a couple of months, and honestly to my irritation at what I saw as the Stoopids' collective dismissal of Duke's NC as luck, NCAA-seeding-bias, various conspiracy theories. Perfectly happy to admit that there were no "classically [?] great" teams in '09-'10, I thought Duke obviously belonged as a top 3-4 going into the NCAAT, and thus thought of course we deserved a #1 seed. Hardly a dominating team, our guys won by... well, we've been over that in multiple threads.

So even if there were any consensus re Kansas, and perhaps less likely UK or Syracuse, as last season's best team, I'd originally [way back in mid-April] have been less irritated had Duke been at least warmly praised as having snuck up on the Stoopids as a by-golly-damn good team. Our guys having being dissed, I became obsessed. I'm calmer now.... sort of.

Apologies to any posters irritated at thread-theft. I'm still as overjoyed as you, on the thing that really counts. And thanks to the many of you who have been willing to sift through issues about "best," measurable or not.

[MarkD83]

It is my goal to keep this thread near the top of the board all off-season. The conversations about "best" team and the tournament are interesting. The quote from 94duke says it all and reminds me I need to go watch another DVD from this season.

As an aside I rewatched the game at Clemson where Dick Vitale nicknamed Jon, Nolan and Kyle as super, scintillating and sensational. What was interesting was the tone of that game. Duke had not won a true road game having just lost at NCSU. The horrible drubbing in LittleJohn from 2009 was brought up and it was apparent that Duke needed to prove themselves before they were considered Nat'l Championship contenders.

Several months later Duke as the only number 1 seed in the Final Four wins it all. Quite a two month ride for Duke fans.

[NSDukeFan]

I think any declaration of "best team" must also incorporate a time period. Players go in and out of form, face injuries, go into shooting slumps, etc. Kenpom would even agree with this as Kansas was the "best team" based on Pomeroy's stats for much of the first part of the season only to be surpassed by Duke as the season went on. Kansas may indeed have been the "best team" for the longest period of time during last season, but by the final third of the season they appeared to be passed by Duke. Does this mean that Duke really was the best team all along or simply that the best team was different depending on when in the season we decide to look?

I agree with your general premise, but, as I think Gumbomoop stated, KenPom seems to be as good a statistical model as I know of (simply because it has been promoted and discussed here the most) but it is not the be-all end-all. I am pretty sure that was not the intention of your post, but by saying that "Duke passed KU" implies a certitude about best team that others have said we don't have the best way of knowing for sure. Interesting discussion and I also agree that I am happy to hijack this thread a bit to keep it near the top.

[DevilDan]

I say again ... DUKE 61 Butler 59. We 'whupped a great team in its own back yard, "what" had won 25 in a row ... !

I left a little get well note for JON on another thread, and while I have him in my thoughts, I want to give him a shout out for his incredible play in the Title Game. He had three huge moments that stand out: the block in the first half on the breakaway layup, that would have made the crowd go thru the roof -- but Jon came out of nowhere to slap it into the stands... the surprise (who?) tipin of Miles' desperation turnaround jumper, with the shot clock expiring ... and the outstanding sideways, fadeaway bank-in 5-footer, that started a three point play.

These three monster plays from our guy, who had no speed, didn't jump very well, and was not that strong (as we heard during his four years as a Blue Devil). Like Coach K said, Jon is neither a "1" or a "2", he's just a good BASKETBALL PLAYER. Thanks for a great career JON !