FEARLESS Predictions: 2014-15 Season

[uh_no]

In the next chat Loran will reiterate that the defense sucks and we just can't compete despite our 9-1 record heading into that game. "Our secondary is terrible, our D-line is terrible, and our LB's suck"

"Just being honest here guys"

someone will reference past years ken-pom ratings indicating that all top performers in the tournament are high in kenpom...and will be promptly reminded that historical kenpom numbers are post-tourney

[OldPhiKap]

In the next chat Loran will reiterate that the defense sucks and we just can't compete despite our 9-1 record heading into that game. "Our secondary is terrible, our D-line is terrible, and our LB's suck"

"Just being honest here guys"

Sometimes Chatbot kicks me out. Sometimes I leave on my own accord.


But I will say -- Loran knows a bit about college football.

[Kedsy]

So what? If Duke scores 44 points in a tournament game and loses, we can't cite lack of scoring as the reason for the loss because one time a team won a tournament game scoring 43 points?

Well, first of all, I said "lots of teams" get past the first round despite poor defense, not "one time." Second, even in your example, you'd be more accurate if you said a reason, rather than the reason. Finally, your statement that "[t]he 2012 and 2014 teams lost because our defense sucked. That's all." isn't correct. There were many reasons we lost, including pure chance. Your statement was an over-generalization.

Mercer lost 9 games in 2013-14. Five of those losses (more than half) came to teams with worse defenses than Duke (according to Pomeroy's post-tourney numbers). I expect that Duke's subpar defense didn't help, but it's clearly inaccurate to say Duke "lost because our defense sucked" and "that's all."

I tried to be proper but I meant high school players who bolt for the nba after a year of college.

The problem is when you're recruiting them you don't necessarily know which players will bolt after one year and which won't. These days, any top 10 player who has a good freshman year also has a decent chance of bolting after that season. Are you suggesting we shouldn't go after any top 10 players? That would probably stop our reliance on one-and-dones, but it would also make it a lot harder to compete.

[uh_no]

Well, first of all, I said "lots of teams" get past the first round despite poor defense, not "one time." Second, even in your example, you'd be more accurate if you said a reason, rather than the reason. Finally, your statement that "[t]he 2012 and 2014 teams lost because our defense sucked. That's all." isn't correct. There were many reasons we lost, including pure chance. Your statement was an over-generalization.

Mercer lost 9 games in 2013-14. Five of those losses (more than half) came to teams with worse defenses than Duke (according to Pomeroy's post-tourney numbers). I expect that Duke's subpar defense didn't help, but it's clearly inaccurate to say Duke "lost because our defense sucked" and "that's all."



The problem is when you're recruiting them you don't necessarily know which players will bolt after one year and which won't. These days, any top 10 player who has a good freshman year also has a decent chance of bolting after that season. Are you suggesting we shouldn't go after any top 10 players? That would probably stop our reliance on one-and-dones, but it would also make it a lot harder to compete.

how many one and dones were on:
uconn 2014
UL 2013
Uconn 2011
duke 2010
UNC 2009

in fact, the only teams that won national championships with significant dependence on one and done talent were UK 2012, and Syracuse, in 2003.

while the tournament is largely chance, and it's true that it's more likely to be won by a team that doesn't have one and done players simply because there are a lot more of them (than teams that do have one and done players), I think it's silly to say "its a lot harder to compete"...as has been proven false by 5 of the last 6 champions.

[OldPhiKap]

how many one and dones were on:
uconn 2014
UL 2013
Uconn 2011
duke 2010
UNC 2009

in fact, the only teams that won national championships with significant dependence on one and done talent were UK 2012, and Syracuse, in 2003.

while the tournament is largely chance, and it's true that it's more likely to be won by a team that doesn't have one and done players simply because there are a lot more of them (than teams that do have one and done players), I think it's silly to say "its a lot harder to compete"...as has been proven false by 5 of the last 6 champions.

To echo - Mercer was a group of upperclassmen. The good Butler teams we're made up of upperclassmen. UVa last year won with grind-it-out talent.

On the college level, experience counts. In the pro ranks, freakish athletic ability counts. That's why the key to our season this year is how Quinn plays, not one of our freshmen phenoms.

[johnb]

I predict much of the country will still hate Duke, and here's a map to prove it:


http://www.washingtonpost.com/blogs/...lly-hate-duke/

[CameronBornAndBred]

I predict much of the country will still hate Duke, and here's a map to prove it:


http://www.washingtonpost.com/blogs/...lly-hate-duke/

Love Rickman's tweet. Also love seeing that the Heels are the most hated in NC.

[pfrduke]

Love Rickman's tweet. Also love seeing that the Heels are the most hated in NC.

Yeah, that's really remarkable. It, the University of Washington, Mississippi State, and Louisville are the only schools most hated in their own state. The latter two make more sense - I'm surprised Oregon didn't win out in Washington, but the Cougar people really don't like the Huskies, as do all of the transplanted Oregon alums, so I guess the outnumber the Husky fans who may have split animosities.

[Kedsy]

how many one and dones were on:
uconn 2014
UL 2013
Uconn 2011
duke 2010
UNC 2009

in fact, the only teams that won national championships with significant dependence on one and done talent were UK 2012, and Syracuse, in 2003.

while the tournament is largely chance, and it's true that it's more likely to be won by a team that doesn't have one and done players simply because there are a lot more of them (than teams that do have one and done players), I think it's silly to say "its a lot harder to compete"...as has been proven false by 5 of the last 6 champions.

I didn't say "one and dones." I said top 10 talent, which are potential one-and-dones. Because if you want to ensure you don't get one-and-dones, the only way to do that is not recruit the top talent.

And while it's true that UConn won some natties with lesser talent (though the 2011 team had a #15 and a #16 recruit, so not so much lesser), your beloved Huskies also won this past year out of a 7-seed, so not exactly a reproducible model. Kentucky in 2012 had 3 top 10 guys plus a #11 guy. Duke in 2010 had a top 10 guy (Kyle Singler), plus a #14, two #18s, and a #20. UNC in 2009 had 3 top 10 guys plus a #15. Kansas in 2008 had a top 10 guy plus a #11 and a #14. UNC in 2005 had 4 top 10 guys plus a #11. Duke in 2001 had 4 top 10 guys plus a #16.

Put another way, I didn't say it's impossible to compete without top 10 talent, I said it's a lot harder to compete. And it is.

To echo - Mercer was a group of upperclassmen. The good Butler teams we're made up of upperclassmen. UVa last year won with grind-it-out talent.

Please take a look at the aggregate amount of post-season success those three schools have had over the past 10 (or 15 or 20, or even 5) years. Yeah, Butler captured lightning in a bottle but had to do it from a #8 and a #5 seed, and hasn't had much success before or since. The other schools haven't had any real post-season success at all, and only sporadic regular season success.

Be honest: how many Duke fans would be satisfied with the records of any of the schools you mention?

[Wander]

Well, first of all, I said "lots of teams" get past the first round despite poor defense, not "one time." Second, even in your example, you'd be more accurate if you said a reason, rather than the reason. Finally, your statement that "[t]he 2012 and 2014 teams lost because our defense sucked. That's all." isn't correct. There were many reasons we lost, including pure chance. Your statement was an over-generalization.

Mercer lost 9 games in 2013-14. Five of those losses (more than half) came to teams with worse defenses than Duke (according to Pomeroy's post-tourney numbers). I expect that Duke's subpar defense didn't help, but it's clearly inaccurate to say Duke "lost because our defense sucked" and "that's all."

Yes, part of the reason why Duke lost is also that Quinn Cook hit 70% of his 3's instead of 90% of his 3's and that Mercer players didn't all contract the bubonic plague right before the game. I'll withdraw the "that's all" part.

More seriously, yes, I agree that of course it was possible for Duke to beat Mercer even with a bad defense, and we'd probably do so 7 out of 10 times. I was speaking more broadly. The team had bad defense from tip-off of the first game all the way through the NCAA tournament, and I don't see any reason to analyze beyond that as to why the team fared poorly. I'll change my statement to "our defense sucking is the primary reason why we lost in 2012 and 2014."

[Kedsy]

More seriously, yes, I agree that of course it was possible for Duke to beat Mercer even with a bad defense, and we'd probably do so 7 out of 10 times. I was speaking more broadly. The team had bad defense from tip-off of the first game all the way through the NCAA tournament, and I don't see any reason to analyze beyond that as to why the team fared poorly. I'll change my statement to "our defense sucking is the primary reason why we lost in 2012 and 2014."

This might be semantic at this point. I agree the team had bad defense for most (if not all) of the season. But I believe we would have beaten Mercer closer to 9 times out of 10, and the same with Lehigh in 2012, so I'm not sure bad defense was even the primary reason we lost. Especially in 2012, when I believe Ryan Kelly's injury was the primary reason we lost.

Because if we agree bad defense was a core characteristic of the team, and we also agree that Duke would have beaten Mercer the vast majority of the time (whether 90% or 70%), then it follows that bad defense was not the primary reason we lost (since we would have had bad defense and still won the vast majority of the time). Like I said, it might be semantic, and I'll be happy to agree to disagree.

[pfrduke]

This might be semantic at this point. I agree the team had bad defense for most (if not all) of the season. But I believe we would have beaten Mercer closer to 9 times out of 10, and the same with Lehigh in 2012, so I'm not sure bad defense was even the primary reason we lost. Especially in 2012, when I believe Ryan Kelly's injury was the primary reason we lost.

Because if we agree bad defense was a core characteristic of the team, and we also agree that Duke would have beaten Mercer the vast majority of the time (whether 90% or 70%), then it follows that bad defense was not the primary reason we lost (since we would have had bad defense and still won the vast majority of the time). Like I said, it might be semantic, and I'll be happy to agree to disagree.

Mercer shot 25-45 from the field, went to the line 28 times, turned the ball over only 8 times, and scored 78 points in a ~64 possession game. We scored 1.11 points per possession and lost by 7. If you don't call that bad defense, I don't know what to tell you.

[flyingdutchdevil]

Mercer shot 25-45 from the field, went to the line 28 times, turned the ball over only 8 times, and scored 78 points in a ~64 possession game. We scored 1.11 points per possession and lost by 7. If you don't call that bad defense, I don't know what to tell you.

Good offense?

[Kedsy]

Mercer shot 25-45 from the field, went to the line 28 times, turned the ball over only 8 times, and scored 78 points in a ~64 possession game. We scored 1.11 points per possession and lost by 7. If you don't call that bad defense, I don't know what to tell you.

I did call it bad defense. I merely said bad defense was only one out of several reasons why we lost.

[Listen to Quants]

while what you say is true, there's hardly enough data to 2 of the last 3 years constitutes a temporary deviation from the norm, or a downward trend. It's impossible to know.

personally I think it is a new trend, and NOT because "K has slipped" or some bull like that. The fact is that the high player turnover in this era lends itself to higher program variability from year to year. In fact I think the fact that K's teams have still be constantly rated so highly (top 10 streak) is a testament to K's ability to continue to coach in the era...IT's just that we can't expect 7 final fours in 9 years...I'm not sure those kind of numbers will EVER be matched , let alone by K. Not to mentin, in 2 of the past 4 years, we've been seriously marred by injury troubles which no doubt inhibited last season performance (kyrie and ryan)

speaking of tredy-ness, I want to recall last year and our defensive efficiency throughout the season...After a couple of good performances in february, I think it was you (and apologies if it wasn't) who claimed that a couple good performances on D indicated an upward trend, and I urged caution lest it be an anomaly. It took the rest of the season to really establish that it was anomolous. I think the same caution needs to be applied here. It will take the next 3-5 years to really establish how much the past few years do or do not indicate a trend.

Makes a lot of sense. Early entry may also mean that the best college teams now are not as good, relative to the lesser teams, as they used to be. If the teams bunch together the variability increases of course. Might be there is additional game-to-game variability from the greater youth of the best teams (early entry again).

[Kedsy]

Makes a lot of sense. Early entry may also mean that the best college teams now are not as good, relative to the lesser teams, as they used to be. If the teams bunch together the variability increases of course. Might be there is additional game-to-game variability from the greater youth of the best teams (early entry again).

Here's some raw data aggregating first round upsets and Final Four composition to test your theory:

Year	10	11	12	13	14	15	Tot Upsets	10 to 12	13 to 15	F4a	F4b	F4c	F4d	Tot #1s	F4 Sum
----	--	--	--	--	--	--	----------	--------	--------	---	---	---	---	-------	------
2014	1	2	3	0	1	0		7	6		1		1	2	7	8	1	18
2013	1	1	3	1	1	1		8	5		3		1	4	4	9	1	18
2012	2	2	2	1	0	2		9	6		3		1	2	2	4	1	9
2011	1	3	1	1	0	0		6	5		1		3	4	8	11	0	26
2010	3	2	1	1	1	0		8	6		2		1	2	5	5	1	13
2009	3	1	3	1	0	0		8	7		1		1	1	2	3	2	7
2008	1	1	2	2	0	0		6	4		2		1	1	1	1	4	4
2007	0	2	0	0	0	0		2	2		0		1	1	2	2	2	6
2006	2	2	2	1	1	0		8	6		2		2	3	4	11	0	20
2005	1	1	1	1	1	0		5	3		2		1	1	4	5	2	11
2004	1	0	2	0	0	0		3	3		0		1	2	2	3	1	8
2003	2	1	1	1	0	0		5	4		1		1	2	3	3	1	9
2002	1	2	3	1	0	0		7	6		1		1	1	2	5	2	9
2001	2	2	2	2	0	1		9	6		3		1	1	2	3	2	7
2000	2	1	0	0	0	0		3	3		0		1	5	8	8	1	22
1999	4	0	2	1	1	0		8	6		2		1	1	1	4	3	7
1998	3	2	1	1	1	0		8	6		2		1	2	3	3	1	9
1997	2	0	1	0	1	1		5	3		2		1	1	1	4	3	7
1996	2	1	2	1	0	0		6	5		1		1	1	4	5	2	11
1995	1	1	1	1	2	0		6	3		3		1	2	2	4	1	9
1994	2	1	2	0	0	0		5	5		0		1	2	2	3	1	8
1993	0	1	1	1	0	1		4	2		2		1	1	1	2	3	5
1992	2	0	1	1	1	0		5	3		2		1	2	4	6	1	13
1991	2	2	1	1	1	1		8	5		3		1	1	2	3	2	7
1990	1	1	2	0	1	0		5	4		1		1	3	4	4	1	12
1989	1	4	1	1	1	0		8	6		2		1	2	3	3	1	9
1988	1	1	0	1	1	0		4	2		2		1	1	2	6	2	10
1987	2	0	1	2	1	0		6	3		3		1	1	2	6	2	10
1986	1	1	1	0	2	0		5	3		2		1	1	2	11	2	15
1985	0	3	1	1	0	0		5	4		1		1	1	2	8	2	12
1984	2	1	1	na	na	na		4	4		na		1	1	2	7	2	11
1983	2	2	1	na	na	na		5	5		na		1	1	4	6	2	12

The first year high school seniors were not allowed to go straight to the NBA was 2006, so the "one-and-done era" began in the 2006-07 season. The 64-team tournament started in 1985. In order to get four 8-season periods (since the one-and-done era has been 8 seasons), I added 1983 and 1984 to the table, but it's not apples-to-apples since in those tournaments the top 4 seeds got byes (meaning there were no 2 vs. 15, 3 vs. 14, or 4 vs. 13 games in those years).

Comparing the four 8-season blocks:

		Tot Upsets	10-12 upsets	13-15 upsets	#1s in F4	F4 Sum
		----------	------------	------------	---------	------
2007-2014	6.75		5.13		1.63		1.50		12.63
1999-2006	6.00		4.63		1.38		1.50		11.63
1991-1998	5.88		4.00		1.88		1.75		8.63
1983-1990	5.71*		3.88		1.83		1.75		11.38

* -- adding 8-year average for 7 to 10 upsets and 6-year average for 13 to 15 upsets

Clearly there has been an increase in upsets over time, though not so much as you might think (basically a one upset per tournament difference between earliest period and latest period) and not consistently (for example, upsets by seeds 13 to 15 was more prevalent in the earlier time periods). Similarly, fewer 1-seeds have made the Final Four, but that trend didn't start in the one-and-done era and it isn't that much of a decrease (a quarter of a 1-seed per season, i.e., one fewer 1-seed every four years). It's hard to see a trend in the sum of the seeds of the four Final Four teams. Three of the past four seasons have been very high, but take away just one improbable run (say, Wichita State staving off UK and beating Louisville in 2014 or Kansas beating VCU in 2011), the average for the period would have been essentially equal to the average from 1983 to 1990.

I'll leave it to those better versed in statistics to tell me if the increase in upsets is truly significant, or if the data from the one-and-done period is really more variable.

[sagegrouse]

		Tot Upsets	10-12 upsets	13-15 upsets	#1s in F4	F4 Sum
		----------	------------	------------	---------	------
2007-2014	6.75		5.13		1.63		1.50		12.63
1999-2006	6.00		4.63		1.38		1.50		11.63
1991-1998	5.88		4.00		1.88		1.75		8.63
1983-1990	5.71*		3.88		1.83		1.75		11.38

* -- adding 8-year average for 7 to 10 upsets and 6-year average for 13 to 15 upsets

Clearly there has been an increase in upsets over time, though not so much as you might think (basically a one upset per tournament difference between earliest period and latest period) and not consistently (for example, upsets by seeds 13 to 15 was more prevalent in the earlier time periods). Similarly, fewer 1-seeds have made the Final Four, but that trend didn't start in the one-and-done era and it isn't that much of a decrease (a quarter of a 1-seed per season, i.e., one fewer 1-seed every four years). It's hard to see a trend in the sum of the seeds of the four Final Four teams. Three of the past four seasons have been very high, but take away just one improbable run (say, Wichita State staving off UK and beating Louisville in 2014 or Kansas beating VCU in 2011), the average for the period would have been essentially equal to the average from 1983 to 1990.

I'll leave it to those better versed in statistics to tell me if the increase in upsets is truly significant, or if the data from the one-and-done period is really more variable.

Kedsy: I didn't verify your calculations, but let me offer two things: (1) the upsets for 10-12 in the past eight years is about two standard deviations away from the first eight years. Interesting, borderline significant, but obviously not definitive. (2) There are a heckuva lot of things that would explain an upward trend in upsets other than one-and-done players. Here's one alternative hypothesis:

As some historians summarized a few years ago, there were about 30 major basketball programs in the 1960s, where winning was extremely important and the schools put a lot of emphasis on recruiting, coaching and player support. Those observers estimated that in the early 1990s there were more than 100 programs that were really focused on winning and determined to do so. Today, I would estimate that there are more than 200 who will fund recruiting, pay really good salaries to coaching staffs, and clean house if the program isn't successful.

Another hypothesis would say that there are a lot more skilled and talented players who play basketball and train year around from the 12-13 on. These players are drawn by a big increase in college scholarships, the huge salaries of a few NBA players (well, more than a few) but by a worldwide basketball industry.

Both alternatives to the onset of the one-and-done rule would tend to explain why competition is fiercer and there are more upsets in NCAAs.

[Kedsy]

Kedsy: I didn't verify your calculations, but let me offer two things: (1) the upsets for 10-12 in the past eight years is about two standard deviations away from the first eight years. Interesting, borderline significant, but obviously not definitive. (2) There are a heckuva lot of things that would explain an upward trend in upsets other than one-and-done players. Here's one alternative hypothesis:

As some historians summarized a few years ago, there were about 30 major basketball programs in the 1960s, where winning was extremely important and the schools put a lot of emphasis on recruiting, coaching and player support. Those observers estimated that in the early 1990s there were more than 100 programs that were really focused on winning and determined to do so. Today, I would estimate that there are more than 200 who will fund recruiting, pay really good salaries to coaching staffs, and clean house if the program isn't successful.

Another hypothesis would say that there are a lot more skilled and talented players who play basketball and train year around from the 12-13 on. These players are drawn by a big increase in college scholarships, the huge salaries of a few NBA players (well, more than a few) but by a worldwide basketball industry.

Both alternatives to the onset of the one-and-done rule would tend to explain why competition is fiercer and there are more upsets in NCAAs.

Thanks. The calculations themselves were performed by a spreadsheet, so they ought to be right. Your explanation makes sense to me, though it also seems odd that even though the 10-12 upsets are up, the 13-15 upsets are down. Again, I have no idea if that difference is statistically significant, but especially since this conversation had its origins in people trying to explain Duke's recent first round losses (both of which came in the 13-15 category), it's an oddity that ought to be explored before any explanation is accepted.

For example, another possible explanation for the increased upsets at the 10 to 12 level is the seeming larger number of high major teams getting low (10 to 12) seeds (I haven't actually run the numbers on this, but it sure seems that way to me). You almost never see a high major in the 13 to 15 seeds, though, so that may explain the dichotomy. And if that's the case, then all the other explanations become minor undercurrents. Because it really shouldn't be so surprising that the 4th place Big 12 team beat the 5th place Big East team (which was an actual 10 over 7 "upset" in 2013).

And if my possible explanation is true, then it may be we need to look primarily at the 13 to 15 range, which would appear to show the exact opposite of the hypothesis that's been advanced (that there are more upsets and competition is fiercer now than in days gone by). Who knows?

[sagegrouse]

Thanks. The calculations themselves were performed by a spreadsheet, so they ought to be right. Your explanation makes sense to me, though it also seems odd that even though the 10-12 upsets are up, the 13-15 upsets are down. Again, I have no idea if that difference is statistically significant, but especially since this conversation had its origins in people trying to explain Duke's recent first round losses (both of which came in the 13-15 category), it's an oddity that ought to be explored before any explanation is accepted.

For example, another possible explanation for the increased upsets at the 10 to 12 level is the seeming larger number of high major teams getting low (10 to 12) seeds (I haven't actually run the numbers on this, but it sure seems that way to me). You almost never see a high major in the 13 to 15 seeds, though, so that may explain the dichotomy. And if that's the case, then all the other explanations become minor undercurrents. Because it really shouldn't be so surprising that the 4th place Big 12 team beat the 5th place Big East team (which was an actual 10 over 7 "upset" in 2013).

And if my possible explanation is true, then it may be we need to look primarily at the 13 to 15 range, which would appear to show the exact opposite of the hypothesis that's been advanced (that there are more upsets and competition is fiercer now than in days gone by). Who knows?

Well,... in any regime, the 13-15 upsets are gonna be rare -- therefore, differences are unlikely to be statistically significant.

When I said I didn't follow your calculations, I meant that I just waded in, calculated some sample means and sample std. deviations without any particular research into what you were calculating, which -- in hindsight -- seems perfectly done.

Quite frankly, one-and-done oughta make a big difference in the success, or lack thereof, of the teams that recruit the top talent. But then -- but then, they have been playing all season with young players. Why should they automatically perform worse in the postseason than in the regular season? I thought, for example, that Kentucky in 2013-14 improved steadily throughout the season, reaching the final game.

[Kedsy]

Well,... in any regime, the 13-15 upsets are gonna be rare -- therefore, differences are unlikely to be statistically significant.

Thanks again. The more I think about it, the more I think there's nothing here. Which is, in itself, something, I suppose. To wit, there's nothing (or at most, very little) to suggest the one-and-done era has had a significant effect on major NCAAT upsets, and nothing to suggest "high player turnover in this era lends itself to higher program variability from year to year." Certainly nothing that would support the hypothesis, seemingly held by many, that recruiting one-and-dones has been or (perhaps more importantly) will be detrimental to Duke's post-season hopes.