RE: KU Tourney Success
by Bensa on Feb 22, 2012 2:57 PM CST
Email
PrintI hinted in Marty's very fine FanPost that I might try to compile some data showing how KU's NCAA tournament losses stack up to the other teams that lead the nation in NCAA tournament wins since Bill Self came to KU. A couple of notes on the data: I only counted NCAA losses. If a team won the national championship, that did not affect their score. Similarly, if a team missed the tournament in a given year (or years, coughUCLAcoughUCONNcoughFLORIDAcough), that fact did not affect their score.
As a recap, here is the data for KU over the same span:
2004: KU 4; GT 3 (1)
2005: KU 3; Bucknell 14 (-11)
2006: KU 4; Bradley 13 (-9)
2007: KU 1; UCLA 2 (-1)
2008: National Champs
2009: KU 3; Mich. St. 2 (1)
2010: KU 1; UNI 9 (-8)
2011: KU 1; VCU 11 (-10)
Total loss-seed disparity: -5.28 per loss
Follows the data I compiled:
UConn
2004: NCAA Title
2005: 2 (10): (-8)
2006: 1 (11): (-10)
2007: DNP
2008: 4 (13): (-9)
2009: 1 (2): (-1)
2010: DNP
2011: NCAA Title
UConn Total: -28 (4): -7
UNC
2004: 6 (3): (+3)
2005: NCAA Title
2006: 3 (11): (-8)
2007: 1 (2): (-1)
2008: 1 (1): (0)
2009: NCAA Title
2010: DNP
2011: 2 (4): (-2)
UNC Total: -8 (5): -1.6
Duke
2004: 1 (2): (-1)
2005: 1 (5): (-4)
2006: 1 (4): (-3)
2007: 6 (11): (-5)
2008: 2 (7): (-5)
2009: 2 (3): (-1)
2010: NCAA Title
2011: 1 (5): (-4)
Duke Total: -23 (7): -3.3
Mich. St.
2004: MSU 7; Nevada 10 (-3)
2005: 5 (1): (+4)
2006: 6 (11): (-5)
2007: 9 (1): (+8)
2008: 5 (1): (+4)
2009: 2 (1): (+1)
2010: 5 (5): (0)
2011: 10 (7): (+3)
Mich. St. Total: 12 (8): +1.5
Kentucky
2004: UK 1; UAB 9 (-8)
2005: 2 (5) (-3)
2006: 8 (1): (+7)
2007: 8 (1): (+7)
2008: 11 (6): (+5)
2009: DNP
2010: 1 (2): (-1)
2011: 4 (3): (-1)
Kentucky Total: +6 (7): +.85
Florida
2004: 5 (12): (-7)
2005: 4 (5) (-1)
2006: NCAA Title
2007: NCAA Title
2008: DNP
2009: DNP
2010: 10 (7): (+3)
2011: 2 (8): (-6)
Florida Total: -11 (4): -2.75
Texas
2004: 3 (7): (-4)
2005: 8 (9): (-1)
2006: 2 (4): (-2)
2007: 4 (5): (-1)
2008: 2 (1): (+1)
2009: 7 (2): (-5)
2010: 8 (9): (-1)
2011: 4 (5): (-1)
Texas Total: -14 (8): -1.75
UCLA
2004: DNP
2005: 11 (7): (+4)
2006: 2 (3): (-1)
2007: 2 (1): (+1)
2008: 1 (1): (0)
2009: 6 (3): (+3)
2010: DNP
2011: 7 (2): (+5)
UCLA Total: +12 (6): +2
A couple of conclusions: UConn's score is slightly skewed by the fact that they only had 4 NCAA losses over the past 8 years (2 titles, 2 DNPs), but still significantly higher than KU's. That said, UConn is the only school that even approaches KU's loss-seed disparity. Of course KU is largely a victim of its own success. It's never been seeded worse than a 4 seed since 2001 (let that sink in here for a bit), so, like Duke, for instance, KU's losses are much more likely to be to a worse seed. A team like Michigan St., on the other hand, is seldom better than a 5 seed, so its tournament success has largely come at the expense of better-seeded teams, and further, its losses are usually to higher seeds.
So what's the point, exactly? I'm not sure. I'm not going to sit here and say that HCBS can't coach in the big game (we all know that's simply not true), but his teams do tend to (dare I say the word?) choke when they're heavily expected to win. Not every time, obviously, but it's happened more frequently than any of us would like to admit or remember.
Anyway, these are the numbers. Discuss!
26 comments
|
2 recs |
Do you like this story?
Comments
I don't think an early exit can possibly be worse than missing the tournament altogether
However frustrating it is to go out early, it would have to be far worse to be stuck in the NIT, right?
We wouldn't know
We’ll have to ask UNC
by jayhawk3r on Feb 22, 2012 3:33 PM CST up reply actions 3 recs
hell yes
I probably wouldn’t even watch an NIT game
I write everywhere. You're probably better off following me on twitter
I wouldn't even know where to look
to watch an NIT game. They’re probably on ESPN but I’ve never watched one because it’s never included the team I would watch it for in my lifetime.
Pay heed...
by pj_hawk4life on Feb 22, 2012 3:58 PM CST up reply actions
hellll yea
there should be some negative point distributor for not making it and a positive for winning it all (probably not possible).
I’d rather be a 2 and lose to a 15 than not make the tourney…..
by I need more Esteban on Feb 23, 2012 6:17 PM CST up reply actions
Also, how is it better to be an 8 losing to a 1 in round 2 than a 5 losing to a 2 in the elite eight?
It seems to me that a loss to a higher seed is a loss to a higher seed. I wouldn’t give extra points for losing to a way-higher seed.
This post is incredibly expertly formatted
I write everywhere. You're probably better off following me on twitter
I have two major problems with this type of analysis
First, it’s impossible for a 1 seed to do anything but break even. It’s incredibly difficult for 2 and 3 seeds to do the same. Granted, these teams should have higher expectations, but given the unpredictable nature of the tournament, it’s unreasonable to punish them (or anyone else) statistically any time they fall short of a Final Four.
My second issue (and the more significant one) is that there’s only one measure being used to judge the quality of the teams here: seeding. For us to assume that this gives us an accurate depiction of what a team “should” do in the tournament, we have to accept as fact that seeding is an accurate depiction of how good a team truly is. Given that the selection committee is famously ignorant, and prone to a lot of questionable tendencies (e.g. reliance on RPI, undervaluation of mid-majors, overemphasis on recent performance, etc.) this is an immeasurably flawed data bank to use when determining what a team is “supposed” to do in the postseason.
Granted, looking at these numbers over a period of time can suggest certain tendencies, but to use them singly to make any kind of definitive point lends itself to inaccuracy.
Use the forks
It also rewards teams that underperformed in the regular season
K-State was preseason #3 last season. They were a 5-seed in the tournament. Should KSU get a +1 for losing to a 4-seed because they were a 5-seed or should KSU get a -3 for losing to a 4-seed because KSU should have been a 1-seed?
this isn't all directed at you, Bensa
since, as you said, you aren’t trying to prove a point. More just my general thoughts on using seeding to form a metric analysis
Use the forks
by PenHawk on Feb 22, 2012 7:18 PM CST via mobile up reply actions
No I agree, wholeheartedly, and no offense taken at all.
Honestly, I’m really baffled why we tend to lose these games in which we’re highly favored. Note, outside of the Bucknell/Bradley years, our losses have all been in the 2nd game of the weekend. You know, the one that gets you to the next weekend? I really wonder why that is.
Criticize FHOFNCHCBS at your own risk...
That's about what you would expect from a team that is extremely well-coached.
The first game of the weekend, you have 4-5 days to prepare for the team you are facing, get plenty of rest and practice time, etc.
The second game, you usually find out your opponent about 40 hours before tip-off, you have to do scouting/prep work for that opponent in a short amount of time, and players are generally still shaking the effects from the previous game.
Such a trend actually leads me to think more highly of Bill Self’s coaching, because more of the factors of the game are out of his control in the second game of the weekend.
by misterbrain on Feb 24, 2012 10:11 AM CST up reply actions 1 recs
My second issue (and the more significant one) is that there’s only one measure being used to judge the quality of the teams here: seeding. For us to assume that this gives us an accurate depiction of what a team "should" do in the tournament, we have to accept as fact that seeding is an accurate depiction of how good a team truly is.
See Bradley loss here
by I need more Esteban on Feb 23, 2012 6:18 PM CST up reply actions
Good work Bensa, it's interesting to see. But....... ;)
I agree with PenHawk and KSinDC that the metric is a bit off – in particular that the best thing that can happen is getting a high number seed and losing.
One way to eliminate that would be to look at the distribution of results for teams seeded 1 through 4 (since that’s KU’s lowest) and lay that over KU’s results for comparison. I also think you have to include wins and losses. If what we’re looking at is “how does/did KU do versus a much lower-seeded team?”, then we want the distribution of seeds 1 through 4 versus seeds in the bottom half (9 and below) – did they win or did they lose. That introduces arbitrary endpoints on both sides, but at the same time, we’re asking a pretty specific question, so that danger may be unavoidable.
Kila's slash for Apr 20 to May 4, 2011, right before he was sent down: .276 / .344 / .448
by SagehenMacGyver47 on Feb 23, 2012 12:16 PM CST reply actions
Just to clear up something:
The only point I’m trying to make (really, I promise) is that when KU loses in the tournament, it loses to substantial underdogs more often than most any other successful program. That’s it. We can debate the whys and the hows somewhere else. I feel like the numbers prove that point pretty decisively.
Criticize FHOFNCHCBS at your own risk...
by Bensa on Feb 23, 2012 6:56 PM CST reply actions 1 recs
That may be true
but by my calculation, Bill Self is 13-4 at KU against mid-major teams in the tournament. I think those four have been so devastating to our psyches that they tend to form the entirety of our opinions on the subject, but 76% against just about any group of teams when you get to something as unpredictable and upset-ridden as the NCAA Tourney is pretty damn good.
Granted, the bottom line is that four teams, who had no business doing so, knocked us out of title contention, and granted, I have no idea what other teams’ records are by comparison. I just think that this subject gets to be a tad overblown this time of year.
Use the forks
that being said
I still, and will always contend that Bradley was not that big of an upset. That was a senior-laden team playing a bunch of sophomores and freshman and I believe they were underseeded. The rest, yeah.
by I need more Esteban on Feb 24, 2012 1:48 PM CST up reply actions
This is actually my only issue with PASE
So much of how you do depends on how well the committee seeds the strong mid-majors. Some committees have been far better than others, and that makes a huge difference in how we’d actually expect high seeds to do.
Balancing the brackets is another issue, but I think it’s harder to do something about that.
I think DNP should be penalized
Those teams basically get a pass for a year when the SUCKED, really can skew the data ;)
Good posts BTW, fun way to look at the numbers
by MichiganJayHawk on Feb 24, 2012 9:28 AM CST reply actions
I don't think it skews the data at all, depending on how you look at it:
All I’m trying to show is, as I said above, that when KU loses in the tournament, it is usually to a heavy underdog. Teams that don’t make the tournament don’t lose to underdogs, so the numbers aren’t skewed, really.
Criticize FHOFNCHCBS at your own risk...
I'm just being snarky hence the winky face
by MichiganJayHawk on Feb 24, 2012 10:13 AM CST up reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
