Reality check from the land of Trust in Bill Self
-
@Blown said:
Its just coincidence but I did notice 12 of 13 titles were won by Eastern Time Zone Teams. 42 of 47, or something has been won by eastern time zone schools since ESPN became televised.
Are the odds of EST teams winning 12 of the last 13 and 42 of 47 since ESPN began televising the games greater, or lesser, than the WTC towers being the first and only incidence of three high rise towers designed to withstand commercial jet impacts falling in their own foot prints in a single day?
To be fair let’s add Red Auerbach’s 10 NBA rings in 12 seasons, Wooden’s 10 rings in 11 years and Self’s 13 titles in 13 years.
What is it about American basketball that makes it so prone to to so many extraordinary streaks?
I am reminded that James Naismith saw his game invented in a Springfield YMCA spread within a couple years throughout Northeast and Great Lakes industrial cities as a professional sport staged by boxing and burlesque promoters with gambling thoroughly a part of the games.
Naismith actually abandoned his game for several years and moved to Denver to get a Doctor of Osteopathy degree. Looking back one wonders if he became disillusioned by his lost control and the sport’s saturation with gambling.
Gambling is about the odds of something happening. Basketball has long been plagued by interference in its outcomes. Perhaps the interference has been more than we suspected?
-
A little late to the game but I decide to give myself a little break from posting to recover from the loss and avoid the acrimony that follows a loss in the tournament.
The coin flip analogy is one that is often used to make a point but in this case, as it is in most other cases where its is used, not an applicable one.
First, the coin flip statistic assumes that you are using ONE perfectly balanced coin, with each side having a perfect 50% chance and the number becomes more accurate with a larger number of coin flips; a small sample will not necessarily show a perfect distribution. Second, every Elite 8 has different teams with different players that change from year to year, so this alone invalidates the comparison. Third, the winning probability of teams playing in the Elite 8 is not a perfect 50% but often something quite different and last, the sample is not big enough to show a definite trend.
Nice try but no cigar…a consolation single malt is OK.
-
@JayHawkFanToo And, by focusing on the tool, you missed the point. I was trying to illustrate by using something familiar (coin flips) that the loss rate by Self is not something statistically so far out of likelihood that, as expressly stated by some here, the only possible explanation is something aberrant about his coaching. I think a 2-5 record with KU, demonstrates no such thing.
Each game has 2 possible outcomes, thus in measuring win/loss statitical probability you are only measuring distribution of all possible outcomes, not the factors that give a team a higher or lower chance of success. Bill’s loss rate in E8 games occurs 21 times out of 128 total possible outcomes, ranging from all wins to all losses. That is approximately one out of six times. The coin flip analogy simply demonstrated where that falls along the probability of an outcome in a test we all are familiar with.
The bigger problem with all of this is the failure to consider where his results fall on graphs that also incorporate, among other things, his seeds, the average success rate of those seed levels in the history of the modern tournament, how he compares to others in other rounds, and the results in all rounds not just the 8. Izzo, by comparison, is great in the E8. But Bill by comparison has Izzo beaten in the semifinal in winning percentage terms, but his sample size is much smaller.
-
@mayjay said:
@HighEliteMajor said:
This relates to how a coach has his team ready to play on a short turnaround vs. an unfamiliar opponent.
Just a thought about this, because it has been mentioned a lot. Does this reflect a flaw in short-turnaround coaching by Self but his 12-5 record in the second round with that same turnaround does not mean the opposite?
I guess I am not willing to make very firm conclusions or find patterns where these things are based on info so highly selective.
And we play a ton of Sat / Mon games. Granted - not unfamiliar teams.
-
No, I am not missing the point. When the #1 seed play the #16 seed the game has two possible outcomes, right? but they are not even close to being equal in probability as it would be when you flip a coin…so far the #1 seed is 100% and the #16 seed is 0%. Again, just because there are two possible outcomes it does not mean they have equal chances.
Again, you are comparing coin flipping where the perfectly balanced coin never changes and the flip is is exactly the same every time to an event in which the teams are different every time and also from year to year and with a different winning possibility every single game based on team personnel at the time of the game. For example, this year we had #1 Gonzaga playing #11 Xavier and #1 UNC playing #2 Kentucky; this is how much variation there is in the Elite 8.
A friend of mine is a top actuary for a well know insurance company and he always laughs at the way sports writers use statistics, more often than not incorrectly…we used to have very interesting conversations on the subject although I mostly listened and learned because even when I have a pretty good working knowledge on the subject he is in a different league than I am.
Nice try…but still no cigar.
-
@mayjay Coin flips are not 50/50. They are biased towards whatever side is face up before the flip.
Let’s say that a coin flip truly was 50/50, there’s no variable. In Self’s E8 games, there are many variables. The school Self is coaching at, the school Self is coaching against, the coach Self is coaching against, the players Self coaches, the players Self coaches against, injured players, the style officiating by the refs. These are all variables that make basketball games very rarely a 50/50 proposition. The only constant in these 9 games has been Bill Self.
So when you have a replicated performance (6 of 9 times) woth only one constant, that constant has some kind of impact that causes these repeated outcomes.
Go.look at Self’s record in the November tournaments KU has had to play 3 games in 3 days in, there’s a lot of losses with poor offensive games in there as well.
-
@JayHawkFanToo I am measuring outcomes after the fact. There are a finite number of outcomes each measured only by win or loss. You are talking about probability of winning, which is a measure of predictability.
Bill could have had WWLLLLL, or LLLLLWW, or LWLLLWL, or any other distribution of 2 wins and 5 losses, whatever the order. etc. If he had zero wins, there is only one distribution where that occurs. If he had 7 wins, also only one. If he had 3 wins, or 4 wins, or 5, or 6, each of these has a finite number of times that occur in the distribution. He had two, which occurs 21 times in a distribution of every possible combination of 2 wins and 5 losses.
You are arguing as if I said he had a 16% chance of winning. That isn’t it. But remember, I am only saying the numbers alone do not mean it is solely the result of coaching.
Forget my argument, because you don’t obviously don’t understand it. Maybe it is me. Do me a favor. Look at the only assertion I am arguing against: they say he could not have lost 5 out of 7 for any reason other than coaching. Use your skills on that one. Tell me why the number of losses vs the number of win cannot just be a coincidence.
That argument depends on an assumption that a coincidence is too unlikely to have occurred. So, where is the proof of improbability?
Remember, I am only saying coincidence cannot be dismissed because a record of 2 W 5 L is not so far out of the realm of possibility as to be virtually impossible.
-
@JayHawkFanToo And, by the way, this is not the use of stats your friend complains about. It drives me crazy when someone says “he should have let Brett hit because he had a 30% chance of getting a hit” based on his batting average.
Take my hypothesis and my posts to your friend. See if he agrees with me that the distribution of all outcomes can be measured to determine if a particular set of outcomes (with no other factors) is statistically not unusual.
-
@mayjay said:
The bigger problem with all of this is the failure to consider where his results fall on graphs that also incorporate, among other things, his seeds, the average success rate of those seed levels in the history of the modern tournament,
So here’s Bill Self’s seed history in his E8 games and the seed he was coaching against
2000^:Tulsa (7) lost to North Carolina (8)
2001: Illinois (1) lost Arizona (2)
2004: Kansas (4) lost to Georgia Tech (3)
2007^: Kansas (1) lost to UCLA (2)
2008^: Kansas (1) beat Davidson (10)
2011^: Kansas (1) lost to VCU (11)
2012: Kansas (2) beat North Carolina (1)
2016^: Kansas (1) lost to Villanova (2)
2017^: Kansas (1) lost to Oregon (3)
^ Year Self’s team had worst offensive game of season in E8
Out of the 6 times Bill Self’s team has had their worst offensive game of the year in the E8, Self has been a 1 seed including all 5 times at KU. KU is 1-5 in those 6 games. Overall, Bill Self has had the higher seed in 7 of his 9 E8 games and is 1-6 in those games. Bill Self has been a 1 seed in 6 of his 9 E8 games and is just 1-5 in those games and his team has had their worst offensive game of the year in 5 of those 6 games. 2008 is the 1 win in all of those records.
1 seeds are historically 38-34 against 2 seeds in the NCAA tournament since the field expanded to 64 teams in 1985. So 1 seeds historically win about 53% of the time against 2 seeds. Self is 0-3 as a 1 seed in 1 vs. 2 match ups and 1-0 as a 2 seed in 1 vs. 2 match ups so 1-3 overall. 25% is much lower than 53%. 1 seeds are historically 20-13 against 3 seeds for a 60% win rate. 1 seeds are 6-1 against 10 seeds , and 1 seeds are 3-3 against 11 seeds. 4 seeds 3-5 against 3 seeds all time and 7 seeds are 1-1 against 8 seeds all time. Here is the source for you (http://mcubed.net/ncaab/seeds.shtml). Bill Self is 1-6 in the E8 as the higher seeded team. I don’t care where you look, you will not find anywhere that a higher seed has 14% winning percentage against a lower seeded team. Just looking through that site, the anomalies are 2 seeds against 4 and 5 seeds (4-5 against 4 seeds and 1-4 against 5 seeds), 4 seeds against 6, 7, and 8 seeds (2-4 against 6 seeds, 2-3 against 7 seeds, and 3-7 against 8 seeds), 5 seeds against 8 and 9 seeds (1-3 against 8 seeds and 1-2 against 9 seeds), 6 seeds against 8 seeds (1-3), and 10 seeds against 11 seeds (1-2).
Self’s record is below average based on what other same seeds have historically done against the same seeds Self has face in the E8. I believe the only other 7 vs. 8 game in NCAA tournament history besides the Tulsa/UNC game in 2000 was the UConn/Kentucky title game in 2014.
-
You still don’t get it. The coin flip analogy is completely not applicable n this case. You are trying to apply a deterministic statistical approach to a process with multiple variables and random outcomes and where a Bayesian Probability Approach is in order. I will not go in detail on what this it because it would bore the bejesus out of everybody here…and probably me too… but you can probably google it and get the general idea. I have myself worked more with Markov Chains since is is more applicable to the work I do but my friend is an expert n Bayesian Probability Analysis, since that is more applicable to the field he is in.
This is quickly getting our of hand and it is time that we get back to just plain ol’ KU basketball.
P.S. I have not talked to my friend in a while but I will see if I still have his phone number; I believe he retired a couple of year ago.
-
@Texas-Hawk-10 Thank you for that info. All of which is more than just his simple win/loss record in one particular round. Knowing more about the losses’ circumstances may tend to implicate coaching, but I also think it implicates incorrect seeding due to our consistent overachieving in the regular season–i.e., I believe we are not as our seeding because our record each year appears to be stronger than we are. (Look how distorted our close game record was vs the average.)
Jayballer keeps citing a stat that only 25% of #1 seeds have gotten to the F4. I haven’t checked that, but (assuming we throw out HCBS’s record to avoid skewing the average toward him) if that 1 over 2 record is as strong as you cite, that must mean most #1 seeds who lost before the F4 (that 75%) either got knocked off early or lost to seeds lower than 2 in the E8.
Seeding drives me crazy because during the tournament everyone knows that it is based on a committee’s collective judgment regarding the relative strength of teams. There is no consensus on how to measure this, RPI vs BPI vs the computer models vs blah blah blah. There is also the monkey wrench thrown in that we don’t have a good working knowledge of how the committee does its work.
A fourth best #1 and the best #2 might easily be switched in almost every tourney with few cases of heartburn. You could probably shuffle all the teams on any particular line without causing undue shock, so the same would also likely be true of shuffling the last two on one line with the best two on the next line.
Would we be as upset if KU had been lower seeded in the years we have lost in the E8? Does the panache of a #1 seed create unrealistic expectations if 75% of them do not live up to their seeding.
-
@JayHawkFanToo So you didn’t even read my last post since I don’t mention coin flips. I am talking finite outcomes (there are only two). You are worrying about how likely those outcomes are (impossible to measure). That is irrelevant when counting things that have already happened.
-
@mayjay Since 1985 when the field expanded to 64 teams (including this season), out of 128 total 1 seeds, 50 have reached the Final Four. That’s 39%, but we can round that up to 40% to simplify the math which says that for every 5 one seeds out there, 2 of them should reach the Final Four. Bill Self has received a 1 seed 8 times (2001 with Illinois, 2007, 2008, 2010, 2011, 2013, 2016, and 2017. The numbers say Bill Self should have between 3 and 4 Final Four appearances from these 8 seasons alone, but he only has 1 so again, he has underachieved to historical performances of other 1 seeds.
-
He’s a hall of fame coach, wouldn’t want another one, and I think he’ll get some more Final Fours and titles. Just wish he’d hurry up and figure it out.
-
I’m for Frank Martin 100%!
I’ve always been a fan of Frank’s… and it just intensified when he said all those great things about Kansas.
Frank is the real deal.
Go, Gamecocks!
-
@drgnslayr I missed it, what did he say about Kansas?
-
@Texas-Hawk-10 Wouldn’t it be more fair to mention 2012? What seed were we? How many times have we had that seed? What are the averages?
Maybe Self has overperformed as a non-1 seed: seems fair to say so…
-
@ParisHawk Self’s been a lower seed in the E8 twice. KU lost as a 4 seed to 3 seed GT in 2004 and beat 1 seed UNC as a 2 seed in 2012. 2 games is not enough to draw any meaningful or trends out of.
-
@Texas-Hawk-10 said:
@ParisHawk Self’s been a lower seed in the E8 twice. KU lost as a 4 seed to 3 seed GT in 2004 and beat 1 seed UNC as a 2 seed in 2012. 2 games is not enough to draw any meaningful or trends out of.
And that 1 seed UNC team was missing their starting point guard.
-
@Texas-Hawk-10 said:
@ParisHawk Self’s been a lower seed in the E8 twice. KU lost as a 4 seed to 3 seed GT in 2004 and beat 1 seed UNC as a 2 seed in 2012. 2 games is not enough to draw any meaningful or trends out of.
Which helps whoever wants to emphasize Self’s postseason shortcomings…
-
When he was at KSU he mentioned KU many times in a very positive light.
-
@ParisHawk Also, just more known info … we did underachieve as a non-#1 seed against Bucknell, Bradley, Stanford, Wichita St. Of course, those sort of stumbles are not unique to high-level coaches and every top coach seems to have similar missteps.
-
The Bucknell game was bad. Langford was on a bad wheel and Wayne seemed really tired at the end of that season.
The Bradley game was the 05-06 team. That team played well at times but was REALLY young and had chemistry issues with Julian Wright. I dont think Self ever figured out the best rotation with him. He was such a hybrid player.
The Stanford game we missed Embiid but Self should have started pressing earlier.
Withcita State loss was a roster of no go to players. Ellis wasn’t that confident yet, Selden never was reliable, and Frank wasn’t really Frank yet. I also don’t think our guys wanted it like they did. The “nobody respects us” is always a huge motivator.
-
If you really want to feel like someone throwing a pity party (pretty much what it is at this point) realize that KU has the 3rd best winning percentage in the NCAA tournament all time at 69% behind Duke 75% and Florida 71%
-
Basketball, because it can be dominated by one or two great players, is prone to dominance you can’t see in other sports because the best players can play such an outsized role.
Jordan’s Bulls won 6 of 8 titles in years he played in the 1990s.
Lebron James has been to the Finals in each of the last six seasons.
If you have one of the top talents, you are automatically good. That’s just what happens with basketball. Magic’s Lakers, Bird’s Celtics, as long as those guys were healthy, their teams were among the best.
College is a bit different because coaches tend to focus on systems more because the talent is always changing, but that means the best coaches have an advantage - it’s no accident that the last 10 titles in college hoops look like this - Donovan, Self, Williams, Coach K, Calhoun, Calipari, Pitino, Ollie, Coach K, Wright. That’s 6 titles from current HOF coaches Williams, Coach K, Calhoun, Calipari, and Pitino, two from likely HOF coaches Wright and Self and one from a coach that probably would have been a HOF coach had he stayed in college (Donovan). We could stretch it back a few more years and we would find Donovan again, Williams again, Calhoun again, Boeheim (another HOF coach), Gary Williams (another HOF coach), Coach K again, Izzo (another HOF coach), Calhoun again, Tubby Smith, Lute Olson (another HOF coach), Pitino again.
Out of the last 22 titles, 15 have been won by current HOF coaches. The only coaches that have won titles in the last two decades plus that likely won’t make the HOF are Tubby Smith and Kevin Ollie, and perhaps Billy Donovan since he moved to the pros. They represent just 4 titles.
That influence is unlikely to ever change.
-
@HighEliteMajor Not to pile on, but Northern Iowa as well. Yes, all coaches have tourney missteps. Looking from the outside, Self has more than his share, though. Getting to the E8 7 out of 14 seasons is pretty darn good…just too bad we couldn’t have won another 1 or 2 of those…resume would look a lot better IMO.
-
@Hawk8086 Right … I had followed up on exits when we were a non-#1 seed. The #1 seed stuff is obviously much more the irritant. As you said, if we’d snuck out a couple of those, we’d be in a different place in our discussion – if he’d lost in the FF, we’d wonder why he couldn’t win the title game (though, odds say, winning one of four FF appearances is par).
@dylans Real discussion of tournament issues has nothing to do with a “pity party.” I guess you don’t feel the discussion has any validity. That’s certainly fine but I don’t see “pity” being the purpose – the purpose is the same as all the discussions here. Love for KU basketball and enjoyment in discussing the same. Sometimes that discussion isn’t all peaches and cream.
-
Great job EVERYONE! I know some feel frustrated at failing to persuade others, but this has been one of the most interesting airing of the basis issues I have read. Tried to Red Rooster everyone!
-
@HighEliteMajor I’ve read 3-4 posts with any ideas on how KU could’ve won the game by doing something differently. Bitching about Selfs record in the elite 8 while ignoring the actual matchups is asinine.
The couch coaches come out of the woodwork this time of year. Gotta prove how they know more than everyone including a hof coach. So where were you when the times were good?
Crying about a week old game while several posters are getting ready to go to chemo. Several more are going to funerals (got one this weekend for the sweetest aunt one could ever have). Others are struggling due to the markets, or related job loss. Etc is just plain silly.
If you want to cry there are good things to cry about. Time to move forward. There are good things on the horizon. Positive recruiting news, surprising player development, Bill Self year two in a new scheme that in year one got to the elite eight.
I’m looking forward to the death of the pity party and the birth of the build up of next season. Endless chances for optimism as long as Bill Self is the coach KU will have a shot.
-
@dylans Sorry, but your post doesn’t make sense to me.
Why discuss KU basketball at all when there are so many more important things to discuss? That’s somewhat you point. Actually, why discuss things that you don’t think have merit, when there are more important things to discuss, might be more accurate. Or when there maybe more important things going on in certain folks’ lives.
No one says this is life and death.
I always wonder the purpose of such replies, like yours … if the discussion so offends, and is so trivial, why participate?
Seems like you’re doing a lot of the trivial complaining while many others are discussing what is a very significant issue with KU basketball … albeit not on the scale of North Korea nukes.
If you want to move forward, then move forward. Start a thread. You’ll probably get a lot of takers.
-
@HighEliteMajor ok. Kindly keep my name out of your mouth. I’ll gladly do the same for you.
-
@dylans I’m sorry, what is it that you mean?
