Reading this thread made me think of the 98 team coming from 10 down against Utah. We also came from 10 down against Duke.
Last edited:
Coach Cal at UK is 0-11 when down 10 or more at the half
- Thread starter catben
- Start date
Kansas fans don't think it is possible for ANY team to win 38 games in a single season.
Thank you for making my case for asking about this stat much better than I could myself.
It is a great indicator of our probability to win a game if we are down double digits at halftime.
Cal is the best coach in the country IMHO, so this stat will not change that. However, it is interesting none the less.
It is a fairly meaningless stat, in my opinion. It's one of those that sounds significant but it's blatantly obvious.
It's like saying "Calipari has lost every game when his team had fewer points than the opponent when the final buzzer sounds." Duh!
Trailing by 10 is a pretty significant margin. It is a rare thing for a Calipari coached UK team to trail by 10 in any game, so for it to be a halftime score tells me that the game is a mismatch of some kind. So of course the team would lose most, if not all, of those games.
It's like saying "Calipari has lost every game when his team had fewer points than the opponent when the final buzzer sounds." Duh!
Trailing by 10 is a pretty significant margin. It is a rare thing for a Calipari coached UK team to trail by 10 in any game, so for it to be a halftime score tells me that the game is a mismatch of some kind. So of course the team would lose most, if not all, of those games.
In case anyone was waiting for it, below is what the cumulative graphs look like.
Basically this is all the same data but looked at three different ways.
For example the blue dots are UK's Win% given a particular halftime margin. For example if they are +5 at halftime they have ~ an 85% chance of winning the game.
The green dots are UK's Win% if they are at the particular margin or higher. So for example if the margin is +5 or greater they have ~95% chance of winning the game.
The red dots are UK's Win % if they are at a particular margin or lower. So if the margin is +5 or lower, they have ~ 55% chance of winning the game.
Not sure if this helps any but that's what it looks like. Again the area of interest is really if UK's halftime margin falls somewhere between -15 to +10. Outside of that and the game's outcome is pretty much a foregone conclusion.
Basically this is all the same data but looked at three different ways.
For example the blue dots are UK's Win% given a particular halftime margin. For example if they are +5 at halftime they have ~ an 85% chance of winning the game.
The green dots are UK's Win% if they are at the particular margin or higher. So for example if the margin is +5 or greater they have ~95% chance of winning the game.
The red dots are UK's Win % if they are at a particular margin or lower. So if the margin is +5 or lower, they have ~ 55% chance of winning the game.
Not sure if this helps any but that's what it looks like. Again the area of interest is really if UK's halftime margin falls somewhere between -15 to +10. Outside of that and the game's outcome is pretty much a foregone conclusion.
Last edited:
I don't know about other teams. One difference is that as I showed earlier the average of UK's distribution is ~+5, whereas if you looked at all teams the average would converge to 0.
But the shape of the curve may indeed be the same, as you suggest.
FWIW, on Kentucky Sports Radio a few days ago there was a caller who mentioned that there was a study done over a large amount of data and they found a linear relationship where IIRC the slope was about the same as what I've found with UK's data. But I don't remember exactly what was said and haven't seen any academic articles about this.
If anyone can locate such an article, let me know.
You know I dont mind you chiming in RC, I am a bit disappointed that you thought UK fans were unaware of UCLA's tainted run without without a glimmer of suspicion that the poster was alluding to a different record...
This isnt a Mizzou board, we know our basketball...
I don't think it is as meaningless as you do. The 98 team did it twice in the NCAA tournament. I guess I never appreciated that fact as much as I do now. Admittedly, there may often be other factors at work. Where a stat like this one takes on meaning is when it is compared to other coaches or teams. Then, it spurs some kind of discussion.
Attempting to compare Cal's results to the average coach is ridiculous because only one or two other coaches in the county have UK's talent. I think it is a reasonable expectation that teams with our talent should be able to win a few of those games.
And of course do not take into account that our talent is 17 and 18 years old with zero chemistry at the start of each year.
You and Zipp really have a hard time with college talents vs NBA potential a few months removed from HS talents.
It is Cal's decision to bring those 17 and 18 years old with zero chemistry to start the season each year. One can't expect a lot of sympathy for a situation that is self inflicted.
As i said, it is silly to argue with the philosophy after such great results.
I like this "self inflicted" term from you - it is Cal's fault we are the way we are this year.
I guess it is Cal's fault we have a title and numerous FF's too eh?
Did you really just admit that?
I think you fully understood my self inflected point. Cal is the god father of the one and done player. He is the one that decided to build his teams around these types of players and worked equally hard to get them into the NBA even when some weren't ready. So, I don't want to hear any more excuses about players' ages or inexperience.
Last edited:
So you also believe Cal pushes kids to the league that are not ready... (We know the answer to that one).
And 'm sorry to inform you, if Cal's self inflicting is the reasoning for the negative results then it should be for the positive ones as well.
I know you hate the OAD and havent liked Cal since he came,
but to deny his results and to deny that the top 2 programs in college are using the same philosophy, and they are both wrong in doing so seems a tad off.
Cal and K > LmdCat
I believe 38-0 is a record for one season. They didn't play up to 40 games back in Wooden's day.
More to the point I guess we definitely don't want to be down by 10 or more at halftime do we.
More to the point I guess we definitely don't want to be down by 10 or more at halftime do we.
As Cal evaluate players season by season, I see nothing wrong with using the same evaluation process for the coaches. Last year success doesn't buy this current team success.
Since I am not being paid 7 or 8 million to coach, Cal and K should be greater than LmdCat. However, if that is your measuring stick, it says a lot about what you know about the game.
Fair enough, but your objections are not worded that way.
You simply criticize and say "Cal should or should not", "Cals' issues are self inflicted", etc, etc...
I cannot find one statement from you that says "this year Cal ...." no your criticism are always blanket statements.
Add to the fact that you have been a negative critic since he arrived lessons the sincerity of this post a bit as well.
Most people live in the present so I assumed that you do as well which is why I didn't think it was necessary to point that out. Most of these discussions have been about this team and not those of the past.
More talented players do not necessarily equate to better players.
For example, Bledsoe is more talented than Meeks. However, Meeks during his junior season was in another stratosphere in ability when compared to Bledsoe as a freshman.
Last edited:
Kentucky was down 17 with less than 10 minutes left against Duke in 1998 in the NCAA tournament and won.
WILDCAT CHAD has a negative comment 100% of the time when he responds on this board.
Back to the anger management siminar Chad, you inmature CLOWN.
Yep! It was 10 at halftime and 17 in the second half. VERY impressive! This thread is more proof of how impressive it was.
One final graph. I did go back and listen to the part of the KSR Thursday podcast where the caller mentioned that a study had been done on win % as a function of halftime margin.
According to the caller: "A team is down by 6 at half they have a 80% chance of losing. It goes up 8% points for 2 points on the scoreboard."
Matt promptly misinterprets this to start counting backwards to down 8, down 10 and down 12 to arrive at a 102% chance of losing.
But I find this to be pretty consistent with my findings, at least in the region where the probability is actively changing (which as I've already noted for UK seems to be -15 to +10). This -15 to +10 range covers 25 points, over which the team's chances of winning for the most part ranges from 0% up to 100%, so as long as it's linear (which it appears to be) the slope is +4%/point. This is exactly what the caller noted.
I dug a little deeper and did find some articles about this from ~2009. The following article by a guy (Brian Burke) who regularly does advanced football stats but looked at basketball included some of his stats.
Link to Article: Modeling Win Probability for a College Basketball Game
Interesting to me that I overlaid his data at halftime, to my UK data and it's pretty close, all things considered.
A couple things to note: for whatever reason he only considered the home team win%, whereas my data is for UK, regardless of where they are playing. Also his data covered 1,782 games over a 3 year period, whereas my data set covers all UK games (i.e. >110 years) for which the halftime score is known (2,782 games), so it seems pretty universal.
Another thing that's interesting, Burke not only looked at halftime margin but the margin at any point during the game. He found that the relationship was largely constant throughout the contest. Only in the final minutes did the slope change appreciably to favor the team leading at the time.
PS, sorry to those who think that this stat has no meaning, but I find it actually pretty fascinating, and something that can provide some pretty useful rules of thumb for the future.
According to the caller: "A team is down by 6 at half they have a 80% chance of losing. It goes up 8% points for 2 points on the scoreboard."
Matt promptly misinterprets this to start counting backwards to down 8, down 10 and down 12 to arrive at a 102% chance of losing.
But I find this to be pretty consistent with my findings, at least in the region where the probability is actively changing (which as I've already noted for UK seems to be -15 to +10). This -15 to +10 range covers 25 points, over which the team's chances of winning for the most part ranges from 0% up to 100%, so as long as it's linear (which it appears to be) the slope is +4%/point. This is exactly what the caller noted.
I dug a little deeper and did find some articles about this from ~2009. The following article by a guy (Brian Burke) who regularly does advanced football stats but looked at basketball included some of his stats.
Link to Article: Modeling Win Probability for a College Basketball Game
Interesting to me that I overlaid his data at halftime, to my UK data and it's pretty close, all things considered.
A couple things to note: for whatever reason he only considered the home team win%, whereas my data is for UK, regardless of where they are playing. Also his data covered 1,782 games over a 3 year period, whereas my data set covers all UK games (i.e. >110 years) for which the halftime score is known (2,782 games), so it seems pretty universal.
Another thing that's interesting, Burke not only looked at halftime margin but the margin at any point during the game. He found that the relationship was largely constant throughout the contest. Only in the final minutes did the slope change appreciably to favor the team leading at the time.
PS, sorry to those who think that this stat has no meaning, but I find it actually pretty fascinating, and something that can provide some pretty useful rules of thumb for the future.
Last edited:
Similar threads
Latest posts
-
-
-
Creepy Crean said we had been Overhelping, but not today- Latest: Dead Cat Bounce
