Pomeroy's Win Probability Graph has given them a 90 percent chance of winning the game at some point in the contest. That's almost a quarter (7 of 30) of MU's losses in the last 2-plus years.
Here's the list:
09-10 Florida State
09-10 North Carolina State
09-10 DePaul
09-10 Notre Dame
09-10 Washington
10-11 Louisville
11-12 Georgetown (MU 90 percent to win last night after Mayo's three with 13:10 to play that put them up 56-39).
Pomeroy's graphs only go back to the 09-10 season.
I have no idea if this is high compared to other teams, but I already sent an e-mail to him to see if there's an easy way to research this or if he'd consider doing a blogpost on the topic. Doing the research myself for all 345 D1 teams would take a rather obscene amount of time I would imagine.
In fairness, MU has won two such games in that timespan, last year at UConn and this year at WI-Madison.
Once we establish if this total is out of the norm compared to other D1 schools (I'm guessing it is), then we might be able to take a look at some causes and possible solutions.
Eye .. thank you for the research. I've been trying to figure an objective choke standard .. I think you've found it.
I also love this information. Thanks for sharing. I'm considering running through and doing it for every BE team. That might be more manageable.
Statsheet.com has "lead safe %age" .. and goes back about 10 seasons. It must use a different formula ..as, for example, the Great Choke of 2011 @ UL hit only 60%.
http://statsheet.com/mcb/games/2011/01/15/marquette-70-louisville-71/flow
I checked 2008-9, didn't see any obvious choke-jobs.
Love the idea, Henry. One wonders in the BE, who has the most Chokes .. who has the most ... Anti-Chokes (??) Vanquishes?
What do you call the opposite of "Pulling a Marquette"?
Not sure a 90% at any point in the game can be used to define 'chokes'.
Ken's number is based on the specific teams playing and the chart for the current year will continue to fluctuate throughout the season until completion. There are many instances when a team starts off with a greater than 90% win probability.
If a team is at a 91% win probability at the beginning of the game, but the big underdog comes in and starts off with an 8-0, and goes on to win the game.. that looks more like a 'crap the bed' game than a 'choke'. Maybe semantics, but if you're going to settle on an objective 'choke' criteria, maybe you need to look at some point in the second half vs. at any point in the game.
As for statsheet, at one point I believe they were using a calculation that neither considered the specific teams playing nor the home team.
These fades are becoming a trademark of Buzz' teams
Absolutely.
A friendly amendment .. 90% win% in the 2nd half? .. Or with 10, 5 mins to go?
I think I'd go with .. 90% win, any time under 10 to go. Maybe 80% under 5, as well.
Quote from: Eye on January 05, 2012, 03:30:26 AM
11-12 Georgetown (MU 90 percent to win last night after Mayo's three with 13:10 to play that put them up 56-39).
I find it very surprising that our win percentage likelihood was only 90% at that point. It seems that teams with a 17 point lead and 13 minutes remaining would go on to lose much less than 1 in 10 times. That makes last night's loss seem to be not as bad as I thought.
Quote from: mu_hilltopper on January 05, 2012, 08:56:43 AM
Absolutely.
A friendly amendment .. 90% win% in the 2nd half? .. Or with 10, 5 mins to go?
I think I'd go with .. 90% win, any time under 10 to go. Maybe 80% under 5, as well.
That seems like a good defining line.
Quote from: mu_hilltopper on January 05, 2012, 08:56:43 AM
I think I'd go with .. 90% win, any time under 10 to go. Maybe 80% under 5, as well.
I'm with it.
Can anyone recall a game where MU was at a 90% chance of loss in the final 10, but pulled it out?
Does KenPom give access to his data in a downloadable format? It wouldn't take too long to write a quick script to answer the 'choke' question. In fact, that sounds like way more fun than doing research today...
EDIT: I'd even pay my $20 if I knew I could get it.
MU can build leads really well, but can't seem to hold 'em. Would like to see the reverse as well, where MU clawed its way back to win from deep. One thing I also saw in Statsheet is the coaches' record in close game (6 or less), Buzz and Jay Wright are about .500, where the masters like Calhoun and Boeheim are over .600.
Talent, defensive stop scheme, end game situationals and coaching experience win close games....we still forget that Buzz is the youngest coach on his staff, and second youngest in the BE. He has a lot of learning yet to do, which he openly admits (unlike the previous). Last night, MU had a great starting game plan...even after half, but Thompson changed up on the switches on defense and it was curtains. Buzz had no answer. And Gtwon continued to shoot lights out against our D and I didn't see any switch to zone or a box to stop Clark. Rinse and repeat.
Good stuff....thanks
http://statsheet.com/mcb/coaches/compare?add=jim-calhoun&c1=buzz-williams&c2=jay-wright&c3=jim-boeheim
If it's 2nd-half only, that actually doesn't help the argument from an MU perspective. Only game of those 9 where the 90 percent threshold was reached in the first half was this year's WI-Madison game. So take that out and it's 1-7 for MU in games where they've had a 90 percent chance of winning/losing at some point in the 2nd half and the opposite has occurred over the past 2-plus years.
Pomeroy had UConn over 90 percent in last year's game at UConn at the peak of the UConn extended run to begin the 2nd half. MU up 34-23 at half, down 49-41 with 9 and change, so at the peak of UConn's 26-7 extended run over 10-plus minutes to begin the half, he had UConn at over 90 percent. Only one he has MU winning in that type of spot in the last 2-plus years.
I'll ask him if the info is downloadable when/if he replies.
Quote from: mu_hilltopper on January 05, 2012, 08:56:43 AM
Absolutely.
A friendly amendment .. 90% win% in the 2nd half? .. Or with 10, 5 mins to go?
I think I'd go with .. 90% win, any time under 10 to go. Maybe 80% under 5, as well.
Games this criteria would exclude:
Florida State - 95% @ 15 min to go
NC State - 94% @ 20 min to go
Washington (NCAA) - 93% @ 14 min to go
LAST NIGHT - 90% @ 13 min to go
I think it should be 90% with 15 min to go. Would also accept the entire 2H just to include games like NC State.
Quote from: mu_hilltopper on January 05, 2012, 09:48:56 AM
Can anyone recall a game where MU was at a 90% chance of loss in the final 10, but pulled it out?
One from last year. UConn 7% with 10 min to go. Expanding to the final 15 min... USF last year (10% with 14 to go)
Going to try crowd-sourcing this one to see how it works.
I created a google docs version of a spreadsheet that I went through. Although the initial question is just restricted to 90% and somewhere between 10-20 min, I believe there's a ton of additional information that could be obtained by looking at the starting/max/min win percentages.
Open access is allowed to edit the doc, and the Marquette template can be copied over. I typically just copy the information from the Pomeroy Game Plan and then put it into a spreadsheet. Anyone interested in helping to fill out the info for the entire league?
https://docs.google.com/spreadsheet/ccc?key=0AqUGYV4K8N_ydGVSeExLLW9FQ3B5WnBRaWp2ekhWQUE
Henry had a couple good questions, though ..
FSU two years ago count as a choke? only 87% with 10 min to go .. but 95% with 15 min to go.
Washington NCAA tourney .. 93% at fourteen min to go .. choke?
While neither hit the 90% at 10 rule .. I imagine if I looked back at MUScoop history, everyone was using the "C-word" after those games.
Truth is, it's a "Choke Continuum" .. 90% at 10 is bad .. 93% at 14 is just a little less bad, but still a choke.
May need to alter the formula. Maybe it should be 80% in the final 15?
Wow, looking at that spreadsheet, Henry .. "Marquette, where we have enough data points to analyze our chokes."
Quote from: mu_hilltopper on January 05, 2012, 09:48:56 AMCan anyone recall a game where MU was at a 90% chance of loss in the final 10, but pulled it out?
Here are a few notables:
USF last year: 90% under 15 to play, 80% under 9
UConn last year: 92% under 10, 80% under 1
Bucknell last year: 73% just over 10
St John's 2 years ago (OT): 80% just over 10
Umm .. that UConn one, 80% under 1 to go .. that exposes a choke factor problem. I mean, UConn was only ahead 2 points. Hard to say UConn choked a 2 point lead away, to lose in OT.
Are we missing something .. maybe there needs to be a double-digit+ lead within that 80-90% win probability to really count.
Quote from: elephantraker on January 05, 2012, 08:56:15 AM
These fades are becoming a trademark of Buzz' teams
This could be an interesting research project too.
Big East standings the past 4 seasons if the games ended at halftime, with 10 minutes to go, and with 5 minutes to go.
Quote from: mu_hilltopper on January 05, 2012, 10:58:06 AM
Wow, looking at that spreadsheet, Henry .. "Marquette, where we have enough data points to analyze our chokes."
I think this says more than any of the data.
Ouch.
I'm hoping that Buzz is just getting outcoached. He's still young with little experience himself or on his bench. Hopefully these are the only growing pains we have to experience with Buzz.
Anyone else somewhat think this?
Must say, the data is most interesting. The question becomes why? Is the team tired because it is over-coached? Is it because Coach Williams has proven himself not to be a defensive coach - as was Al McGuire? Or, is it simply the coach who is "choking?" This Warrior fan is a big supporter of Coach Williams, but he needs to calm down a bit a become more consistent.
Quote from: mu_hilltopper on January 05, 2012, 11:42:35 AMUmm .. that UConn one, 80% under 1 to go .. that exposes a choke factor problem. I mean, UConn was only ahead 2 points. Hard to say UConn choked a 2 point lead away, to lose in OT.
Are we missing something .. maybe there needs to be a double-digit+ lead within that 80-90% win probability to really count.
With that one, I'd definitely say the 92% under 10 was more indicative of their choke.
Quote from: UticaBusBarn on January 05, 2012, 12:28:27 PM
Must say, the data is most interesting. The question becomes why? Is the team tired because it is over-coached? Is it because Coach Williams has proven himself not to be a defensive coach - as was Al McGuire? Or, is it simply the coach who is "choking?" This Warrior fan is a big supporter of Coach Williams, but he needs to calm down a bit a become more consistent.
I'm sure there are statistics for this, but when Buzz's teams go bad, every facet of their game seems to go bad. Shooting will go hot and cold, but defense, for example, is something that should be able to be maintained for the most part.
When they start blowing leads like last night, not only do they stop scoring, they also seem to forget how to play defense, rebound, pass and dribble. I don't know if it's a lack of mental toughness, maturity, or something else, but when things start going bad their whole game falls apart.
Quote from: MUeagle05 on January 05, 2012, 02:34:12 PM
I don't know if it's a lack of mental toughness, maturity, or something else, but when things start going bad their whole game falls apart.
Part of that is that we generate a lot of offense from our defense in transition. When we're not getting stops and steals we aren't getting easy baskets in transition, which is a staple of ours.
Quote from: brewcity77 on January 05, 2012, 11:10:19 AM
Here are a few notables:
USF last year: 90% under 15 to play, 80% under 9
UConn last year: 92% under 10, 80% under 1
Bucknell last year: 73% just over 10
St John's 2 years ago (OT): 80% just over 10
Problem is, other than UCONN, those teams weren't that good. And as 'Topper pointed out, although the likelihood of victory was low against UCONN, the deficit was not.
However, MU has come close a couple of times in Buzz's tenure, both against 'Nova.
- The 09-10 team @Nova - Down 22 with 13 to go, got it down to 2 with 30 secs to go and had the ball. Ended up losing by 2.
- The 08-09 team, against Nova, Big East Tournament (The McNeal game). Down 16 at half, rally to take a lead with a minute to play... Lose by 1.
Quote from: marqptm on January 05, 2012, 12:22:36 PM
I'm hoping that Buzz is just getting outcoached. He's still young with little experience himself or on his bench. Hopefully these are the only growing pains we have to experience with Buzz.
Anyone else somewhat think this?
I don't know what to think about Buzz's teams on the road. Interesting fact... In Buzz's tenure, MU has had only one true road win by more than 10 points. It was the 08-09 team vs. DePaul (a truly terrible team).
It seems no matter the opponent, win or lose, the games are always extremely close.
For the most part, his teams always seem to hang with much better squads. Similar to yesterday, against teams that are higher rank, have more talent, play in tough venues, MU plays well for the most of game, but comes up just short.
The easy answer is that it is simply a talent difference... MU plays hard and outworks the other team, but in the end, the opponent picks up the intensity and talent wins out. Makes sense, until you examine how MU plays on the road against teams where MU is the more talented squad.
Loses against teams like LSU, DePaul, USF, or even extremely close wins against UWM, St. Johns, USF (last year) doesn't jive with a team that takes UConn, West Virginia, Georgetown, Louisville, etc. to the limit.
Quote from: RJax55 on January 05, 2012, 03:10:01 PMProblem is, other than UCONN, those teams weren't that good. And as 'Topper pointed out, although the likelihood of victory was low against UCONN, the deficit was not.
I'm not saying they were all as bad, but someone asked for examples and I gave them. As far as UConn...I think it's a lot tougher to come back on the road than it is at home. Granted, 8 points with 9:32 to play isn't the same as 17 with 13:10, but GTown had more than 3 more minutes for their comeback and a crowd that was behind them.
I think what's sometimes lost on people is that basketball is a game of runs. Maybe that's why I never felt secure last night against Georgetown. Because despite the lead, we were really only two runs up on them. They go on a 10-2 in a 3:40 minute span (not a monstrous accomplishment), suddenly the margin is single digits with 10 minutes to play. We play even for 3 minutes, then they go on a 12-4 run over 4 minutes, now it's a 1-point game with 3 minutes to play.
Here's my issue with everyone teetering on the ledge. What if yesterday goes like this...
- At 3:30 of the first half, after the Mayo 3, Georgetown goes on an 12-4 run to close the half (the same run they went on from 3:30 to 0:00 of the second) Halftime: MARQ 39 - GT 37
- Marquette extends their lead over the first 7 minutes of the second half 13:10 to play: MARQ 52 - GT 47
- Georgetown makes a 11-4 run from 13:10 to 9:38 to take the lead 9:38 to play: MARQ 56 - GT 58
- Despite a quick Crowder basket, GT runs again and extends their lead 3:56 to play: MARQ 62 - GT 69
- MU storms back with an 8-2 run to close the gap 0:21 to play MARQ 70 - GT 71
- Blue fouls Sims to send him to the line, he makes them both 0:21 to play MARQ 70 - GT 73
- DJO misses a good look with 4 seconds to play, Mayo misses another shot at the buzzer Final Score: MARQ 70 - GT 73
.
All I did was move the final 3:30 of each half around. Change the location of two runs and sure, Marquette blows a lead, but it's a 2-point halftime lead they blow and a max 5-point second half lead. Instead, Marquette is the team storming back late and comes up one three-ball at the buzzer short of taking the #9 team in the country to overtime after a back-and-forth game.
It's a game of runs. Unfortunately, they strung a few runs together at the end. But this isn't the disastrous loss everyone is making it out to be. It's a 3-point loss to a top-ten team on the road. Yeah, it sucks because a lot of people thought we had it, but we didn't. Sack up and move on, let's hope to get Syracuse on Saturday. And if not, there's another 15 Big East games and we'll be the favorite in probably at least 10 of them. Win the ones we should, pull an upset or two, and we'll be just fine.
I think i can keep this simple. Every game mu won, at some point they were a 90% fav.
22 + 22 + 12 = 56 games they won + 7 games they had 90% but lost = 63 games total where their chances to win were 90%.
56/63 = 89%
So when they have a 90% chance of winning, they win 89% of the time.
Wow, what colossal choke artists! Fire Buzz!
Quote from: UticaBusBarn on January 05, 2012, 12:28:27 PM
This Warrior fan is a big supporter of Coach Williams, but he needs to calm down a bit a become more consistent.
Agreed, and that's actually a good thought I hadn't considered yet. Was covering a baseball game a few years back and heard a coach tell his team that execution will win out over rah-rah every time. There's a time for rah-rah; there's a time when execution wins out. The rah-rah is fun and nice, but maybe there's a little bit too much of it. Maybe you don't totally have the player's proverbial ear when it's really needed in big spots; when a good team (or less than a good one in the case of DePaul on the road two years ago) makes a big run at you away from home.
Once again someone (Pomeroy said he may take a shot at it at some point, I may be able to get to it Monday using Pomeroy's numbers; would only do the BE) needs to figure out if an 89 percent close rate in games where you're 90 percent at some point to win the game is normal or abnormal. What you're not taking it account RD is that your total of wins includes 18 cowboys in that timespan where MU started at higher than 95 percent to win the game and never dropped below that. Take those out, and it's down to 38 of 45, or much closer to 80 percent.