Kenpom "Luck"

[pj]

It's not though. You can explain away a bad bounce in a game or two but Hurley's record is 9-22 in 5 points or less games since he's been here. We're good at blowing teams out and we're awful in close games.

He's 89-52 at UConn overall so that would make him 80-30 (73% winning percentage) in the non-close games.

Jim Calhoun's career winning percentage was 70%. If Hurley can be as good at winning close games as he is at winning blowouts, he'll be a Hall of Famer.

 

[Echo Hawk]

Interesting topic. Ken Pom is a little over the top with this statistic because you are dealing with human behaviors. Teams get hot then go cold for no apparent reason. Players have injuries that you never know about. They get sick and play. Players choke. Players get hot. And very human coaches make really bad decisions. Bench players come off the bench and have the game of their life. All of these things are opaque, among others, and are not part of any data set that Ken Pom or anyone else can see.

What is interesting to me, using this team as an example, is how a team that looks unbeatable can lose 5 of 6 in a row when the bookies say they should win every one easily. You can't point to any non injury reason. It just happens. You can say that another coach under the same exact conditions would have stopped the bleeding and had a different outcome. Ask anyone in Buffalo this week if they should have won that game.

 

[Chin Diesel]

It's not though. You can explain away a bad bounce in a game or two but Hurley's record is 9-22 in 5 points or less games since he's been here. We're good at blowing teams out and we're awful in close games.

And several/many of those games UConn held leads that slipped away. I'd venture to say out of those 31 games, more of those games were games where UConn led and the lead flipped than games where UConn was down 10-15 points and clawed their way back to a "good" back door loss.

 

[JustbrewitMan]

It's not though. You can explain away a bad bounce in a game or two but Hurley's record is 9-22 in 5 points or less games since he's been here. We're good at blowing teams out and we're awful in close games.

Thank god Coach K retired. And other teams are giddy that Coach B soldiers on...


"...Mike Krzyzewski, Cliff Ellis, and Jim Boeheim have a combined 137 years of coaching experience between them; they have combined to go 73-86 in close games in the last five seasons. Coaching experience had even less relevance to the story than player experience; we should immediately throw this narrative out the window..."

What matters most in winning college basketball’s closest games?

Coin-flip games in sports: a land of contrasts. Coin-flip games in college basketball: probably both luck-based and skill-based.

statsbywill.com

 

[Tenspro2002]

And several/many of those games UConn held leads that slipped away. I'd venture to say out of those 31 games, more of those games were games where UConn led and the lead flipped than games where UConn was down 10-15 points and clawed their way back to a "good" back door loss.

The other side of this, however, is the fact that we had a lot of games that were 3-5 point margins in the second half where we expanded the lead late, turning what would have been "close" wins into blowout wins.

Last month we led at Butler by 4 points midway through the second half. If we had held on to win by 4, would that be a better win, more indicative of Hurley's coaching prowess, than blowing them out by 22? Obviously not. It's unfair to penalize Hurley by winning by "too much."

What would be more informative is to look at games that were within 5 points at, say, 5:00 left in the game, and see what the outcome was. (You could add any games that were outside of that margin where there was a lead change, let's say.)

 

[Jbirdp5]

Interesting topic. Ken Pom is a little over the top with this statistic because you are dealing with human behaviors. Teams get hot then go cold for no apparent reason. Players have injuries that you never know about. They get sick and play. Players choke. Players get hot. And very human coaches make really bad decisions. Bench players come off the bench and have the game of their life. All of these things are opaque, among others, and are not part of any data set that Ken Pom or anyone else can see.

What is interesting to me, using this team as an example, is how a team that looks unbeatable can lose 5 of 6 in a row when the bookies say they should win every one easily. You can't point to any non injury reason. It just happens. You can say that another coach under the same exact conditions would have stopped the bleeding and had a different outcome. Ask anyone in Buffalo this week if they should have won that game.

I think your points about unexpected players getting hot is what the luck feature is trying to account for. When you play a team with a 50% free throw shooter and he goes 9-10 from the line, or the 20% three point shooter going 5-6 from 3 that day. The outliers in teams performances I think are what goes into the luck category both good and bad.

 

[JustbrewitMan]

He's 89-52 at UConn overall so that would make him 80-30 (73% winning percentage) in the non-close games.

Jim Calhoun's career winning percentage was 70%. If Hurley can be as good at winning close games as he is at winning blowouts, he'll be a Hall of Famer.

And as a little more context, Coach Calhoun is in this scatterplot of "Winning % in Games decided by <5 pts" versus Number of Games. (Included all games in KenPom era from 2002-2018).

Even with a sample of Calhoun's best/most-talented teams, his win % in those games was ~52-53%. The best-of-the-best in these close games max out at ~62-64%.

 

[Chin Diesel]

The other side of this, however, is the fact that we had a lot of games that were 3-5 point margins in the second half where we expanded the lead late, turning what would have been "close" wins into blowout wins.

Last month we led at Butler by 4 points midway through the second half. If we had held on to win by 4, would that be a better win, more indicative of Hurley's coaching prowess, than blowing them out by 22? Obviously not. It's unfair to penalize Hurley by winning by "too much."

What would be more informative is to look at games that were within 5 points at, say, 5:00 left in the game, and see what the outcome was. (You could add any games that were outside of that margin where there was a lead change, let's say.)

Not disagreeing. Just going off memory it seems like this season was the first time UConn was delivering the kill shot in the last 5-10 minutes and removing any doubt about who would. Over the past few seasons it seemed like more games went from big leads to close endings than close leads to big endings.

Earlier this year there were some graphics showing at what point in the game UConn scored which essentially sealed the deal. It was the point where UConn's point total at that point was greater than the opponent's final score and there were some crazy big clock numbers on it.

 

[JustbrewitMan]

The other side of this, however, is the fact that we had a lot of games that were 3-5 point margins in the second half where we expanded the lead late, turning what would have been "close" wins into blowout wins.

Last month we led at Butler by 4 points midway through the second half. If we had held on to win by 4, would that be a better win, more indicative of Hurley's coaching prowess, than blowing them out by 22? Obviously not. It's unfair to penalize Hurley by winning by "too much."

What would be more informative is to look at games that were within 5 points at, say, 5:00 left in the game, and see what the outcome was. (You could add any games that were outside of that margin where there was a lead change, let's say.)

The other thing that makes logical sense is that any coach whose total body of work has a fairly high % of "rebuild a program" years should have more close losses than someone who has established a powerhouse program and is maintaining it for many years.

I don't have the time to look into these things at the moment, but it would be a fascinating stats dive to look at "program rebuilding years" versus "stabilized program" years across a sample size of 10 coaches or so.

 

[methodology]

And as a little more context, Coach Calhoun is in this scatterplot of "Winning % in Games decided by <5 pts" versus Number of Games. (Included all games in KenPom era from 2002-2018).

Even with a sample of Calhoun's best/most-talented teams, his win % in those games was ~52-53%. The best-of-the-best in these close games max out at ~62-64%.

And Hurley is at a measly 29% based on the stats superjohn provided which makes Calhoun 82-83% better than DH in those same situations

 

[JustbrewitMan]

And Hurley is at a measly 29% based on the stats superjohn provided which makes Calhoun 82-83% better than DH in those same situations

So by making this post, what you’re also saying is that the team talent level on Hurley’s 2018-present teams is equivalent to the 2002-13 Calhoun years? Since it’s all about the coach?

Sheesh.

 

[Hans Sprungfeld]

You watch the games. Do you ever feel like it's a coin flip when we're in tight games at the end? I sure don't.

Agreed.

Throughout this thread, nobody has offered any argument against this POV, let alone a strong one. I ritually refer to this as one among several clearly identified "growing edges" for Dan Hurley.

Correspondingly, I find little value in rehashing this well-trodden ground. It offers little insight and no assistance beyond its sharp observation.

If the Hippocratic Oath first counsels a physician to, "Do no harm," then I can choose not to shine the spotlight repeatedly on what must already be seen & known by all involved & caring parties.

I strongly root for the coaches, players, and fans to all perform at their focused, calm, prepared, composed, and capable best throughout each game.

My personal efforts, including this comment, have minimal-to-no direct effect on team performance. I don't draw up the game plan or coach to maximize skills. I don't make the shots or dive for the loose balls or do anything on the court.

But I also don't feed undue concentration on what's wrong to the detriment of creating space for a frame of mind that best supports the team in playing to its highest capability and being relaxed & having fun while doing so for the fans' leisure-time entertainment.

I create no distractions, I feed no doubts, and I do no harm.

 

[methodology]

So by making this post, what you’re also saying is that the team talent level on Hurley’s 2018-present teams is equivalent to the 2002-13 Calhoun years? Since it’s all about the coach?

Sheesh.

Big east was also a lot stronger/more talent/better coaches from 2002-2013. That creates more opportunities/chances to lose those close games.

If you are going to bring up the difference in talent on the roster you have to acknowledge the talent of opponents also

 

[Echo Hawk]

I think your points about unexpected players getting hot is what the luck feature is trying to account for. When you play a team with a 50% free throw shooter and he goes 9-10 from the line, or the 20% three point shooter going 5-6 from 3 that day. The outliers in teams performances I think are what goes into the luck category both good and bad.

Maybe. My lifetime observation about luck is that it's usually 50/50 over a long enough timeline. I think most coaches in all sports would say the same thing. Breaks (luck) fall both ways and even out. Trying to rank team's luck is an impossible stat.

Pom is really an opinion writer making a statistical case in real time which is always unverifiable. That's the beauty of opinion. You never have to verify. People forget. But it's a living. Now if he could say this time of the year statistically that a 7th seed in the NCAA tournament would play an 8th seed for the championship and the 7th seed would win, now you have something.

 

[JustbrewitMan]

Big east was also a lot stronger/more talent/better coaches from 2002-2013. That creates more opportunities/chances to lose those close games.

If you are going to bring up the difference in talent on the roster you have to acknowledge the talent of opponents also

I agree with this. To a degree. A conference with 1-2 powerhouses and 8-9 mediocre teams could actually have just as many, if not more "close game opportunities".

Would love to be able to dive into this "close game outcome" metric while sorting for things like:

• Games where KenPom rankings were +/- 20 between the two teams
• Games where the betting line was < "x" points or > "y" points, etc.
• Games where the margin was <5 pts with "x" minutes left, etc.

Because then you could account for the end-game a bit better and get at what we're all trying to quantify ("Is Hurley a good/bad endgame coach?" "Are his teams clutch or chokers?")

There's lots of stats massaging you can do to support your narrative. For example, I could make a perfectly valid argument that "games decided by <=7 pts" are not really much different than "games decided by <=5 pts" There are garbage pts scored that make nailbiter games into 6-7 pt MOVs all the darn time.

I also could say that "it's completely unfair to look at ANY coach's data from year 1 of a major rebuild", and "the COVID19 year was messed up. There was no homecrowd advatange, teams' rosters were hit-or-miss based on who was in COVID protocols, etc."

With those parameters in mind, Hurley's/Hurley's teams' records in "close games" has been 11-10 (52.3%). Very respectable & really "normal".

 

[navery12]

Maybe. My lifetime observation about luck is that it's usually 50/50 over a long enough timeline. I think most coaches in all sports would say the same thing. Breaks (luck) fall both ways and even out. Trying to rank team's luck is an impossible stat.

Pom is really an opinion writer making a statistical case in real time which is always unverifiable. That's the beauty of opinion. You never have to verify. People forget. But it's a living. Now if he could say this time of the year statistically that a 7th seed in the NCAA tournament would play an 8th seed for the championship and the 7th seed would win, now you have something.

He's not an opinion writer, he's not even a writer. These aren't even opinions he's posting. He's not sitting down ranking teams 1-353 by how lucky he thinks they are

 

[FfldCntyFan]

I have a strong suspicion that with a veteran, skilled point guard (junior year Kemba, senior year Shabazz) a team that is relatively equal to their opponent in all other phases will appear to be exceptionally lucky by these metrics.

Down the stretch in a tight game their is nothing better than an experienced leader who can create his own shot, make it at a rate above the mean and also set up teammates to score. This was clearly was set us apart from the field in 2011 & 2014 and is the biggest piece that we've been missing in our subsequent NCAA tournament games.

 

[ClifSpliffy]

I agree with this. To a degree. A conference with 1-2 powerhouses and 8-9 mediocre teams could actually have just as many, if not more "close game opportunities".

Would love to be able to dive into this "close game outcome" metric while sorting for things like:

• Games where KenPom rankings were +/- 20 between the two teams
• Games where the betting line was < "x" points or > "y" points, etc.
• Games where the margin was <5 pts with "x" minutes left, etc.

Because then you could account for the end-game a bit better and get at what we're all trying to quantify ("Is Hurley a good/bad endgame coach?" "Are his teams clutch or chokers?")

There's lots of stats massaging you can do to support your narrative. For example, I could make a perfectly valid argument that "games decided by <=7 pts" are not really much different than "games decided by <=5 pts" There are garbage pts scored that make nailbiter games into 6-7 pt MOVs all the darn time.

I also could say that "it's completely unfair to look at ANY coach's data from year 1 of a major rebuild", and "the COVID19 year was messed up. There was no homecrowd advatange, teams' rosters were hit-or-miss based on who was in COVID protocols, etc."

With those parameters in mind, Hurley's/Hurley's teams' records in "close games" has been 11-10 (52.3%). Very respectable & really "normal".

lots of beagling for that 9-22 record.
now, aboot this o'fer on ncaa tourney wins in the years since his arrival?

and, oh, 7 points is three possessions while 5 is two, so there's that.

 

[HuskyHawk]

Sometimes luck really is luck, like the rebound at the end of the Seton Hall game. Some teams make their own luck. Ultimately, guards have the ball, if that guard can score or get fouled trying, it's going to help you close out games. If we want our luck to increase, we can't try to run designed plays that rely on (a) feeding the post or (b) freeing up a jump shooter off of a screen. It's going to be a shot coming off the dribble. At the moment, that's not Hawk's strength. He's no Ben Gordon or Kemba Walker. We need Newton to take the reins in those situations.

Hurley mentioned a week or so ago that the offense is too scripted, and I think that's a big part of it. If the "play" isn't there we freeze up and don't know what to do.

 

[Tenspro2002]

Sometimes luck really is luck, like the rebound at the end of the Seton Hall game.

If Seton Hall's 3 attempt was closer to the mark, hit squarely off the rim and bounced up towards the corner, instead of a near-airball straight down to one of their guys, Hurley is called clutch for winning a close game.

 

[HeGotGame]

If Seton Hall's 3 attempt was closer to the mark, hit squarely off the rim and bounced up towards the corner, instead of a near-airball straight down to one of their guys, Hurley is called clutch for winning a close game.

I know based on all the posts after the game and the hilarity of them blaming DH for the loss, he was actually at home in his living room when that shot went up.