Blown leads

[SparzWizard]

Saw this on Reddit:  After the Nuggets effectively clinched the 1 seed last year, they went 8-10 with a -1.4 NETRTG.  Let's relax a bit and hope that the team is healthy going into the playoffs.

But that's the Nuggets. With the Celtics, there seems to be a running history of it in regards to blown leads.

[libermaniac]

To blow a 10+ point lead, you have to be up by 10+ points. And my guess is the C's have had more 10+ point leads than anybody.

Excellent point.  I’m embarrassed I didn’t think of that. TP

[Kernewek]

We blow more leads because we shoot the most 3s. 1-18 from 3 yesterday after starting 10-20. If they shoot a paltry 5-18 we aren’t having this conversation.

My guess is that from an expected points per possession perspective, shooting lots of good 3s maximizes that. But there is probably more variance in the outcome - so you are going to have stinkers.

I think this is why most Cs fans criticize Joe when we blow leads because when you are up big you should change your approach to minimize that variance - eg go away from so many 3s. Drive and get to the line. But they don’t seem to change their system.

I've posted this before, but it bears repeating that variance works oppositely to this. It's a reasonable mistake to make, because lots of basketball writers and sports-talk types will use 'variance' as a synonym for 'low-percentage shot, when actually it means the exact opposite.

For sake of simplicity:
Suppose you have a team that shoots midrange shots on every possession and hits them 50% of the time.
And you have another team that shoots threes on every possession and hits them 33% of the time.

Even though both teams are scoring the same points per possession, the team shooting the midrange shots has a higher variance, because the variance we're actually measuring is the expected points per possession compared to the total shot attempts. In other words, you don't get more points for a higher FG%, but you do get more points for shooting a three.

I used this example last time and it's still pretty good: check out the game we played against the Wizards at the beginning of February - we absolutely outplayed them on both ends of the floor, but it was 'only' a four-point game because they shot (and hit) way more threes.

[hwangjini_1]

SoSH has been posting about this topic as well. Here are a few interesting point/info.


Dan Greenberg
@StoolGreenie
·
Follow
Blown leads this season:

BOS: 14
DEN: 18
OKC: 18
MIL: 19
MIN: 22

Blown 4th quarter leads this season:

BOS: 8
DEN: 9
MIL: 9
OKC: 6
MIN: 13

and........

Does Losing Lead to Winning? An Analysis of the Winning Paradox in Sports


https://www.youtube.com/watch?v=aqk9MPfjKSw

[scaryjerry]

Saw this on Reddit:  After the Nuggets effectively clinched the 1 seed last year, they went 8-10 with a -1.4 NETRTG.  Let's relax a bit and hope that the team is healthy going into the playoffs.

I see your point in bringing this up and it’s plenty valid but unless Tatum takes a massive leap immediately we simply don’t have a player of jokics caliber.

Wouldn't expect any other kind of response from you. Celtics are just a bunch of g leaguers that simply have no chance in the playoffs.

I know you are counting the days down to tell everyone how right you were all season.

Again, I think the Celtics are overwhelming favorites. Championship or absolute bust. I think there’s 3 to 4 teams in the entire league that can beat them in a 7 game series and these blown leads are part of the reason why. I guess in your mind that is me calling them “g leaguers”

[Celtics2021]

Saw this on Reddit:  After the Nuggets effectively clinched the 1 seed last year, they went 8-10 with a -1.4 NETRTG.  Let's relax a bit and hope that the team is healthy going into the playoffs.

But that's the Nuggets. With the Celtics, there seems to be a running history of it in regards to blown leads.

Probably because we don’t watch the nuggets every night.

[CptZoogs]

We blow more leads because we shoot the most 3s. 1-18 from 3 yesterday after starting 10-20. If they shoot a paltry 5-18 we aren’t having this conversation.

My guess is that from an expected points per possession perspective, shooting lots of good 3s maximizes that. But there is probably more variance in the outcome - so you are going to have stinkers.

I think this is why most Cs fans criticize Joe when we blow leads because when you are up big you should change your approach to minimize that variance - eg go away from so many 3s. Drive and get to the line. But they don’t seem to change their system.

I've posted this before, but it bears repeating that variance works oppositely to this. It's a reasonable mistake to make, because lots of basketball writers and sports-talk types will use 'variance' as a synonym for 'low-percentage shot, when actually it means the exact opposite.

For sake of simplicity:
Suppose you have a team that shoots midrange shots on every possession and hits them 50% of the time.
And you have another team that shoots threes on every possession and hits them 33% of the time.

Even though both teams are scoring the same points per possession, the team shooting the midrange shots has a higher variance, because the variance we're actually measuring is the expected points per possession compared to the total shot attempts. In other words, you don't get more points for a higher FG%, but you do get more points for shooting a three.

I used this example last time and it's still pretty good: check out the game we played against the Wizards at the beginning of February - we absolutely outplayed them on both ends of the floor, but it was 'only' a four-point game because they shot (and hit) way more threes.

The issue leading to the confusion is that people don’t specify the random variable being discussed.  Variance is the expected squared distance a random variable falls from its mean.  We usually convert this to a standard deviation by taking the square root so that the units are more interpretable.  If we are talking about points per shot, the 33% 3-point shot has a variance of 2.  The 50% 2-point shot has a variance of 1.  However, if the random variable of interest is simply the binary make (1) or miss (0) of a given shot, then we have a basic Bernoulli RV.  In this case, the 2-point shot has the higher variance at 1/4 compared to the 3-point shot at 2/9.

[Donoghus]

Interesting stuff from Grande

Sean Grande
@SeanGrandePBP
While the '24 compares favorably to '08  in most places, the dramatic changes in the game make a lot of it moot. The '08 record is better with a 14-point lead, they were 49-2, '24 team is 47-4.  But leads are FAR more fluid now.  When down 13+, '08 team was 2-5, '24 team is 4-6.

[Kernewek]

We blow more leads because we shoot the most 3s. 1-18 from 3 yesterday after starting 10-20. If they shoot a paltry 5-18 we aren’t having this conversation.

My guess is that from an expected points per possession perspective, shooting lots of good 3s maximizes that. But there is probably more variance in the outcome - so you are going to have stinkers.

I think this is why most Cs fans criticize Joe when we blow leads because when you are up big you should change your approach to minimize that variance - eg go away from so many 3s. Drive and get to the line. But they don’t seem to change their system.

I've posted this before, but it bears repeating that variance works oppositely to this. It's a reasonable mistake to make, because lots of basketball writers and sports-talk types will use 'variance' as a synonym for 'low-percentage shot, when actually it means the exact opposite.

For sake of simplicity:
Suppose you have a team that shoots midrange shots on every possession and hits them 50% of the time.
And you have another team that shoots threes on every possession and hits them 33% of the time.

Even though both teams are scoring the same points per possession, the team shooting the midrange shots has a higher variance, because the variance we're actually measuring is the expected points per possession compared to the total shot attempts. In other words, you don't get more points for a higher FG%, but you do get more points for shooting a three.

I used this example last time and it's still pretty good: check out the game we played against the Wizards at the beginning of February - we absolutely outplayed them on both ends of the floor, but it was 'only' a four-point game because they shot (and hit) way more threes.

The issue leading to the confusion is that people don’t specify the random variable being discussed.  Variance is the expected squared distance a random variable falls from its mean.  We usually convert this to a standard deviation by taking the square root so that the units are more interpretable.  If we are talking about points per shot, the 33% 3-point shot has a variance of 2.  The 50% 2-point shot has a variance of 1.  However, if the random variable of interest is simply the binary make (1) or miss (0) of a given shot, then we have a basic Bernoulli RV.  In this case, the 2-point shot has the higher variance at 1/4 compared to the 3-point shot at 2/9.

Yeah very good point, TP for the clarification.

[CelticSooner]

Three point variance will lead to blown leads. C’s couldn’t miss in the first quarter last night and cooled off dramatically each quarter afterwards. The biggest problem with this team is they don’t seem to be able to right the ship and close teams out once they lose momentum. Once the game gets close in the 4th here comes the horrible pace (C’s are last in clutch pace btw) that has doomed them for years. It begins and ends with the person holding the ball on that 8 second call.

The jury is out, just keep it close because they can be had late.

[tenn_smoothie]

Interesting comments by Brian Windhorst on the NBA page of the ESPN website today ......

He talks about every point the fans here discuss in frustration on a daily basis, including the awful lack of execution on important, late-game possessions, Mazzulla's refusal to call a timeout to set up something other than an iso-dribble fall-back jumper from Tatum or Brown, the constant habit of playing fast and aggressive in the first half of games only to slow the offense to a walk in the 4th quarter. He talks about the overwhelming talent this team has and how it is not utilized to win games. Very interesting clip he shows of the ridiculous possession we had at the end of the Atlanta game when Brown stands on the right wing at the 3-point line dribbling the clock away with zero movement from the other players, all while Joe stands clueless on the sideline with two timeouts in his pocket.

Seems that certain members of Celtics Strong are not the only ones who have serious concerns about the ability of this team and coach to succeed in the playoffs. I just keep thinking about the great things a quality coach could do with this talented team.

Note: Also submitted on the "Fire Joe" thread when I couldn't decide where to post it. Hope I'm not out of line, but it was really interesting to hear the opinion of someone outside of the Celtics fan base pointing out the exact concerns some of us have.

[Kernewek]

If anyone wants to find the comments on ESPN, by the way, it's a video called "What 'infuriates' Brian Windhorst about Celtics' late-game play".

[Who]

This nugget was in a recent ESPN piece republished today:

The frequency of 10-point and 15-point comebacks has increased as well. In 1997-98, teams that fell behind by double digits had an .181 winning percentage. That climbed to .250 a season ago and is at .229 this season, meaning nearly one in every four games in which a team takes a double-digit lead ends with the other team winning.

The inverse means that teams with 10-point leads have a .771 winning percentage in such games this year.  I’ve got to imagine the Celtics are close to that, if not above it.  My guess is, at least this year, we’ve blown a lot of 10-point leads because we’ve gotten a lot of 10-point leads, and the league trends make comebacks, and thus blown leads, more likely.

EDIT:  Did the research — the Celtics are 51-8 in games in which they’ve had a 10-point lead at some point.  That’s an .864 winning percentage in such games, which is actually very, very good.  No other team is likely close to our number of games with 10-point leads, so it stinks when we lose them, and this year it’s more than half of our losses, but it really does seem that we’ve lost as many as we’ve had because we’ve just had more big leads to begin with.  New Orleans, first on the list of big leads blown, is second to the Celtics in number of 10-point wins (30 to our 36), giving another datapoint to having a lot of large leads means you’ll blow more.

That is cool. Well done finding those stats. Thank you.

[radiohead]

https://twitter.com/TA1297/status/1772621938579357969

This is interesting.

[celticinorlando]

Bucks blew a 19 point lead at home vs a Lebron less lakers and LA forces OT.

Don’t worry, Scary Jerry will just tell us that this is character building for the bucks and the fact they blew a big lead doesn’t apply to them.