12/9: MBB vs Alabama A&M

[slugboy]

LOL. I doubt ANY of our problems are about tiredness.

They’re running hard. Lack of effort doesn’t seem to be an issue

 

[Root4GT]

They’re running hard. Lack of effort doesn’t seem to be an issue

It does matter on shooting if you have neem playing harder/faster than you previously did. That will settle out over time as the players become accustomed to the standard.

 

[Connell62]

It does matter on shooting if you have neem playing harder/faster than you previously did. That will settle out over time as the players become accustomed to the standard.

Anyone that has ever played knows that when your legs go, your shot goes.

When you exert energy in other phases of the game, that you may not have had too in the past, it effects your shots.

 

[Steverc]

This is definitely modern basketball strategy, but I’m with alagold on this one. It seems misguided when taken to the extreme. Watching guys give up shots from the low post area gets tiring, especially for a team that only shoots 29% from 3.

44% from 2 would beat 29% from 3, and 44% should be doable from the low post.

Actually, 29% from 3 is the same as 44% from 2.

 

[gte447f]

Actually, 29% from 3 is the same as 44% from 2.

No, actually, 29% from 3 is the same as 43.5% from 2, so 44% from 2 would beat 29% from 3.

 

[cpf2001]

If we want to get into the probability/statistics-weeds, 29% from 3 is higher variance than 43.5% from 2 so you'll expect to be less consistent and have more games where you score a lot more and more where you score a lot less.

 

[lv20gt]

If we want to get into the probability/statistics-weeds, 29% from 3 is higher variance than 43.5% from 2 so you'll expect to be less consistent and have more games where you score a lot more and more where you score a lot less.

That isn't how variance works. You need more than just the mean to determine that.

Taking each game as an individual data point (which yes has it's own issues but it's the quickest way to get a decent picture) we get the following for our team this year

for 2 point our fg% average is 47.15% with a standard deviation of 10.4.
for 3 point our fg% average is 29.325% with a standard deviation of 6.88.

So for our team we have significantly less variance in our 3point shooting than our 2 point shooting. Now you could go further and account for 3s being worth 50% more than twos which would make a swing in 3% account for a bigger swing in points than 2s, but we also shoot about 50% more twos than we do threes which would make the 2fg% swing have a bigger swing in baskets made meaning a bigger swing in points. Without actually looking at it more closely I'm tempted to just stick with the above and say our variance on 3 point shooting is less than that of our 2 fg% right now.

 

[forensicbuzz]

So I went back to watch theA&m game.I don't quite get why we are passing up 5-10 ft shots by then passing to a guy outside 3 pt line when we don't shoot 3s well? Especially vs a weak team like AM.I saw one REAL pass inside in the first 10 mins of that game.We've got to develop an inside game more than just Off rebs.

The goal is never to beat Alabama A&M. That's a given. the goal is to work on things that you want to do against Duke and North Carolina.

 

[orientalnc]

The goal is never to beat Alabama A&M. That's a given. the goal is to work on things that you want to do against Duke and North Carolina.

Is sleep optional in Chicago? A post on the Swarm at 3:20 AM!!?!!

But, I agree with the post.

 

[AUFC]

Speaking of testing rotations and sets, did we at one point have Sturdy, Abram, and George on the court at one time or am I hallucinating that? I know we definitely went with some 2 PG rotations at least towards the end of the first half.

 

[orientalnc]

Speaking of testing rotations and sets, did we at one point have Sturdy, Abram, and George on the court at one time or am I hallucinating that? I know we definitely went with some 2 PG rotations at least towards the end of the first half.

I do not remember George and Kyle on the court together, but both shared time with Abram late. Thanks for reminding of that.

 

[forensicbuzz]

Is sleep optional in Chicago? A post on the Swarm at 3:20 AM!!?!!

But, I agree with the post.

I'm in Seattle today.

 

[cpf2001]

That isn't how variance works. You need more than just the mean to determine that.

Taking each game as an individual data point (which yes has it's own issues but it's the quickest way to get a decent picture) we get the following for our team this year

for 2 point our fg% average is 47.15% with a standard deviation of 10.4.
for 3 point our fg% average is 29.325% with a standard deviation of 6.88.

So for our team we have significantly less variance in our 3point shooting than our 2 point shooting. Now you could go further and account for 3s being worth 50% more than twos which would make a swing in 3% account for a bigger swing in points than 2s, but we also shoot about 50% more twos than we do threes which would make the 2fg% swing have a bigger swing in baskets made meaning a bigger swing in points. Without actually looking at it more closely I'm tempted to just stick with the above and say our variance on 3 point shooting is less than that of our 2 fg% right now.

I think you're right that I was sloppy/unclear - at shot-rate, variance is lower for 3's since the `p` is further from 0.5 based on the binomial variance that's based on n*p*(1-p).

Studies at the NBA level have been done like here for the effect on scoring: https://www.sloansportsconference.com/research-papers/live-by-the-three-die-by-the-three - not just 'made bucket' variance.

The formatting of the math doesn't work to quote it but it's https://assets-global.website-files.com/5f1af76ed86d6771ad48324b/654bf324570bf68832c728b1_SSAC13 - Live by the Three, Die by the Three The Price of Risk in the NBA.pdf in the footnote that includes "Consider a case where" on the third page. Why they square the point value isn't immediately clear to me from the binomial stuff I remember, but even without it you get higher point variance in 3pt case. (Even if you plug in GT's 0.29 and 0.435 for their example.)

Edit: I'm rusty but I don't know that looking at just 8 game-level samples and comparing variance across those would be a reliable way of trying to establish an empirical measure instead.

 

[orientalnc]

I think you're right that I was sloppy/unclear - at shot-rate, variance is lower for 3's since the `p` is further from 0.5 based on the binomial variance that's based on n*p*(1-p).

Studies at the NBA level have been done like here for the effect on scoring: https://www.sloansportsconference.com/research-papers/live-by-the-three-die-by-the-three - not just 'made bucket' variance.

The formatting of the math doesn't work to quote it but it's https://assets-global.website-files.com/5f1af76ed86d6771ad48324b/654bf324570bf68832c728b1_SSAC13 - Live by the Three, Die by the Three The Price of Risk in the NBA.pdf in the footnote that includes "Consider a case where" on the third page. Why they square the point value isn't immediately clear to me from the binomial stuff I remember, but even without it you get higher point variance in 3pt case. (Even if you plug in GT's 0.29 and 0.435 for their example.)

I think the math works, but the defense can be working hard to deny 3-point shots. Or make you take them from farther out or with close defenders. Whatever the math, you have to take what the defense is giving you.

 

[cpf2001]

Yeah I’m a big believer in “take the first good look you get” vs holding out for a “perfect” one.

 

[gte447f]

The math in our case, even accounting for variability, does not favor 3 point shooting. All the more reason to take opportunities around the basket when they present themselves.

 

[BeeRBee]

Speaking of testing rotations and sets, did we at one point have Sturdy, Abram, and George on the court at one time or am I hallucinating that? I know we definitely went with some 2 PG rotations at least towards the end of the first half.

My first thought was that you were right, but I don’t see it in the lineup breakdown

 

[Steverc]

A team taking fifty 3-point shots at 29% is likely to lose to a team taking fifty 2-point shots at 44% because on 44% of its possessions, it can get back and set up its defense.

 

[78pike]

A team taking fifty 3-point shots at 29% is likely to lose to a team taking fifty 2-point shots at 44% because on 44% of its possessions, it can get back and set up its defense.

But the chance of an offensive rebound on a 3 point shot is probably higher than getting the rebound on a 2 point shot. Three pointers often bounce much further from the basket increasing the chances for an offensive rebound. Not sure how that factors into the analytics above.

 

[YlJacket]

All the analytics are nice but recognize that CDS is committed to a spread offense focused on ball movement and 3 point shots. Hang on for the ride