Help me with 3-point analytics

[gamecockcat]

The common wisdom among basketball people is a 3-pointer is a better shot than a 2-pointer if the team hits roughly 1/3 of their 3s as most teams don't hit 50% of their total shots, right? But, shooting statistics INCLUDE 3-pointers. So, I wanted to see if the 3-point analytics hold water.

I looked at the stats on ESPN of the top four scoring NBA teams. Here's what I found:

Memphis
48.5% FG
37.5% 3Pt ( so far, so good)
55.4% FG on 2-point shots (subtracting 3 pointers from the total stats)
Conclusion: basically a push, slightly better shooting 3s

Cleveland
49.8% FG
41.2% 3Pt
59.2% 2Pt FG
Conclusion: Cleveland is better off shooting 3s, but not by much

Denver
45.8% FG
31.2% 3Pt
57.6% 2Pt FG
Conclusion: Denver is much better off shooting 2s

Knicks
49.6% FG
34% 3Pt
56.4% 2Pt FG
Conclusion: Knicks are better off shooting 2s

That's just the top 4 scoring teams. I would assume the lower scoring teams would skew more towards 2s being better than 3s, but that's a guess.

But, among the best, we have one clear winner for 3s, one pretty much a toss up and 2 teams for which shooting 2s is quite a bit better.

Am I missing something? In addition to making the NBA less enjoyable to watch, the preponderance of three points shooting is detrimental to winning games, at least from what I see the stats telling me.

Surely I'm missing something but I don't know what.

 

[storm1507]

How does UK turn out with those percentages this season?

 

[Seth_C]

What you aren’t seeing in your analysis is how much layups and dunks are pushing up your 2p shots. No one is saying a 3 pointer is better than a dunk. Still plenty of dunking and layups. The question is whether a 3 pointer is a better shot than a 2 point jump shot and it almost always is.

According to my Googling, the average percentage of made 2 point jump shots in the NBA is only 38%. That means even if you shot 26% from 3 you would be better off than the average 2 point jump shot. Across 100 shots you would score 14 more points even if you only averaged 30% on them.

 

[JonathanW]

How does UK turn out with those percentages this season?

UK is shooting I think around 58% from 2 this year. 37% from 3

 

[rick64]

Look at Alabama shot chart dispersement. Practically everything is either a 3pt shot or 3ft shot. Nothing mid range.

 

[GoAndRunIsGoneAndDone]

and how many dunks and layups are created by spreading the floor with 3pt shooting

 

[Cats192]

The common wisdom among basketball people is a 3-pointer is a better shot than a 2-pointer if the team hits roughly 1/3 of their 3s as most teams don't hit 50% of their total shots, right? But, shooting statistics INCLUDE 3-pointers. So, I wanted to see if the 3-point analytics hold water.

I looked at the stats on ESPN of the top four scoring NBA teams. Here's what I found:

Memphis
48.5% FG
37.5% 3Pt ( so far, so good)
55.4% FG on 2-point shots (subtracting 3 pointers from the total stats)
Conclusion: basically a push, slightly better shooting 3s

Cleveland
49.8% FG
41.2% 3Pt
59.2% 2Pt FG
Conclusion: Cleveland is better off shooting 3s, but not by much

Denver
45.8% FG
31.2% 3Pt
57.6% 2Pt FG
Conclusion: Denver is much better off shooting 2s

Knicks
49.6% FG
34% 3Pt
56.4% 2Pt FG
Conclusion: Knicks are better off shooting 2s

That's just the top 4 scoring teams. I would assume the lower scoring teams would skew more towards 2s being better than 3s, but that's a guess.

But, among the best, we have one clear winner for 3s, one pretty much a toss up and 2 teams for which shooting 2s is quite a bit better.

Am I missing something? In addition to making the NBA less enjoyable to watch, the preponderance of three points shooting is detrimental to winning games, at least from what I see the stats telling me.

Surely I'm missing something but I don't know what.

1. The layup/Dunk is the dream shot. Every team would dunk it every possession if that option existed.

The argument isn't "3's are better than 2's." 3's are better than long 2's. You almost need to take break 2 point shots down into subsections.

2. Spreading the floor, maintaining a 3 point thread, creates spacing. Spacing leads to layups and dunks.

There are a lot of 3's shot in the NBA. But people act like every team is just out there chucking up 3's in the name of "analytics." You'll see a lot of 2 point shots every game. Most of the 3's are the result of good offense to create quality looks.

 

[JonathanW]

I am a Statistician.

It’s not as simple as many people want to believe. First of all, many teams shoot 50%+ from 2. FG% include 3pt shots. A better estimate is 36% from 3 which equates to the number of points you would get making 54% from 2’s.

But….there are other factors as well. These include:
- the more 3’s you make, the more teams will guard you closer out there, making 2’s easier to score at a higher %
- missed 3’s and missed 2’s are not offensively rebounded at the same rate
- not all 2’s (and their %’s) are alike; a 15-18’ jumper is very different than a n 8-10’ floater, is very different from a basket at the rim
- attempting to shoot 2’s vs 3’s probably have different turnover rates
- attempting to shoot 2’s vs 3’s definitely have very different foul and FT rates (this one is very much under considered!). And 2-shot FTs are the most efficient way to score (1+1 FTs are not). A 70% FT shooter, fouled 100 times while shooting a 2pt shots, will average 140 pts at the line. But if those 100 times are all 1+1 then he would average 119 points (which is still good, equates to 39.7% from 3, but not as good as 140 pts which equates to 46.7% from 3).
- there are probably a few other factors I’m not thinking of

 

[electricat2]

i just woke up man......too many numbers makes my brain hurt

 

[gamecockcat]

What you aren’t seeing in your analysis is how much layups and dunks are pushing up your 2p shots. No one is saying a 3 pointer is better than a dunk. Still plenty of dunking and layups. The question is whether a 3 pointer is a better shot than a 2 point jump shot and it almost always is.

According to my Googling, the average percentage of made 2 point jump shots in the NBA is only 38%. That means even if you shot 26% from 3 you would be better off than the average 2 point jump shot. Across 100 shots you would score 14 more points even if you only averaged 30% on them.

Agreed.

But, if teams concentrated more on creating open 12-foot jumpers, would those shooting percentages increase? The two point shots may be a bit longer due to trying to work for a three, may be deeper in the clock, may be a step or two longer because the offense's goal is to NOT shoot mid- to long-range jumpers. A lot of things could be affecting that stat, including kids playing today shoot 3s in middle school rather than jumpers.

So much of the NBA offense is built around shooting a 3. If it was built around getting open looks for midrange jumpers or more layups, perhaps the corresponding shooting accuracy would increase. Perhaps.

And, yes, spreading the floor leads to more layups and dunks. But the floor spacing could remain but the end result goal could be a short jumper or layups rather than a 3.

I cringe every time I see a guy within 5' of the basket with an open look throw the ball out to the arc for a 3.

I clearly remember the Walton-led Trailblazers who would shoot layups and short jumpers after moving the ball around. The 80s Celtics would certainly take some 3s but shot a lot more 2s.

I just don't think it's cut and dried to the extent that many people feel it is. But, the NBA style bores the beejeezus out of me so I'm probably letting my distaste for their game color my analysis somewhat.

 

[gamecockcat]

I am a Statistician.

It’s not as simple as many people want to believe. First of all, many teams shoot 50%+ from 2. FG% include 3pt shots. A better estimate is 36% from 3 which equates to the number of points you would get making 54% from 2’s.

But….there are other factors as well. These include:
- the more 3’s you make, the more teams will guard you closer out there, making 2’s easier to score at a higher %
- missed 3’s and missed 2’s are not offensively rebounded at the same rate
- not all 2’s (and their %’s) are alike; a 15-18’ jumper is very different than a n 8-10’ floater, is very different from a basket at the rim
- attempting to shoot 2’s vs 3’s probably have different turnover rates
- attempting to shoot 2’s vs 3’s definitely have very different foul and FT rates (this one is very much under considered!). And 2-shot FTs are the most efficient way to score (1+1 FTs are not). A 70% FT shooter, fouled 100 times while shooting a 2pt shots, will average 140 pts at the line. But if those 100 times are all 1+1 then he would average 119 points (which is still good, equates to 39.7% from 3, but not as good as 140 pts which equates to 46.7% from 3).
- there are probably a few other factors I’m not thinking of

Wow. That's a lot to consider, which is what I originally thought when I started this exercise. It's way oversimplified to just look at FG%, multiply the 3Pt by 1.5 and draw a conclusion.

Good stuff, JW.

 

[G-PIP]

The common wisdom among basketball people is a 3-pointer is a better shot than a 2-pointer if the team hits roughly 1/3 of their 3s as most teams don't hit 50% of their total shots, right? But, shooting statistics INCLUDE 3-pointers. So, I wanted to see if the 3-point analytics hold water.

I looked at the stats on ESPN of the top four scoring NBA teams. Here's what I found:

Memphis
48.5% FG
37.5% 3Pt ( so far, so good)
55.4% FG on 2-point shots (subtracting 3 pointers from the total stats)
Conclusion: basically a push, slightly better shooting 3s

Cleveland
49.8% FG
41.2% 3Pt
59.2% 2Pt FG
Conclusion: Cleveland is better off shooting 3s, but not by much

Denver
45.8% FG
31.2% 3Pt
57.6% 2Pt FG
Conclusion: Denver is much better off shooting 2s

Knicks
49.6% FG
34% 3Pt
56.4% 2Pt FG
Conclusion: Knicks are better off shooting 2s

That's just the top 4 scoring teams. I would assume the lower scoring teams would skew more towards 2s being better than 3s, but that's a guess.

But, among the best, we have one clear winner for 3s, one pretty much a toss up and 2 teams for which shooting 2s is quite a bit better.

Am I missing something? In addition to making the NBA less enjoyable to watch, the preponderance of three points shooting is detrimental to winning games, at least from what I see the stats telling me.

Surely I'm missing something but I don't know what.

Good thread, and a nice analytical starting point for thinking about it logically, rather than conventional wisdom clichés (way too much of that in the world in regard to everything). Posters responding have already added a number of good points which, I think, show you have to think about it deeper than just the raw percentages (which you already know). The fouls and free throws accumulated with scoring around the basket on post-ups and drives is another factor to figure in also.

 

[Eagles_Ball]

What you aren’t seeing in your analysis is how much layups and dunks are pushing up your 2p shots. No one is saying a 3 pointer is better than a dunk. Still plenty of dunking and layups. The question is whether a 3 pointer is a better shot than a 2 point jump shot and it almost always is.

According to my Googling, the average percentage of made 2 point jump shots in the NBA is only 38%. That means even if you shot 26% from 3 you would be better off than the average 2 point jump shot. Across 100 shots you would score 14 more points even if you only averaged 30% on them.

This is the moral of the story. If you shoot 3s well enough, defenses have to get out on your shooters. This opens the lane, thus allowing for more 2s at the rim. This pushes your 2pt % much higher obviously. Long 2s are typically shot around 6-8% higher than all 3s. Those two factors, shoot 3s well to space/create opportunities for easy baskets and shoot 3s instead of long 2s, is the recipe.

OP would be better off looking at season long shot charts for the most efficient (not highest scoring, think points per possession) offensive teams. I’ll give you a hint, it’s basically like looking at Bamas shot chart. Everything in the paint or beyond the arc.

 

[Eagles_Ball]

Agreed.

But, if teams concentrated more on creating open 12-foot jumpers, would those shooting percentages increase? The two point shots may be a bit longer due to trying to work for a three, may be deeper in the clock, may be a step or two longer because the offense's goal is to NOT shoot mid- to long-range jumpers. A lot of things could be affecting that stat, including kids playing today shoot 3s in middle school rather than jumpers.

So much of the NBA offense is built around shooting a 3. If it was built around getting open looks for midrange jumpers or more layups, perhaps the corresponding shooting accuracy would increase. Perhaps.

And, yes, spreading the floor leads to more layups and dunks. But the floor spacing could remain but the end result goal could be a short jumper or layups rather than a 3.

I cringe every time I see a guy within 5' of the basket with an open look throw the ball out to the arc for a 3.

I clearly remember the Walton-led Trailblazers who would shoot layups and short jumpers after moving the ball around. The 80s Celtics would certainly take some 3s but shot a lot more 2s.

I just don't think it's cut and dried to the extent that many people feel it is. But, the NBA style bores the beejeezus out of me so I'm probably letting my distaste for their game color my analysis somewhat.

If you design an offense around getting those 12-15’ looks, the floor is compressed or shrunk. End result? No more easy dunks or layups. Everything at the rim would be contested. Sure, you’d probably shoot a higher percentage on those 12-15’ shots than from 3, but you’d have far less opportunities and far less makes at the rim. That’s the thing most don’t consider when this topic comes up. It’s why teams don’t run Flex, Triple Post, or the Triangle any longer. It’s not like offenses don’t exist to get these shots. They’ve been around forever. Nobody runs them anymore because they’re wildly inefficient.

 

[Cats192]

I am a Statistician.

It’s not as simple as many people want to believe. First of all, many teams shoot 50%+ from 2. FG% include 3pt shots. A better estimate is 36% from 3 which equates to the number of points you would get making 54% from 2’s.

But….there are other factors as well. These include:
- the more 3’s you make, the more teams will guard you closer out there, making 2’s easier to score at a higher %
- missed 3’s and missed 2’s are not offensively rebounded at the same rate
- not all 2’s (and their %’s) are alike; a 15-18’ jumper is very different than a n 8-10’ floater, is very different from a basket at the rim
- attempting to shoot 2’s vs 3’s probably have different turnover rates
- attempting to shoot 2’s vs 3’s definitely have very different foul and FT rates (this one is very much under considered!). And 2-shot FTs are the most efficient way to score (1+1 FTs are not). A 70% FT shooter, fouled 100 times while shooting a 2pt shots, will average 140 pts at the line. But if those 100 times are all 1+1 then he would average 119 points (which is still good, equates to 39.7% from 3, but not as good as 140 pts which equates to 46.7% from 3).
- there are probably a few other factors I’m not thinking of

Really good breakdown

 

[rick64]

I remember when Pitino had his first team att UK. During practice if you took a long 2pt shot he would stop practice and make you run. If you took a 3pt shot but you touched any part of the line you ran.

 

[Lempface]

and how many dunks and layups are created by spreading the floor with 3pt shooting

Yep, things like this are easy to miss with this type of bottom line number analysis. Very similar to the pencil pushers in corporate offices that make brain dead decisions because it’s all based on black and red and they miss the nuance.

 

[Aike]

1. The layup/Dunk is the dream shot. Every team would dunk it every possession if that option existed.

The argument isn't "3's are better than 2's." 3's are better than long 2's. You almost need to take break 2 point shots down into subsections.

2. Spreading the floor, maintaining a 3 point thread, creates spacing. Spacing leads to layups and dunks.

There are a lot of 3's shot in the NBA. But people act like every team is just out there chucking up 3's in the name of "analytics." You'll see a lot of 2 point shots every game. Most of the 3's are the result of good offense to create quality looks.

This is the main thing OP is missing, imo. None of these shots are taken in a vacuum.

 

[megablue]

I don't know if these are accurate, but something to consider in this discussion:

 

[Tubbyfan78]

The common wisdom among basketball people is a 3-pointer is a better shot than a 2-pointer if the team hits roughly 1/3 of their 3s as most teams don't hit 50% of their total shots, right? But, shooting statistics INCLUDE 3-pointers. So, I wanted to see if the 3-point analytics hold water.

I looked at the stats on ESPN of the top four scoring NBA teams. Here's what I found:

Memphis
48.5% FG
37.5% 3Pt ( so far, so good)
55.4% FG on 2-point shots (subtracting 3 pointers from the total stats)
Conclusion: basically a push, slightly better shooting 3s

Cleveland
49.8% FG
41.2% 3Pt
59.2% 2Pt FG
Conclusion: Cleveland is better off shooting 3s, but not by much

Denver
45.8% FG
31.2% 3Pt
57.6% 2Pt FG
Conclusion: Denver is much better off shooting 2s

Knicks
49.6% FG
34% 3Pt
56.4% 2Pt FG
Conclusion: Knicks are better off shooting 2s

That's just the top 4 scoring teams. I would assume the lower scoring teams would skew more towards 2s being better than 3s, but that's a guess.

But, among the best, we have one clear winner for 3s, one pretty much a toss up and 2 teams for which shooting 2s is quite a bit better.

Am I missing something? In addition to making the NBA less enjoyable to watch, the preponderance of three points shooting is detrimental to winning games, at least from what I see the stats telling me.

Surely I'm missing something but I don't know what.

That’s not the way analytics work. You’ve removed all variables from the equation. An 18 foot jump shot is not the same as a dunk and a challenged three as the shot clock expires is not the same as an inside out 3 in the flow of offense. You’re talking about stats not analytics.

 

[runt#69]

Look at Alabama shot chart dispersement. Practically everything is either a 3pt shot or 3ft shot. Nothing mid range.

THIS

 

[Col. Angus]

UK is shooting I think around 58% from 2 this year. 37% from 3

EFFICIENCY

 

[Cats192]

I don't know if these are accurate, but something to consider in this discussion:

View embedded media

View embedded media

Admittedly, didn't watch the video.

But being titled "the value of the mid-range 2" I think I can assume what it's going to say.

There's absolutely value in any "good shot." As teams are fighting for playoff seeding, and certainly in the playoffs, teams will try to force you into the midrange. They'll run you off the 3 point line and try to clog/catch up in the paint. So that 18-20 foot jumper is left open a lot. Especially as it gets late in a possession, an open mid range shot can be a very good shot. And the best players will have that in "their bag."

You'd still rather have a good look from 3 or at the rim. But an open shot is a good shot. There's a reason teams don't just stop guarding you at 15-20 feet and dare you to shoot the "ineffient" shot.