The value of 3's

[TankOverlord]

This analysis is meant for the regular season as each individual game means very little in the overall scheme of things. The goal should be staying fresh and healthy, not maximizing wins. In the playoffs, reducing variance is far more important since there's little margin for error.

Advanced metrics such as eFG% take into account FT's as well as 2's and 3's. Here's a very simple breakdown of points per shot (PPS) on 2PA and 3PA. I only included our players with at least 100 attempts on 2's and 3's. First column is PPS on 2's, second column is PPS on 3's.

Butler .99 .99
Rose .92 .90
Hinrich .78 1.03
Brooks .90 1.37
Dunleavy .98 1.25
Mirotic .91 1.07

Now everyone loves to bash Rose for taking too many 3's, but as you can see even with his rather poor %, he produces about the same PPS. Same with Butler. Everyone else is producing much better offense taking 3's. In fact Brooks shoots a higher % from 3 than 2! And he gets an extra point!

Many seem to think attacking the basket is optimal. It is not. Not only do the #'s not support this, but the opportunity cost of wear and tear over the course of a season make it an even worse proposition. Furthermore, this is a business and players are taking a huge financial risk with any potential injury.

Now of course there is a balance like everything in life, so I'm not suggesting only doing one or the other. This would be highly exploitable, but you can still bias it a bit without the opposition being able to take advantage.

Certainly this doesn't take everything into account. If there's any interest, I can come up with an algorithm that quantifies some of these other concepts such as :

- taking 2's results in more FTA's
- 3PA's produce more variance than 2PA's
- although taking 3's can can lead to long rebounds and fast breaks, missing a shot in the paint or under the basket is worse since the player shooting is now stuck 94 feet from the other basket. Setting up your transition D is easier coming off 3PA's.
- it's much harder to defend someone taking a 3 as there is more space outside the arc and fewer defenders.
-taking shots in the paint is considerably more dangerous to the player, both injury and fatigue.

As an aside, it's extremely irritating listening to Funk spew nonsense regarding stats. The other day we were 1-10 at one point shooting 3's. He of course says we should stop taking them since we're missing. This is completely false as the sample size is statistically insignificant. There's a reason shooters keep shooting.

And last night, when Gasol was 0-3 from the FT line, he asked Stacy if this was due to fatigue. Again missing 3 in a row is insignificant and trying to come up with some reason for this is laughable.

Anyway, this is a very basic look at why teams are correctly taking more 3's every year.

[micromonkey]

A healthy Rose is going to get FTA's on 2s or we are totally doomed with him getting .92 PPS on a 2s. Doomed.
eFG does not include FTs--just value per shot:
Effective Field Goal Percentage; the formula is (FG + 0.5 * 3P) / FGA.

TS% includes FTs and value per shot
True Shooting Percentage; the formula is PTS / (2 * TSA). True shooting percentage is a measure of shooting efficiency that takes into account field goals, 3-point field goals, and free throws.

Dun .592
Brooks .583
Butler .581
Niko .566
Gasol .541
Gibson .536
Rose .497
Snell .496
Moore .496
McD .488
Noah .485
Kurt .473
Nazr .384
Cam .352

.55 is average league I think
Rose is killing us using possessions especially 3s early in the clock which freeze out more efficient options--which is everyone else just about except Snell/Noah/Kirk. Catch and shoot, late in clock 3s are fine the other stuff is horsecrap.

Rose has to drive and kick to the efficient 3 options ideally.

[TankOverlord]

A healthy Rose is going to get FTA's on 2s or we are totally doomed with him getting .92 PPS on a 2s. Doomed.
eFG does not include FTs--just value per shot:
Effective Field Goal Percentage; the formula is (FG + 0.5 * 3P) / FGA.

TS% includes FTs and value per shot
True Shooting Percentage; the formula is PTS / (2 * TSA). True shooting percentage is a measure of shooting efficiency that takes into account field goals, 3-point field goals, and free throws.

Dun .592
Brooks .583
Butler .581
Niko .566
Gasol .541
Gibson .536
Rose .497
Snell .496
Moore .496
McD .488
Noah .485
Kurt .473
Nazr .384
Cam .352

.55 is average league I think
Rose is killing us using possessions especially 3s early in the clock which freeze out more efficient options--which is everyone else just about except Snell/Noah/Kirk. Catch and shoot, late in clock 3s are fine the other stuff is horsecrap.

Rose has to drive and kick to the efficient 3 options ideally.

Thanks for the correction, I meant to say TS%.

[kingkirk]

The most efficient way of scoring, or at least bumping up your advanced metrics, is getting points from:

- The 3 point line
- The free throw line
- Dunks

It's why the mid range 2 is near extinct in the modern NBA.

With that in mind, if we want Rose taking a high percentage of his shots from deep, obviously, it's going to limit the amount of free buckets we get at the line.

If he isn't penetrating, we're not getting easy looks around the rim, particularly dunks.

Whilst taking 10 shots from 2 and making 5 mathematically equates to the same amount of points as taking 10 shots from 3 and making 3, it's not a balanced offense.

The 3 point game should be supplemented by the role players. The stars, they are the one that should be getting to the line and throwing it down, as they are often the ones that have the natural and physical ability to do so.

If Rose becomes more and more a jumper shooter, with less variation to his game simply because he gets more PPS on 3s, the entire offense will suffer, and he will never be more than an average TS% player, meaning he will never lead the Bulls anywhere.

[pb-ceo]

did I read that right snell and rose have = efficiency?

[kingkirk]

did I read that right snell and rose have = efficiency?

Based on TS%, Derrick Rose is at 49.7% and Tony Snell is at 49.6%.

So yes, they're virtually equal in that regard. That said, Derrick is obviously taking tougher shots with more defensive coverage, but yes, their TS% currently is the same.

[Rerisen]

Threes are generally efficient offense - IF - you can get looks. But you can't just force them up.

And this post has a bit of a Daryl Morey-ish vibe to it, in almost elevating the three to a messianic like level.

There are a lot of downsides to chucking too many regardless efficiency. Your bigs, like Noah and Gasol, are going to stop running up and down the floor as hard, setting picks, and such if they never touch the ball and its just perimeter guys jacking threes.

Your ball movement and teamwork might also suffer.

Further its just hard to generate a lot of good three point looks if you don't first have credible drivers to get defenders off their man. Then you also need enough shooters and spacing to make the necessary secondary and beyond, follow up passes, when the first shooter is run off his shot by the initial help. This happened to Korver a lot in 2011 in the playoffs, and we just didn't have the next reliable shooter lined up to finish the breakdown of the defense's rotations.

[Stratmaster]

did I read that right snell and rose have = efficiency?

Just another indication that advanced stats fail at their intended goal.

[Stratmaster]

I have been through this discussion so many times. Respectfully, there are bad assumptions in the OP (e.g. it is easier to set your defense coming off a missed 3 than a shot under the basket) and a ton of variables missing. But rather than go through the whole list again I would like to ask a few questions.

Against most teams, particularly the Western teams but even some of the east, do you think the bulls improve their chances by getting into a track meet, particularly defensively?

Are you willing to simply take Gasol out of the equation and make him into a wasted acquisition?

The Bulls started the season hot but are slowly moving down the stats charts in team 3 point FG%. Do you think the Bulls have an advantage over most other teams in 3 point shooting?

If the answer to any of those questions is no you need to seriously consider reducing the number of 3 point attempts; or at any rate certainly not increasing them.

If you are willing to trot out a lineup of Brooks, Hinrich, MDJ, Niko and whoever for meaningful minutes, then by all means fire away at every opportunity.

[DanTown8587]

did I read that right snell and rose have = efficiency?

Just another indication that advanced stats fail at their intended goal.

Nope that tells me exactly what I need to see but different outcomes

Rose = His low TS% is indicative of taking shots he's not going to make (threes) at a high percentage plus a large amount of drives not ending in fouls.

Snell = His low TS% on the volume+shots he takes means he probably doesn't have a spot in this league when you pair that with his poor defensive acumen.

[MrSparkle]

3P% Playoff Leaders:

13/14 Spurs, Heat, Pacers
12/13 Warriors, Heat, Spurs
11/12 Spurs, Clippers, Mavs (then Thunder, Heat)
10/11 Celtics, Mavs, Sixers

I wouldn't over-rely on the statistic, but it's clear that to a win a chip, you need to shoot the 3P shot a lot, and efficiently... unless you're loaded with multiple, best mid-range/slash players in the game (Lebron, Wade, Bosh).

Then you look at reg. season... and most the time, the team with the best record is shooting the best 3P%: Hawks, Spurs last year... and top-3 ranked shooters in prior years.

[Stratmaster]

did I read that right snell and rose have = efficiency?

Just another indication that advanced stats fail at their intended goal.

Nope that tells me exactly what I need to see but different outcomes

Rose = His low TS% is indicative of taking shots he's not going to make (threes) at a high percentage plus a large amount of drives not ending in fouls.

Snell = His low TS% on the volume+shots he takes means he probably doesn't have a spot in this league when you pair that with his poor defensive acumen.

But the quote was comparing Snell to Rose based on TS%. You can't do that, as you just outlined. Too many people use these stats to compare players with different roles. they just aren't of any value for that. If you want to compare Rose's TS% to another Guard who is the main scorer on their team to assess SCORING effectiveness, sure. but that still doesn't assess his overall offensive role and value.

The stat gives you a rating of one very focused aspect of a player's game. But many people use it as a catch-all for all that is offense. The goal of the stat is to provide a combined assessment of scoring effectiveness. If used in a very specific comparison based on roles and only looking at scoring effectiveness it has value. I see it used more often as some type of magic number tot ell you if a player has any offensive value.

I worded it badly in saying it fails at it's intended goal. I should have said many don't seem to understand it's intended goal.

[veji1]

What is now crazy, is that some defense schemes try to get opposing players to drive into the lane, rather than let them shoot 3s ! With the verticality rule, guys like Hibbert, Gobert, Whiteside, Chandler, Bogut, Asik and others are now efficient enough layup stoppers that it is almost better to let the opposing player drive and try to shoot a tough floater / high arching lay up / or stop-jumpshot than to let them shoot even half contested 3s... Crazy to think that.

[DanTown8587]

But the quote was comparing Snell to Rose based on TS%. You can't do that, as you just outlined. Too many people use these stats to compare players with different roles. they just aren't of any value for that. If you want to compare Rose's TS% to another Guard who is the main scorer on their team to assess SCORING effectiveness, sure. but that still doesn't assess his overall offensive role and value.

The stat gives you a rating of one very focused aspect of a player's game. But many people use it as a catch-all for all that is offense. The goal of the stat is to provide a combined assessment of scoring effectiveness. If used in a very specific comparison based on roles and only looking at scoring effectiveness it has value. I see it used more often as some type of magic number tot ell you if a player has any offensive value.

I worded it badly in saying it fails at it's intended goal. I should have said many don't seem to understand it's intended goal.

I tend to agree that you can't really glean much just on TS% about players and their value to both offense and defense. I just disagree that TS% fails at it's intended goal: it's a stat that measures how efficiently one scores with the shots they take where as all other stats (points, FG%, etc) are raw stats and don't paint a whole picture.

The fact Tony Snell and Derrick Rose have an exact same TS% tell me that they are both inefficient scorers; the difference is that they are inefficient for completely different reasons. I think the stat is completely fine, I think that 90% of people use it poorly.

[Ctownbulls]

i think you are excluding free throws from your calc....simply multiplying his 2p% by 2 and his 3p% by 3 is not a good analysis

[coldfish]

One thing that is lost in the analysis of what shot to take is the 24 second clock. In an optimal offense, you would work the ball around until you get a high value shot. The best shots are ones right at the rim but a wide open 3 taken by a competent shooter isn't much behind that.

That said, if you work the shot clock down to 5-8 seconds and you haven't got something you want, you have to take what is left. An open long 2 is a better shot than a clock violation.

Also, the long 2 is a set up shot. If you absolutely positively refuse to take it, teams will start playing super tough defenses on you. For example, a triangle and 2 where you have 3 big men packing the paint and 2 other people chasing around the competent distance shooters.

At the end of the day, we are better off having Rose shoot 30% from 3 versus taking a step in and hitting 40% of his long 2's. That's where this analysis has validity and its why Thibs forced Deng to shoot 3's instead of long 2's several years ago. That said, you can't take it too far. Basketball fundamentals still wins.

......

The example I always use is football run versus pass. By the numbers, passing always gives you more yards and a better chance to score. That said, you rarely see pass only teams do well because of the non statistical impacts. Their defense is on the field more and tires them out. Teams load up on the pass rush and beat up your QB. LB's drop back into coverage quicker because they have no fear of the run. By the numbers, running is rarely a good idea but in practice, you have to do it enough to keep teams honest.

Same thing with shooting 2's versus 3's.

[DanTown8587]

One thing that is lost in the analysis of what shot to take is the 24 second clock. In an optimal offense, you would work the ball around until you get a high value shot. The best shots are ones right at the rim but a wide open 3 taken by a competent shooter isn't much behind that.

That said, if you work the shot clock down to 5-8 seconds and you haven't got something you want, you have to take what is left. An open long 2 is a better shot than a clock violation.

Also, the long 2 is a set up shot. If you absolutely positively refuse to take it, teams will start playing super tough defenses on you. For example, a triangle and 2 where you have 3 big men packing the paint and 2 other people chasing around the competent distance shooters.

At the end of the day, we are better off having Rose shoot 30% from 3 versus taking a step in and hitting 40% of his long 2's. That's where this analysis has validity and its why Thibs forced Deng to shoot 3's instead of long 2's several years ago. That said, you can't take it too far. Basketball fundamentals still wins.

......

The example I always use is football run versus pass. By the numbers, passing always gives you more yards and a better chance to score. That said, you rarely see pass only teams do well because of the non statistical impacts. Their defense is on the field more and tires them out. Teams load up on the pass rush and beat up your QB. LB's drop back into coverage quicker because they have no fear of the run. By the numbers, running is rarely a good idea but in practice, you have to do it enough to keep teams honest.

Same thing with shooting 2's versus 3's.

I just wanted to add to this

Pau Gasol shoots about 36% on jump shots were he is guarded (2-6 feet closest defender) but that percentage sky rockets to 57% when he's open. It's the same shot and it goes from woefully inefficient to efficient solely based on how open Gasol is. I have no problem with open mid range twos (i bet you wouldn't find a coach who does as long as the player has some ability to shoot); it's the number of contested mid range twos that concerns me. But I do understand that you have to take some guarded twos to get the open ones.

[pb-ceo]

did I read that right snell and rose have = efficiency?

Just another indication that advanced stats fail at their intended goal.

i think the advanced stats are very useful in evaluating how our pg is doing. because he is well liked, and he produces lots of youtube highlights that people enjoy and tend to equate with good basketball. i think an analysis of the advanced stats goes a long way in in explaining why our pg was left off the all star team.

[Stratmaster]

did I read that right snell and rose have = efficiency?

Just another indication that advanced stats fail at their intended goal.

i think the advanced stats are very useful in evaluating how our pg is doing. because he is well liked, and he produces lots of youtube highlights that people enjoy and tend to equate with good basketball. i think an analysis of the advanced stats goes a long way in in explaining why our pg was left off the all star team.

I would say his turnover totals, assist to turnover ratio, and basic shooting percentage tell you all you need to know, and more than his TS% tells you.

[TankOverlord]

Good discussion going on, but no one addressed my most important point. The statistical value of 2's and 3's are very close no matter how you look at it. However no one can debate the risk of injury between the two, this is why 3's are so valuable during the regular season.

You can achieve very similar results while greatly reducing wear and tear. All you're giving up is higher variance, but this evens out over 82 games. Going into the playoffs beat up is not a winning strategy as we've seen.

Also, interesting point on Gasol % contested vs. uncontested. Teams are likely funneling players into the lane as it's more difficult to get off an uncontested spot in this situation. Whereas we've seen with the top 3 point shooters they can get themselves good looks pretty much whenever they want it. After watching Brooks this year, do you really think he can't get high % looks?

Morey does some crazy things in Houston, but their style will become the league norm at some point. While I agree the Bulls are not ideally constructed for this, I'm arguing they should be. When everyone on the floor can shoot, running a guy off the 3 point line is ineffective as he just swings it to the next shooter.