APBRmetrics

[magicmerl]

Ok, so over at BasketBall Reference there is a cool season summary and the interesting thing to me is how for for EVERY single team (except Philly) in the league their eFG% is higher than their 2P%.

This suggests to me that if teams were somehow able to take their most inefficient 2PT shots and take them as inefficient 3PT shots their offensive efficiency would be increased. And their wins.

Take the surprise good teams this year (whose record ended up being better than we thought it would be). To my mind they are:
New York
Orlando
Indiana
Cleveland (I thought they would be good, but didn't expect them to be as good as they have been)
Atlanta
Miami
Denver

Now look at all of the under-performing teams:
Utah
Philadelphia
Detroit
Toronto
Phoenix
New Orleans

With the exception of Denver, which I put down to the Chauncy trade, the overachieving teams are all in the top half of 3PA. While the under-performing teams are in the bottom half of 3PA.

So irrespective of 3P%, it would appear that teams can increase their offensive efficiency (and also wins) by taking more three pointers.

Is there something to this? If NBA coaches want to win why are their teams not jacking up more 3s?

[BobboFitos]

you would think in an optimal offense, a team would have equal 2pt% to eFG%... if one was higher, then you should shift to your strength (until the defense reacts/adjusts/game plans to the point that whatever shot you moved away from became easier, etc etc). that said you don't find this happening in practice whatsoever. There was a good Hollinger article a few weeks (10 days?) back about predicting results based on 3PA, and hopefully the trend continues season by season (more 3s).

really if the solution was as simple as "shoot more 3s" it's surprising this hasn't caught on at either the coaching level, or I suppose the free agent market. (Bigger contracts to 3pt specialists, "shooters" in general.. But perhaps we'll see this happen this summer).

Philly is a black sheep, but also somewhat of an interesting team to study. They're somewhat of a throwback team, I'm not sure if that's by design or that they simply had no other options entering the season.

edit:

[Mike G]

In a league where fouls are not called, you are probably better jacking up a 3 every time. Because we have fouls, we have FT (75% made), foul trouble for the opponent, and-one 3-pt plays, etc.

On almost every 2-pt attempt there's a chance for a make, a chance for a foul, a chance for an and-one, a chance for an offensive rebound. There are few fouls on 3pt attempts.
_________________
`
36% of all statistics are wrong

[Ryan J. Parker]

I'm sure there is some sort of game theory in all of this that I'm not familiar with, but jacking up more 3s isn't the answer. In fact, it could hurt rather than help.

My take is that a solid inside threat will open up your 3pt shooting. I'm sure this is obvious to most here on the board. Thus the teams that have this sort of advantage inside that smartly look for their best 3pt shooters on the arc are going to be some of the best teams in the league.

Maximizing 3FGA/FGA isn't the answer here.
_________________
I am a basketball geek.

[Mike G]

In the season, Philly put up 13 3PA/G, and hit just 32%.
In the playoffs, they've shot 14/G, hitting 42%.
They're getting an extra 1.85 PPG, not by shooting more but by hitting more.
_________________
`
36% of all statistics are wrong

[Ben]

It's the marginal shot that we need to compare efficiency on. It's not obvious how one gets at that, but the average efficiency might not be equal in the optimal scenario. Otherwise, one would just funnel the ball the guy with the highest 3p% - presumably more efficient than anybody else's TS% taking 2's.

[Mountain]

A lot of this has been covered before, but I'll recap and suggest some research opportunities.

The 3 point shot is better (straightaway and corner) or at least as good (wing) as any other shot except the deep paint by the 82 games floor location article per points per possession- including the different free throw effects and turnovers. http://www.82games.com/locations.htm

Teams have equalized eFG% on 3s versus all 2s almost exactly and may think they have good shot distribution balance but the efg% of 2s is the mix of inside and mid-range shots and by a 3 part split shot distribution appears off optimal with way way too many mid-rangers but the question is how much can teams really reduce them? I'd think by a lot and primarily by deciding in advance to go for more 3s. More transition 3s, walk up 3s, screen 3s, drive and kick out 3s.

I think in today's game one main answer is indeed to increase 3 ptA / FGA but as I've said before it is also to rigorously reduce mid-range shots. Two sides of the same coin.

Get to the rim or get fouled as much as possible but "settle" for the midrange pullup as little as possible.

Shooting the 3 appears to make sense from what I see, especially instead of mid-rangers, though we'd need to know how many 3s are contested and how much and how 3pt FG% varies by degree of contest and how easy it is to create additional uncontested or modestly contested / still more efficient 3 pt shots. I'd guess most teams have the ability to find additional uncontested or modestly contested and still more efficient 3 pt shots. They have kept doing it year to year pretty steadily. If you say increasing 3 attempts isn't "the answer" then has the league has been doing the wrong thing for 25 years? I don't think they have maxed out yet, though there is probably some ultimate limit. And defenses have some ability to adjust further if teams adjust shot distribution further.

If a contested above a certain level 3 is no better on offensive efficiency than a contested 2 than that is where you stop emphasizing it. And of course the stop points will vary by player and team.

If you can adjust the mid-range / inside ratio you certainly should. Some mid-range shots essentially happen as bailout while trying to get inside and are acceptable price to get the juice of the greater payoff shots. But discretionary 2s are generally a poor decision. Mid-range shots flat out on average are the worst shot. At last resort take the mid-range but preferably only if you are one of the team's best mid-range shots and open. Contested 2s are a defensive victory.

Are offenses that emphasize the 3 point shot more heavily more volatile? I would not be surprised if they are but somebody should check it. Maybe it opens up the floor and makes things work better more often.

Pairing a 3 pt attack with a strong, reliable defense can certainly help the win equation. Adding strong inside game of course helps too. Post-ups that lead to shot in close but also add as much penetration as you can get that gets to the rim or the foul line. Jumpshots by bigs are generally not that high-powered offense and should be considered a decent plan B, C or D, not plan A.

A good 3 point game probably makes it easier for second-tier interior players to score, directly with the benefit of more space / less help defense or off offensive rebounds.

De-emphasizing the number of players who think they can beat a team with their mid-range game probably helps too. You probably need a Kobe or Brandon Roy penetrating to set the table for the high powered inside shot and 3 shots. You don't need below star-level or average and below players taking mostly very inefficient mid-range shots. Teams that design their roster to minimize run of the mill "ballers" who go 1 on 1 and end up in a lot of inefficient contested mid-ranger shots are doing the right thing. Even self-styled mid-range specialists who shot the mid-range better still end up shooting their mid-range shots at an eFG% below the league average on eFG% for all shots Not a single player who shot enough to qualify at 82 games shot 50% from mid-range this season and typically it is only a handful. And only 19 shot better than 45% from mid-range overall. Tough, have to take it, end of shot clock shots contribute to this but probably doesn't fully explain. Though if 82games or another play by play source wanted to detail mid-range shooting quality by part of shot clock that would be helpful.

Coaches who actively manage the team thru play calling and play enforcement benchings to reduce mid-range shots more than others are doing it right. Teams that design their roster to emphasize high percentage shots - inside or 3- and offensive rebounding and passes that "create" high percentage shots are doing the right things.

Penetration can help too, as noted above and previously, but it has to to end up pretty often in inside shots or foul shots by them or other higher percentage shots for others. It would be good to try to isolate team offensive efficiency when players put the ball on the floor and try to penetrate. Green light those who produce efficient team results; restrict or eliminate those who don't.

[Ryan J. Parker]

The 3pt shot is a very good way to maximize eFG%, but the point I'd like to make is that we can't just start suggesting that teams throw up a lot more 3pt shots.

Developing plays that have some probability of creating an open/lightly contested 3pt shot is more along the wording I prefer. If this shot is available for a player with appropriate skill then the shot should be taken (unless, of course, game situation calls for something else).

I'm sure we're thinking along the same lines here, but I just want to make sure we don't overlook the fact that open 3pt shots don't create themselves.
_________________
I am a basketball geek.

[Mountain]

In the Philly example the point gain is entirely because of improved FG% as Mike said; but this is an extreme example of vast improvement from league worst in regular to a rate above any had in regular season.

For many teams, shooting more 3s is quite valuable and usually very competitive with shooting them better as a way to improve point yield, though of course it varies. Total increased point yield for 3 point shots above what the team yielded on a similar number of 2 point shots or just mid-range shots is the way to evaluate the strength of team 3 point games. In most cases quantity and quality of 3s are both important.

To gain .16 points an average FG% and 3 point frequency team can either move total team 3 point FG% up by about 2 percentage points (moving from average to 3rd best in the league) or they can simply shoot one more if their FG% is steady and get about the same increase point yield.