Posted by: natehughart | February 19, 2011

True Shooting Percentage, Offensive Rating, and why it’s probably a good idea to have players who do well in this area

Awhile ago, in an attempt that didn’t go over well with a few people (which resulted in some reasonable, and not so reasonable, frustrations), to point out that the reason the Kings offense isn’t very good because they don’t have an efficient offense. How do I know that? Offensive rating, or points per 100 possessions, is one area that helps give a nice overview of a team’s total offensive capability. The Kings are currently 26th in ORtg at the All-Star break. I think there are 2 main reasons for that. One, that’s not necessarily measured in statistics but observation, is that they don’t do a great job creating easy shots, and the shots they do get are often contested. That makes it much harder to have a quality offense. The second part, however, we very much can measure. Right now, the Kings have 2 players over an average efficiency level: Beno Udrih and Francisco Garcia. With Garcia hurt, that’s a problem. And even Beno’s vacillates between being very good and great in this area. Of everyone else, only Carl Landry is at an average efficiency (which is below what he did in Houston) right now. Thus, because the Kings don’t use their possessions efficiently, they have a worse offense. So, why do I use TS% for players and ORtg for teams? I’m getting to that. Hold on.

I bring this particular point, about TS% for players, and ORtg for teams for a reason. I don’t really use TS% for teams because how well they make or miss their shots doesn’t tell me much. It tells me if they make shots at a high level, but it doesn’t do much else. But what does ORtg take into account? Everything. By everything, I mean pace, quality of shots, how many shots you get in a game (if you’re missing a lot of shots but getting a lot of putbacks, that reflects in your ORtg and poorly in your shooting percentages), and you can see where you ORtg rank’s among other teams. With players, ORtg doesn’t tell you as much and it’s useful to break their shot selection, and shot location, along with how adept they are at making the most of their possessions. That’s why I use TS%, (eFG% less because it only uses 3pt FG’s and 2 pt FG’s) because we can see a players total offensive sum worth is. I also USG% because it’s a quality estimation about how many possessions (remember a possession isn’t just shots but turnovers and FT’s too), From several years Mr Ziller waxed poetic about why TS% is so important (in this case about Kevin Martin):

True Shooting percentage, a metric I use often, measures shooting efficiency. Field goal percentage is the basic metric to measure shooting efficiency, but as most of you know it’s not the best use because it doesn’t take into account three-pointers. Three-pointers are a big deal, obviously, and shooters who make them should be credited with all three points a make earns. Using FG%, going 2-4 from three is the same as going 2-4 from two-point range. On the scoreboard — which is what matters — going 2-4 from three means six points; going 2-4 under the rim means four points. As such, the TS% for the three-point shooting scenario is 75% (six points in four shooting possessions) and the TS% for the two-point shooting scenario is 50%.

But what about free throws? Those matter too! How many times this season has someone referred to Martin as inefficient because of his 42% field goal percentage? Of course, Martin is among the most efficient top scorers … because of his foul-drawing ability. True Shooting percentage says that points scored from the line should be credited in your shooting efficiency, too. If a player shooting 3-10 from the field (all twos), but also draws 8 FTAs (and makes them all), he has scored 14 points and used 14 of his team’s possessions (assuming the FTAs didn’t come on and-1s or technical fouls). On the scoreboard, this counts the same as a player who shoots 7-14 (all twos again) and draws zero FTAs. Again, if it counts the same on the scoreboard, it should count the same in our shooting percentage metric. With TS%, it does.

Think about this for a moment. Some players don’t shoot well very often, but play very little. How important is their TS% to the overall level of the offense? It’s not very important. What’s most important are the players who play frequently and use the most possessions (or USG% as Ill note below). Your rotation players in otherwords.

The whole point of the post I linked to at the very beginning was to show that efficient offensive players (players with an excellent or better TS%) and having multiple players like that make up an efficient offense. Now, I never concluded (or wished to) how teams do that. I knew better. There are so many factors as to why teams get quality shots, and convert them at different shots. It’s really hard to find one tell all method to figure that out. For instance, why the Kings are inefficient offensively are not necessarily why the Bucks are inefficiently offensively.

As to why I didn’t really explain this earlier, my feeling is this: If I have to explain something pretty simple like whether a player is good at making shots or not, you might not get the more complicated or nuanced parts of it. I recognize why people don’t accept formula’s easily: They don’t understand how formulas break down. But if you look at TZ’s post that I linked to, that’s why I use TS% for formula’s. I’m not even going to re-iterate what TZ said or try to emulate it: He’s the stathead not I. He’s the Dean Oliver-lyte, not I. But I do know that it works and I’ve seen it work in a game to game (although he stresses against that in the post and it’s not hard to see why) basis too. The real reason is that while offense’s have many components to them overall, players individually shoot shots to get their points. Looking at a team’s overall percentage in making and/or taking shots tells me little. If you shoot 20 shots at a 40% clip, that’s a lot different than shooting 100 shots at a 40% clip. That’s why I look at ORtg. (Obviously it isn’t that extreme, but, hopefully the point is made.) The latter is how Red Auerbach (and not just the Celtics but teams in that era generally) operated. Get as many shots up and as often as you can. Seriously, check out teams from the 56-70 era on Basketball-Reference. As the game evolved, a need to become more efficient was apparent.

Having said all that, I’m explaining this for a reason: I want people to understand why, and go from there. Complicated stuff tends to be something people either get or don’t; there’s little reason to explain it much differently. But something simple? There shouldn’t be any confusion about TS% or ORtg and yet there is. I think that’s why I’m writing this as much as anything.

So this is what I’ve been leading up to (and will go on further): TS% is a great way to measure a players efficiency and compare them to each other because TS% doesn’t discriminate. You can compare Dwight Howard to Kevin Martin and conclude that they were both efficient offensive players. (Both players TS%’s historically have been excellent. It’s also nice as Howard almost plays exclusively near the basket to get his shots and his league FTA, where Martin is a perimeter player who shoots well and gets to the line–hence the high TS%) One other thing here: If you look at the 20/60 list that TZ listed (which is almost 2 years old now), you’ll notice that almost exclusively the players on that list were post players from the older times (George Gervin being the exception), or players from the 80’s on. Why? 3 pointers weren’t available until 1980-81, and that changes TS% dramatically. Also the game changed a bit from the 70’s to the 80’s.

Now, does any of this explain why Kevin Martin was traded, or why the Kings offense is so bad? Not exactly.

First Martin (even though I’m tired of discussing it truthfully). Yes, Sir SpeedRacer was ridiculously efficient in almost all of his time in Sacramento, with the exception of last season. So did it make sense to keep him? With what I know about TS%, efficient (or inefficient offense in the Kings case–excuse me in my homage to the Baby Royals, I have to waste as many words as possible and do more extraneous things I ever used to make a point that I could have made with 500 words less, but I do this because I’m a young stupid writer who thinks that writing a paragraph with sarcasm and how it might be analogy to a young team, and their ridiculous, at times, decision making and overall offense might be a bad idea because some might mistake it as me pontificating for no reason; Oh well) and from what I saw about Kevin’s performance (based on a relatively small sample), I didn’t believe it would be a very beneficial relationship for team or player. That’s why, as much as I love Kevin Martin–it’s gonna take a lot to bump him out of my top 5 favorite players–I feel its’ better that he went to a different team that had the right kind of talent and could allow him to do what he does well. In Sacramento, with him & Tyreke, that just wouldn’t have worked. Geoff Petrie just jumped the gun too quickly for the Martin support contingent. That said, this is not about Martin.

The Kings are inefficient offensively. Why? Well, they have a lot of inefficient offensive players and as I noted above, only 3 players are even average or above average: Beno Udrih, Francisco Garcia and Carl Landry.

One problem is that DeMarcus Cousins and Tyreke Evans, the 2 fellows who use possessions at a very high rate, aren’t very efficient. They’ve gotten better in January and October with Cuz posting a 50.1 TS% for Jan and a 50.5 TS% for Feb (so far). Reke has posted a 51 TS% for Jan and a 53TS% for Feb (so far). Sammy Dalembert has a 57 improved remarkably with a 57.3 TS% for January and a 55.3 TS% for February (so far). (This is very close to his career numbers.) Even Carl Landry has shot 58.6 TS% in Jan and 56.9 TS% in Feb (so far).

So, moving forward, looking at the players with the 6 most amount of minutes averaged per game (Evans, Udrih, Cousins, Casspi, Landry and Garcia)

Just for the unaware: Turquoise is for over league average and red is for below league average. (Silver is neutral.) Hoopdata has all those figures, but for posterity, this is what it says today: USG% 18.85, TS% 54.2%


Sacramento Kings
Minutes Per Game Usage Percentage True Shooting Percentage
Tyreke Evans 39.1 (1) 25.7% (2) 48.1% (6)
Beno Udrih 34.7 (2) 17% (5) 57.7% (1)
DeMarcus Cousins 27.1 (3) 28% (1) 48.3% (5)
Carl Landry 26.5 (4) 20.5% (3) 54.5% (3)
Omri Casspi 25.4 (5) 17.5% (4) 51.3% (4)
Francisco Garcia 23.3 (6) 17% (6) 56.2% (2)

If you can’t see the problem already, it’s that Udrih and Garcia probably don’t get enough shots comparatively to Cousins and Evans. Except, that is what the Kings need Evans and Cousins to be doing. They just aren’t doing it very well. I’ll make this very simple and plain for you: The Kings franchise has essentially created a situation where the 2 best talents on this team (Evans & Cousins) are getting the situation to play star. It hasn’t gone well, and the team has suffered as a result, but the risk isn’t too great when you haven’t made the playoffs since 2006 and you feel that your 2 talented young players can grow with the result of the accelerated playing time. The risk is that Cuz & Reke never grow up; the upside is that they do and become the stars we all hope. The Kings are rolling the dice and gambling that they will become those stars.

But, let’s not go down this road at this point. The reason the Kings struggle is that Evans and Cousins take too many shots and don’t came away with success often enough. As I showed with the monthly breakdown, Landry’s excellent month and a half in TS% has shown him to be close to the level that he was at with Houston, Dalembert has improved dramatically over the course of the season and the last month & half is right at his typical career levels, and even Evans/Cousins have improved slightly as well. Even Donte Greene and Pooh Jeter have had higher TS% than they had earlier in the season. Only Beno Udrih, Francisco Garcia and Jason Thompson (who I left off the list but I could have easily included instead of Garcia) have remained similar (Udrih well above average, Garcia a bit above average, and Thompson a bit below average) throughout the season. One sad note about Omri Casspi is that he’s gotten worse as the season has gone on. (I’m not sure why. It’s maddening for both him, and us as well.) The bad part is that he was average to slightly below average efficiency at his best times during the season.

At any rate, what is the point? The point is that the Kings have a significantly below average offense because there isn’t enough possessions being used at an efficient rate. The most efficient players don’t get as much touches as their efficiency should suggest, and guys like Landry should touch the ball a bit more if anything. What is clear is that the Kings are trying to speed up the education of their youngsters by giving them the minutes (and they are among the best players so it’s tough to say sit them until they learn) now and feel the pain with the tradeoff of glory later. Like I said, it’s a tough process.

I’ve left something out, and I’m glad I’m ending this portion with this: How does this effect the overall Kings situation? Well, like I said they are 26th in ORtg (look in the miscellaneous section), and play at the 8th fastest pace at 94.2. Does this mean the Kings are that fast? Well, the Timberwolves pace is 96.7, and the Golden State Warriors are currently 4th at 94.5. So really there is not much difference in that respect than the Kings (other than the Dubs are more efficiently offensively–which I’ll get to), and the Trail Blazers (the other team) is currently 29th in pace at 88.5 The League average is (right now) at 92.1.

******

I bring this up because I’d like to point out 2 teams, the Warriors and Blazers, are currently not in the top 10 of efficient offensive teams, but just below that. Let’s do the top 6 min’s, USG% and TS% and see how it rolls. I’ll think you’ll notice something.

For the Warriors:


Golden State Warriors
Minutes Per Game Usage Percentage True Shooting Percentage
Monta Ellis 41.1 (1) 28.7% (1) 54.2% (5)
Dorell Wright 39.2 (2) 19.1% (4) 54.7% (3)
David Lee 36.2 (3) 21.3% (3) 53.5% (6)
Stephen Curry 33.5 (4) 24.2% (2) 59.6% (2)
Andris Biedrins 25.5 (5) 10.8% (6) 54.4% (4)
Reggie Williams 21.8 (6) 18.8% (5) 60.6% (2)

Now, it’s pretty easy to see why the Warriors have a better offensive team than the Kings. Everyone on this list is at worst a slightly below average efficient offensive player (David Lee & Monta Ellis), average efficiency (Biedrins and Wright), and Curry & Williams are exceptional when it comes to offensive efficiency. But there is one big difference between this list and the Kings list. The Kings have 3 players in their rotation that were drafted in 2009; only Williams and Curry have less than 2 years experience in the NBA and both are significantly older than Evans and Cousins. (Casspi and Curry are both 22.) Also, Curry/Williams played 3 years in college and that helps them in some ways. (Mostly playing with veterans around them is what does it.) Both Biedrins and Wright have been in the league since 2004 and Lee and Ellis came into the NBA in 2005. Hopefully you’re seeing where I’m going with this.

The TrailBlazers work at a snails pace, 2nd slowest in the NBA (the slowest team is the Pistons), and are identical to the Warriors offensively. They just do it differently. (I’m going to ignore Brandon Roy for this exercise. In years past when he was better, they were better offensively.) Here’s the table

Portland Trail Blazers Minutes Per Game Usage Percentage True Shooting Percentage
LaMarcus Aldridge 39.1 (1) 26.3% (1) 55.1% (2)
Wesley Matthews 34.2 (2) 22.2% (2) 57.1% (1)
Andre Miller 32.8 (3) 21.2% (3) 53.1% (5)
Nicolas Batum 31.9 (4) 17% (5) 54.8% (3)
Marcus Camby 28.7 (5) 12.6% (6) 45.8% (6)
Rudy Fernandez 23 (6) 19.3% (4) 53.6% (4)

The Blazers are unique in that the guys who use a heavy bulk of their possessions consistently are all somewhere in between slightly below or above average with the exception of Wesley Matthews who has a very good TS%. (It’s not excellent or great, but it’s very good.)

You know why I don’t look TS% very much in the case of the Blazers? This is one way TS% can get dangerous to rely on, and this is where the 4 Factors come into play. The 4 factors are an unique thing, and here is why it’s considered important. The 4 Factors were developed by Dean Oliver, author of Basketball on Paper. These are the most important factors in order: Shooting is the most important factor, followed by turnovers, rebounding, and free throws. (Note: Editor of B-Ref agree’s with the order but not the weighting of the formula’s; you’ll have to click on the link to do all that stuff as there is too much here to go into that as well.)

Let’s take a look at all 3 teams again.

Pace ORtg DRtg Offense 4 Factors eFG% TOV% ORB% FT/FG
League Average 92.1 107 107 49.7% 13.5% 26.3% 23.2%
Sacramento 94.2 (8) 102.9 (26) 108.4 (17) 47.3% (26) 14.2% (25) 30.2% (2) 21% (28)
Golden State 94.5 (5) 108.3 (11) 111 (27) 51.4 (8) 13.8 (21) 27.1 (9) 18.4 (30)
Portland 88.5 (29) 108.1 (11) 107.2 (15) 48.2% (25) 12.8% (4) 29.9% (4) 22.5 (18)

(H/T to Evan Dunlap at OPP) for the color coding idea.)

So why do I use eFG% for teams and not players? Because in general, eFG% doesn’t tell the whole story. It just shows how you shoot from the field, and not every player is equal in that regard. (Some players aren’t great FG% or 3pt FG% like Devin Harris but still have a slightly higher than average TS%. Why? He gets to the line.) But for teams to show whether they shoot well? eFG% works great in that regard.

Wanna know why the Kings are inefficient? Well, you can see the table right there. The Kings hit a low amount of their shots, turn the ball over too much, don’t get to the line enough, and as a result get a ton of offensive rebounds.

Portland is a team that doesn’t turn the ball over much, gets a lot of offensive boards and plays slow to accentuate this. That’s why they’re the 11th team in ORtg.

Golden State is a team that runs, hits a lot of their shots, gets a surprising amount of offensive boards (which helps them score more points), and doesn’t have a godawful turnover% (they are just a bit below league average). They just don’t get to the line primarily because they shoot a ton from the outside.

Now Sacramento? We know they don’t shoot well, turn the ball over too much and probably get to the line less than they should. Some of that has to do with Evans’ injury, and some of that is that a lot of the Kings players just don’t get to the line often enough (Omri Casspi especially). What’s the fix? Well……….

That’s an answer for another day. Isn’t this long enough?


Responses

  1. [...] said, I’ve talked about shot creation in the not so distant past and how the Kings are so bad at it. Last night illustrated how much easier it is for the Kings when Beno Udrih doesn’t have to [...]

  2. [...] at the All-Star break I had a post about offensive efficiency for each team and why the Kings need more efficient offensive players. Well, Marcus Thornton, after 15 games granted, is leading the Kings in TS% on the season. With [...]


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

Follow

Get every new post delivered to your Inbox.

%d bloggers like this: