True Shooting% and Effective FG%

The incredible stat is that Martin's eFG% is only 41.8%, yet his PPFGA(Hollinger method) is the highest on the current team. Only Carl Landry's (ok, and Mike Harris' :grin was higher, but Carl's eFG% was 13% higher. If Martin consistently gets his FG shooting to be good, his numbers are going to be downright silly. We're dealing with a small sample size, but still, one can dream I suppose.

Here's a thought... when was the last time the Rockets had a player like this? I hope I didn't just jinx him. Based upon tonight's game, I don't think I did.

 

was von wafer close? just curious.

 

Is Shane's eFG the highest because he shoots almost exclusively 3s?

 

Your work is greatly appreciated. Thank you.

 

For reference here are tmac and landry's numbers:

tmac
efg .474
ts .532
per 18.5

landry
efg .494
ts .588
per 16.7

Landry eff numbers have dropped since he left the rockets.

 

Are these season numbers or career numbers?

 

They're season numbers - look at the basketball-reference.com link at the bottom of his post. If you want the career numbers, you can get them by clicking on the individual player names on basketball-reference.com.

 

hey durvasa,

do you know where I can get the actual field goal attempts including the ones with fouls? I searched in hoopdata but no luck, maybe I overlooked it?

thanks

 

You have to just go through the play by plays and count the number of plays where there is a shooting foul called but the shot doesn't go in (or a defensive foul with the team in the penalty). .44*FTA is believed to be a good enough estimate for these scenarios.

 

Here's the thing though, if you look at shooting % in the clutch on 82games, you'll find that Landry is shooting above 60%, brooks is in the 30s, and EVERYONE else who matters is below 30%.

Looking at averages doesn't reveal the true story, clutch shooting % is more accurate of what happens when the defensive intensity picks up and teams force players away from their favourite spots/plays/sets etc...

 

Maybe I'm missing something but I don't know why they are making an estimation for something that we can have the exact value for.

I'm trying to understand the 0.44

If the player shoots two FTs that means he missed, so 0.44*2=0.88 shot attempts
if the same guy shots another one that means he made, 0.44*1=0.44 shot attemps (but the attempt is in the stat since he made it).

real shot attempts is 2.

So shot attempts after the 0.44 adjuster is (1 already in stat) + 0.88+0.44 = 2.32.

So, is the extra 0.32 because they are saying that he will miss more than make when he is fouled? Ex

Misses 2: 4FT*0.4=1.76
Makes 1: 1FT*.44=0.44
Total = 2.2
Added atempt for make =1

Grand total = 3.2
compared to actual = 3.

So the calculated number is approaching the actual the more he misses. Also I'm neglected 3's for simplicity. I don't think technical free throws should be added because almost always only one player shoots the technical and it doesn't depict his play making skills of creating a shot.

 

Because it is easier to use the boxscore than to parse the play-by-plays for information.

The extra 0.32 would be estimation error. If there exists a player who gets an and-1 50% of the time he gets fouled on a shot over a decent sample of games, then TS% will probably not do a good job of estimating his true shot attempts. But that is very unlikely.

Consider just the Rockets.

http://www.hoopdata.com/scoringstats.aspx?team=HOU&type=tot&posi=%&yr=2010&gp=0&mins=0

You can see and-1 stats there. Here is a summary:

Player              G      MP     FG     FGA    FT     FTA    And1   PTS    ts     ts1    diff
Aaron Brooks        60     2152   423    979    188    233    19     1185   1082   1086   -4
Trevor Ariza        53     1997   301    785    115    175    18     817    862    864    -2
Shane Battier       59     1953   162    403    70     96     8      484    445    447    -2
Luis Scola          60     1816   372    727    136    176    18     881    804    806    -2
Carl Landry         52     1415   302    552    235    280    34     839    675    675     0
Chuck Hayes         60     1280   122    265    23     49     7      267    287    286    +1
Kyle Lowry          48     1186   125    308    145    182    18     422    388    390    -2
Chase Budinger      52     1066   164    393    53     69     5      444    423    425    -2
David Andersen      59     859    141    328    46     65     5      353    357    358    -1
Kevin Martin        7      250    43     108    57     62     2      153    135    138    -3
Jermaine Taylor     20     142    21     54     19     23     3      61     64     64      0
Garrett Temple      9      118    13     29     16     24     2      45     40     40      0
Jared Jeffries      6      113    15     30     5      11     1      35     35     35      0
Joey Dorsey         7      54     5      11     1      2      0      11     12     12      0
Tracy McGrady       6      46     7      19     4      6      0      19     22     22      0


For the table above:

ts = FGA + 0.44*FTA
ts1 = FGA + (FTA - and1)/2
diff = ts - ts1

You can see from the last column that, typically, ts < ts1 but still its pretty close. And ts1 should overestimate a bit on average, because it doesn't take into account free throws off technicals or the extra free throw when a player is fouled on a 3-point attempt.

 

Great work there. However, I hate to be devil's advocate, but that 1994 team was an exception. Very, very rarely does a team with one superstar win a championship. Also, I believe that in the playoffs, nearly every Rockets shooter was shooting above his season average for a stretch.

As a result, I think it would hard to extradite good comparison from the '94 team data. I think looking at the '95 team data, mainly playoffs would give a better barometer.