Latrinsorm
09-22-2013, 02:09 PM
Here are the top 10 NBA players in points per game in 2013.
28.7 - Carmelo Anthony-NYK
28.1 - Kevin Durant-OKC
27.3 - Kobe Bryant-LAL
26.8 - LeBron James-MIA
25.9 - James Harden-HOU
23.2 - Russell Westbrook-OKC
22.9 - Stephen Curry-GSW
21.2 - Dwyane Wade-MIA
21.1 - LaMarcus Aldridge-POR
19.4 - Brook Lopez-BRK
We all know Durant is a much, much better scorer than Anthony, but how might we make this case concisely? I've mentioned before (http://forum.gsplayers.com/showthread.php?84403-NBA-Advanced-Stats-and-Efficiency) how advanced metrics penalize players to varying degrees for missing, but each carries a lot more information than just scoring. We can follow the same reasoning, though:
1. Missing a FGA is bad, because most (~70%) of the time it gives the other team possession.
2. Missing the last ordinary FTA is bad for the same reason. How many FTAs are the last? We can take a cue from Usage rate and say ~44% of them. We'll also take into account that it's a lot easier to get defensive rebounds on FTAs (~85%).
3. Turnovers are a tough one. Passing turnovers nominally have nothing to do with scoring ability, but a lot of these young whippersnappers in the NBA today will get caught in the air and desperately pass out of it: that's a scoring move that went really bad just like a charging foul. Then there's traveling or 3 seconds or 5 seconds. Because it's hard to decide which way to go and because they are such relatively infrequent events (more than 6 times as many scoring possessions in categories 1 and 2) I'm going to leave them out at least for now.
So our provisional formula is PTS - (FGA - FG) * .7 - (FTA - FT) * .44 * .85. Like PER, it's still very easy to get positive results here: FG% only needs to hit 26% (!), FT% only needs to hit 27% (!!!). This is how the top 10 shakes out for 2013:
21.8 - Kevin Durant
20.7 - LeBron James
19.6 - Carmelo Anthony
19.2 - Kobe Bryant
18.6 - James Harden
15.9 - Stephen Curry
15.3 - Russell Westbrook
15.3 - Dwyane Wade
15.2 - Kyrie Irving
15.0 - Tony Parker
Pretty interesting! We have a very high incidence of point players and playmakers (I count 7), which could suggest we should squeeze turnovers in, or it could suggest that these particular playmakers happen to be very good scorers. Also, Irving and Parker didn't qualify for the scoring title, and LA and Lopez only fall to 11 and 12 in the new system, so we haven't really changed the order too terribly much. But how would this change the historical number of scoring titles? Until next post!
28.7 - Carmelo Anthony-NYK
28.1 - Kevin Durant-OKC
27.3 - Kobe Bryant-LAL
26.8 - LeBron James-MIA
25.9 - James Harden-HOU
23.2 - Russell Westbrook-OKC
22.9 - Stephen Curry-GSW
21.2 - Dwyane Wade-MIA
21.1 - LaMarcus Aldridge-POR
19.4 - Brook Lopez-BRK
We all know Durant is a much, much better scorer than Anthony, but how might we make this case concisely? I've mentioned before (http://forum.gsplayers.com/showthread.php?84403-NBA-Advanced-Stats-and-Efficiency) how advanced metrics penalize players to varying degrees for missing, but each carries a lot more information than just scoring. We can follow the same reasoning, though:
1. Missing a FGA is bad, because most (~70%) of the time it gives the other team possession.
2. Missing the last ordinary FTA is bad for the same reason. How many FTAs are the last? We can take a cue from Usage rate and say ~44% of them. We'll also take into account that it's a lot easier to get defensive rebounds on FTAs (~85%).
3. Turnovers are a tough one. Passing turnovers nominally have nothing to do with scoring ability, but a lot of these young whippersnappers in the NBA today will get caught in the air and desperately pass out of it: that's a scoring move that went really bad just like a charging foul. Then there's traveling or 3 seconds or 5 seconds. Because it's hard to decide which way to go and because they are such relatively infrequent events (more than 6 times as many scoring possessions in categories 1 and 2) I'm going to leave them out at least for now.
So our provisional formula is PTS - (FGA - FG) * .7 - (FTA - FT) * .44 * .85. Like PER, it's still very easy to get positive results here: FG% only needs to hit 26% (!), FT% only needs to hit 27% (!!!). This is how the top 10 shakes out for 2013:
21.8 - Kevin Durant
20.7 - LeBron James
19.6 - Carmelo Anthony
19.2 - Kobe Bryant
18.6 - James Harden
15.9 - Stephen Curry
15.3 - Russell Westbrook
15.3 - Dwyane Wade
15.2 - Kyrie Irving
15.0 - Tony Parker
Pretty interesting! We have a very high incidence of point players and playmakers (I count 7), which could suggest we should squeeze turnovers in, or it could suggest that these particular playmakers happen to be very good scorers. Also, Irving and Parker didn't qualify for the scoring title, and LA and Lopez only fall to 11 and 12 in the new system, so we haven't really changed the order too terribly much. But how would this change the historical number of scoring titles? Until next post!