Average players

[Mike G]

Using the b-r.com Player Season Finder, there were 25 players in 2015-16 with PER between 14 and 16, WS/48 in the .085 to .115 range.
Of these, 10 also fell within -1 and 1 in BPM

Code: Select all

? avg player      tm   PER   WS/48  BPM
Ersan Ilyasova   tot   14.2  .113   0.1
C. J. Miles      Ind   14.0  ,103  -0.2
Darren Collison  Sac   15.9  .113  -0.4
Tyson Chandler   Phx   14.7  .112  -0.5
Jeremy Lamb      Cha   15.9  .107  -0.5

Toney Douglas    NOP   14.7  .094   0.5
Goran Dragic     Mia   15.5  .096   0.7
Rodney Hood      Uta   14.1  .112   0.7
Evan Fournier    Orl   14.6  .106   0.7
Richan Holmes    Phl   15.9  .103  -0.9

Are these all average players? Three metrics agree they are.

For whole career, in the 3-point era, if we shrink the range of 'avg' to 14.66 < PER < 15.33, and .095 < ws/48 < 1.05, we get these (ranked by absolute BPM)

Code: Select all

avg player       years     PER  WS/48  BPM
Khris Middleton  '13-16   14.8  .096   0.0
Leandro Barbosa  '04-16   15.3  .101  -0.2
Vern Fleming     '85-96   15.1  .101   0.3
Jameer Nelson    '05-16   15.0  .100   0.4
Edgar Jones      '81-86   15.1  .103  -0.4

Vinnie Johnson   '80-92   15.2  .100   0.5
Kenny Carr       '80-87   15.1  .102  -0.6
Terry Mills      '91-01   14.7  .097  -0.7
Steve Johnson    '82-91   15.3  .095  -0.7
Kenyon Martin    '01-15   15.0  .096   0.7
G Antetokounmpo  '14-16   15.3  .096   0.7

[Crow]

For last season RPM had Ilyasova as below average with an estimate of about -2.3.
CJ Miles at -1.6.
D Collison at -0.7 is closer to "neutral'.
Chandler at -0.5.
Lamb +1.

Neutral on RPM is mean performance, I believe. Not sure what "average" as median is on RPM but it might be -1, give or take a bit. So some difference between the boxscore metrics (with team shot defense or no shot defense and no measure of help / hurt of team offense beyond the boxscore), but not a huge amount. Some of the difference is real and some may be "error" on the part of RPM, the boxscore metrics or both.

fwiw, RPM had Ilyasova halfway back to neutral the previous season. Miles was estimated about as positive the previous season compared to last season so the average would be about 0. Collison was +1.7, so his average for 2 seasons would be only +0.5. Lamb was estimated real bad in 2014-15, Chandler real good. Maybe they changed a whole lot but maybe errors broke both ways. Maybe Lamb was neither that bad or that good and below average on average. Maybe instead of great or near average Chandler is in between. These are guesses, as players can indeed improve or decline- especially in different environments- and not always in same direction or as expected; but maybe you can guess better (or at least have more reason for caution and compromise) knowing boxscore metrics and RPM instead of just one.