RealGM

Can't speak for methodology, analysis or presentation, but some diligent RealGM poster (SCassell19) has put this site up, containing some per-game metrics (for Milwaukee only at this stage) sorted by referee.

http://www.nbarefstats.com/

This is certainly statistics with an agenda, but a bigger dataset including other teams might allow greater objective scrutiny of referee performance without resorting to "OMG WE WUZ ROBBED!" He's also asking for input to the site...

wow

edit: dunno how accurate it is just amazed someone is actually doing this

Id love it if it was done from a less biased perspective and across the whole NBA. Using objective data too and not "questionable win" type stuff. Over a couple years I imagine the trend of a paticular ref not giving calls to a paticular player would come out if it exsisted.

Probably too difficult to break down on a player basis, but team-by-team I can see it as a worthy endeavour.

Jimmy76 wrote:wow

edit: dunno how accurate it is just amazed someone is actually doing this

Id love it if it was done from a less biased perspective and across the whole NBA. Using objective data too and not "questionable win" type stuff. Over a couple years I imagine the trend of a paticular ref not giving calls to a paticular player would come out if it exsisted.

I made the site. If you scroll under the stats you will see a key.

QUES GAMES: A QUESTIONABLE GAME IS WHERE THE WINNING TEAM SHOOTS AT LEAST 10+ MORE FTS AND THE FT DIFFERENTIAL IS EQUAL TO OR GREATER THAN THE WINNING MARGIN.

All of the stats are based on a statistical criteria and not opinion.

The stats should be pretty accurate. In the process of logging FT counts from hundreds and hundreds of games its possible a mistake or two could have been made somewhere but the error margin here if any is very small because I put great care into making sure the numbers were correct.

The Philadelphia stat guy used to do a nice one showing referee bias toward the home team (pretty much every ref in the NBA except Mendy Rudolph if I remember with some having big swings) but I understand the NBA made him quit publishing it. Fun stuff scassell, thanks!

SCassell19 wrote:

Jimmy76 wrote:wow

edit: dunno how accurate it is just amazed someone is actually doing this

Id love it if it was done from a less biased perspective and across the whole NBA. Using objective data too and not "questionable win" type stuff. Over a couple years I imagine the trend of a paticular ref not giving calls to a paticular player would come out if it exsisted.

I made the site. If you scroll under the stats you will see a key.

QUES GAMES: A QUESTIONABLE GAME IS WHERE THE WINNING TEAM SHOOTS AT LEAST 10+ MORE FTS AND THE FT DIFFERENTIAL IS EQUAL TO OR GREATER THAN THE WINNING MARGIN.

All of the stats are based on a statistical criteria and not opinion.

The stats should be pretty accurate. In the process of logging FT counts from hundreds and hundreds of games its possible a mistake or two could have been made somewhere but the error margin here if any is very small because I put great care into making sure the numbers were correct.

might just be that the Bucks dont take many fts

i dunno maybe you took that into account

interesting regardless

Ten free throws aren't that much if one team is constantly attacking the rim and manages to draw quick fouls on the other team early in the quarter. A foul equals 2 free throws, so 10 FT are just 5 fouls more for the entire game. Very possible to happen.

Jimmy76 wrote:

SCassell19 wrote:

Jimmy76 wrote:wow

edit: dunno how accurate it is just amazed someone is actually doing this

Id love it if it was done from a less biased perspective and across the whole NBA. Using objective data too and not "questionable win" type stuff. Over a couple years I imagine the trend of a paticular ref not giving calls to a paticular player would come out if it exsisted.

I made the site. If you scroll under the stats you will see a key.

QUES GAMES: A QUESTIONABLE GAME IS WHERE THE WINNING TEAM SHOOTS AT LEAST 10+ MORE FTS AND THE FT DIFFERENTIAL IS EQUAL TO OR GREATER THAN THE WINNING MARGIN.

All of the stats are based on a statistical criteria and not opinion.

The stats should be pretty accurate. In the process of logging FT counts from hundreds and hundreds of games its possible a mistake or two could have been made somewhere but the error margin here if any is very small because I put great care into making sure the numbers were correct.

might just be that the Bucks dont take many fts

i dunno maybe you took that into account

interesting regardless

That could be true, but it's a system that can be employed for any game between any two teams with any three refs. If there is intentional or unintentional bias, it seems like a system that might help reveal it, regardless of any one team's tendancies.