>The difference between Matt Boland and Mark Lindsay is 0.34 errors per game, which would equate to an extra error every third game more or less.
I feel like when people disparage a referee, it's usually not because of their perceived average error rate, but rather their *inconsistency* e.g. "If so-and-so called that on one end, they should call it the same way on the other". In statistical terms, wouldn't that make the variance of a referee's errors in a L2M report more enlightening than their average errors per game?
Do you have access to the raw data so we could see the average standard deviation of errors per game and how each referee's individual variance compares?
>The difference between Matt Boland and Mark Lindsay is 0.34 errors per game, which would equate to an extra error every third game more or less.
I feel like when people disparage a referee, it's usually not because of their perceived average error rate, but rather their *inconsistency* e.g. "If so-and-so called that on one end, they should call it the same way on the other". In statistical terms, wouldn't that make the variance of a referee's errors in a L2M report more enlightening than their average errors per game?
Do you have access to the raw data so we could see the average standard deviation of errors per game and how each referee's individual variance compares?
Thanks for your comment, and that's a great point that you're making.
I do have the raw Data, and have made the quick calculations to get the standard deviation for each ref. Here is the table these exact refs:
OFFICIAL ERRORS GAMES ERROR_STD ERROR_G
Matt Boland 274 127 1.342375 2.157480
Josh Tiven 278 132 1.629545 2.106061
James Williams 311 150 1.178115 2.073333
Tony Brothers 281 137 1.160928 2.051095
Kevin Scott 264 129 1.219102 2.046512
Eric Lewis 267 132 1.400573 2.022727
Zach Zarba 258 130 1.767156 1.984615
Courtney Kirkland 260 132 1.127745 1.969697
Marc Davis 265 135 1.021309 1.962963
Brian Forte 249 128 1.048128 1.945312
Gediminas Petraitis 250 130 1.042301 1.923077
Scott Twardoski 251 131 1.063181 1.916031
Mitchell Ervin 264 140 1.000612 1.885714
Nick Buchert 234 128 0.919488 1.828125
Mark Lindsay 233 128 1.078436 1.820312
And here is the table for top 15 refs sorted by of standard deviation descending (and at least 100 games):
OFFICIAL ERRORS GAMES ERROR_STD ERROR_G
Eric Dalen 218 102 2.039027 2.137255
Leroy Richardson 227 101 1.982128 2.247525
Zach Zarba 258 130 1.767156 1.984615
Michael Smith 246 113 1.684022 2.176991
Scott Foster 254 117 1.650126 2.170940
Josh Tiven 278 132 1.629545 2.106061
Tre Maddox 280 114 1.622973 2.456140
Justin Van Duyne 232 120 1.526070 1.933333
Curtis Blair 235 119 1.457880 1.974790
Eric Lewis 267 132 1.400573 2.022727
Derrick Collins 267 124 1.362187 2.153226
Scott Wall 228 115 1.351216 1.982609
Matt Boland 274 127 1.342375 2.157480
Marat Kogut 257 127 1.276465 2.023622
Sean Wright 239 118 1.265325 2.025424
I apologize if the formatting is odd, I can't post images here unfortunately