THE BOOK--Playing The Percentages In Baseball

Will do!  Probably not tonight, though.  Happy New Year!

Done!  New post up.

Good stuff:

http://sabermetricresearch.blogspot.com/2012/01/do-nhl-referees-call-make-up-penalties.html

Only had time to get through about half of Phil’s new post, but I’m curious about why winning teams take more penalties. This seems counterintuitive. Winning teams have to take fewer chances, fewer risky plays, and a winning team is often winning because they’re better (and it seems reasonable to assume better teams commit fewer penalties than lesser teams).

Leading teams don’t TAKE more penalties (necessarily).  More penalties are CALLED on the leading teams.  That’s the only point of fact there is.

I wonder how this works as the quality of refs decreases (i.e., do collegiate/juniors/amateur refs show this bias more?).  You’d have to compensate for what’s presumably a larger skill disparity in player talent on lower-level teams though…

I played hockey through high school, and I really think I have memories of refs explicitly telling us that there would be/was a make-up call.  I wonder if as refs get worse, they have more errors to make up for.

Nice piece, Phil.

Phil gives us the regression equation:

http://sabermetricresearch.blogspot.com/2012/01/do-nhl-referees-call-make-up-penalties_06.html

The key point is when the score is tied.  If it’s a tie game, the next penalty called is against the home team ___ often, if the previous penalty was against ___:

35%, if previous against home team
61%, if previous against visiting team

That is one huge even-up effect.

Now, whether it’s the compassionate referee, or the teams play more careful, or everyone is aware of the differing standards, who knows.

But, there is an even-up effect for SOME non-random reason.

Or to put it more clearly, in a tie game, the team that gets the next penalty will have gotten the previous penalty only 37% of the time.  63% of the time, the penalty called will have been an “evened-up” call.

We’ve all know that the even-up effect exists (again, we don’t know if it’s the ref, the style of play, or an understanding that the implicit rules have changed).

What is a huge finding here is to have it quantified finally.

It must be something real ... it’s so big compared to the other effects, like PP goal and home/road.  The only thing that competes with it is score.

Still, having taken the last penalty is even bigger an effect as going from 3 goals down to 3 goals up.

Maybe I should check my programming again ...

How hard would it be to compare the rates by referee?

If we’re only looking at pre-1998, then at least we don’t need to worry about referee pairings.

Did that in the first post.

It’s definitely SOMEthing real.  Now we have to figure out WHAT it is, be it:
a. compassionate referee
b. previously penalized team being more careful
c. team not previously penalized being more aggressive

Or any combination of the above.