Comments on: The Remains of Walk-offs

By: Chad

Chad — Fri, 22 Feb 2013 00:17:02 +0000

I believe the Brewers blew somewhere around 20 games last year. As the article points out, Jim Henderson had a 3.23 era but a lot of walk off losses most of which were were with lots of baserunners on because of Axford and Veras.

By: Tim

Tim — Thu, 21 Feb 2013 02:35:20 +0000

This is rather silly given the drastic tactical differences present in potential walk-off situations.

By: Zac

Zac — Thu, 21 Feb 2013 00:14:46 +0000

This is a reminder that all the old stats that old writers love are subject to their own biases that can pervert their value in evaluating true talent. Average and OBP in small sample sizes are subject to the luck of BABIP. ERA in small sample sizes is subject to this, where a player isn’t penalized enough for the runners he allows on base because he gave up the game winning hit. Wins aren’t even worth discussing, and everyone should be aware of how pointless a 3-run save is. Runs and RBI (excluding those coming from home runs) require the assistance of a player’s teammates, and Home Runs are affected by the ballpark in which a player plays.

This may sound like nitpicking, but this is the same thing writers do when they tell you that WAR is useless (probably with a headling that references an Edwin Starr song).

By: TKDC

TKDC — Wed, 20 Feb 2013 22:00:37 +0000

Using ERA for relievers is generally a bad idea, largely for this reason (along with the small sample size). I think K% and BB% are good, but you can also look at FIP or xFIP.

By: John

John — Wed, 20 Feb 2013 21:32:34 +0000

Right. Distinguishing between an outing that ends due to walk-off and one that ends due to a pitching change seems somewhat arbitrary, except that in some cases the game ending prevents runs that a pitching change can’t.

Is there an advanced pitching metric that produces an ERA-like (or xFIP-like) number assigning some penalty to runners left on?

By: John

John — Wed, 20 Feb 2013 21:27:43 +0000

Really interesting article; thanks for the great work.

By: sambf

sambf — Wed, 20 Feb 2013 21:10:09 +0000

In a similar vain to this, starters are going to have slightly worse ERAs than they “should” because they’re often pulled with 1-2 outs–missing parts of innings that generate artificially good ERAs (because runners you allow are less likely to score); similarly that’s going to artificially make relievers’ ERAs slightly better on the whole.