PDA

View Full Version : Tips for evaluating Best Single Season a player has had



powersackers
07-10-2012, 02:36 PM
Can you guys give me a few tips for how you would evaluate the best single season a player as had in his career? Consider Def as it relates to the value of the season as well.

Let's use Bobby Abreu as an example.

http://www.baseball-reference.com/players/a/abreubo01.shtml

Using WAR or OPS+ seems simplest, but which is best?
OPS+ says 2002, 1999, 2004 are his best seasons
WAR says 2004, 1998, 1999 and 2000 are his best seasons

Thanks.

dougdirt
07-10-2012, 02:41 PM
WAR tries to account for everything he does. OPS+ only tries to account for what he did as a hitter. It doesn't account for base running or defense. WAR is an easy answer here. With that said, WAR isn't perfect and has some problems of its own, but if I had to choose between the two, I would go with WAR even if it was the inferior BR version of it.

powersackers
07-10-2012, 02:48 PM
I think I have to use b-r version for my study because fWAR isn't available for comparing players that played long ago. Is that still true?

nate
07-10-2012, 02:56 PM
I think I have to use b-r version for my study because fWAR isn't available for comparing players that played long ago. Is that still true?

If you're just finding a player's best season, why do you need to compare him to other historic eras?

I like fWAR because unless you're only looking at DHs or pitchers, defense is a factor and you have to measure it somehow.

jojo
07-10-2012, 03:09 PM
The first thing you need to do is define what "best" means for your season. Then you can tackle the issue of which metric/metrics to use for the evaluation.

Probably summary statistics will be most useful for you. For instance, WAR will give you an estimate of a position player's total value (offense+defense). In the case of Abreu, fWAR suggests his best overall season was 2000. wOBA suggests his best offensive season was 1999. UZR suggests his best defensive season was 1998. That said, Abreu is an interesting case study to pick as an example because he has had some extreme seasons defensively as measured by UZR. So you have to consider some type of error or uncertainty in the metric. For instance is his 7.2 WAR season really different than hi 6.7 WAR season? It's half a win and that seems large but fangraphs rates his defense between +28 runs at it's best and -21 runs at it's worst which is a huge range. One could ignore the UZR data and argue Abreu was a +10 run glove over years '98 thru '03 and a -10 run/yr glove over years '04 thru now. That would make either '99 or '02 his "best" seasons. Nonetheless, advanced metrics suggest his peak years were defined by 1998-2004.

powersackers
07-10-2012, 03:58 PM
If you're just finding a player's best season, why do you need to compare him to other historic eras?

I like fWAR because unless you're only looking at DHs or pitchers, defense is a factor and you have to measure it somehow.

Compare was the wrong word. I just want the same metric applied to the best single season analysis. Whether it be Abreu or the other 999 players and pitchers/DHs I am looking at dating back to the 1920s.