Quote:
Originally Posted by jaa36
Thanks for publishing all this.
...
HOWEVER- in practice, I also found that it was extremely difficult to use these numbers, because "real life" OOTP teams have such varied defenses and you aren't comparing to 50 or 55. Not to mention that for many players in practice, they might have 70 range but not be maxed out on experience at the position, which impairs their performance. So in modeling defense for my projection system, I actually found it considerably simpler (and reasonably accurate) to just base it on the position ratings rather than the individual component ratings. This was all calculated in OOTP 25 in a "real life" league, essentially figuring out how many runs per game a 25 SS vs a 55 SS would end up with in ZR (and incorporating arm and benefit to runs allowed for catchers). The position ratings are heavily driven by range anyway, which is the most important component. I found that from a 20 to 80 defender you would end up with the following ranges of values, in terms of runs per inning, from worst to best:
C .039 (53 per 150 games/1350 innings)
1B .009 (12)
2B .033 (45)
3B .025 (34)
SS .045 (61)
LF .024 (32)
CF .040 (54)
RF .028 (38)
In most cases, the "cut point" for where a player is accruing positive defensive value is around 55- so a player less than 55 at that position will end up with negative defensive value, while a player above that will end up with positive value. Of note, this is likely not true in an online league where savvy players are not running out 40 center fielders and are prioritizing good defensive players- in that setting, you probably need at least 60-65 at a position to end up with positive defensive value.
|
I am trying to utilize this (and the OP's work) to optimize a lineup, was hoping you could explain a bit further. Specifically, how do I translate the above into defensive runs or WAR? For example, let's say I have 4 different players:
1B #1: 50 skill
1B #2: 60 skill
CF #1: 45 skill
CF #2: 65 skill
How would the above translate to runs / WAR?
EDIT:
Not sure if my math is right here, but assuming that a 80/80 catcher would be worth +53 runs (and 5.3 wins), and a 20/80 catcher would be worth -53 runs (and -5.3 wins), I come up with the following:
1B, 50 skill: 0 WAR (Replacement Level)
1B, 60 skill: 0.4 WAR
CF, 45 skill: -0.9 WAR
CF, 65 skill: 2.7 WAR
EDIT 2: Unless my math is wrong. It looks like if I take 1,350 innings and divide by 53, I get 0.039. So that indicates the range difference between a 20 and 80 catcher. Since this is a bell curve, does that mean you can have +26.5 runs at the top (+2.6 wins), and -26.5 runs at the bottom (-2.6 wins)?