Sunday, June 7, 2009

Awkard Position - Comish with a .781 Wpct

It's a love/hate thing for sure. I love ATB and I try my darnedest to win as many games as possible each year, but at the same time I hate it when my team is successful. As comish in a centralized league - I am the only one who has full access to all the data and am in 100% control of just about everything. No one has accused me of any impropriety, but I always fear the minds of a few owners veer down this line of thought anyway.

Nothing I can do about it normally, but after a 4-0 sim and another stellar run of pitching success, I do think it's time I share the belief that I've gone a long way to cracking the DMB code on how to value pitchers.

Let me first say I have absolutely no hard evidence of this, I just know that the last two seasons I have found gems such as Bobby Shantz, Taylor Bucholz, Ellis Kinder, Derek Lilliquist, and Joe Neale.
  • Last year, battling injuries, Shantz was 12-3 with a 3.22 ERA.
  • Bucholz pitched 50 relief innings with a 2.90 ERA
  • Kinder 41 relief innings with a 2.18 ERA (1.64 ERA this year)
  • Lilliquist 70 innings of 4.24 earned runs given up every 9 innings (0.46 ERA this year)
  • And Neal, my find of this year, has a 0.79 ERA in 23 innings so far.
Maybe it is dumb luck, maybe it is the park I am playing in, or maybe it is my formulas used. Likely it is a combination of all.

Anyway, here's my formula for drafting pitchers:
  • Normalize the park factors. Maybe normalize isn't the right word, but I think it too much to take a 90 park factor and simply reduce the number of hits given up by a pitcher by 10%. So, I cut the factors in half. A 90 becomes 95, and 80 becomes 90, a 110 becomes a 105 etc.
  • Recalculate hits given up using the park factor score, only for the opposite hand. By this I mean, if it is a righty, adjust his hits given up only taking into consideration the left handed park scores. And the opposite for lefties, only use the right handed scores. For example, using righties: Take hits given up, multiply it by .4, and multiply that number by the left handed singles park factor. Then, multiply hits given up times .6 and add it to the first number. Notice, I did use the right handed park factor at all, and I use 40% and 60% as a quick swag of the number of lefty vs righty batters a typical pitcher faces. The formula: Hits*0.4*LH Singles Park Factor + Hits*0.6. 1931 Bill Walker, who played in the Polo Grounds with 70 singles park factor versus lefties, sees his 212 real life hits soar all the way up to 231
  • Repeat this same process for HR's, using the HR park factors
  • Calculate new WHIP's and HR/9 rates using these numbers
  • Compare this WHIP vs the league average WHIP (Lg Avg - New WHIP)/Lg Avg
  • Do the same for HR's
  • Do the same for K's
  • Do the same for ERA, just the real life ERA compared to league average ERA
  • Then, come up with an overall score: WHIP Score *10 + HR Score+ERA Score *0.25 + K Score * 0.05
In other words, my formula counts WHIP 10 times as much as HR rate, 40 times as much as ERA, and 200 times as much as strike out rates.

Create your own formula's anyway you see fit, but there is no more important stat than WHIP, and that WHIP must be adjusted for park factor, and this must be compared against the league average.

There, I said it. I am not cheating, just spend way to much time playing diamond mind baseball! (Now of course, I will go on a 10 game losing streak after this major jinx.)

No comments:

Post a Comment