baseball systems....

[adrian_tman]

What system are you using...

my current system is best described by this formula for SP

(ip)*3+k's-H-BB/GP

15=best
13-14=average
12=worst

[Wheell]

So you are using a dips based system for evaluating pitchers. I prefer a park effect based OPS allowed relative to OPS faced. To each his own.

[rake922]

So you are using a dips based system for evaluating pitchers. I prefer a park effect based OPS allowed relative to OPS faced. To each his own.

hows that working out for you

[Wheell]

51-37 for 10.58 units. Went 0-7 today.

[youngsc]

if you don't mind telling, where did you come up with the formula you listed above? Just curious as I'm trying to develop my own system and am intrested as you are to know what others are doing. TIA

[BadBeatBodog]

So you are using a dips based system for evaluating pitchers. I prefer a park effect based OPS allowed relative to OPS faced. To each his own.

I'm not sure if you could find it but ever consider using GPA instead of OPS?

[Wheell]

I'm not sure if you could find it but ever consider using GPA instead of OPS?

Not all of these guys went to college.

Also, I can't find any sabermetric system that allows me to deal with historical anomolies like Tom Glavine.

[BadBeatBodog]

Not all of these guys went to college.

Also, I can't find any sabermetric system that allows me to deal with historical anomolies like Tom Glavine.

I'm going to assume you know what I meant and move along, avoiding making a fool of myself

[Wheell]

I know what you meant, and the formula is a good one. OPS is a simple equation that gets you 90+% of the way there. GPA gets you 99% of the way there. However, it works for hitters since you are dealing with an average of opposition. In other words, it can tell you how many runs a Wade Boggs or a Dave Kingman are creating, but it doesn't tell you how many runs a pitcher should give up when facing a half Boggs half Kingman monster. It gets worse when you have faced hundreds of different hitters and all the OBP and Slug regress to the mean. GPA might STILL be better, but first I need to work the kinks out of OPS. After that I can test more advanced formulas.

I'm not kidding, Glavine really is a statistical anomaly, much more so than a CM Wang.