garf112 wrote: Zito is God wrote: garf112 wrote:
Zito is God wrote:The thing is, I agree that Ks are arguably the most solid stat because you are not giving the hitter even a chance to beat out a grounder or get a sac fly etc. However, there are many pitchers who simply are not K pitchers, yet do a solid job on the mound. This is where WHIP comes in. There are so many sinkerballer pitchers who just get infinite groundouts during their starts but have a low K ratio so they get passed by.
Doesn't this make gb/fb ratio the most important stat? Not WHIP
It actually could be debated that it is.
I think it is debatable, but you can have a very good ratio and still give up a ton of hits and runs being a sinkerballer. You don't HAVE to give up HRs to be a bad pitcher. GB/FB is a big reason I traded Vazques away (I am regretting it now because the guy I got for him...well, lets just say I REALLY needed a shortstop). Yet I will say again that WHIP means you aren't giving up the hits period, GB/FB could still have a high WHIP and is simply giving up a lot of grounders.
Basically to be a successful pitcher you need to do three things:
1) Strike people out
2) Keep people from hitting the ball in the air
3) Limit your walks
In that order. If you can do all three, you are an elite pitcher. If you break pitchers into categories, the best strike guys out and have a high gb%.... There are about 8 categories and the worst are flyball pitchers that don't strike anybody out.
WHIP has too much luck involved in it. Any pitcher will allow balls to fall in for hits at about a 30% rate.
If you are unable to find these stats, you might be better off just going with WHIP, but knowing that it is a flawed statistic.
First I completely agree with this post and that will save me a lot of the typing I needed to do.
Second I agree with Zito is God in one regard - if I had to pick one number from the 'back of the baseball card' type statistics it would be whip. It's certainly better than era or wins. If I was allowed to make one minor calculation I think I'd almost rather go with k/bb than whip but it's close.
Lastly I use two numbers when I'm actually evaluating pitchers. They're both projected numbers - one projects era and the other projects whip.
= 3.00 + ((13 * HR + 3 * BB&HBP - 2 * K) / IP)
= ((0.4 * (3 * IP - K)) + BB + HR) / IP
DICE was created by Voros McCracken. pWhip I created myself but if you read the article it's more common sense based on the work that others have done than any huge leap that I made myself. These numbers can be spit out to you en masse by dumping projections/past year's data into a spreadsheet and making an extra column for both of the numbers. If you want to delve even deeper I would recommend looking at gb/fb percentages (ground ball pitchers will give up fewer extra base hits overall which isn't reflected here...only the home runs are reflected here) and also look at the hr/f ratio normalized to park (The Hardball Times
is a good place to get stats like that) as anything significantly different from 11-12% would be a cause to expect a regression back to the mean in home run totals.