The problem with Howards BABIP (people should just call it H% like Shandler does since its so much easier~) was that it was too 2nd half heavy. He had a 42% H% in the 2nd half, nobody sustains that, nobody.
H% is a tool that really relies on past data to be useful. Its also not a predictive stat, its a stat that you use to look for the real ability of a player from the previous year. You don't use H% trends to decide if a player is improving or getting worse etc, its useful for figuring out if a jump in ability the previous year was a fluke.
A good example is Gary Mathews last year. His H% went from 28 to 35 and his AVG jumped from .255 to .313. You can pretty much be assured that the .313 is just a fluke. His normal H% is around 30-31 though so most likely the .255 was a fluke too. His real ability both years was probably in the .270-.275 range.
However last year his contact rate actually jumped 3% over his career norms. Usually contact rate predicts future success so I expect him to hit higher this year than that .270-.275 range, probably closer to .280-.285 assuming he doesn't lose AVG from the park switch.
Also, someone should determine some ratio for "pitcher's tendency to not have his best stuff." It would have to be based on the pitcher's own admission, though, at post-defeat press conferences.
Actually Baseball HQ has developed a stat called qERA. It rates each start by a pitcher from 1-5 based on how good it was. Then uses a ratio of good starts vs bad starts to define ERA. Generally speaking what they have found is pitchers who are consistently mediocre almost never improve but those who are dominant at times and then disasters at other times do end up becoming better pitchers a lot more.
When looking at whether or not qERA was predictive the results were pretty good. From 2001+2002 the system predicted big improvements in 27 starters and 22 of them did improve. It predicted dips in production for 62 pitchers and 46 of them declined.
I'm fairly skeptical of judging players potential performance base solely on stats. I think there's just too many variables that don't totally show themselves in numbers. To really judge a player, I think you need to watch him play and understand the context of his performance, particularly for pitchers. The one stat I do trust and watch for, be it pitcher or hitter, is the walk rate and how it's trending.
Big Pimpin wrote:Anyway, there's a league average number, but some pitchers do have their own norms, usually based on Ground Ball/Fly Ball tendencies and the defense behind them. For instance, Felix Hernandez, with his groundball tendencies and a great IF defense, should have a lower number.
Half right. Good defence certainly helps, but groundball pitchers should have higher BABIP than flyballers. Someone like Peavy or Patterson (pitchers' park, flyballer) ought to be an example of a guy you'd expect to have a low BABIP.
Interesting. I would have sworn that I just read something that had line drives turning into the highest percentage of hits, then fly balls, then ground balls. Maybe that included HRs though in fly balls, which mixed me all up.
Big Pimpin wrote:Interesting. I would have sworn that I just read something that had line drives turning into the highest percentage of hits, then fly balls, then ground balls. Maybe that included HRs though in fly balls, which mixed me all up.
Or maybe I'm just crazy.
You are not crazy, but you are off by a little. About 75% of line drives become hits. About 70% of ground balls become outs. About 80% of flyballs become outs (and that includes homers). Take out the HRs and almost 90% of flyballs become outs.
So, FB pitchers will tend to have lower BABIP. But, they may not be better pitchers to have, because they will also have dingers that boost their ERA.
flbtank wrote:So, 1) Does the collective wisdom of the Cafe think this is a valid strategy to find improvements (ie BABIP correlates to lower ERA's & WHIP's)? 2) Any ideas on where to find easy to use 2004-2006 data? and 3) Has anybody already done this?
Thanks in advance.
why not just look for DIPS abnormalities?
Like a Bonderman or Vazquez, whose DERA was lower than their ERA?
it's also worth while most pitchers have little control over BABIP, the best pitchers do. so that's really the trick in separating the lucky from the unlucky.
Did anyone apply the BABIP strategy when drafting pitchers last year? It's great to hear how it works and see some examples for 2007. But did anyone do it last year, and, if so, for which pitchers and were you right?
they took a simple stat that was used for 100 plus years, one that everybody could relate too, Batting Average, removed the homeruns, (WHY?) and called it ..
BABIP, BABP or H% .. depending upon who you talk too ..
I can look at Gary Matthews 2006 Batting Average and see it wasn't the norm, and expect regression. I can look at Chris Youngs Batting Average Allowed and compare it to his career norms (minors and Majors) and expect some regression ...
so what exactly does this stat tell me different?
I will agree that their have been some interesting finds using sabermetrics over the last couple of decades, but this stat here is simple Batting Average, repainted so that somebody can look smart ..
His H% went up, so one should expect his Batting Average to go up ...
boy, you stat heads sure are funny sometimes ...