The Whiffing Machine
Per innings ratios are often misleading. Identical pitchers can face the same number of batters and produce the same number of walks, strikeouts, and home runs while one completes more innings than the other. Defense plays an integral part in shutting down an opposition. If one of those pitchers has the Seattle Mariners defense behind him and the other has the Boston Red Sox, then you would imagine the pitcher benefiting from Franklin Gutierrez and Adrian Beltre will complete more innings.
The more defensive independent option is to look at total amount of strikeouts or walks on a per batters faced ratio. The ratios are easy to get a grasp on and have less room for outliers (at least in theory), which is why when you see a line that features 40% strikeouts and 6% walks you feel your eyes protrude and heart beat quicken.
Then you realize those percentages belong to a hitter and the magnetic field alters in the opposite direction.
Believe it or not, that line is real, and it belongs to the M’s Greg Halman. The 22-year-old has a history of striking out a ton and not walking very much. The differences between strikeouts and any other type of out is usually overplayed, but there’s something seriously flawed about an approach when the ratio is this high at the Double-A level. His other raw tools – mostly power and speed – are impressive. Impressive enough that Baseball America named him the Mariners number one prospect entering this season. Although that ranking came with this warning:
Weaknesses: For all his upside, Halman presents more risk than most No. 1 prospects. His pitch recognition is below-average, resulting in many swings and misses and mis-hits as he chases pitches out of the zone. He’s too aggressive at the plate to execute much of a plan, and as a result he strikes out too much and walks too little.
Our minor league leaderboards only track back to 2006, but as best as I can tell Halman’s rate represents the highest in the Southern League in this time period by a decent margin. For some reason, I don’t think that’s an honor Halman cares to hold.
Print This Post
.278 OBP in AA for a #1 prospect? No thanks. This is like the AA version of Chris Davis.
So he’s Clint Barmes?
This is something I’ve been thinking about for a while, but have never really put into words until I saw this post.
There is something fundamentally flawed about per IP or per 9 IP rate stats. Think about the K/9 stat for example. K/9 is tantamount to K/out; as in, somebody who has a 9.0 K/9 gets .333 K/out. This is obviously because IP relies solely on outs recorded. Why should a K rate rely on outs? It should rely on batters faced. An extreme example: if Jon Lester has a 3-strikeout, 1-2-3 inning, he has a 1 K/out ratio for that inning (27 K/9), and a 1 K/batter faced ratio; if Jonathan Sanchez has a 3-strikeout, 3BB, 3H, 5 ER inning, he also has a 1 K/out ratio for that inning (27 K/9), but a .333 K/batter faced ratio. The K/batter faced ratio is much, much more telling of a) both a pitcher’s proficiency and efficiency, and b) much more importantly, how often he strikes dudes out.
These stats are (somewhat) widely used for hitters (K% and BB%); why not for pitchers? It’s pretty intuitive.
I agree with this. I think the reason they endure is that it gives people some idea of how many strikeouts they’d expect in an average appearance. It’s misleading of course, for the reasons you point out.
I would also like to mention that Fangraphs (and other sites) has K% as K/AB and BB% as BB/PA. They really should have the same denominator. And it shouldn’t be AB; AB was invented as a denominator for BA, and it’s really pretty useless for most other things (except maybe SLG). Why should walks and sac flies not be counted towards K%? Why should a batter that walks slightly more than he strikes out have a higher K% than BB%?
There are two good articles at THT about this issue.
http://www.hardballtimes.com/main/fantasy/article/the-great-strikeout-debate/
http://www.hardballtimes.com/main/fantasy/article/the-great-strikeout-debate-part-ii/
It’s not quite as clear as you make it out to be due to differences in BABIP and other issues.
The BABIP thing is why K% is better than K/9. The other thing he mentions, that walks shouldn’t be counted towards K%, is, to me, insane (hopefully he’s at least using a “balls in play” count that counts home runs, sac flies, etc). The skill of striking out batters is that you have to either put the ball over the plate or, if you don’t, at least entice the batter to swing and miss/foul it off. If you can’t make this happen three times to a hitter you walk him, hit him, or he puts it in play. The ability to strike batters out and the propensity to walk batters are clearly linked. Similarly, a hitter’s ability to lay off a 3-2 curve in the dirt causes him both to walk more often and strike out less. Unless someone can show me that K/AB or TrueK% (they’re almost the same thing) are more predictive than K/PA, I don’t buy that article for a second.
There were reports well before a deal was actually made that the Mariners were talking with the Pirates and Halman was a player being discussed. If this deal was the eventual Jack Wilson/Ian Snell for Jeff Clement/Ronnie Cedeno/pitching prospects deal, then I wounder which side decided to back off on Halman and go with Clement.
This year is particularly disappointing for Halman because he was pretty much across the board better in 2008 at the same level (AA) and obviously a year younger. He was showing progression in K% and BB% during his 08 stint at AA. While he played about half the games at AA in 08 (61 games and 230 ABs in 08 vs 121 games and 260 AB in 09), it’s a decent sample size, albeit a little of the small side but he had a decreased K%, increased BB% while still showing strong power and a decent AVG (which was inflated due to a .345 BABIP). It showed good progress but the wheels fell off the bus this year in pretty much every measurable category…