Pitching Outside the Box (The Less Wrong Remix)
As we await the stirring finale of R.J.’s Contact Tales series (tentatively called “Who Shot R.J.?”), I thought it might make sense to revisit a post I submitted in these electronic pages just before the start of this shiny new decade.
In said post, I took a look at Matt Hanna’s work on Expected Strikeout Percentage (eK%) — a study that first appeared at DRaysBay and which resulted in this equation:
eK%=(ClStr%*.9)+(Foul%*.5)+(InPly%*-.9)+(InZSwStr%*1.1)+(OZSwStr%*1.5)
As I noted then, what’s particularly revealing about Hanna’s work is that it allows us to see (for the first time, so far as I know) the relative importance of inducing in-zone versus out-of-zone swinging strikes. The latter, if you can believe it, are actually about 2.5 times more important than the former.
Using this information, I then provided a contact leaderboard of starters from last season, weighting O-Contact% about 2.5 time more strongly than Z-Contact%.
The error I made in that post — and which reader dyross brought to our attention in the most agreeable way imaginable — is that I submitted for the reader’s consideration only the O-Contact% (and Z-Contact%). The problem with only providing the contact percentages is that it presupposes a constant O-Swing% among all pitchers — which, obviously, that’s not the case. From just the sample I picked, there’s a range of 12.30% (Sidney Ponson) to 32.80% (Hiroki Kuroda). Certainly, pitching outside of the zone in such a way as to induce swings in the first place — that’s important.
Without further ado, here’s that Top 10 list as it should’ve appeared in place of the one I actually provided. Here swinging strikes are posted as a percentage of all pitches. [TOZSS% = Total Out-of-Zone Swinging Strikes Percentage / TIZSS% = Total In-Zone Swinging Strike Percentage / SDTOZ = Standard Deviations from the Mean, Out-of-Zone / SDTIZ = Standard Deviations from the Mean, In-Zone / wAVG = Weighted Average of Standard Deviations, or (SDTOZ*2.5+ SDTIZ)/2.]
| Name | TOZSS% | TIZSS% | SDTOZ | SDTIZ | wAvg |
|---|---|---|---|---|---|
| Rich Harden | 8.09% | 7.47% | 2.71 | 3.46 | 5.03 |
| Felipe Paulino | 7.53% | 5.13% | 2.27 | 1.29 | 3.41 |
| Javier Vazquez | 7.01% | 5.61% | 1.87 | 1.73 | 3.15 |
| Chad Billingsley | 7.91% | 3.23% | 2.56 | -0.47 | 2.89 |
| Francisco Liriano | 7.08% | 4.74% | 1.92 | 0.93 | 2.81 |
| Billy Buckner | 7.48% | 3.68% | 2.23 | -0.06 | 2.69 |
| Ryan Dempster | 7.24% | 4.06% | 2.04 | 0.30 | 2.64 |
| Jorge de la Rosa | 7.29% | 3.81% | 2.08 | 0.06 | 2.57 |
| John Smoltz | 7.27% | 3.72% | 2.07 | -0.01 | 2.51 |
| Randy Johnson | 6.96% | 4.12% | 1.83 | 0.35 | 2.41 |
Understandably, we get some of the same names here as on the other list: Harden, Paulino, Vazquez, Dempster, de la Rosa. Those are pitchers who are able to induce swings and misses — not just misses, as in my last effort.
Newcomers to the list are: Billingsley, Liriano, Buckner, Smoltz, and Randy Johnson.
Omissions? Norris, Freddy Garcia, Gio Gonzalez, Lincecum, Sanchez.
So, like I say, that’s using all pitches as the sample.
My one concern about such a list, though, is that it could end up favoring pitchers who throw out of the strike zone with greater frequency. This, for example, is what we see Lincecum fall of the list: he throws too many balls in the zone to register a very high overall percentage of out-of-zone swinging strikes.
Really, the thing we’re probably most interested is this: pitchers who (a) throw a pitch out of the zone, (b) induce a swing with said out-of-zone pitch, and (c) receive a whiff on said swing. In other words, we’re probably most interested in how efficient a pitcher is with his out-of-zone pitches. Ideally, if he could, a pitcher would induce swings and misses on every pitch out of the zone. That’s unlikely. But here are the guys who got closes to that mark:
| Name | OZSS% |
|---|---|
| Rich Harden | 15.80% |
| Felipe Paulino | 14.70% |
| Chad Billingsley | 14.56% |
| Freddy Garcia | 14.36% |
| Ryan Dempster | 14.25% |
| John Smoltz | 14.14% |
| Billy Buckner | 13.88% |
| Javier Vazquez | 13.86% |
| Tim Lincecum | 13.82% |
| Randy Johnson | 13.59% |
Ultimately, it would probably be right and good to adjust such a list for league, as it’s very likely that NL pitchers derive quite a few out-of-zone whiffs from batting pitchers. As I have very likely exhausted my skills in this particular area, I’ll just pause and reflect on the few things I’ve gotten right.
Carson Cistulli has published a book of aphorisms called Spirited Ejaculations of a New Enthusiast.
Couple questions quickly come to mind:
Does this stat fluctuate game to game (umps), month to month (luck), year to year (pitcher ability)?
When can you say a pitcher is above/below average in this regard? (Assuming when hitters see someone for the first time, advantage goes to the pitcher.)
The idea makes complete sense. I remember watching a lot of Kazmir last season, just amazed at how few bats he was missing. Would get 2 strikes and they just kept fouling off pitch after pitch. Winds up either a walk or a ball in play…with high pitch counts per batter faced.
Surprised with his video-game breaking stuff that Gavin Floyd didn’t make the list. Any chance of seeing a top 40??
I agree-there’s a lot left to be learned through use of a larger sample. The inclusion of names like Paulino, Buckner, and Garcia ask the question “Are these guys perhaps better than we originally believed?” But we don’t have enough additional information to know whether they’re the effect of random chance. For the most part, that list of names seems to make perfect sense.
This is really good work, though, to refine the data down to where we can see something very telling.
Jimbo, provided I have some idea of what I’m doing, you should be able to get to the relevant spreadsheet by clicking here.
Let me know if it doesn’t work.