Limiting Damage from the Longball

In continuing our look at pitcher sequencing from last week, today I want to focus on limiting the damage from Home Runs Allowed. Naturally, if a pitcher is going to surrender a longball, the best time to do it is with the bases empty, rather than with runners on. That is probably an all-too-obvious maxim to pitch by, of course, but the question is, do pitchers adjust their approach on the mound accordingly? And if so, does that adjustment actually produce favorable results?

As recently as 2011 we saw Ivan Nova of the New York Yankees take full advantage of this brand of "home run sequencing" when he allowed 13 longballs during the regular season-- yet all 13 of them were solo shots. That sort of incredibly low "Runs via HR" rate ranks #1 since the turn of the century among pitchers with at least 500 Batters Faced.

Lowest Single-Season R/HR Rate Since 2000*

Nova's incredibly low R/HR rate that season was at least .4 runs below the league-average. Now, if you do some quick math and assume his rate had actually been closer to league-average (say 1.4) over the course of 13 HR-- that's 13 x 1.4 = 18.2, which is 5.2 runs more than the 13 he actually allowed. Tack those extra 5 runs into his ERA and suddenly the difference between Nova's ERA (3.70) and his FIP (4.01) has completely evaporated. Just one example how sequencing can contribute to beating out your FIP.

Repeatability

Now, you may be thinking to yourself, way to go, Ivan, you FIP-destroying stud. Except for Nova, in his very next season he allowed 1.5 runs/HR helping to contribute to a miserable ERA- of 120.

Naturally, then, this leads one to wonder, how repeatable is this skill of favorably sequencing your home runs allowed? So I ran a correlation on year-to-year R/HR rates for 865 pitcher-seasons from 2002-2012 with at least 500 Batters Faced. The results were not encouraging, with just an r of just .08.

Yet at #6 we have a season from Tom Glavine, who was quite famous for this sort of pitching strategy. Glavine was known to attack batters with the bases empty and then "nibble" significantly more with runners on. So it should be no surprise that the alleged sequencer of sequencers himself shows up here. Perhaps only particular pitchers, those like Tom Glavine, have this talent?

If we return to the same 1993-2011 time period we used in last week's run_ex queries, we find Glavine shows up again amongst the top ten lowest R/HR with a minimum of 5,000 Batters Faced:

Lowest R/HR Rate Since 1993

But it is important to remember that skill is very much related to sequencing success. In other words, a pitcher who limits his baserunners is generally going to limit his Runs/HR as a consequence. Now, Pete Harnisch at #1 may not exactly fit the description of a pitcher with an elite talent for limiting baserunners (lifetime WHIP 1.3), but Pedro Martinez at #2 fits that description precisely (lifetime WHIP 1.054). So, we have to ask ourselves, Is Pedro actually limiting home runs based with baserunners on, or did he simply limit baserunners all the time?

In order to seek out which pitchers have excelled strictly in terms of sequencing, we need to examine how their HR-rates change according to the base states. In the following table we'll look at which pitchers lowered their HR/PA with bases empty vs runners on:

Greatest Drop in HR% with Runners On Since 1993

As it turns out, Steve Traschel may be a bit craftier than I remember. His FIP- is a full 10 points below his ERA-, he ranks 5th in Fangraphs LOB-Wins since 1993 with 7, and his change in HR-rate with runners on ranks 5th overall in that same timeframe. Wandy Rodriguez places first overall, but if you'll remember, he fared particularly poorly with regards to sequencing his walks, which contributes to a negative LOB-Win value and virtually no ERA-FIP split to speak of.

Tom Glavine places 9th overall, which is expected, but Pedro Martinez fares even better than that at #3, with a drop in his HR rate of over 34% with runners on during his career. Pedro, fascinatingly, may have relied on 'cerebral pitching' just as much as his unearthly peripherals to post those legendary, historic seasons.

There are a few interesting names that show up just outside the top-ten as well. One of them is Pete Harnisch (who curiously also did very well in sequencing his walks) at #13. Pete wasn't a particularly outstanding pitcher, managing an ERA- just below average over the course of his career. Yet, in the 1993-2011 timeframe we've been dealing with, Harnish's ERA/FIP discrepancy (as measured by E-F) is second greatest at -0.58, just a sliver above Glavine's -0.57. You may not be as excited about this as I am, but Harnisch may be exactly the type of pitcher I am trying to hunt down in this series. A pitcher who's sequencing skills are not as obvious as a Tom Glavine, but are prominent throughout his career nonetheless.

Next week, we'll look at situational K-rates.

*Does not include post-season.

All table data courtesy of retrosheet, includes post-season unless otherwise noted.

Follow @JDGentile on twitter.