23 January 2015

Looking at Miguel Gonzalez's Clutch Performance

Perhaps no current Oriole fascinates me to a greater extent than Miguel Gonzalez. Over the past three years, he's magnificently outpitched his peripherals, to an incredible extent. Among starting pitchers with 400 innings since 1871 (of whom there are 1,885), he has the lowest ERA-FIP differential, by a massive margin — the distance between first and second is the same as the distance between second and 17th. Unsurprisingly, others at this site share my enamorment for this man.

For Gonzalez, the discrepancy between expectations and reality comes as the result of two things: preventing hits on balls in play, and preventing base runners from scoring. While he's done pretty well in the former regard — in that aforementioned sample, his .264 BABIP ranks 203rd — he's predicated his accomplishments on the latter, in which his 80.3% mark bests every other hurler in the sample. Because of that, I want to analyze the causes of his LOB%, to see if he'll sustain his otherworldly stranding abilities.

Looking at his situational splits, we can see one thing that stands out:

Situation TBF K% uBB% BABIP HR% wOBA
Bases Empty 1111 17.4% 7.1% .258 4.2% .325
Men on Base 706 16.4% 7.7% .275 2.1% .305

When runners get on, Gonzalez's walk rate, strikeout rate, and BABIP decline, but the absence of home runs prevents those runners from scoring.

His batted-ball profile supports this discrepancy, to some extent:

Situation LD% GB% FB% IFFB% HR/FB
Bases Empty 20.6% 37.2% 42.2% 11.0 % 13.5%
Men on Base 22.5% 37.6% 39.9% 14.2 % 7.4%

Fewer fly balls, as well as more popups, will generally beget a lower home run-fly ball rate.

While a difference as large as the 2-to-1 that Gonzalez currently sports will regress in the future, it will probably continue to exist, because his pitching style supports it. Depending on the circumstances, Gonzalez has a different repertoire:

Situation Fourseam Sinker Slider Curve Split
Bases Empty 43.1% 16.3% 13.3% 11.2% 15.9%
Runners On 40.6% 21.3% 13.1% 8.8% 16.0%

His four-seamer — which hitters have knocked out of the park far more than his other pitches — goes down in usage (along with his curveball) when the going gets tough. His sinker steps in to replace it, and as you'd expect, that pitch doesn't lend itself to long balls.

The difference doesn't stop at pitch types, either — he also changes his location:

Bases Empty:

Runners On:

Lower pitches don't go airborne as often as higher ones, which means an approach that prioritizes the latter will lead to fewer long balls.

Again, are any of these monumental differences? No, so going forward, three-fourths of his home runs allowed won't be of the solo variety. But his ability to avoid damage with the long ball seems legitimate, and it probably means that he's a true-talent overachiever.

Steamer projects a 72.2% LOB% for Gonzalez in 2015; for reference, the major-league average in 2014 was 73.0%. While I don't think he'll continue to leave on 80% of his baserunners, I don't see why it'll regress that far, and a solidly above-average rate — and another year of overperformance — strikes me as the likeliest outcome.


Anonymous said...

Is there a reason he shouldn't just do his men-on-base approach all of the time?

Jon Shepherd said...

Great post, Ryan.

Yeah, we have kicked this around in various forms with MiGo and Tillman for a while. Basically...high strike zone pitching results in higher strikeout rates and flyballs (which typically are more easily converted into outs than groundballs). However, with men on base those flyballs that rarely turn into home runs become more meaningful, so it may make sense to throw down in the zone at that point and take advantage of lower zone pitching (except with batters with extreme upper cut swings). In other words...MiGo does not have a silver bullet, so a context based mixture of approaches might be ideal.