Beyond The Box Score's staff Ace Glenn DuPaul has been doing some very interesting work regarding the differences between starters and relievers over at THT the past few weeks. In the first article of the seires, Glenn asked if relievers across baseball were consistently improving, citing an ever-narrowing RA9 gap between RP and SP as evidence.
1.) More same-handed match-ups for relievers.
2.) Shorter outings.
I decided to draw up a few quick queries on the matter, starting with the history of same-handed match-ups.
I began by querying the retrosheet database for the percentage of match-ups where the batter and the pitcher were of the same handedness dating back to when this information was first recorded in 1960 (I eliminated from the calculation all Plate Appearances where handedness data was not available). Remember a higher percentage here creates an advantage for the pitcher, a lower percentage favors the hitter:
Now, I'll admit this was not exactly what I was expecting. I was expecting the percentage of same-handedness PA's to increase as time went on. But it is important to remember that as the pitcher's managers grew wiser over time and began opting for same-handed match-ups, so did the opposing manager in charge of drawing up that line-up card. The ultimate effect of all this is a sort of neutralization of the pitcher's advantage. And while it's true that same-handed match-ups have been on the rise since the early 1990's, they've really only been restored to their 1970's levels in recent seasons. In fact, baseball saw much more of a pitcher's advantage in the 1960's than at any other time.
But perhaps, the relievers have been disproportionally accounting for more of the same-handed at bats than the starters? To test this I then broke up the percentage of same-handedness at-bats into the two roles-- starters and relievers. Surprisingly, the trends remained consistently parallel with one another:
Maybe the effect of LOOGYs in particular has allowed the relievers to make up ground on their game-opening counterparts? I queried the percentage of LHP vs LHB and RHP vs RHB plate appearances in each season, but found no dramatic increase in lefty on lefty match-ups for relievers:
So if relief pitchers haven't been gaining their advantage from increasingly favorable match-ups, I had to turn to popular theory two.
The next logical explanation to test for was shorter outings. Perhaps by employing more relievers, and delegating fewer responsibilities upon the reliever in terms of batters faced, the modern RP is therefore allowed to exert more energy into his abbreviated performances? To test this I separated all Relief Appearances dating back to 1950 into 5 groups: appearances involving one Batter Faced, two BF, three BF, four BF, and five or more Batters Faced.
Interestingly, the percentage of relief-appearances that consist of facing just one or two batters has remained consistently steady over the last half-century. While, as you might expect, those legendary long relief appearances have become increasingly rare events in recent years, verified here by the sharp drop in the "5 of more BF" RP outings.
To visualize this another way, I've also put together a similar chart based on appearances by their length in Innings Pitched. What immediately stands out is the meteoric rise of the singular 1 IP reliever outing as baseball entered the Era of the Save Opportunity.
To what degree these changes explain the widening gap in performance that Glenn described between starters and relievers I'll leave for another article. I would expect, however, that the effect of shorter outings from relievers in recent seasons plays more of a role in that discrepancy. But at the very least I thought this sort of information might be useful to you.
For more on RP/SP differences you can check out this mega-google doc I put together this summer.
Follow @JDGentile on twitter.