Taking the Platoon Advantage
Fernando Rodney has three saves and a win so far this season. Fernando Rodney has gotten eight outs so far this season. As strange as it may first seem for a late-inning reliever to have four decisions with so few batters faced, it’s business as usual in Tampa Bay. Here’s a box score that is fairly typical for the Rays:
It certainly appears that the Rays are micro-managing their bullpen. Perhaps the aim is to gain the platoon advantage in as many situations as possible — teams do that all the time. But which ones are doing it most often?
Let’s take a look at the teams put their pitchers in lefty-lefty and righty-righty situations the most times in 2011. This should help us figure out which teams pushed for the platoon advantage most often.
| Team | LHPvLHB PA | Team | RHPvRHB PA | Team | Combined Rank | ||
|---|---|---|---|---|---|---|---|
| TEX | 868 | SLN | 2800 | CHN | 1 | ||
| TOR | 862 | COL | 2797 | TBA | 2 | ||
| CHA | 840 | CHN | 2768 | HOU | 3 | ||
| KCA | 741 | ATL | 2719 | SLN | 4 | ||
| OAK | 653 | FLO | 2702 | BOS | 5 | ||
| MIN | 646 | ANA | 2645 | DET | 6 | ||
| PHI | 516 | CIN | 2620 | NYN | 7 | ||
| TBA | 512 | CLE | 2610 | ATL | 8 | ||
| BAL | 498 | HOU | 2541 | BAL | 9 | ||
| WAS | 495 | NYA | 2435 | KCA | 10 | ||
| SEA | 493 | NYN | 2394 | NYA | 11 | ||
| DET | 489 | ARI | 2324 | CIN | 12 | ||
| BOS | 478 | BOS | 2314 | CHA | 13 | ||
| LAN | 452 | DET | 2313 | COL | 14 | ||
| NYN | 447 | TBA | 2294 | OAK | 15 | ||
| HOU | 428 | SDN | 2282 | PHI | 16 | ||
| NYA | 419 | MIL | 2243 | TOR | 17 | ||
| SDN | 406 | BAL | 2206 | WAS | 18 | ||
| ARI | 382 | WAS | 2147 | TEX | 19 | ||
| CHN | 380 | SEA | 2122 | ARI | 20 | ||
| CIN | 333 | LAN | 2100 | MIN | 21 | ||
| MIL | 332 | PHI | 2029 | SEA | 22 | ||
| ATL | 312 | KCA | 1995 | ANA | 23 | ||
| SLN | 302 | OAK | 1937 | CLE | 24 | ||
| SFN | 300 | MIN | 1834 | FLO | 25 | ||
| CLE | 299 | CHA | 1790 | SDN | 26 | ||
| COL | 269 | TOR | 1783 | LAN | 27 | ||
| ANA | 249 | PIT | 1462 | MIL | 28 | ||
| FLO | 242 | TEX | 1442 | SFN | 29 | ||
| PIT | 174 | SFN | 962 | PIT | 30 |
Tampa Bay shows up in the top third for lefty-lefty situations, and exactly half-way when it comes to righty-righty. That may not seem like a team fighting hard to find these situations, but look at the combined rank. Since the teams mostly show up on the leaderboard for one platoon advantage, but not the other, Tampa ends up near the top of the combined rankings. The other teams atop the combined leaderboard are largely unsurprising.
Tony La Russa’s legendary bullpen micro-management in St. Louis was the subject of more than one good article on the subject. This leaderboard seems to pass the sniff test. Then again, the White Sox have long had an excellent bullpen full of hard-throwing lefties and righties. Until this season, the team also had an unconventional manager who even platooned his closer role on occasion. They only show up half-way up the leaderboard.
This analysis is not yet complete. These matchups are subject to the vagaries of each team’s starting rotation. Starters take up more than half of these plate appearances, and they aren’t as fungible as relievers. If you have two lefties in your rotation — as the Rays do now — you’ll come out ahead on these charts.
Is there another way to focus in on this? What about total batters faced by relief appearance? The lower the number, the fewer batters each reliever sees. This would point to a team looking to push the envelope and find the best matchups for their relievers.
| Team | Relief TBF | TBF/Relief G |
|---|---|---|
| TBA | 1645 | 3.76 |
| LAN | 1878 | 4.07 |
| ARI | 1888 | 4.08 |
| SFN | 1958 | 4.08 |
| NYN | 2101 | 4.09 |
| COL | 2133 | 4.13 |
| SDN | 2031 | 4.14 |
| HOU | 2094 | 4.16 |
| PIT | 2298 | 4.19 |
| SLN | 1969 | 4.21 |
| CIN | 2120 | 4.22 |
| ATL | 2177 | 4.27 |
| FLO | 2167 | 4.27 |
| OAK | 1964 | 4.27 |
| CLE | 2089 | 4.33 |
| MIL | 1888 | 4.35 |
| WAS | 2193 | 4.36 |
| CHN | 2164 | 4.37 |
| NYA | 2032 | 4.37 |
| PHI | 1751 | 4.43 |
| MIN | 2034 | 4.45 |
| TOR | 2109 | 4.45 |
| TEX | 1881 | 4.51 |
| DET | 1936 | 4.6 |
| CHA | 1889 | 4.61 |
| ANA | 1812 | 4.69 |
| BOS | 2161 | 4.87 |
| SEA | 1743 | 4.97 |
| BAL | 2428 | 5.08 |
| KCA | 2168 | 5.16 |
Look at that. The Rays come out ahead once again; they know exactly which batters they want their relievers to face, and they pull their relievers quickly.
But there are still some starting pitching effects. As you can see, the Rays got so many innings from their starters that they faced roughly 200 fewer batters than the teams that used their relievers the second-least. That sort of workhorse production from your starters makes it easier to mix-and-match your relievers. In a way, the Mets and Rockies have worked harder to get their relievers in the perfect matchups, considering that they counted among the league leaders in overall relief appearances.
We’re still not done. Baltimore — at the bottom — encapsulates two problems with this approach. Their starters were not very good, so their relievers were forced into the most innings. Also, as a poorer team, they may not have had the resources to spend on LOOGYs. The Giants have two relief-only lefties making more than $5 million each. The Orioles had Pedro Viola, Clay Rapada and Troy Patton making the minimum after Mike Gonzalez left town.
Then again, the Rays only paid their lefty relievers a bit more than $2 million combined last year and still managed to come out on the top of both lists. Their platooning may also not stop at lefty-righty, as I posited in 2011: they may not only look for the correct matchups based on handedness, but also based on pitcher strengths and batter strengths. In any case, it certainly appears that -— more than any other team in baseball —- the Rays succeed in taking the platoon advantage.
Thanks to Jeff Zimmerman for some of the numbers in this piece.
Maddon = the true MLB MVP for the last 5yrs or so
Haha
After TB, teams 2-12 on the second chart are all NL teams. I’m guessing it has to do with the pitcher’s spot in the lineup coming up and forcing removal. But it makes the TB result even more “impressive.”
It also has to do with late-and-close situations. With TB not having a great offense, but great starting pitching, they are in a lot of 3-2 games in the 7th inning, so they feel the need to be particular with matchups. Same goes with San Fran, unsurprisingly high on the list as well.
Must. Put. Offense myth. To bed:
http://www.fangraphs.com/leaders.aspx?pos=all&stats=bat&lg=al&qual=0&type=8&season=2012&month=0&season1=2010&ind=0&team=0,ts&rost=0&age=0&players=0&sort=16,d
I am. A Stabbing Robot.
(The Rays have the No. 5 offense in the AL through the last 2+ seasons. Their home park plays a lot like Petco, so they get a bad rap, but the truth is they can score and score often. In 2012, they have a 130 wRC+ so far, good for 3rd in the MLB.)
I said “not great”, Brad. 5th out of 14 is not great, it is decent to boarderline-good. I guess by lumping in San Fran I made it seem they were deficient on the same level, they are not. TB is also 6th in the AL in wRAA in that time frame. The Red Sox they are not.
Would the home park playing like Petco also lead to a lot of late-and-close situations? DD’s theory could work whether the low scoring is due to poor offenses or park effects.
Seriously producers?? This is the best title you can come up with?? Holy s*** what is this world coming to??
And which of those teams boast the most Bullpen wins?
C’mon fangraphs-data on vulture wins is the fantasy grail :)
it’s always tyler clippard
Wouldn’t isolating it to lefty vs. lefty and righty vs. righty to relievers only solve this better?
And then say normalizing it to total batters faced? (to account for different bullpen workloads)
And of course maybe just adding up rankings to get a “combined” (read – somewhat arbitrary) ranking is not a good thing given the different size of the 2 sets of data.
It should be:
Sum of lefty-lefty and righty-righty matchups in relief / total batters face in relief
This analysis was written to justify the smartness of the Rays; their actual bullpen performance with the advanced stats (FIP,, xFIP, fWAR) was not very good. The other odd thing is the team 2nd on the list in terms of fewest batters/relief appearance was 27th on the combined platoon advantage ranking so using batters faced/relief appearance to talk about smartly/optimally using the pen and finding the platoon advantage is a bit of a stretch.
The other thing not mentioned is the unbalanced schedule and how the opponent lineup comes into play… Every team in the AL east is in the upper half of the platoon advantage, except Tor which is just outside of it at 17. Every AL West team was in the bottom half, except OAK at 15. Which division has lineups which tend to be more set and a higher quality of hitters who don’t tend to get pinchit for to avoid the platoon disadvantage, and which division has more mix and match lineups where it will be harder to get a platoon advantage against?
I did do the research because of looking at a Rays box score, but I was honestly trying to find out who takes the platoon advantage most. Given my limitations, I think it’s a start.
I could normalize the TBF/Apperance for total appearances, that’s an idea. Getting platoon relief appearances only was a bit harder than I expected.
I’m not calling this definitive by any means, but I do think it’s interesting that the Rays relievers saw the fewest batters per appearance and there wasn’t another AL team close. I think they might be platooning in a different sense, as I alluded in the text.
I could have done the math/sorting wrong but even if you just add up the 2 platoon columns (left-left and right-right) in your table and divide by total batters faced by that team the Rays end up nowhere close to 2nd overall on the list (of course this still has the starter vs reliever issues).
The following teams have a greater % of PA’s with the platoon advantage while pitching:
STL, HOU, CLE, CHN, ANA, ATL, FLA, COL, CIN
I assume this is mainly because of the greater magnitude of right-right PA’s as opposed to simply giving the left and right platoons equal weight like you did. Since the TBF data is available on this site, I’m not sure why you would just equally weight the two columns to get a “combined” ranking
The 2 AL teams (CLE, ANA) on that list are in the middle and bottom of the TBF.per relief appearance list, which also makes me question the link between platoon advantage and TBF/ relief appearance that you are suggesting. The Giants, who have far and away the lowest % of PA’s with the platoon advantage are still near the top of the list in terms of lowest TBF per appearance.
Without going through the data I’d intuitively think that the shorter relief appearances has a lot to do with the bullpen having to throw fewer innings – the manager being able to pull relievers quickly because the bullpen is not burned out from previous days’ usage. The other factor would seem to be the quality of the bullpen arms. The Rays aren’t exactly stacked with a bunch of quality arms in the pen where you wouldn’t care about platoon advantage (for example Girardi is going to let pitchers like Rivera, Robertson,and even Soriano pitch to both righties or lefties and not really be managing to platoon situations… )
by looking at B-R you can figure that LHB faced relief pitchers in 917 PA last year. Of the 917 PA 460 were by LHP and 457 by RHP, and of the 457 108 were by closer Kyle Farnsworth.
Take out those, 57% of the time a LHB faces a non-closer reliever its a Lefty
The same methpd with East rival Yankees shows just 15% of the time a LHB faces a non-closer reliever its a Lefty ( out of 1089 PA)
Is it accurate to say that “the Rays succeed in taking the platoon advantage.” based on this data? What it shows is that they tried to take the platoon advantage. They were 6th worst in MLB last year in reliever FIP, 4th worst in reliever WAR, and 2nd worst in reliever xFIP. Maybe they would have been even worse if they hadn’t utilized the bullpen in this manner but the article doesn’t prove that.
The “CHN” at the top of the first leaderboard is the Cubs, not the White Sox, so you might want to update the section about the “sniff test”
wow. My bad. fixed.
Is Ozzie Guillen Voldemort now?
The Rays bullpen was 29th in xFIP and 26th in FIP last year
Given the results, is the approach working?
“Fernando Rodney has three saves and a win so far this season. Fernando Rodney has gotten eight outs so far this season.”
That should headline an ESPN segment on why popular pitching statistics are useless.