It’s not completely set in stone, but we can be pretty sure that next Tuesday, the Pirates and Reds are going to square off in the NL Wild Card game. If the Pirates sweep the Reds and the Cardinals can’t win a single game against the Cubs, then Pittsburgh and St. Louis would tie for the division title, but let’s be honest, that’s probably not happening. Without giving up entirely, the Pirates can start planning for next Tuesday’s win-or-go-home game.
So, let’s talk strategy. Last year, I appealed to the Braves to just skip the starter entirely, using Craig Kimbrel to open the game, and then mix-and-match behind him to take advantage of platoon match-ups and let the bullpen carry them to the division series. I still believe in in the theory of skipping the starter in an elimination game, but the reality is that the participants in these games are people, and they are used to set routines, and divergence from their established roles might have a negative impact on their performance. It is probably, in reality, too crazy of a strategy for any team to actually adopt.
But that doesn’t mean that the team has to just stick with the conventional in-game approach that they used for the first 162 games of the season. The modern starting rotation is entirely a function of keeping 11 or 12 pitchers healthy and productive over a six month time frame, and it is simply not the most effective way to limit the opponents from scoring in any single game. Staring at nine innings that decide the fate of a team’s season, full year strategy should go out the window.
The rules surrounding the Wild Card Game rosters even incentivize a significant turn away from standard operating procedure. The 25 man roster that will be set for Tuesday’s contest can be completely overhauled for the division series, so Tuesday’s roster only has to play on Tuesday. After that, the winner can go right back to a more normal construction for a best-of-five series. So, instead of carrying several starting pitchers who won’t actually pitch in that days game — as every team does in the regular season — the Wild Card participants can simply load up their bullpen for a war of match-ups.
Last year, for instance, the Cardinals carried 10 pitchers and 15 position players into the Wild Card Game, but 8 of those 10 pitchers were relievers. Kyle Lohse started for St. Louis, and then Lance Lynn was around if the game went extra innings, but their ability to omit three other starters from the game allowed them carry eight relievers and still have room for a designated pinch runner and extra bench depth. And I think you could even argue that a team like the Pirates might be better off punting one of those bench players — the Cardinals carried three catchers, for instance, which isn’t really necessary — and adding a ninth reliever, going with a 14/11 position player/pitcher split.
One of those 11 pitchers should be a starter held in reserve in case of extra innings, leaving the team with 10 pitchers that they actually want to use in the Wild Card Game. And with 10 arms available to get through nine innings, there’s simply no compelling reason to manage the game like a regular season affair. In particular, the Pirates have a strong incentive to make numerous pitching changes.
Francisco Liriano is almost assuredly going to start the game, and even if he hadn’t been the team’s best pitcher this year, he would make the most sense given the Reds reliance on left-handed hitters at the top of their batting order. With Shin-Soo Choo, Joey Votto, and Jay Bruce as the primary offensive forces for the Reds, tough left-handers are going to be extremely valuable for Pittsburgh, and there is no left-hander that was more of a problem for opposing lefties than Liriano this year.
Liriano held left-handed hitters to a .130/.175/.146 line on the season, good for a .151 wOBA; Major League pitchers, as a group, put up a .153 wOBA at the plate this year. Basically, Liriano turned left-handed batters into the offensive equivalent of a pitcher forced to hit. This is the guy you want on the mound against Choo, Votto, and Bruce.
However, he has to get through six other batters — almost certainly all of whom will be right-handed — to face those guys, and Liriano is roughly an average pitcher against right-handed batters, as they posted a .310 wOBA against him this year, and are at .325 against him for his career. That’s not awful, but it’s actually the second highest wOBA allowed to right-handers on the Pirates staff, with only departed pitcher James McDonald faring worse against them this year. The Pirates are crazy good against right-handers, so they can do better than letting Liriano face them too often. Instead of asking Liriano to make a traditional start, they should simply tell him that he’s going to face 14 hitters.
That’s one time through the order, plus the first five batters in the line-up a second time, which presumes that the Reds will bat something like Choo-RHB-Votto-RHB-Bruce. Under that plan, six of the 14 batters Liriano would face would be left-handers, and a seventh would be the opposing starting pitcher, assuming Dusty Baker won’t be aggressive enough to pinch-hit for his pitcher the first time he’s due up. Given the profile of the batters he’s facing, it’d probably be reasonable to expect Liriano to get somewhere between 9 and 11 outs while facing those 14 hitters, so you’re asking him to go something like 3+ innings, knowing he’ll be removed after facing Bruce the second time.
The aggressiveness of that hook might seem unnecessary, especially if Liriano is pitching well, but the data is very clear about the deterioration in performance for a starting pitcher upon repeat viewings within a game. For 2013, here are the league averages based on time through the order, and then Liriano’s numbers by the same split:
|1st PA vs SP||0.250||0.310||0.390|
|2nd PA vs SP||0.259||0.319||0.411|
|3rd PA vs SP||0.270||0.331||0.429|
|1st PA vs SP||0.196||0.276||0.269|
|2nd PA vs SP||0.233||0.296||0.329|
|3rd PA vs SP||0.252||0.333||0.355|
There is a huge times-through-the-order effect on the pitcher/hitter match-up, to the point where only the absolute best pitchers in the game should even be considered for a third trip through the batting order in a critical game. In a National League elimination contest, where there is the complicating factor of the pitcher having to hit for himself, there probably isn’t a pitcher alive good enough to justify letting him face hitters for a third time. Liriano, as good as he against left-handers, shouldn’t be asked to roll through a series of right-handed batters more than once, especially with eight or nine relievers available behind him.
If the Pirates go with eight relievers, those seem likely to include some combination of Jason Grilli (RHP), Mark Melancon (RHP), Tony Watson (LHP), Justin Wilson (LHP), Vin Mazzaro (RHP), Kyle Farnsworth (RHP), Jeanmar Gomez (RHP), and then either Brandon Cumpton (RHP), Bryan Morris (RHP) or Kris Johnson (LHP). If they go with nine relievers, they can carry two of those final three, and Johnson would make some sense if they want a third left-hander to serve as a specialist to go after Choo, Bruce, or Votto once.
But, to be honest, I don’t think they’ll likely need that third left-hander, because several of their right-handers are pretty great against lefties too. For instance, Mark Melancon’s .163 wOBA allowed against left-handers is nearly identical to the mark Liriano held lefties to, and both Justin Wilson and Tony Watson are quality left-handed relievers. Between Liriano and those three, the Pirates should be able to cover the Reds three lefties for five trips the order.
So, let’s lay out the whole plan and look at how Pittsburgh might be able to use their arms to get 27 outs. We’re going to presume that the game is always fairly close, so there’s not really such a thing as a low leverage situation in game like this.
Liriano: 14 batters faced, 9-11 expected outs.
The full order through one time, plus five batters the second time to get through Jay Bruce.
Farnsworth: 4 batters faced, 3-4 expected outs.
The four right-handers that should make up the #6-#9 spots in the batting order. Farnsworth has historically been pretty decent against right-handers, and allowing him to face the weakest hitters in the line-up gives the Pirates their best chance of using him to get outs. If Liriano scuffled the second time through the order, this becomes a much better reliever to put out the fire, but if there’s no rally brewing, this is the time to use him to steal a few outs against RHBs.
Wilson: 5 batters faced, 2-3 expected outs.
You want Wilson to face the Choo/Votto/Bruce group once, so he has to get through the two right-handers in order to get those match-ups.
Mazzaro: 3-4 batters faced, 2-3 expected outs.
At this point in the game, you have to expect a pinch-hitter for the pitcher’s spot, so Mazzaro is probably only in for the 6-7-8 hitters, and there should be a pretty quick hook here if he gives up a baserunner or two, since the game should theoretically be in 5th-7th inning at this point and the Pirates would still have all their best relievers available. In fact, I’d probably suggest having Tony Watson warming with Mazzaro and throwing for his entire appearance, just in case he’s needed before the top of the order comes up.
Watson: 5 batters faced, 3-4 expected outs.
His turn to roll through the lefties. You have to really hope that he doesn’t have to hit in order to stay in to face more than one lefty, but even if he does, the value of having him go after Votto/Bruce might be worth letting him lay down a bunt and hope for a throwing error.
Grilli: 4 batters faced, 2-3 expected outs.
If everyone before him had been perfect, this would actually be a save situation, but that’s pretty unlikely, so this is probably more of a 7th-8th inning appearance. Grilli is nominally still the Pirates closer, and a very good one, but he is still best utilized against right-handed hitting. He held RHBs to a .222 wOBA this year, but gave up a .314 wOBA to LHBs. That’s not awful, and you don’t have to yank him if a left-hander gets sent up to pinch hit, but ideally, he’s mostly facing righties, and the #6-#9 batters should be mostly right-handers or pinch-hitters.
Melancon: Up to 9 batters faced, as many outs as needed to close it out.
Grilli’s the closer, but Melancon’s the best arm they have for rolling through the entire batting order thanks to his ability to put lefties away. If Grilli gets in trouble against the bottom of the order, you’ve got Melancon ready to go after Choo/Votto/Bruce and hopefully put out any fire that may arise late in the game.
Obviously, this is all very speculative, and it’s impossible to know ahead of time when Clint Hurdle may need to pinch-hit or what kind of double-switches will be pulled that move Cincinnati’s line-up around, but this kind of blueprint could be the target, with adjustments made as needed.
It only calls for using seven pitchers, so there’s room for a little more specialization if pinch-hitting or ineffectiveness cuts short a reliever’s appearance, and they’d still have a starter in reserve to handle extra innings should things get beyond 27-30 outs.
This is the kind of pitching staff a team can run in the Wild Card Game, and a side benefit is that no individual pitcher should be taxed to a level that they shouldn’t be ready to contribute in the NLDS should the Pirates advance. Even Liriano, having faced 14 hitters, could theoretically be ready to start a Game 2 if needed, since an effective 14 batter outing might only take 50 pitches or so.
The Reds line-up is tailor made for exploiting platoon splits, and the Pirates have the pitchers to take advantage of frequent pitching changes. With their season on the line, this would represent their best shot at winning the Wild Card and advancing in the postseason.
Dave is the Managing Editor of FanGraphs.