MASTER NOTES: The Romo Experiment

OK, quick: Which pitcher would you rather have? Pitcher A? Or Pitcher B?

  • Pitcher A’s ERA is 3.13, Pitcher B’s is 7.25.
  • Pitcher A’s WHIP is 1.16, Pitcher B’s is 1.58.
  • Pitcher A’s K rate is 23% of batters, Pitcher B’s is 18%.
  • Pitcher A’s hard-hit rate is 34%, Pitcher B’s is 42%.
  • And Pitcher A’s soft-hit rate is 18%, while Pitcher B’s is 14%.

Clearly you want Pitcher A in this comparison. But here’s the kicker: They’re the same pitcher.

Pitcher A is MIN RHP Jake Odorizzi, when he pitches through the opposing batting order the first two times. Pitcher B is also Odorizzi, when he faces the order the third time.

I started thinking about this earlier this week, when I heard the news that the Tampa Bay Rays had reliever Sergio Romo do something that hasn’t been done since Babe Ruth was stealing hot dogs, or however that story goes.

Romo, as you probably heard, started two games on consecutive days against LAA as the Rays experiment with using matchups in a whole new way.

In the first game, the Rays’ were going to use rookie LH starter Ryan Yarbrough, but not in the first inning. The first inning is the highest-scoring, and the Angels’ scheduled batting order was very righty-heavy, with Zack Cozart, Mike Trout and Justin Upton hitting 1-2-3, and more RHHs after that. Romo is a monster vs RHH—his K% against RHH is over 35% of hitters faced, roughly twice Yarbrough’s K rate vs righties. The Rays thought they stood a better chance of getting through the first inning unscathed with Romo, and that Yarbrough would benefit from starting the second inning with a zero on the scoreboard and having avoided the Angels’ top hitters.

Indeed, that’s exactly what happened. After TAM didn’t score in the top of the first, the game-state was a 55% Win Probability (WP) favoring the home Angels. But Romo struck out Cozart, Trout and Upton, which flipped the game back to a 51% WP favoring the Angels—basically a coin toss, with a small edge to the home team. Yarbrough came in pitched 6.1 effective innings, giving up just one run on four hits and a walk, and striking out four. The Rays won.

A similar story ensued on Sunday. Romo started, facing RHHs Ian Kinsler, Trout and Upton. He walked Kinsler, but fanned Trout and Upton and got Andrelton Simmons (another RHH) to ground out. Romo stayed in to start the second, walking Cozart and then fanning Jefrey Marte (still another RHH). Romo then turned the scoreless game over to Matt Andriese, who finished the inning with a strikeout and a groundout. Andriese was a little unlucky in giving up some runs, and the Rays lost.

Still, the experiment worked well enough that the Rays have already said they’ll start Romo twice again this weekend, against the predominantly right-handed-hitting Orioles.

This could become a thing. From the start of last season through Tuesday, 46 pitchers with at least 50 batters faced the third time through an order have seen their ERAs jump by three full runs or more when facing the opposition for the third time. And these are not all bums and castoffs. Regardless of what you think of Odorizzi, he’s a pretty capable starter, and, as we saw, he’s really good those first two times through. The same can be said of other third-time strugglers, like Charlie Morton, Jameson Taillon, Zack Greinke, Jake Faria, Yu Darvish, Masahiro Tanaka, Lance McCullers, Brad Peacock and Jonathan Gray. Pitchers who are close to three-run increase in ERA include Aaron Sanchez and current AL Cy Young contender Gerrit Cole.

Now, if you were a manager or a General Manager, and you know you had a solid pitcher for those two times through the order, why wouldn’t you want to maximize the effect by letting him start his games lower down the opposing order?

So let’s suppose that’s what happens. How will fantasy be affected? I can think of a few effects that would change valuation methods and roster strategies:

Lower-value starters could see an uptick in value. Go back to Odorizzi. Assume the Twins had a Romo-like short man (a bold assumption, given the parlous state of the Twins' bullpen) who could have started Odorizzi’s games this season. Odorizzi’s ERA in first innings this year is 5.40, his WHIP 1.30, and his OPS against almost .800. In the second, third and fourth innings combined, his ERA is 1.80, his WHIP 0.93, and his OppOPS a mere .527. Give him a pass on the top of the order in the first, and he could have had far better results, in the decimals and, it seems likely, in Wins.

(Of course, it could also be that the first-inning blues are masking some other effect. Odorizzi’s weak spot might not be the first inning per se, but his poor performance in his first 25 pitches (a .781 OppOPS in those early tosses vs .565 in pitches 26-75). Whenever he comes into a game, he has to throw pitches 1-25 before he can throw #26. On the other hand, maybe the poor first 25 pitches are poor because they're in the first inning. Chicken-and-egg, so it's hard to say.)

Leagues with minimum- and maximum-starts rules would have to consider revising them. A guy like Romo, in a fulltime role as a one-inning specialist game-"opener," could rack up 50+ starts all by himself. And all the starters joining games in the second inning would see their start totals crater or even disappear. Since starts requirements are meant to require fantasy teams to roster a representative staff including a fair number of “starting pitchers,” which actually means “several-innings pitchers,” leagues would have some arguing to do. And we all know how much fun that is.

Ditto for innings requirements. If this experiment or other new pitcher-usage and pitcher–management methods take root, it might be nearly impossible to get 950 innings out of a nine-man roto pitching staff. Which also means...

...Leagues would have to reconsider their overall hitter-pitcher roster requirements. The Romo Scheme, if fully implemented, would seem to mean a lot of pitchers on each MLB team’s roster, maybe even more than we have now. That would mean even shorter benches, more Marwin Gonzalez-type multi-utility guys—and fewer injury replacements in the free-agent pool in AL- and NL-only formats. Those leagues would have to reconfigure their 23-man rosters, maybe all the way down to 11-12 or even 10-13.

Pitchers who can consistently navigate the third time through the order will become more valuable. In this short study, 28 pitchers actually improved in ERA, and 61 improved WHIP. Sometimes this is because their baseline first- and second-times-through decimals were horrible, but in other cases, pitchers lived up to the broadcast-booth ninny narrative of getting stronger as the game moved on. Some examples: Luis Castillo, Sean Manaea, Rick Porcello, Chad Bettis, Corey Kluber, Noah Syndergaard, and Reynaldo Lopez, who held his ERA just under 4.00 but crushed his WHIP from 1.35 the first two times through to 0.96 the third time.

Finally, to capture the real value of the Romo-type pitcher, leagues might need a new category, or might want to classify a successful one-inning start as a “Hold” for scoring purposes. The current rules wouldn’t have a classification for a Romo-type “start,” and a "Hold" requires the pitcher to enter with a lead which could only happen for an "opener" in a road game where his team did score in the top of the first. But a successful appearance would sure look like a Hold in spirit.

I don’t know for sure if this Romo Experiment will work out over a longer test. Since baseball adopts new methods about as readily as your grandma welcomed the microwave oven, I’m not optimistic—unless someone uses the Romo Scheme and starts winning playoff rounds.

I’m also not sure if this method, in particular, is a panacea. Maybe it can be stymied by the simple step of alternating left- and right-handed hitters at the top of the order. I’m not sure if teams will start looking to see if their Andrew Millers and Josh Haders, who dominate either-hand hitters, might be more useful pitching as openers on a set schedule, perhaps to manage their workloads more efffectively while giving their teams optimal starts to games and shortening games for what used to be the starting pitchers.

There are probably other ramifications that haven’t yet come to mind. Let me know what you think in the comments field under this story at, or on my Twitter feed @patrickdavitt.

Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.