Platoon Bats for Ottoneu

I don’t think I am breaking new ground by suggesting that platoon bats are useful for Ottoneu. The combination of daily lineups (and most Ottoneu leagues are daily, though not all) and deep rosters makes it possible to roster players you know you can’t start daily, but who can still provide value.

It’s pretty easy to look at a player’s platoon splits but platoon splits don’t tell the whole story. Because you aren’t starting a player for his PA vs. RHP or LHP. You are putting a guy into your lineup or on your bench based on who the opposing starting pitcher is, and a lot can happen after that.

Some guys crush one side and are so bad against the other side that they give the production back. Some guys get pulled for a PH as soon as they face a same-sided pitcher, so they get far fewer PA, driving down their opportunity to pile up points per game. Some guys can hold their own just enough against same-handed pitching that they get to stay in the game, or play good enough defense that they stay in the game, or aren’t the biggest threats on their team so other teams don’t optimize their pen to face them.

All of these differences go beyond a simple L/R split and add up to potentially meaningful changes in a player’s P/G performance.

Using Baseball Reference’s Stathead, I pulled stats and calculated P/G for players based not on their L/R splits, but based on the handedness of the opposing starting pitcher. As an example, if a righty starts a game, a hitter who scores 5 points in that game would get credit for 5 points in a game started by a RHP, even if they only faced that RHP one time and faced lefties the rest of the time.

I only pulled players with 200+ PA in games started by RHP (297 hitters) and 100+ PA in games started by LHP (237 hitters). That gave me 336 total players, with 195 on both lists, 102 on the RHP list and 39 on the LHP list. And from that, I filtered down to four groups: Hitters who only got regular playing time against RHSP, hitters who got regular playing time vs. LHSP, hitters who should be benched vs. RHSP, and hitters who should be benched vs. LHSP.

1) Only Play in Games Started by an Opposing RHP

These are the players who were either already platooned by their teams or, for some other reason, simply didn’t get to 100 PA in games started by LHSP. Of the 102 players who had more than 200 PA in games started by RHP but less than 100 in games started by LHP, 23 put up more than 4.5 P/G in those games against RHSP.

Some of these guys weren’t on the LHSP list simply due to injury and missed time, while others were actually platooned by their teams. These guys, in general, make for good platoon bats. Pasquantino needs to take a step forward – you can’t start a platoon 1B with 4.56 P/G against RHSP – and some of the OF are borderline.

Nootbaar and Wade stand out as guys who, even with the platoon advantage vs. the SP didn’t do enough last year to be sure-fire starters. An OF who puts up 4.6-4.7 P/G might be one of your best five on any given day (they probably will be) but it isn’t a guarantee. That means that if you had Nootbaar or Wade on your roster last year, it wasn’t as simple as saying “start them against all righty starters.” Butler and Raley, on the other hand, could be pretty much locked in as 4th or 5th OF options on days they faced RHSP and you would be in a good spot.

The guys at the top of the list are worth watching because their P/G when a RHP starts are so good they might be swamping really poor performances when a lefty starts. We’ll talk more about that on list #3 below.

2) Only Play in Games Started by an Opposing LHP

These are the players who were either already platooned by their teams or, for some other reason, simply didn’t get to 200 PA in games started by RHSP. This is less common than the other side – it’s simply harder to make a career (or even a season) as a hitter who hits lefties but never plays against righties. Of the 39 players who had more than 100 PA in games started by LHP but less than 200 in games started by RHP, 13 put up more than 4.5 P/G in those games against LHSP.

The bottom of this list is way less interesting. Even a guy like Robles is more challenging to roster than someone like Winker or Abreu from the previous list. Last year, pitchers made 4,858 starts (162 games per team time 30 teams minus two teams that only played 161) and 27.2% were started by lefties. So if you have a guy like Robles, there are only about 44 games a year that he is even a reasonable candidate for you to start. But at 4.78 P/G, there will be some of those 44 games when you have five other OF you would rather start. Or some of them he will face a tough lefty and you will prefer to bench him anyway. Suddenly you are rostering a guy who you can only use 20 or 30 times.

The guys at the top of the list you can at least reliably count on using when they get those starts, but even then, 40ish games only has so much value.

3) Started Games Against Both RHP and LHP, but Performed Better vs. RHSP

These guys were not platooned by their teams, but maybe should have been. Or maybe you should platoon them. Because they hit when they started against RHP, but not when they started against LHP. Of the 195 players who had both 200+ PA in games started by righties and 100+ in games started by lefties, some of the biggest gaps between performance in those games came from guys like Shohei Ohtani (10.23 P/G in games started by righties, 7.48 P/G in games started by lefties) and Brent Rooker (6.38 P/G in games started by righties, 8.88 P/G in games started by lefties). There is nothing informative there (Ohtani is still good against lefites; Rooker should still start vs. righties) so I filtered this list and the next to players who were over 4.5 P/G against their better side, under 4.5 P/G against their worse side, and had a gap of at least 1 P/G between them. Here are the 23 hitters who meet that criteria and favor facing RHSP:

Some of these names are not surprising and aren’t that interesting. If you rostered Bleday, you knew not to start him against LHP. Same with Castro, Burleson, and others.

But some of these are interesting. As great as Jarren Duran was last year and as much as his defense kept him in the lineup, he was actively hurting you in games started by LHP. You should have benched him last year against lefties. Austin Riley, James Wood, Jackson Chourio, and Cal Raleigh are others who many managers might have treated as set-and-forget everyday starters; all would have been more useful to you if you had benched them against lefty starters.

I want to note, though, that your threshold for starting a player might be different at OF vs. MI or 3B, and not just because replacement level is different. Compare Duran and Burger. Most Ottoneu teams carry 7-10 OF, meaning that on a day Duran faces a lefty starter, you very likely have a better option than his 4.15 P/G. And if you are rostering Duran, you can plan to only start him vs. RHP, knowing you might sometimes slot him in vs. a lefty starter.

But most teams do not carry a ton of 3B depth. You may have Burger and a 2B/3B guy you use to play some MI, or a 3B/OF guy you slot into the OF often. Carrying an extra 3B specifically to bench Burger vs. LH starters might impact how you want to build your roster. Taking his 4.22 over 40ish games might be a better decision than rostering Jose Miranda in the hopes that when Burger faces a lefty, Miranda will face a righty so you can squeeze out another 0.29 P/G.

4) Started Games Against Both RHP and LHP, but Performed Better vs. LHSP

Same filters here, but the other direction, and you end up with 29 names this time.

The message here is mostly the same, with different names. Heliot Ramos, Starling Marte, Randy Arozarena, Ozzie Albies, and others simply should not have been in your lineup against RHSP last year. Jorge Soler, Steven Kwan, Brenton Doyle, among others, might require more thought than just plug-and-play.

Concluding Thoughts

I am rethinking how I value a number of these players. In some cases, especially some of the guys who crush righties, I might increase my valuation on them because I think I can optimize their usage more with this data. Others I might decrease because I think I can use them less.

Having said that, there is a big caveat here which is that I have no idea how sticky this is year-to-year. Some of what drives this is platoon splits and those can be sticky. But some might be team usage, how opposing teams treated them, the noise inherent in partial-season sample sizes, etc. Is Duran really that much worse when a LHSP is on the mound, or did he happen to face a number of particularly tough LHSP this year? So while I am rethinking some guys, I am doing so cautiously and with a grain of salt.

But this is something I’ll be watching going forward.