Everybody Except the Yankees Is Doing Wild Card Roster Construction Wrong
Big league front offices are, for the most part, populated by cowards. They say they want to win, but what they want most of all is to keep their jobs and not get yelled at. OK, maybe I came on too strong. That’s not cowardice, that’s just life under capitalism. A clever observer might note that one of the major sources of “getting yelled at” for major league GMs is the media — i.e., people like me. Maybe I should be nicer.
Two decades after the sabermetrics revolution shook up baseball’s norms, 30 GMs and/or Ps of BO have converged on a new conventional wisdom: You can never have enough pitchers.
During the regular season, sure. The 162-game slate is a siege by design. Teams play six times a week for six months, and most of the innings of most of those games are pretty pointless. Let’s say the average team plays about 55 defensive innings per week; the team’s top three starters and top two relievers might combine to throw, what, 25 of those innings?
Michael is a writer at FanGraphs. Previously, he was a staff writer at The Ringer and D1Baseball, and his work has appeared at Grantland, Baseball Prospectus, The Atlantic, ESPN.com, and various ill-remembered Phillies blogs. Follow him on Twitter, if you must, @MichaelBaumann.
Aaron Boone: Everyone knows that players get hurt from time to time and shouldn’t be playing. What my managerial style pre-supposes is “maybe they aren’t”?
As I’m part way through the Custer season of the rest is history
This analysis makes a lot of sense! But it doesn’t account for how much a fanbase clamors to see their ace starter pulled early in favor of a series of no-name relievers who immediately sprint face-first into the very worst-case scenario the manager was ostensibly trying to avoid. 😅
Boston fans didn’t have to wait long for another one…
This all raises excellent points but karma now decrees that today’s Yankee-Red Sox game goes 25 innings and everyone gets mad at Boone/Cashman for the roster
C’mon man! This is Fangraphs, not Fan Duel, and, while both rely and live on probabilities, the chances of another Game 3 of the 2018 World Series are very long and to fear, and prepare, for such an unlikely situation does not fit that criteria. I would like to find a place that would offer 100/1 on such a thing happening and I would be happy to oblige.
I’m not sure the white sox have any position players to call up tbh.
Or pitchers who can go deep into games. They’re kinda stuck having to go at least 4 pitchers per game.
A 5-and-dive plus three relievers in each game means you need 9 pitchers in a three game series.
Call it 10 or 11 to be safe.
Feel the same about the Padres. None of the 4 players on the 40 man have a difference making tool for playoffs.
I often have a hard-time telling the difference
I mean, I’m not sure if was that different under Stalin or whatever
Yeah I bet hockey managers in the Soviet Union didn’t like getting yelled at either.
I follow and am persuaded by most of this but I don’t think it is quite so easy to dismiss the dangers the Yankees strategy has if there is a long extra-inning game. Not saying games like Game 3 of the 2025 World Series are something you should plan for, but an 18 inning game would be extremely dangerous for the Yankees, even if Rodon and Warren each go 3 innings in relief.
… why? Rodon and Warren could easily soak up those innings
Why can’t either of them go 5, or even 7, like Eovaldi did in Game 3 of the 2018 WS? They started all year and have made numerous starts of several innings. It’s not likely, actually impossible, that, if used, they are going to be needed again until the next series
You generally do not expect or want starting pitchers who are pitching in relief to throw that many innings. On a day they sitting in the bullpen, starters have not prepped to throw that pitches. Eovaldi is an entirely different case because he was the Game 4 scheduled starter and would have been going through preparation to go length that Rodon/Warren will not. Especially when the Wild Card round just started and they have not prepared in the bullpen for very long.
A 15/11 split would have been much safer with little downside. There are Yankees not named Ali Sanchez that are going to sit on the bench and get at most a single at bat this round. Was that really worth the risk of running out of pitchers?
The only reason I agree with you is I find Amed Rosario more or less worthless at this point. Then again, not only is he on the roster, he started last night.
The Yankees have two starters in their bullpen and a third guy in Yarbrough who’s built up to pitch multiple innings. Not a concern.
Yarbrough actually feels like a guy who could just throw unlimited innings.
Yes, but not only is the probability of this low, but if they carried one extra one inning pitcher, how much difference does it REALLY make?
If you account for occasionally using a throw-day starting pitcher as a SIRP that’s one less reason to carry that 12th pitcher.
Furthermore if you wanted to be really cutting-edge, you could name a pitcher as “injured” and replace him mid-series if you wanted to. Now, that’s not going to happen with anyone high-leverage, but if the game goes 18 innings and you blow through both of your bulk contingencies there’s usually an SP6 in the org somewhere who could take the place of one of them.
If he isn’t hurt, this is dicey. MLB plays a role in allowing a replacement. Are you counting on a conspiracy where a player just willingly gives up any chance to be on the roster for at least the next series? Seems like a huge risk to count on this to work and not end up being a giant cluster-you-know-what.
I do think it’s widely accepted that during the regular season, at least, clubs play the “sore arm” game to shuffle pitchers back and forth from AAA to keep the MLB bullpen fresh all the time and no one thinks twice about it.
My comment was a bit tongue in cheek, mostly because planning to use a fake injury to restock a bullpen is a shit strategy. It does, however, at least present as an emergency option if the organization is worried about running out of pitchers. Which … clearly … most organizations do seem to be worried about.
One interesting thing to me is that Cashman has only 7 players on multi year contracts. And he does not seem to blow relatively big money on non-core players the way some other rich teams tend to. Stanton and McMahon, who they traded for, come off the payroll after 2027.
I think blowing relatively big money on McMahon types in an attempt to maintain luxury tax flexibility has played a massive role in how long it’s been since they’ve won a proper playoff series against a team from outside the AL Central…but that’s pretty firmly coming from over Cashman’s head.
Sure, the Yankees – with three frontline, workhorse starters and a deep reserve of useful position players – seem perfect for this strategy, but I don’t know that I otherwise agree, especially for teams with shakier starting pitching.
I could absolutely see going 15/11 with an another pinch runner as being a better option for most teams, but beyond that, unless you’ve got an extra plus platoon bat lurking, or a roster with a bunch of interchangeable players with varied skillsets, I don’t really see the upside in exchanging your insurance against a 20 inning marathon for a bench guy you likely don’t want to use anyway.
four frontline, workhorse starters to be honest. Rodon was CY6 last season.
Game 1 played out perfectly in this regard: the Yanks used 3 pitchers, and Blackburn didn’t exactly wear himself out (18 pitches). Boone aggressively pulled Rosario when it was clear the lineup wouldn’t turn over to him a third time with Tolle still in the game.
That having been said, the fact that the roster includes a guy who is almost literally DH-only (Stanton), a guy who should basically be DH-only (Rice), a guy who for some reason Aaron Boone is convinced can actually field a position but cannot (Rosario), two first basemen (Goldschmidt/Garcia), and a third-baseman/emergency SS (McMahon), the sheer number of guys is necessary to have anything resembling flexibility.
Also, it’s worth mentioning, that in the year of our Lord Twenty Twenty-Five, Giancarlo Stanton played actual games in the actual outfield. So presumably, he still could find his OF glove. That’s not to suggest he should, but it’s less preposterous than it sounds.
Ive always felt the playoffs should reflect the regular season more in format. Fewer rest days, no breaks between rounds. Those 5 and 7 game series would hopefully lead to more true talent results.
For the 3 games though, this makes a ton of sense. Maybe even call up the fastest guy on in your org just for a pinch run even if he’s completely unplayable in any other situation.
So you want worse baseball played by the lesser known players in the most important situations in the most high profile games? That doesn’t sound like presenting your best product (after all, the point is still entertainment). I get not setting it up to allow a team to ride 2 pitchers to a title but ensuring that the best players have sufficient rest to play the biggest games on the biggest stage seems to be the right answer. Besides how does forcing the bottom 10%-15% of a roster to play help arrive at a “true talent” result? I doubt that anyone would say that the Braves putting JR Ritchie on the mound in the 7th with their season in the balance is either compelling or representative of their true talent level, yet he did soak up 70+ innings for them this year. If anything that gives you a greater chance at random variance. If you want to increase the likelihood of true talent rising to the top then play 15 or 19 game series instead of 5 or 7.
“True talent” reflects the ENTIRE roster, or do you not get that?
Actually, it doesn’t. True talent level is a term used in regard to random variance in performance versus true skill over a larger sample, which comes back to my point. You can play the entire 40 man roster and not reflect a team’s true talent over 7 games. Even in a scenario in which teams play 6 games in 7 days (essentially impossible due to the extra travel in the postseason) you might get 4 innings from the two guys at the end of the bullpen. To get to true talent outcomes (which is impossible as variance comes into play over 162 games too) you’d need larger series. If we aren’t going to play 15 or 19 game series (and we aren’t) ensuring that the upper 90% of the best teams determine the outcome is a perfectly reasonable, and more compelling, way to determine a winner. Otherwise, just crown the regular season champ and move on with our lives.
I really liked that about the format of the 2020 playoffs. It required more starting depth and “looked” to me more like normal baseball than something totally different. I’m biased as a cardinals fan who watched his team get bulldozed by the 2019 Nats, who only had like four good pitchers, but still would love to see fewer off days in the series themselves. Let the playoffs feel like a rush!
the Cards put up a .374 OPS over 4 games in 5 days. Y’all had two baserunners in Game 1 against Anibal Sanchez. I’m not sure what fewer off days does for that.
If you played six times a week in the playoffs, then you could also go to a five or seven game wildcard and seven game divisional round.
Short series are random.
Hot take: the wild card should be random, it’s right there in the name!
Yep – if you don’t like it, win your division!
One thing I wonder about is whether some of those unused guys still have some impact just by way of strategy. One team’s willingness to use the next reliever/pinch hitter depends partly on the expected countermove. I think more often than not, teams just make the move assuming there will be no counter (or not caring about it) and adjust accordingly. But not always. If you sub in the lefty killing hitter against the LHP who is beyond 3 batters, you have to consider whether the opponent may bring in a RHP (which depends on whether there are RHP, how good they are, how good the LHP is, who else is coming up in the batting order, etc.). If there are no more options in the pen, then you don’t worry about it. And of course if there are no more options on the bench, the pitching team doesn’t have to worry.
A lot of this depends on the credibility of the threat — do you have surplus bats/pitchers that a team would bother game planning around? I think there are more teams who trust their marginal relievers than those who trust their marginal bats. On the other hand, your extra wheels and gloves are probably the most trustworthy in the sense of knowing what you’ll get, it’s just that the marginal gain may not be much to write home about.
“Big league front offices are, for the most part, populated by cowards. They say they want to win, but what they want most of all is to keep their jobs and not get yelled at. OK, maybe I came on too strong.” NO! You didn’t come on too strong. You nailed it! I puked at the the roster construction of the teams that thought they needed 12 pitchers for a three game series. The Red Sox fit the template Michael mentions about the Yankees as well. It hurt them a lot when Mickey Gasper went down but they still had IKF and Monasterio and should have called up Siegler. Jahmai Jones and Romy Gonzalez have also been solid contributors at various times. There could have been opportunities for any of them to help. I find it hard to believe that these teams could be so ignorant, foolish, stupid or any other synonym that fits this scenario.
Siegler and Gonzalez are both hurt, and Jones is leading off tonight.
It seems part of the roster decisions for both Boston and NY were in carrying guys returning from injury and hoping they could start to work them back in for hopeful use in later rounds – Crochet for the Sox, Stanton and Grisham for NY. Hence the BoSox had an extra pitcher, the Yanks an extra player.
You’d have never gotten to a “sabermetric revolution” if not for capitalism.
There’s nothing “capitalist” about the MLB.
It’s a classic example of a cartel.
…which is what the fans want, as expressed through the politicians that allow the anti-trust exemption.
If we’ve learned anything from recent history, it’s that politicians are very beholden to the wishes of their constituents.
Fans want data centres and high gas prices, too!
It wouldn’t be a Baumann article without a dig at capitalism or “the rich”
Great “point” (decimal point???)
Luckily for the Phillies, their games were during the afternoon east coast time, so hardly anybody noticed.
EDIT: two minutes after I posted this, the Phillies came back from being down 3-1 in the 8th to tie it lol
I’ve been talking about how rosters in the playoffs should be different than in the regular season for years. You probably aren’t going to use your 5th starter (unless he can play up in relief) or your 2 (3?) worst relievers, so why are they here?
The example I always use is for position players is Billy Hamilton. You can’t really afford to carry that guy on the roster for 162 games because your bench players need to be able to play at least once a week to give your starters rest. But in a short series where you might start the same 9 bats every game? Having someone who can take an base a plodder couldn’t or take away an extra base hit hit in a close game late might be the difference between champagne and tears even if he never comes to the plate.
There does need to be someone to pitch in unwinnable (or losable) games. Like down 10 up 10 don’t need your best arms, I’d go as far as to say up or down 5-6 runs should be your worst rostered arms but the coaches aren’t gonna do that. But overall, agree.
“Ghost runner” my eyes!
It seemed surprising to me that they carried Stanton in the playoffs then didn’t start him vs a lefty. I mean I get he probably isn’t ready, but then why carry him at all?
He’s basically there to take one shot a game — hit a HR, make an out, or get pinch run for if he gets a hit. It is at least something to give Chad Tracy another variable to think about.
Let’s put it this way: it is highly unlikely that Stanton’s presence on the roster will matter negatively, and it has a puncher’s chance of being the difference in a close game.
EDIT: it’s certainly going to factor into the Red Sox’s decision making a lot more than whatever 11th arm the Yanks stuck on the end of the roster.
I don’t have a strong preference in this game, I guess I’d lean Red Sox but as a sentimental guy I’d sure love to see a vintage “Playoff Stanton” at bat.
The White Sox care not for this.
And eventually they are going to run out of middling teams to play.
It’s hard to understand why the playoffs aren’t reseeded after round 1. The ChiSox should be facing the Rays now, not Cleveland. As a division winner the Guards should still have home field but ‘cmon. Is it fair to the Rays to have to now face NY as the best team in the AL?
How much time do opposing teams spend preparing for certain specific pitchers?
(If any?)
I’m wondering if having extra pitchers might be like in Football when teams occasionally run out weird formations/trick plays just to force opponents to spend time thinking about it.
Might be on a player-to-player basis, but I imagine some guys study each pitchers’ arsenal and throwing motions and develop some personal plan of attack.
Along the same lines, I bet that teams have more advanced analytics that show optimal matchups in more nuanced ways than just lefty-righty. This is less a factor with the 3 batter minimum, but still could be – just because you only use 10 pitchers doesn’t mean you know for sure which 10 guys.
Of course I think the mere possibility of an infinitely long game in a short series is the real driving factor.
I think there were a few relevant details omitted.
First, just empirically, how many PAs has the last man on the bench gotten across these 32 best-of-three series? Maybe there’s an asymmetry here, but maybe teams only use 20 or 21 players and it doesn’t much matter who the rest are.
Two, why do we think the optionality provided by having extra position players is more than that provided by having extra pitchers? Maybe I’d believe that the 3 batter rule skews it but it’s not straightforward that a team couldn’t create a Mason Miller in the aggregate in the same way as the Elly De La Cruz example.
Three, pitchers get hurt more often, which means that if all else is equal, you’d rather carry extra pitchers. It’s a longtail risk, but I wonder how many relievers getting hurt in a row would be rarer than losing two catchers, for example. So you need to actually have a clear strategic advantage to the position player optionality for it to skew that way.
I mean, it’s obviously not a sample size, but the Yankees had 3 pitchers and 2 hitters not get into a game and the Red Sox had 2 pitchers and 2 hitters not get into a game. Not sure I see much competitive advantage either way.