UTK Special 7/15/26

Never Be Another, Unless There Is

After Justin Verlander announced his retirement just before the All Star break, there were the inevitable “There will never be another Justin Verlander” pieces, just as we will get when Max Scherzer retires and that we got when Greg Maddux, Randy Johnson, Roger Clemens, Curt Schilling, Pedro Martinez and I could go on, but you get my point. Greatness is isolated, but seldom singular. One piece from Bradford Doolittle did more than just the panegyric and pointed out that the structure of the game has changed so much that getting to 200 innings is now rare, let alone 250, just as 300 was beyond Verlander’s reach - not because he couldn’t but because he was a decade late to get the chance, probably for the better.

The five-man rotation has become one of those things baseball accepts without remembering why it exists. The Mets, and then the Dodgers, had five good starting pitchers, so they used them that way and that makes sense. The problem is that the teams now seldom have five good ones even when they use eight or more over the course of a season. They steal starts from their best and give them to the guy they hoped they wouldn’t have to use in equal measure. The funny thing is that the game has never been shy about reinventing itself everywhere else. Bullpens went from an afterthought to a collection of specialists. Fireman became a thing, then firemen became closers, and now managers are rediscovering that your best reliever doesn’t always pitch the ninth. Defensive shifts were remembered from the 1940’s and weaponized against everyone, not just Ted Williams, before the league legislated them away. The opener arrived being ridiculed until it quietly stopped looking ridiculous. Through all of that, one assumption has remained remarkably untouched: starters throw every fifth day because that’s what starters do.

Watching Paul Skenes and Jacob Misiorowski this season got me wondering whether we’ve backed ourselves into a corner. These are the kinds of pitchers organizations dream about, the rare combination of elite stuff, elite athleticism, and the ability to dominate a lineup from the first inning. They’re also exactly the pitchers everyone is terrified of breaking, even knowing that the Brewers have been researching just this kind of workload monitoring and biomechanical correction for almost two decades. The result is a strange compromise where we spend years developing a thoroughbred only to ask him to run fewer races. Modern aces make roughly 32 starts if everything goes well and “everything goes well” has become an increasingly ambitious assumption. The instinctive response to suggesting a four-man rotation is that you’d be asking pitchers to throw themselves into reconstructive surgery. The more I started running the numbers, though, the more I realized that instinct might be aimed at the wrong target.

I am certainly not proposing a return to 1972 or even 1992. This isn’t about complete games, 140-pitch outings, or asking someone to face the middle of the order for a fourth time because he “looks strong.” Baseball has already answered those questions. Starters are leaving after five or six innings whether they’re on a four-man rotation or a five-man rotation, largely because lineups improve dramatically the third time they see the same pitcher. The bullpen already owns a third of every game. If that’s the reality we’re living in, then maybe the better question isn’t how long a starter should pitch. Maybe it’s how often.

A true four-man rotation across 162 games would theoretically produce about 40 or 41 starts per pitcher, but nobody actually schedules baseball on the back of a cocktail napkin. Off-days interrupt the rhythm. Rainouts happen. There are stretches where using a fifth starter - what we used to call a swingman - makes perfect sense, and stretches where the calendar practically begs you to skip him. In practice, the frontline starters would probably wind up making something like 37 to 39 starts. Suddenly the innings look different. Five innings across 38 starts produces 190 innings. Average five and two-thirds and you’re over 200. Let someone work six efficient innings and you’re in the neighborhood of 225 without ever asking him to approach the kind of pitch counts that have become taboo. The extra innings don’t come from asking a pitcher to survive longer. They come from asking him to appear more often.

That’s where the conversation usually shifts toward workload and understandably so. Baseball has become fluent in pitch counts because pitch counts are easy. They’re objective, they’re public, and they fit neatly into a television graphic. The problem is that elbows don’t own calculators. They don’t know whether the scoreboard says 87 pitches or 93. Tissue responds to stress, recovery, adaptation, sleep, nutrition, previous workload, mechanics, and probably another dozen variables we still don’t understand very well. Counting pitches is useful, but confusing the measurement with the phenomenon has always made me a little uneasy.

One of the more influential attempts to quantify that phenomenon came from sports scientist Tim Gabbett, who developed what became known as the Acute:Chronic Workload Ratio. If you’ve never run across it before, Gabbett’s own original research is worth reading because he explains both what it is and what it isn’t better than anyone. The basic idea is elegantly simple. Compare what an athlete has done over the last week with what he’s been conditioned to handle over roughly the last month. If the recent workload dramatically exceeds the established workload, risk may increase because the athlete is operating outside the environment his body has adapted to. The ratio became enormously influential not because anyone believed it could predict injuries with precision, but because it gave coaches and sports scientists a common language for discussing workload instead of arguing exclusively about minutes, miles, or pitches.

Like almost every useful idea in sports science, ACWR has accumulated critics as quickly as admirers. Researchers including Franco Impellizzeri have argued that the mathematics behind the ratio can exaggerate relationships or produce misleading conclusions, particularly when rolling averages overlap. They’re not wrong. ACWR has never been a crystal ball and treating it like one probably delayed more productive conversations about workload than it advanced. Even so, it’s still a useful framework because it forces us to think beyond pitch counts. Baseball has become so conditioned to asking how many pitches a starter threw that we’ve almost forgotten to ask how much work he actually performed, and that we’re very bad at measuring even pitch counts.

So I built a simple model. Assume a modern starter averages five and a half innings* while throwing roughly 16 or 17 pitches per inning. That’s about 90 game pitches. Add the bullpen session before the game, the warmup pitches between innings, and the various throws required simply to prepare for competition, and the start day probably represents something like 130 meaningful throws. Recovery day includes a light game of catch. The following day features a bullpen session. The day before the next start becomes another relatively easy throwing day. By the time everything is added together, the pitcher accumulates roughly 245 throws during a four-day cycle.

*Yes, I know a half inning doesn’t exist. Baseball works in thirds, but if a pitcher throws one start of 5 1/3 and one of 5 2/3, you can see what the average is. I made this round and simple because, well, I am also round and simple.

Here’s the part that surprised me. Once the pitcher settles into that routine, the workload ratio hardly budges. Running a traditional seven-day acute window against a rolling 28-day chronic window produces ratios that hover remarkably close to 1.0, drifting somewhere between roughly 0.85 and 1.10 depending on where the calendar catches the cycle. In other words, the model doesn’t produce repeated workload spikes. It produces a consistently elevated baseline, and those are two very different physiological questions. The chronic workload rises along with the acute workload, meaning the pitcher is continually operating inside the environment his body has already adapted to rather than repeatedly jumping outside it.

That distinction matters because adaptation has always been the point, even if baseball sometimes forgets it. Nobody worries about a marathon runner covering fifty miles this week if he’s covered fifty miles every week for the last two months. They worry about the guy who ran fifteen, got inspired by something motivational on Instagram, and decided Saturday looked like an excellent day to attempt twenty-six. The body isn’t offended by hard work nearly as often as it’s offended by surprise. Consistency is usually kinder than volatility and that’s one reason workload models exist in the first place.

Spring training, interestingly enough, becomes far more important than Opening Day. If pitchers spend six weeks preparing for a five-man schedule and then suddenly begin taking the ball every fourth day, the acute workload jumps before the chronic workload has time to catch up. If, however, the entire spring is built around gradually increasing the frequency of starts while keeping individual outings relatively modest, the transition into the regular season becomes much smoother. That’s a coaching problem, not a mathematical one, and baseball has become awfully good at solving coaching problems when it decides they’re worth solving.

The other interesting wrinkle is that the increase in actual game workload isn’t nearly as dramatic as people imagine. Take a pitcher making six starts over four weeks at 98 pitches per outing. That’s 588 competitive pitches. Now compare him with a four-man starter making seven starts at 88 pitches each. That’s 616 pitches, barely a five percent increase despite adding another appearance. The pitcher reaches 200 innings not because every outing becomes longer, but because the innings are distributed across one additional start every month. That’s a fundamentally different workload than asking someone to grind through eight innings every fifth day, and it changes where the stress actually lives.

Of course, none of this answers the biological questions and pretending otherwise would be irresponsible. I don’t know whether Skenes would recover beautifully every fourth day while Misiorowski needed the extra twenty-four hours, but the Pirates likely do and I know the Brewers do. I don’t know how velocity, biomechanics, sleep, travel, nutrition, and individualized recovery programs interact with a compressed schedule. Modern pitchers throw harder than any previous generation and they do it with an efficiency that places enormous stress on the elbow and shoulder. Those realities don’t disappear because a spreadsheet looks encouraging. What the spreadsheet does suggest, however, is that workload itself may not be the barrier we’ve assumed it to be.

That’s really the point. Baseball has spent the last decade optimizing around inherited assumptions instead of testing whether the assumptions still deserve to exist. We’ve accepted that the path to protecting elite arms is making fewer starts, even though the evidence supporting that conclusion has always been surprisingly thin. At the same time we’ve watched injuries continue to climb, velocity continue to rise, and teams become increasingly dependent on bullpens that are carrying workloads of their own. Perhaps the five-man rotation truly is the optimal solution. If it is, fantastic. Keep it. But it ought to survive because we tested the alternatives, not because nobody felt like asking the question.

I’d love to see one organization treat this as a genuine research project instead of a philosophical debate. Pick the right pitchers. Build the spring around the schedule. Measure recovery markers, biomechanics, strength, soreness, sleep quality, and performance. Follow the workload every day instead of just the pitch count after every game. Maybe after two years the answer is that four-man rotations are a terrible idea and the team goes back to the status quo. So be it. I’d be perfectly happy with that outcome because at least we’d know. For a team like the Rockies or Angels or the suddenly pitching-light Cubs, it can’t be much worse. Right now, though, baseball has convinced itself that the answer is obvious without ever performing the experiment and history has a funny way of reminding us that the game’s strongest traditions are often just yesterday’s best guesses.