Beware the Six-Inning Minimum

Gaming out the impact of MLB's latest proposed rule change

Over the last couple years, I have been largely critical of Major League Baseball’s efforts to reset the balance of the game via changing the rules. Stemming the tide of ever-escalating strikeouts, I have argued, must start with a game-theory-informed analysis of the pitcher, hitter, and team incentives — which the league has not always shown much interest in.

Take the new rules about defensive positioning that the league implemented last year, commonly known as the shift ban. Despite my initial skepticism, the change has achieved its primary goal of helping more balls in play fall for hits. Yet the hopeful second-order impact of reducing strikeouts did not materialize. It’s true that outlawing the shift gives batters extra incentive to make even weak contact in hopes of sneaking a dribbler through the infield. The flipside is that pitchers have just as much additional motivation to prevent the hitter from putting the bat on the ball. What’s more, I found suggestive evidence that the positioning restrictions have actually increased strikeouts, since players who were most hurt by the shift (and are therefore more valued now than they were two years ago) are also susceptible to swinging and missing a lot.

The Shift is Hitting the Ban

So I will offer credit where it is due: the new rule change being floated around the Commissioner’s Office reads like a genuine attempt to reckon with the underlying causes of the strikeout scourge. ESPN’s Jesse Rogers reported this week that MLB is considering a proposal to require starting pitchers to go at least six innings except in extenuating circumstances. The suggested exceptions would be if the pitcher reaches 100 pitches, gives up more than three earned runs, or is hurt badly enough to warrant an automatic stint on the injured list.

Mandating that the starter pitch two-thirds of the game would upend modern pitching philosophies, arguably for the better. Within my lifetime, teams have gained better understandings first of how valuable strikeouts are, then of how to create them. As a result the league now values quality of pitches far more than quantity. Faster fastballs, more sliders, and higher-effort deliveries: all in the service of inducing whiffs, at the incidental cost of durability. It’s not that teams or especially the pitchers themselves don’t care about going deep into games. It’s that shorter stints are considered acceptable tradeoffs in the pursuit of ever-more-electric stuff.

We don’t know how exactly teams would approach the new constrained-optimization problem of sustaining dominance long enough to record 18 outs, but it’s a safe bet that these trends would get dialed back if pitchers had to pace themselves more deliberately. You’d see a return to pitching to contact, to modulating one’s effort level over the course of an outing, to sacrificing some nastiness in exchange for better control. This would lead to more-hittable pitches, more-aggressive hitting approaches, and more balls in play. In short, it is one of the best proposals I have heard from the perspective of reducing strikeouts.

But as always, the devil is in the details. And the rumored specifics of this plan, particularly the suggested conditions under which a starter could get pulled early, open the door for major unintended consequences and perverse incentives.

Is Catcher Framing Cheating?

Consider the first potential exception: A pitcher could be taken out before the end of the sixth inning if they had already reached 100 pitches. Per Sports-Reference’s incredible Stathead database, starting pitchers have reached 100 pitches but failed to go six innings 157 times so far in 2024 — an average of more than once per day throughout the MLB season.¹ It happened 246 times last year, and peaked at 414 instances in 2013. We’ve seen hurlers reach the century-pitch mark over five innings or fewer 82 times this season, and three have done so in no more than four frames.²

What happens if a team decides 100 pitches is an easier goal for a starter to attain than six innings? The incentives could get muddled. A pitcher would still need to pace themselves to throw that many times, especially if they were condensed into fewer innings. But if being inefficient with one’s pitch count were seen as a sunk cost, you might see yet another increase in breaking balls thrown to induce chases. This could lead to more whiffs, more walks, and less of the ball-in-play action that the league wants to inspire.

The next extenuating circumstance by which a team could remove a pitcher early is allowing four or more earned runs. I am struggling to bring myself to write this paragraph, as the mere thought of intentionally allowing the other team to score feels so unsportsmanlike and antithetical to the ethos of professional sports that I don’t want to speak it into existence. And yet! Football players routinely stop short of the endzone in order to milk the clock. It’s considered smart basketball strategy for a team to commit fouls when they’re down by a few points at the end of the game. Would we get used to it as a tactical tradeoff, like load management or rebuilding?

Play Harder, Not Smarter

Pitchers are already willing to balk runners to third to stop them from stealing signs at second. Could it make sense to balk a runner in to bring in a fresher arm? If the game already looked like a blowout, would it be smart to intentionally walk in a run to give your starter some extra rest? Might a runner on third choose to hold up on a base hit so their teammates can get another crack or two at the struggling starter?

Then there’s injuries. On the one hand, something like the proposed mandatory IL stint for players who are unable to pitch but have not reached their innings quota would be necessary for these rules to have any teeth. Yet it would encourage more of the “phantom IL” placements that MLB is trying to crack down on — some guaranteed extra rest for a pitcher who doesn’t have their best stuff may not be a deterrent in practice.

There is also risk in the opposite direction. Imagine that a pitcher has lost some velocity by the fourth inning. They could just be gassed, or experiencing their usual amount of arm fatigue earlier than usual. An attentive manager would take them out as a precaution, and there’s a decent chance they’d be back to normal by the next time their rotation spot came around. But if they’re pushing through soreness and fatigue because the rigid rules don’t allow for respite? That’s when you’d be really worried about the pitcher getting hurt. The six-inning minimum would thus create both more fake and real injuries.

Whoever is pushing this plan in the league office would probably tell you that these concerns are the byproduct of the playing style that they’re trying to disincentivize. There’s some truth to this — if going max-effort for six innings puts players at injury risk, they’ll have to dial back the intensity.³ Yet repeatedly throwing, say, 90 miles per hour instead of 95 still takes its toll on the body. The rules ought to be nuanced enough not to discourage employers from protecting their employees from physical harm.

When Mike Trout Became the Best

Even before the game started, a minimum-innings threshold would give teams pause about sending certain players out to the mound. Consider a starter returning from the IL, whose team wants to be cautious in their return to action. They might make an extra rehab appearance or two in the minors to work up to full strength instead of risking overexerting themselves in the big leagues. Or a top prospect whose stuff is big-league ready but is still building up to a full rotation workload. MLB is specifically trying to incentivize teams to promote their promising youngsters earlier; these rules would have the opposite impact. Rookie phenom Paul Skenes was a starting pitcher in this year’s All-Star Game, just two months after throwing four innings in his MLB debut. Under this proposal, he might still be in Triple-A.

And despite what the Back In My Day crowd would tell you, starters have been failing to go deep into games since long before anyone knew what FIP and spin rate were. I again turned to Stathead to pull what these rules consider “unqualified starts” — games in which the team’s first pitcher failed to reach six innings, 100 pitches, or four earned runs — going back as far as pitch-count data is reliably available.⁴ It turns out they are not just a modern phenomenon:

To be fair, there are a lot more ostensibly unnecessarily short starts than there used to be. The 42 percent peak in 2020 (unsurprising, given that everyone’s preparedness was hampered by the shutdown and abbreviated “summer camp” preseason) is over four times the nadir of 10 percent from 2000. This year’s 33 percent unqualified-start rate is less dramatic, though more than double what it was just ten years ago (13 percent). That a third of all games started (do not) meet these criteria indeed feels like too much.

Still, this chart shows that such outings are part of the game. In 1988, the first year in the dataset, 1 in 9 starters failed to reach any of these thresholds. By 1989 it was 1 in 7. In 1990 it was 1 in 6. Surely these are more-palatable ratios than our current 1 in 3. But there is no remotely recent or well-attested precedent for eliminating short, effective starts altogether. And yes, I checked — it even happened with Nolan Ryan.

Which means we can only guess at how that would shift the balance of the league. My prediction is that offense would explode. It seems obvious that hitters would benefit — it is an explicit goal of the proposal — but I see this as an enormous overcorrection in the quest to rebalance the equilibrium between batting and pitching. Encouraging pitchers to throw more hittable strikes, nudging teams to select for stamina over stuff, tying managers’ hands from preemptively removing struggling starters, allocating more innings to lesser pitchers to cover for more IL placements: Any of these alone would be a boon for offense. All of them at once? I fear the league run environment would resemble that of Coors Field at the height of the Steroid Era, or the NCAA before the BBCOR bat standards. And that’s before considering the negative ripple effects spurred by potential edge-case incentives, like intentionally wasting pitches, walking in runs to reach the removal thresholds faster, or whatever loopholes teams discover as they solve this new constrained-optimization problem.

The good news is, this idea is still in the speculative stages. It may be significantly revised before it is implemented, or never tried at all. (I suspect the league leaked the proposal in part to get public feedback, in which case, you’re welcome!) I also respect this plan’s proponents for taking a game-theory-informed approach to baseball’s swing-and-miss problem and confounding the incentives behind our current high-strikeout equilibrium. There are some common threads in the thinking behind this proposal and my argument that we should increase balls in play by ejecting pitchers after multiple hit-by-pitches.

Plunking Down a Plan to Curb Strikeouts

Still, banning starts of under six innings is a drastic move that would completely upset the balance of the game, and if I could identify the aforementioned ripple effects of the exceptions in a couple hours, imagine what the think tanks of smart folks inside every front office would come up with over weeks or months of analysis. It’s a good starting point for future conversions, but it’s time to go to the bullpen for a fresh look.

1. Not counting players who reached 100 pitches in less time but still went on to complete six innings.

2. If you’re wondering: Cooper Criswell, Freddy Peralta, and Carlos Rodón.

3. Consider how the new runner-on-second rules for extra innings have the opposite effect: In addition to being an affront to the established rhythm and aesthetics of baseball, reducing the likelihood of a drawn-out tie game means managers can be more aggressive in burning through their bullpens.

4. I am not accounting for injuries here, so some of these early pulls may have been permissible under these proposed rules. On the other hand, a pitcher who was belatedly charged with a fourth earned run after the reliever came in would be excluded from these numbers, even though they would have been ineligible for removal at the time.