The Atlantic League Is Testing New Rules

Gregory Fisher-USA TODAY Sports

The new rules introduced to MLB this season have been the subject of much discussion. From the larger bases and pickoff rules leading to a rise in stolen bases, to the shift restrictions resuscitating the pulled groundball, to the potential impact of the pitch timer on the fan and player experience, the 2023 rule changes have been the most significant in recent memory.

Whether or not new ideas end up in the final rulebook, the league is constantly innovating to determine whether further changes would improve the game. In 2019, MLB reached an agreement with the Atlantic League, considered to be the highest non-affiliated league in the country, to “test experimental playing rules and equipment during the Atlantic League’s Championship Season.” Recent rule changes like the larger bases and limited infield shifting were piloted in the Atlantic League before coming to MLB, but not every rule tested there has been or will be implemented in the majors. In 2021, the Atlantic League moved the pitching mound back by a foot, breaking a century-long custom. The goal was to give hitters a better chance against premium velocity while also allowing breaking balls to shine more, but testing found that the change hadn’t accomplished its intended effects. Last week, MLB announced its new experimental rules for the 2023 Atlantic League season, which begins today:

While the single disengagement rule is a furthering of the two step off limit currently in place in the majors, and a different version of the double hook has already been in use in the Atlantic League for two seasons, the designated pinch runner is a brand new (and very unorthodox) proposal. The rule allows a non-starting position player to pinch run for any baserunner without removing the baserunner from the game exactly once during each game. (The second time the designated pinch runner enters, it’s treated like a normal substitution and he must remain in the game.) The intent behind this rule is clear – faster baserunners make action on the basepaths more exciting. Stolen bases are great! Batters taking an extra base on a single or a fly out is fun! And it’s equally exciting for a catcher or outfielder with a cannon arm to throw out a would-be base swiper. What could go wrong?

A lot, actually. The adoption of the designated pinch runner rule may increase the incentive for teams to start slow, lumbering hitters. After all, why worry about having a Miguel Cabrera or a Luke Voit on the bases if you can simply pinch run for him without penalty after a walk or single? This rule could prove counterproductive, reducing the amount of in-game action on defense and the basepaths. Furthermore, slower hitters also tend to put fewer balls in play. Looking at the slowest hitters down the line last season (excluding catchers), most have outlier Three True Outcome rates as part of their offensive profile:

Slowest MLB Hitters in 2022
Name Home-to-1st Time TTO%
Miguel Cabrera 5.13 31.0%
Albert Pujols 5.08 30.5%
Giancarlo Stanton 5.05 48.3%
Jesús Aguilar 5.01 32.2%
Kyle Lewis 4.97 43.6%
Luke Voit 4.96 45.1%
Jed Lowrie 4.92 31.0%
Daniel Vogelbach 4.90 44.4%

Batsmen like Voit and Stanton avoided a sprint to first nearly half the time, and even when hitters with average Three True Outcome rates like Cabrera and Pujols put balls in play, their five-second journey to first often looked more like a warmup jog than a sprint for all parties involved.

In addition to affecting the quality of play on the field, the designated pinch runner violates one of the most important rules of baseball: “A player once removed from a game shall not re-enter that game.” Rule 5.10(d) isn’t just a law of the game, it’s a primordial axiom of baseball that distinguishes it from other sports that allow re-entry after a player is substituted out. Rule 5.10(d) is the principle that underlies many of the game’s fundamental strategic decisions, like when to pinch hit or pinch run, and inspired incredible and bizarre innovations like the Waxahachie Swap. Allowing a baserunner to be substituted in and out without penalty doesn’t just contradict existing rules, it defies the thing that makes baseball baseball.

If the designated pinch runner were to be implemented, the league’s best baserunners could provide considerable value to their clubs. While baserunning specialists like Terrance Gore and Tim Locastro have never had full seasons on a big league roster due to their lackluster hitting, having them available to pinch-run every game without making them grab a bat could make them valuable bench players. In Gore’s 112-game regular season career (of which he entered 67 as a pinch runner), he’s generated +8.2 BaseRuns through a combination of an 83% success rate on steals and his ability to take extra bases at an elite rate. If he kept up that pace over 150 games as an everyday pinch runner, he could generate 11 BaseRuns on the bases. In ideal circumstances, a team could get themselves a 1-WAR bench player who doesn’t even own a bat. Of course, Gore is an outlier even among the upper echelon of runners, and most hitters couldn’t generate that kind of value. But a player’s ability to gain a few extra runs with their legs in high-leverage moments could be the difference between a spot on the big league bench or the minor league shuttle.

The early returns on MLB’s two disengagement rule have been tremendous. As the end of April nears, the league is on pace for 3,407 stolen bases, nearly a thousand more than in 2022 and just a couple hundred fewer than the post-integration record of 3,585 set in 1987. In addition to the few extra inches afforded by the larger bases, forcing pitchers to ration their throws over to first has given speedsters the power to take off with significantly less fear of being picked off. According to Baseball Savant’s catcher throwing metrics, the median expected success rate has climbed from 75% in 2022 to 80% this month. Much of this increase in success has been driven by steal attempts after one or more disengagements, as runners have an 81% success rate after one step off and haven’t been caught after two throws over. With only one disengagement allowed in the Atlantic League, pitchers may be afraid to exhaust their only pickoff attempt in (logical) fear that the batter will take off after their attempt has been used. Basestealers will continue to attempt to find the exact lead distance to bait a throw from the pitcher without risking a pickoff, and should they succeed, they could swipe bags with impunity.

The current pickoff limit has already brought steals near an all-time high. But there’s a critical difference between 2023 and 1987. In 1987, catchers threw out attempted thieves about 30% of the time. So far this year, batteries only have a success rate of 20.4%. In the sabermetric era, the break-even point on steals is well understood, and teams have pushed in favor of efficiency on steals ever since. In 1977, Dave Parker stole 17 bags, but was also caught 19 times. Had he played today, he likely wouldn’t have more than a couple of attempts. If the rules are pushed too far in favor of steals, the intrigue and excitement of a bang-bang play could fade as the outcome becomes more and more predetermined. Watching a runner take off towards second becomes far less suspenseful when you know he’ll slide in safely nearly every time.

Stolen bases are an area of the game where teams’ pursuit of efficiency has directly clashed with the amount of action taking place on the field. The 2010s were a dark age for the swipe after the “run wild” era of the 1980s and ’90s, and while teams gained more runs with their abilities on the basepaths, the number of opportunities for a thrilling moment dropped precipitously. The disengagement rules have brought back the stolen base, but further restricting pitchers to one step off might not bring back the excitement along with the steals.

Finally, the Atlantic League has modified its “double-hook” rule. Under the original double hook, a team’s DH was tied to its starting pitcher – the starting DH would have to leave the game or play the field to remain in as a hitter. This proposal, supported by MLB managers like Buck Showalter and Dave Roberts, was intended to bring a balance between the AL and NL styles of baseball before the universal DH was implemented, as well as encourage teams to keep their starters in the game longer. The new rule, which allows teams to keep their DH if the starting pitcher finishes at least five innings, was clearly designed with the latter goal in mind.

When the double hook was proposed in 2021, the average length of a start had decreased by a full inning and 14 pitches compared to just the decade prior. In 2018, the Rays famously (or infamously) reintroduced the use of the opener, where the likes of Sergio Romo, Ryne Stanek, and Diego Castillo started a combined 45 games, making single-inning appearances to support bulk arms on an injury-ravaged squad. On that same Tampa Bay team, Blake Snell won the Cy Young while averaging just 5.8 innings per start, the lowest of any award winner in MLB history. While the opener has somewhat fallen out of favor (likely because it’s been observed not to be a miracle, revolutionary tactic), openers are still used on occasion and many starters enter games with the understanding that they won’t be asked to go deep into them. In 2022, 25 starters averaged fewer than five innings per start – as recently as 2011, that number was zero.

To examine the potential effectiveness of the revised double hook, I looked at all starts in 2022 where the starting pitcher recorded 14 or fewer outs. There were 1,471 of them, comprising 30.4% of starts. In many of these cases, the pitcher was removed simply because he got rocked, like when Dallas Keuchel allowed 10 runs in a single inning of work on April 20. The double hook clearly isn’t meant to keep guys like that in the game – it would be downright cruel to force someone with a 9.20 ERA to keep going after allowing 14 baserunners in a single inning. With that in mind, let’s just look at games where the starter was pulled after completing between three and 4.2 innings while allowing three runs or fewer – not necessarily “cruising,” but certainly in a position where they would traditionally be left in.

There were 554 of these occurrences, or roughly once every nine starts. Leading the way in these outings was Chris Archer, who on a dozen occasions was removed before his fifth frame was complete without a blowup start. Six teams did so in at least a sixth of games, including (unsurprisingly) the Rays, who often gave swingmen like Jeffrey Springs and Drew Rasmussen the quick hook. In fact, Rasmussen was removed after exactly 4.2 innings twice in the same month – perhaps the sort of start that would disappear with the modified double hook.

When deciding whether to implement new rule changes, MLB often has to strike a balance between producing a more entertaining product for fans and creating a better environment for team success. So let’s consider how each group would be affected by the modified double hook. Teams clearly don’t want the starters they pull early going longer – otherwise they would’ve just been left in. Given the ever-increasing quality of relief arms and eight-men bullpens the new norm, teams would much rather go to their stable of nasty relievers than let their starter face the top of the order a third time.

Many fans despise the death of the traditional starting pitcher, the ace who could consistently deliver eight or more innings and rest the bullpen for a day. Regardless of individual prescriptive opinions on the matter, imposing a “minimum” start of five innings won’t bring the league anywhere near the starter innings totals of yesteryear. When we talk about the re-definition of the starter role, we’re generally referring to the fact that the most durable starters no longer throw 230-plus innings or regularly throw complete games, not that the Archers and Ryan Yarbroughs of the world are throwing four innings a night instead of five. If MLB wants to encourage more 2011-style starter usage, they’ll have to go much further than this – perhaps a double hook that requires six innings to keep the DH, limiting the number of pitchers on a roster to 11 or 12, or forcing relievers to throw multiple innings, which will likely cause them to limit or abandon the maximum-effort slinging that makes them so effective relative to starters.

The new double hook could even produce a worse experience for fans. We can all relate to the experience of watching a clearly laboring pitcher stay in far too long after he’s become fatigued, can’t throw strikes, or has simply been “figured out” by his opponent. All of us can think of a time we yelled “Get him out of there, [manager’s name]!!”, but the pitcher was forced to stay in, often continuing to struggle. Pitchers like Archer and the Rays’ hodgepodge of swingmen get pulled in the fourth or fifth specifically because they have well-documented struggles after a certain number of pitches or batters faced. Making a struggling “five and fly” guy finish up his inning in order to keep the DH could make for a considerably worse viewing experience, especially for those rooting for the team with the sputtering hurler on the mound. Three True Outcomes baseball clearly reduces the amount of action in the game, but balls in play should ideally be coming against rested pitchers who aren’t pushing past their limits to meet an arbitrary innings cutoff.

I find it rather unlikely that any of these new rules, at least in their current form, will ever make their way to MLB’s rulebook (if they do, come back to this piece and tweet at me). That said, there’s still use in experimenting. Moving the mound back was certainly more extreme than any of these proposed changes, and two years of that told us that we need to find better ways to reverse the Three True Outcomes trend. Each of the 2023 rule changes were tested in the Atlantic League, and while opinions may vary on whether or not the new rules have made the game better, their use in sanctioned games ensured that they achieved MLB’s intended outcomes – more stolen bases and significantly shorter games. In short, having a Lab League is good for baseball – even if every experiment doesn’t go quite to plan.





Kyle is a FanGraphs contributor who likes to write about unique players who aren't superstars. He likes multipositional catchers, dislikes fastballs, and wants to see the return of the 100-inning reliever. He's currently a college student studying math education, and wants to apply that experience to his writing by making sabermetrics more accessible to learn about. Previously, he's written for PitcherList using pitch data to bring analytical insight to pitcher GIFs and on his personal blog about the Angels.

5 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Chip Lockemember
10 months ago

Good write-up. Agreed on all points, but the real thing that stuck out to me here is Kyle Lewis was the fifth-slowest runner last year?!?