Writing about trends from the first two weeks of the MLB season is a fraught exercise even in a typical season. This is not a typical season though, not with baseball adopting so many significant rule changes that officially took effect on Opening Day.
We want to know: What does it all mean? With the obligatory caveat that the samples are small, here are five observations from the first slice of this historic 2023 season, in response to the new rules and otherwise.
1. The stolen base resurgence is just beginning
The rise in steals has gotten a lot of attention, as it should. The early rate of steals per game (0.69) is the highest in baseball since 1999. If the number increases in a typical fashion, it could end up at around 0.73, which would put us on par with the early-1990s.
That said, typical trajectories might not tell us much in this category, because I’m not too sure teams have figured out what they are dealing with in this new era of limited pickoff throws and bigger bases.
You can see this in the spread of steal attempts between teams. Sure, the makeup of a team’s roster plays into this, but the difference is tremendous. Teams such as the Baltimore Orioles, Arizona Diamondbacks and Cleveland Guardians are running wild. Others like the Minnesota Twins, Los Angeles Angels, St. Louis Cardinals and Kansas City Royals are barely running at all.
Others have pointed this out, but I’m going to repeat it here in hopes that straggling managers take notice: Teams are not stealing enough. We know this because the league success rate (81.3% through Thursday) would easily be an all-time record, but it’s a number that’s higher than it should be.
The old records in league steal percentage have all been set in recent seasons, as teams have fine-tuned their running games so that success rates settle in just above the 75% level, where the risk assessment of any potential theft favors the attempt. If the success rate is higher than that, teams aren’t stealing enough.
Ten teams have success rates of more than 90%, which is outrageous. No team has done that over a full season. So there are lots of steals out there being left on the table. As the summer warms, hopefully teams will figure that out.
Let them run!
2. Hits are back, baby
Last season, the collective MLB batting average (.243) was baseball’s lowest since the Year of the Pitcher (1968), even though it marked the first full campaign with the universal DH. While this season’s early climb to the high .240s looks like modest progress, we could really be seeing a return to classic batting average normalcy in the months to come.
Batting average is another category that tends to change over the course of the season. The weather gets better. Hitters find their rhythms. Teams have to dig into their minor league depth to replace injured hurlers.
The uptick in average varies from year to year, but the typical jump is around nine points from where we are now on the calendar. If we hit that target once again, then we’re looking at around a .257 league batting average. That would be our best since 2009.
Incidentally, the improvement in average is entirely a function of a higher average on balls in play. League BABIP (.299 through Thursday) is way ahead of last season’s end-of-season mark (.290), which was our lowest since the early 1990s.
This, too, is a number that should go up a few points as the season progresses, and if that happens, we could be looking at historic BABIPs. (The modern record for BABIP, by the way, is .303 — posted in 1930 and 2007. It’s a figure that has not been surpassed since 1897.) Now if we could just do something about all those strikeouts.
3. Veteran pitchers are having a harder time adjusting to the new pace of play
I saw this mentioned on a broadcast, and then I attended a New York Mets–Milwaukee Brewers game where Max Scherzer gave up back-to-back-to-back homers for just the second time in his career. So I started tracking pitcher performance by age, using 34 years old as the cutoff. Here’s where it stands through Thursday’s games:
Age 34 and up: 5.53 ERA
Age 33 and under: 4.34 ERA
Is that unusual? You bet it is. Going back to 2000, older pitchers have had a better collective ERA than their younger counterparts 17 times out of 23 seasons, including the past five. This is almost assuredly a product of selection bias. That is, older pitchers don’t really get a chance to fail. They succeed or they get replaced. So the vets who are still going tend to be good, or else they wouldn’t be on a roster.
But it’s not just that the younger pitchers have been better, it’s how much better they’ve been than the older pitchers. That 1.19 gap in ERA is way beyond any gap we’ve seen between the two age groupings during this century. The average gap has been just 0.23 runs.
The split is unlikely to stay this extreme, but it remains an area to watch. What we can’t say for sure is that it’s the pitch clock, simply because the samples are so small. But it would be an awfully big coincidence if that wasn’t at least part of the story.
4. The spike in walks and wild pitches probably isn’t a pitch clock thing
Walks and wild pitches are both occurring at higher rates so far than the end-of-season rates of recent years. Some have noticed it and wondered if harried pitchers have collectively gotten a bit wilder because of the pitch clock.
Well, maybe. It’s far too soon to rule it out. But that’s probably not going to turn out to be real. The thing is, walks and wild pitches both occur at higher rates in the early going of a season than they do as the season goes along. The rates should and most likely will drop.
Focusing now on walks, the league rate is 9.1% through Thursday’s games. If that held up, it would be the highest rate since 2000. But given a typical rate of decline from the early games to the eventual end-of-season number, we can expect it to land at around 8.5% by October. That would be a little higher than last season, but lower than 2021 and right on target with the three seasons before that.
5. The games are fast indeed, and we are still sorting out the consequences
We know the games are shorter, down roughly 25 minutes over last season for a nine-inning game. But if you haven’t been to a game in person yet, hang on to your hat. These contests can really zip by, and you notice in person most of all, probably because there is no pause button for reality.
The average game time gets a lot of ink (2 hours, 37 minutes), as it should. But that’s just the tip of the iceberg:
• About 30% of games thus far have ended in the range between 2 hours and 2 hours, 30 minutes. Over the past few years, that number has hovered around 3%.
• From 2018 through last season, the percentage of nine-inning games that have gone 3½ hours or longer has ranged from 9% all the way up to 19%. Through Thursday, there has been only two nine-inning games that long in 2023.
• The standard for nine-inning games ending in the range of 3 to 3½ hours has generally been around 46-47%. This year, only 11% of games have ended up in that range.
• The game-to-game standard deviation in the total minutes needed to play nine innings has typically been around 21. This year, that has dropped to 16.8. What that means is the length of games has not just been shorter — it has also been more predictable, the durations more uniform from one contest to the next.
The change is really startling. What’s weird about that is that it does indeed feel so different. After all, this is simply how long games used to be, up to about 25 years ago. Younger fans have never experienced this, but it’s jolting even for those old enough to remember how things were.
It’ll be a while before we have enough data to judge the on-field consequences of the faster pace, if indeed there are some. One area to watch: One-two-three innings can now pass by in the blink of an eye if hitters are swinging early in the count. Thus a pitcher can come off the mound, stop for a drink of water, and then find himself headed right back out to the field.
Is that a problem? So far, managers and pitchers that I’ve encountered haven’t been willing to point to it as an issue or an excuse. If it is having an effect, we should eventually see it in the data.
