clock menu more-arrow no yes mobile

Filed under:

Where has baseball’s run scoring gone?

Here’s an overview of assorted hitting statistics from 1995 through 2022.

(060709 Wrentham, MA) Longest baseball ever played at Wrentham Little league field. League president Terry McGovern adds a three to the scoreboard as the game near 300 runs. Home team had 290 at this pint in the 83rd inning. See the web site www Photo by Ted Fitzgerald/MediaNews Group/Boston Herald via Getty Images

There’s been a lot of attention lately on the dearth of runs being scored in baseball, and not just by Detroit Tigers fans. Naturally, the list of possible culprits is as long as your arm: a deadened baseball, the season always starting slow for hitters, teams are getting ever better at baseball’s version of combined arms, coordinating pitching and defensive positioning, and so on.

So, I decided to gather a whole bunch of data going back to 1995, throw them into graphs, play with the axes and scales a bit, and see if I could find anything interesting, insightful, or unusual. When I started this exercise, I imagined I’d see relatively flat lines until this year and then see a huge change but, spoiler alert, that’s not what I saw.

Data sources

For the full-season data, I used Baseball Reference. For the March/April data, I used FanGraphs because I couldn’t figure out how to tease-out monthly data splits league-wide for anything less than a season. (It’s probably possible on B-R, but I have enough trouble finding stuff on that site, with its myriad options and pages and whatnot.)

Data for the 2022 season was through all games played on May 4. We’ll try to circle back after the second month of play and see if things have come back toward recent norms.

The “Early in the Season” Effect

It’s a mantra around baseball: for various reasons, lowered offense in March and April, before the weather heats up on a good part of the continent, means lower-scoring games. Is the air denser then? Are pitchers generally ahead of hitters? Are hitters all big Star Wars fans and get all zazzed-up around May the Fourth? Could be a mix of all of those.

To examine this effect, I compared several stats for full seasons against those from only March and April. I figured the past five full seasons should be enough to see a trend, and I think it was. (The 2020 season was left out for a very obvious reason.)

As you can see, yes, there’s always a gap of some sort between March/April games and where things end up. Here’s a summary showing how much lower these offensive categories compare, start vs. full season.

Percentage Differences for March/April vs. Full Season

Year % Lower for BA % Lower for SLG % Lower for OPS % Lower for BABIP
Year % Lower for BA % Lower for SLG % Lower for OPS % Lower for BABIP
2016 2.4% 3.1% 2.2% 1.7%
2017 3.1% 4.5% 3.5% 3.3%
2018 1.6% 2.2% 1.1% 0.7%
2019 2.8% 3.4% 2.2% 2.0%
2021 4.9% 5.4% 4.1% 3.1%

What this means is that, for the graphs that follow, we’re going to have to take 2022’s data with a grain of salt. The trends may look to be going a particular way, but that’ll probably tick upwards by the time September rolls around. In the meantime, there’s still plenty of interesting stuff going on in the past few years.

Batting Average and Runs Scored

These are the ones that get the most attention these days, so let’s just see what’s happened over the past 28 seasons.

Again, we can expect the 2022 figure to finish higher than it’s currently sitting. We can see a pretty clear downward trend starting in the 2017 season. But since the team with the highest batting average doesn’t win the World Series (the last time I checked, anyway), let’s look at runs scored by each team per game. These are runs by each team, not total runs scored in a game.

These shapes looked pretty similar to me, so what I did is put both on the same set of axes and adjusted the scales so a good chunk of the data overlapped.

Well, now we’ve got something. For 21 seasons shown here from 1995 through 2015, batting average — admittedly a pretty crude statistic, but it’s something — and runs scored were in absolute lockstep. In the years since, though, their divergence is striking: the peaks and valleys still tend to march together, but there was a split starting in 2016... the first full year of Statcast and Trackman data, when people started looking more at things like launch angles and exit velocities. So let’s compare runs per game and slugging percentage, perhaps a little more indicative of run-scoring.

There we go: a much better fit from ‘95 through ‘21... but for ‘22, that’s a pretty big gap. While offense is generally weaker as we saw above, that’s going to be reflected in both runs scored and slugging, one would think. So, here, it looks like we might have something going. Exactly what this is, I’m not too sure.

The Quality of the Hits

A really useful statistic, which you can grab onto even if you’re not a full-on stat-head, is Batting Average on Balls In Play (BABIP). It only looks at actual balls that have been put into play in some way, and as you’ll see, it’s been very consistent — until now.

For over two decades, BABIP hovered between a low of .293 and a high of .303, and never outside that .010 window. Starting in 2019, though, it sat at .292 for two seasons, and this year is an unheard-of .282. I assume this is due to defensive positioning. However, as we’ve seen above, BABIP can be expected to rise by an average of 2.2%, which would put the final value for the season at .288 — which is still pretty far outside that narrow .010 window we saw for 20-plus years.

I wondered if there was a trend in the percentage of balls hit which were counted as fly balls, line drives or ground balls.

There was a strange three-season blip from 2000 through 2002 for ground balls and line drives, but my guess is that there was just something strange about how contact was classified for those three seasons, as it sat back down for 2003. But for 2013 through 2020, line drives outnumbered fly balls, before becoming roughly equal last season and this season. Oddly, ground-ball percentage has stayed roughly constant since the Clinton presidency (minus the blip, of course).

How do Plate Appearances End?

Batting average is one thing, but if we take all plate appearances into account, what percentage of the time, when a batter comes to the plate, does he reach base safely via a hit? As you might imagine, the graph looks a lot like batting average, which is no surprise. But we’ll break this apart in a little more detail later.

Not all hits are created equally, of course, so I took a closer look at extra-base hits.

Home runs were so “2015 through 2021,” guys; we might as well just get home runs a MySpace account while we’re at it! The rate has settled back down in 2022... so far. Doubles very slowly and steadily increased up to about 2007, then very gently decreased before taking quite a pronounced dive for the past three seasons. And it might be hard to see this on the scale above, but triples have decreased from a high of 0.53% in 1995 to 0.37% in 2021 and a meagre 0.32% in 2022 so far.

And now, the least surprising graph of this entire article.

I guess the thing that jumps out for me here is that walks have been relatively consistent. I’m not sure if this data includes intentional walks, but I imagine it would, in which case I’m curious as to what percentage of that ‘99-’00 uptick was due to the likes of Barry Bonds, Mark McGwire and Sammy Sosa getting the ol’ four-wide. But, yes, strikeouts, oof — however, if this trend from 2021 through 2022 continues, maybe we’ll get back to a little sanity. I couldn’t help but notice that long upward trend started in 2006, so maybe we should blame Joel Zumaya for all this madness.

Are Fly Balls Going Over the Fence?

A lot of talk these days seems to be centered around the baseball: is it mushy? How much did MLB and Rawlings change its composition? Did their overcorrection go too far?

I figured a good way to examine this would be to compare the percentage of contact as fly balls and line drives varied (or didn’t) along with the percentage of plate appearances ending in a home run.

(Remember, there’s that three-year blip in ground balls and line drives in 2000-2002.)

I’m not entirely sure how to read this, but here’s what I see.

From 2000 through around 2013, home runs generally trended downward. During that same stretch, both fly ball and line drive percentage very gradually increased (blips notwithstanding).

Then things went pretty haywire between 2015 thorough 2019: home runs shot way up, line drive percentage went up, and fly ball percentage went down. Is this the effect of launch angle? I can’t help but imagine people focusing more on hitting laser-beam line drives rather than moonshots, and as a result more of these laser-beams ended up over the fence.

What I did notice from 2021 through this season so far is that the percentage of fly balls and line drives stayed pretty constant, while the home run percentage plummeted. As mentioned before, it’s early, so that might come up later in the season. Still, this jumped out at me.

Obviously, this is worth more study, and with more sophisticated stats like launch angle, exit velocity, and all that jazz. This article wasn’t meant to be a terribly stats-y dive, and you can take this further if you like.


Here’s what I’ve got. Feel free to add your insights in the comments... as I have no doubt you will.

  • Things have been pretty weird in the past 7 or so years.
  • Strikeouts are actually down a tick in the last couple of years but are still crazy-high.
  • Where did the triples go? I love triples.
  • The ball might be mushy.
  • Yes, offense picks up some as the year wears on.
  • There’s a big gap between slugging percentage (way, way down) and runs scored per game (only “way” down) this year, even though slugging is usually an excellent predictor of run-scoring.

I’m not sure if you could consider baseball to be “broken” just yet, but I’d say overall that things have definitely, if you’ll pardon the pun, shifted in recent years. A friend of mine is convinced that bigger bases next year will have a huge impact, but... well, colour me sceptical on that one. A more well defined automated strike zone, and plans to limit the shift may have a real impact, but we’ll have to wait and see how those items play out.