What Strikeouts Have Taken from Baseball
It shouldn’t be news that strikeouts have increased at a pretty alarming pace over the last decade. From the end of the last strike through 2007, the league-wide strikeout rate was pretty steady, averaging 16.7%. That is, roughly one in every six plate appearances ended in a strikeout. Over the last decade, the average has reached nearly 20%, including a high of 21.6% this year. Now, more than one in every five plate appearances ends in a strikeout. A strikeout is now 30% more likely than it was a little over a decade ago.
This is a problem with many possible solutions: raising the bottom of the strike zone; lowering the mound; or, my personal favorite, expanding the league. This piece isn’t prescriptive, however. The focus of this piece is to show exactly what the strikeout has replaced, and it isn’t actually all bad.
Because I started this piece by doing some research on the home-run record, we will focus here on the year 2000 as it compares to the present. That season represents what was probably the height of the PED era; it was also the season responsible for the league-wide high in home runs until this year. Because the power numbers between the two periods are similar, a comparison of the seasons creates an interesting vantage point from which to view the role of the strikeout.
Generally, we imagine that players have to sacrifice contact for power. It’s notable, then, that the power numbers of today are equivalent those of the 2000 season even though players are striking out 30% more often now than they did back then. To provide some background, here are some standard numbers from 2000 and 2017.
| Season | G | PA | HR | BB% | K% | ISO | BABIP | AVG | OBP | SLG | wOBA |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 2000 | 62083 | 190261 | 5693 | 9.6 % | 16.5 % | .167 | .300 | .270 | .345 | .437 | .341 |
| 2017 | 54957 | 173900 | 5753 | 8.5 % | 21.6 % | .172 | .300 | .255 | .325 | .427 | .321 |
As the 20-point difference in wOBA illustrates, overall offensive levels were higher back in 2000. The ISO and BABIP figures are all roughly similar. When a batter hits a ball in play, that batted ball is just as likely to become a hit as it was before. When it lands fair, it’s leading to roughly the same amount of extra bases. All that’s basically the same.
In terms of differences, one finds that this year’s walk rate is a bit lower than in 2000. That has some influence on run scoring, but not at all to the same degree that the increase in strikeout rate has. In effect, 5% of potentially positive plate appearances have been turned into strikeouts. That’s significant. However, while the main complaint about strikeouts is that they lead to fewer balls in play, it isn’t accurate to suggest that every extra strikeout has actually had that effect.
We can take a deeper look and see exactly what events those strikeouts have replaced. The table below shows the percentage of plate appearances that have ended in particular outcomes between the two seasons.
| Event | 2017 | 2000 | Difference |
|---|---|---|---|
| 1B | 14.53% | 15.61% | -1.08% |
| 2B | 4.54% | 4.68% | -0.14% |
| 3B | 0.43% | 0.50% | -0.07% |
| HR | 3.31% | 2.99% | 0.32% |
| UIBB | 8.01% | 8.95% | -0.94% |
| IBB | 0.53% | 0.64% | -0.11% |
| HBP | 0.96% | 0.83% | 0.13% |
| SH | 0.50% | 0.86% | -0.35% |
| SF | 0.63% | 0.80% | -0.17% |
While strikeouts have increased by five percentage points since 2000, hits have decreased by less than one percentage point. We’re still seeing almost as many hits as we saw before. One positive byproduct of teams getting smarter over the past 15 years is that we are seeing fewer intentional walks and fewer sacrifice bunts. While half a percent of plate appearances might not seem like much, if we’ve taken 10% of the current increase in strikeouts in place of intentional walks and sacrifice bunts, that seems like a positive 10%.
There are a couple types of events missing from the table above, notably most outs in play as well as outs due to caught stealing. Even though caught stealings aren’t recorded as plate appearances, we can express them as a percentage of plate appearances to get an idea of their frequency. In 2000, there were 1323 caught stealings, which was 0.7% of plate appearances. This season, there have been just 872 caught stealings, which is 0.5% of plate appearances. I would also add that grounding into a double play adds an extra out around 2% of plate appearances, but that is virtually unchanged from 2000. I would also note that, unfortunately, hit-by-pitches are up slightly from 2000 by roughly one-tenth of 1%.
The graph below takes the above information and divides the added strikeouts in to portions based on what has been taken away in the game compared to 2000.
To get a sense of the consequences of an increased strikeout rate, we can use the graph here to identify which trends have been positives and which items have been negatives. Events that create action are generally regarded as a net positive for the game. Unfortunately, the two outcomes that have been replaced most frequently by the strikeout are two that lead to action: ball-in-play outs and hits. Add in sacrifice flies and that makes up for roughly 70% of the losses due to strikeouts. That’s not great.
On the other hand, getting rid of walks isn’t a big deal, as strikeouts tend to offer more action than walks. Bunts can be exciting, but sacrifice bunts take the bat out of a hitter’s hands. It isn’t as though the batter is trying for a hit in these scenarios. I’d rather see a pitcher swing than try to bunt if we are talking about pure aesthetics. A runner caught stealing is exciting, but if the goal is to put more balls in play, then having fewer players — particularly players not good at it — steal is probably a net benefit for the game. Intentional walks are no fun even if they take less time now than they did a few years back.
Increased strikeouts in the game isn’t a positive. While hitters might be hitting the ball farther than they used to, the homer trend is somewhat recent while strikeouts have been on the rise for some time. The increase in fly balls seems more likely a response to the increasing dominance of pitchers and not necessarily a permanent evolution. It’s important to note that not all that five-point increase in strikeouts has taken away desirable elements of the game, as some of the events that have been pushed out are even less desirable than strikeouts. This isn’t to say that fewer bunts and intentional walks wouldn’t have happened without an increase in strikeouts, but in terms of desirable outcomes, MLB is not suffering quite as big of a loss as it might seem. Fewer strikeouts along with fewer walks and bunts would be even better, and that should be the goal, though continued debate is likely needed on how to accomplish those aims.
Craig Edwards can be found on twitter @craigjedwards.
I’m an Astros fan. I’m sure most of you are well aware the “worst to first” transformation the Astros offense has taken this year in basically every category since the down years, and especially in terms of K%. Let me tell you something: what makes watching baseball fun is watching a good offense. Watching Carlos Beltran roll over weakly on groundballs is not any more “fun” than watching Chris Carter strike out 200 times a year. Both are equally frustrating. Not one Astros fan would trade having a 120 wRC+ as a team for a 0% K rate and a 100 wRC+. Baseball is fun when you get to watch good teams. Baseball is less fun watching bad teams. In my humble opinion, how you make your outs is a very, very small factor in the “fun” equation.
Of course you want a great team. It is fun to watch your team win. I find watching innings of K, K, HR, K are not that exciting relative to balls in play that fielders need to field. Defense and baserunning provides some of the more exciting plays in baseball and the additional strikeouts just detract from that for me. Personal opinion.
Balls in play are for peasants.
It really boils down to what’s entertaining – this is my argument for an NL DH, I just hate that a potentially fun AB in, for example, a bases loaded situation, is largely a snoozefest if the pitcher spot happens to be coming up next. You’re basically making a conscious decision to make that moment less entertaining.
Strikeouts are there in the same sense – no fun to watch a rally get killed by some hitter who flails at three strikes that aren’t even close to the zone, but it’s also harder to fix. You would have to somehow change the current incentives that are driving players/teams to this player type by doing something extreme like mandating that all the fences get moved back.
That John Lacky v. Carlos Martinez at bat a few weeks ago was one of the most exciting at bats of the year. Then again, a lot of that had to do with factors outside what actually happened in terms of baseball play.
A more fruitful analysis might be evaluating the relative rates of difficult defensive plays vs routine ones, on the theory that BIP outs that come from defensive highlights are more desirable. But since statcast data is only a couple years old, I’m not sure exactly how we would measure whether difficult defensive plays have become more or less likely over a longer period of time.
If the narrative didn’t exist, would people *really* notice that 2 PA per game per team were ending in K instead of BIP outs or the occasional single? And if they did, would they care?
With an “un-BABIP” of .700 (think of it as a placeholder for BIP outs for my crappy explanation), is it surprising that Ks have replaced mostly BIP outs? That seems intuitive.
BABIP is batting average on the balls in play. Doesn’t speak to the amount of balls in play.
In my opinion K’s are actually generally less exciting in the aggregate than walks for the following reason: Walks generate baserunners which leads to generally more interesting situations and more drama, whereas strikeouts tend to remove motion and interest from the game.
If people want to have more balls in play it is super simple. Expand the field. A 5% increase in size on all the dimensions would make for more balls in play, and also make it is so faster players would be more valuable. Pitchers and hitters have never been as large and explosive as they are today. Growing up kids out grow the little league dimensions. It’s not crazy to believe that after over 150 years that our baseball players have out grown the dimensions of a baseball field.
“A runner caught stealing is exciting, but if the goal is to put more balls in play, then having fewer players — particularly players not good at it — steal is probably a net benefit for the game.”
I would argue that the goal isn’t to put more balls in play. The goal is to find a way to increase action, or the perception of action. A stolen base, either successful or not, increases both of these. 1. it creates action on a pitch there otherwise wouldn’t have been, and most of the time, the stolen base is a close play. 2. Stolen bases, by the fact that they are often close plays lead themselves to replay, which reduces the amount of time that the camera is focusing on 4 dudes just kinda standing there looking at each other for 20 seconds.
While it probably isn’t good for the actual game in terms of winning strategy, more steals, even if the players get caught slightly more often, would create a more exciting fan experience. It could easily replace the excitement of a 2B diving into the hole to snag a grounder and flipping to first just in time.
I think the goal is more along create more action plays, rather than just pitches with no action involved.
When you put a bunch of slow no hustle sluggers on the field who walk or k you get a game of three true outcomes when unless the game is a high leverage situation are boooring to watch. Not to mention any high leverage situation after the 5th inning is met with a pitching change. Why is the game longer…more pitching changes more long at bats ending in walks or strikeouts. Duh
Ahh, the latest “My opinion about the state of the game is objectively correct” article.
If Fangraphs writers really find themselves incapable of writing about the differences in the offensive environment without making a qualitative determination which one they like better, maybe they should stop writing about it.
If this article was just written about what has changed instead of making a decision about whether those changes are good (“What Strikeouts Have Taken from Baseball” and “Increased strikeouts in the game isn’t a positive.”), it’d be interesting.
Every time one of these is written I am reminded of that excellent line by the official MLB historian: Everyone’s idea of the way baseball should be is exactly coincidental with the way baseball was played when they were 8.
Comparing 2000 to 2017 without mentioning shifts is wrong. Any changes in outcomes, such as absolute number of BIP, could be explained by shifts as well.
good point. My belief is that some ruling will come down regarding the spike in defensive shifts… no ‘fourth outfielder’ positioning would be my guess.
Most comments seem to come from the fans of offense. Remember, there are two phases to this game and to have your ace pitcher over-match a hitter by ‘blowing three strikes’ by him in very exciting in my book.