The First Pitch Strike Game
The best result from a first pitch? Has to be a dribbler to the mound, right? We spend all this time chasing the swinging strike and drooling on triple-digit velocity, and there’s a future Hall of Fame pitcher who made his living getting first-pitch sawed-off million-hoppers to the second base side — Greg Maddux.
But if you’re not Greg Maddux, the first strike is the nexus for a game of cat and mouse. We’ve found that throwing a first-pitch strike is one of the best ways to get your walk rate down. But if the league throws too many meatballs on 0-0 counts, batters should swing more. It might be the best pitch they see. If the league then throws fewer strikes for the first pitch, batters would find themselves looking more. I don’t know if it’s game theory, but it’s certainly a theory about this game.
But the first-pitch out was the beginning of this string. J.D. Gentile lamented the drop in first-pitch outs recently, even as he showed that Greg Maddux was, indeed, the master of this craft. In that piece, Gentile brought up the batting average on balls in play on the first pitch:
It might be easy to say, hey, that’s where first-pitch outs went, but to Gentile’s credit, he does point out that league BABIP also jumped at the same time. Let’s see those two effects on top of each other, no?
Apologies for that. Anyway, two things might be visible in that mess. One is that something strange happened in 1992, maybe. That’s the only year where the league BABIP and first-pitch BABIP were so close. The other thing is that, while the two have tracked each other well, the BABIP on the first pitch is usually about fifteen-to-twenty points higher than the league BABIP.
That seems like a decent reason to swing at the first pitch. It’s going to be a strike, and (probably since it’s going to be in the strike zone) it’s more likely to find grass than the average pitch. Was the league-wide first-pitch swing rate steady over this time? Thanks to Jeff Zimmerman, we have this answer:
Wait what. We’re in the golden age of first pitch BABIP, and batters are swinging at fewer and fewer first pitches? Pitchers must be throwing it outside the zone more often in response, right? They must be reacting to the BABIP, playing that first-pitch strike game, adjusting to the adjusters. Right?
Oh come on now. Something happened in 1996, yes. It looks like that jump in first-pitch swing rate wasn’t brought on by BABIP changes, too — 1992’s small separation between first-pitch BABIP and league BABIP might have sent the swing rate down, but in 1996, the difference between the two BABIPs was less than it was before 1992.
Maybe instead it was the first-pitch strikes that made a difference in 1996. Maybe that year, front offices looked around the league, saw that first-pitch strikes were down, and sent that information down to the field, where coaches passed it on and batters decided to swing more. And then pitchers maybe saw what was happening, and the batters swung less often in order to correct back to their correction.
But whatever possible here-and-there that happened in the late 1990s didn’t last, and now batters are swinging less at first pitch strikes, even though there are more first pitch strikes. It doesn’t make any sense.
There are some confounding issues that might be messing up these graphs. In 2000, the American League and National League umpires merged their crews. In 2001, a larger strike zone was implemented along with an evaluation system. Those factors might explain some of the oscillation to the left of the graph.
But on the right of the graph? You still see a fat BABIP, and some of the best first-strike percentages the modern game has seen… and some of the lowest first-pitch swing rates we’ve seen recently. It makes sense from a pitcher perspective — if they aren’t going to swing, you might as well get on the way to a strikeout instead of slipping towards a walk.
Hitters? Don’t know why they aren’t swinging more. After all, Greg Maddux is out of the game.
With a phone full of pictures of pitchers' fingers, strange beers, and his two toddler sons, Eno Sarris can be found at the ballpark or a brewery most days. Read him here, writing about the A's or Giants at The Athletic, or about beer at October. Follow him on Twitter @enosarris if you can handle the sandwiches and inanity.
Hitters want to work the count because that’s how the Yankees won all those World Series (not the plethora of all-stars and future HOFers on their rosters).
Despite the sarcasm, this was my thought too. It’s the baby boom, baseball style. The effects don’t show up immediately but you see it in a generation of baseball players (and batting coaches) later.
I’m no statistician, but are these differences meaningful? The graphs look like there is a large variation, but the ranges on each of them are really quite small. Is there any chance that this is subtle nonsense?
Small differences but big samples and the curve is decently smooth… I feel comfortable saying batters are swinging less AT&T the first pitch at least. Subtle nonsense, though, that could be on my tombstone.
Hell of an autocorrect. AT not AT&T obv.
they’re probably swinging more sprint because of the unlimited data plans
yuk yuk yuk, i’ll show myself out
We’re in the golden age of first pitch BABIP, and batters are swinging at fewer and fewer first pitches?
Or, we’re in the golden age of first pitch BABIP because batters are swinging at fewer and fewer first pitches – they’ve become more selective about what they swing at and as a result are making better contact.
This was my thought too. The larger strike zone may have contributed to an increase in first pitch strike %, but if hitters are selective beyond strike/ball and are looking at a smaller zone for a certain type of pitch then the data makes a lot of sense.
I would imagine first pitch ISO has increased rather steadily too and would illustrate a rise in power in exchange for that strike. The real story may be that pitchers and batters have come to a game theory solution of first pitches.
This is the onvious common cause hypothesis.
Onvious? That’s a perfectly cromulent word…
Just like Scott Hatteberg
Could you have causality backwards? I.e., could we flip it around and hypothesize that an increased focus on OBP starting in the 2000s led to lower 1st pitch swing rates. This factor then led to higher 1st pitch BABIPs as hitters were sitting on 1 pitch or were more selective?
Could simply be that the overall hitting strategy to NOT swing at more 1st pitches has resulted in an increased batting result for the full count, while the relatively offsetting values for 1st pitch strikes (increasing for pitcher) and 1st pitch balls-in-play (increasing for batter) are offsetting…
Smart questions from you two. I think both versions make sense — not all strikes are worth swinging at — but testing it… that might be beyond me.
maybe by calculating run expectancy by year for different counts? That sounds really time-consuming because the reasonsing for patience is to work into more batter-friendly counts. I think you’d have to weight the run expectancy per count with the percentage of the count in some way.
Someone also mentioned first-pitch ISO which makes some sense considering the hypothesis that batters are only swinging if they think they have a really good pitch to hit.
would it be possible to look at first-pitch swings at pitches in vs. out of the zone, maybe at pitch types as well? It wouldn’t shock me to see a lot of hitters sitting on a first-pitch fastball in the zone, swinging at that if they see it, and that causing what JeffMathisCera described above (if i understand what he’s saying, that is. I’m not a smart man)
I think there must be something to the fact that the changes started so clearly when the larger strike zone was implemented.
even in the post-2002 era, swing rate has gone down and f-strike has gone up. no changes in that era.
“even in the post-2002 era, swing rate has gone down and f-strike has gone up. no changes in that era.”
From 2001, 1st pitch swing rate drops steadily but 1st pitch STR% does not…
Bigger strike zone = smaller % of strikes are hittable.
The hittability of strikes, if it did change, would completely alter the game-theory framework you set up. In fact, that would explain the lower SW% despite the higher babip overall.
MLB implemented a change to bring offense down and it took more than a decade for the full effect to play itself out, including hitter adjustment cycles.
Another thought: by using BABIP you’re excluding home runs. What happens if you use BACON?
You get fat and die of heart disease and deliciousness.
Also the neckline of your shirt gets all bunched up.
Seriously, why exclude home runs? And shouldn’t we look at swings rather than balls in play? Maybe guys are just trying to bash the ball and are swinging and missing more?
Actually, in my chart at least, BABIP included HR (hence the asterisk in the title). I’ve seen the ‘BACON’ label before, but I didn’t know it had caught on.
This opens up a great question on how aggressive hitters should be. It looks as if hitters are losing about 15 batting average points (0.015) by not swinging at the first pitch, but, on the other hand, they’re reducing their chances of walking by some %. (The overall walk % is 80 points (0.080). What the tradeoff is depends on a probability tree kind of analysis, which also would be interesting.
The trick to that is also considering the idea that working lots of counts gets a pitcher out of the game earlier, which is terribly hard to quantify.
The other question here, is what if there hasn’t been any actual change in the percentage of pitches in the strike zone, but instead a change in the way the game is called.
The fact that a pitch is called a strike doesn’t mean its a hittable pitch.
I talk about that in the piece, but things have been fairly steady in that regard in the 2000s.
There have also been many new stadiums built. New stadiums tend to have less foul ground. Balls in stands are a strike but not a knock against the BABIP. I’m not sure how big of an impact these stadiums are or even what stadiums factor in.
We live in an era where advanced stats say that hiting HR’s and taking walks make you a great value player. (even if you strike out 200 times and can’t field a lick and drag the proverbial safe around the bases) Swinging at the first pitch doesn’t fit that mold.
I don’t buy that for one second. Folks here recognize the value of defense and speed. There are plenty of ways to contribute to your club outside of hitting HR’s.
Nomar Garciaparra was one of the greatest hitters of his generation, and was well known for first pitch swinging.
I suspect that BABIP in general went up because of increased recognition of the value of being ahead in the count, which would necessarily lead to lower first-pitch swing percentages.
Is F-Strike% just for called 1st Strikes or does it all strikes (In Play, Foul, etc.)?
Would it be better if you replaced this with Zone% (granted you don’t have that data for the entire timeframe)
It includes all first strikes. And yeah, Zone% is cool, but only back to 2002.
I also found that Called_Strike% on the first pitch was the *type of first-strike* that was significantly on the rise, while FOUL% and SwStr% were largely static from 1988-2010.
Also, the BIS plate discipline data doesn’t have a consistent definition of “in the zone” over the years, so using Zone% to look for trends is a bad idea.
“I also found that Called_Strike% on the first pitch was the *type of first-strike* that was significantly on the rise”
… this suggests the strike zone change is playing a role.
Right. According to Baseball Almanac, in “1996
The Strike Zone is expanded on the lower end, moving from the top of the knees to the bottom of the knees (bottom has been identified as the hollow beneath the kneecap).”
http://www.baseball-almanac.com/articles/strike_zone_rules_history.shtml
So that’s that.
Maybe they’re throwing more first pitch strike offspeed pitches and batters are looking fastball?
Where can I find stats for 1st pitch strikes, for each game and the number of called strikes for each game played?
why is a great pitcher avg the 3rd time through the lineup? Because batters are seeing pitches. the more they see the harder it is to get them out later in the game.
when batters work the count there is a better chance they will see a pitchers 3rd or 4th pitch the 1st or 2nd time through the line up.
that’s why verlander is good 3rd time through because he pitches at 92 and throws at 99. when they time 92 it’s like another pitch at 100. when he gets in trouble early and tries to throw everything by them, they time it at he isnt as effective 3rd time through.