Welcome Back, Offense
Do you remember last offseason? Last offseason, it seemed like Major League Baseball had two primary goals. One, it wanted to make the game move faster. And two, it wanted to do something about the ongoing drop in run-scoring. They didn’t really have a plan in mind, but they said it was a thing they were going to monitor. Last September, runs scored per nine innings dropped below 4 for the first time this millennium. And last year, also for the first time this millennium, there was a month with a league-wide OPS below .700. Actually, there were three of those months, in just a four-month span. Offense was going down; everyone was aware. Baseball needs a certain minimum level of offense to survive.
There’s your setup. This year, to some extent, baseball has succeeded in trimming unnecessary delays. The game does go faster, even if it’ll never go fast. And then there’s the matter of run-scoring. Here’s one plot, of year-to-year R/9 averages:
Hey, look, a little rebound. Relative to last year, this year is up 0.21 runs per nine, getting back closer to 2012 levels. That’s interesting on its own. But this gets more interesting the closer you look.
That, right there, is a plot of R/9 over each individual year. But we don’t need to stay at the year level. This is what happens when you break things down on a month-to-month basis.
In August, around the league, teams averaged 4.50 runs per nine innings. I don’t know if that seems like a lot to you, but again, last September, offense dropped to 3.96. The significance of that 4.50 — it was the highest mark for the league in any month since May 2010. So, that gets your attention, but it’s only one month. Well, we’re through the majority of this current September. And in September, offense has increased even more, to 4.71 R/9. We haven’t been here since August 2009. One partial month added on to one full month doesn’t make this inarguable proof of an offensive comeback, but this is becoming some really compelling evidence.
Run-scoring, of course, does depend in part on sequencing. So let’s try to strip that away. Instead of plotting monthly R/9, here’s a plot of monthly OPS.
The pattern stays true. The OPS in August was .736, the highest since August 2011. The OPS in September is .749, the highest since August 2009. So this isn’t just a weird sequencing fluke. There really has been more offense, for at least a month and a half, and given how many games and plate appearances there are, that seems like a difficult thing to just happen totally randomly. It can happen randomly! But one hypothesis is always “random noise.” Doesn’t mean we can’t investigate other things.
Here’s a plot of monthly BABIP:
August came in at .302; September is running at .305. Those are indeed slightly elevated. But then, last July, we had a .304. It was bookended by a couple .297s. BABIP could be a little part of this. BABIP is a little part of this.
How about strikeouts and walks? Maybe there’s been something up with the zone? Here’s K% – BB%:
No, nothing there. It’s hung at or around 13%, increasing slowly over time. A few weeks ago, at the Hardball Times, Jon Roegele talked about the strike zone, and observed the elevated offense in August. But he found no evidence that anything had changed with the zone. He found no evidence that pitchers had meaningfully changed their behavior. The zone has stayed big. Umpires haven’t become more batter-friendly. Maybe in time they will, but then, they won’t have to if offense stays up.
Here is, presumably, the biggest explanation. The rate of home runs on contact:
In August, there was a league rate of 4.1%, the highest since August 2009. And now in September there’s a league rate of 4.4%, the highest since August 2004. Again, yes, I know, it’s a partial month, but let’s put some stuff together. In this season’s second half, which is nearly complete, the league rate of home runs on contact has been 4.1%. Since 2000, there have been 32 individual “season halves.” This half’s rate would rank third-highest, behind only the first half of 2000 and the second half of 2004. The first half’s rate this year, incidentally, ranks 24th. The seasons below it are fairly recent. In the second half, homers have taken off, and they’ve taken run-scoring with it. It’s not just your favorite team that’s scoring more runs lately. It’s most of the teams. Sorry if that’s discouraging?
What you’ll notice is that’s an explanation that needs an explanation. Offense, all of a sudden, is up. The strike zone hasn’t changed. A small amount of this is probably BABIP-related, but it’s mostly about the return of power. Why has there been a return of power? And if I’m going to be completely honest with you, I don’t know yet. As it happens, I’ll be exploring this deeper in the forthcoming Hardball Times Annual, but I don’t yet have answers. No one I’ve spoken to has easy answers. It’s a weird and potentially complicated thing, but it’s playing out on a league scale.
Maybe it’s not sustainable. Homers are notoriously fluky, and the difference between strikeouts and walks remains a factor. It’s too soon to say that offense is back. But offense has been back, for a couple of months. It’s been back at the levels it used to occupy. The expanding zone seems like a problem, but if this is what we have now, maybe it’s less in need of a change than we thought.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
The offensive explosion has largely been a 2nd half phenomenon. Since its so widespread one must assume the balls are behind it. Perhaps just a batch of balls with a higher COR than normal, all within spec of course. Manufacturing tolerances are not 0 after all.
Of course, Selig juiced the ball when he first became commissioner to increase attendance, and its been done in Japan and Korea, so maybe Manfred is behind it. Attendance was pretty flat in the first half, actually a bit down. Its bounced back a bit in the 2nd half.
Getting at the reasons why is very important for teams who plan on signing a FA as a result of his good second half, or in the case of pitchers, their bad 2nd half
Here in the Northeast we have had a drier than normal spring followed by a very dry summer. Is it possible that just a bit less humidity is allowing more balls to leave the park?
Higher humidity actually makes balls travel farther. Water vapor is less dense than air. This is why the ball goes farthest in hot, humid weather and you see more home runs in summer than in spring.
No, water vapor is denser than air (which always has some water vapor in it anyway). An analogy of a baseball traveling further through the greater density of humid, moist air would be that things float more easily in saltier water (i.e., denser water) than in fresh water.
Nope, I’m wrong and take it all back: https://www.washingtonpost.com/blogs/capital-weather-gang/wp/2013/08/05/why-dry-air-is-heavier-than-humid-air/
OK, so it is not the air, but it just may be that with lower humidity for the summer the balls have dried out, losing weight and gaining just a bit more bounce. The opposite of the Coors Field Humidor.
The ball carries further when it is humid because humid air is lighter than dry air.
But the balls are more elastic when they’re drier. They bounce more, and thus leave the bat at higher speeds.
If we control for the ball, yes, the balls carry more in humid air. But we’re not controlling for the air.
Remember the Denver humidor. They moisten the balls there for a reason.
It could just be hotter weather in general. It’d be great to see a plot of Runs per Game Temp.
Consider ‘LIFT’ that keeps an airplane in the air (and think baseball).
At 30,000 ft it takes more energy (jet fuel) to keep the airplane at that altitude. The air is thinner(agreed?)so to maintain a constant velocity and lift requires more energy.
Flying over the Arctic Ocean airplanes go down in altitude to possibly 5000 ft where the air is heavier. It is counter-intuituve but a fact that the airplane gets the same lift using less fuel (which is economical and saves on fuel costs).
Apply ‘lift’ to a baseball in humid air. One might think there would be more resistance. No. Being counter-intuitive there is more lift in humid air than dry air so the baseball will go futher upon the application of an equal amount of energy (when the ball is hit). Spin (rpm’s) applies here as is does to a curve ball which is ‘pushed from high to low air pressure at 5,000 rpm’s. Air is heavier at the bottom and lighter at the top. Simple aeronatical principles applied to a baseball in flight.
Let me make sure I understand what you are saying. This is what you wrote: “..there is more lift in humid air than dry air”. I have to respectfully disagree with you. As other have already noted in the comments, humid air is less dense than dry air. That means there will be less drag in humid air than in dry air, all other things equal. And for exactly the same reason, there will be less lift in humid air than in dry air (for a given speed and spin). The Magnus force that creates the lift is proportional to the air density. The effect is not very large on either drag or lift, at least at the types of temperatures that occur in the ballpark.
If I am misinterpreting what you are saying, then I apologize.
Clearly a pilot you are not, much less an aeronautical engineer.
For sure I am neither a pilot nor an aeronautical engineer. However, I think we may be talking past each other. When the air is less dense, the ball carries farther, mainly because there is less drag. There is also less lift (for a given amount of speed and spin), but the lift plays a much less important role than drag in determining the carry of a fly ball. A good example is Denver, where the ball carries significantly farther than it does at sea level (at comparable temperatures). Data from various sources help us quantify the Denver effect. Moreover, lots of PITCHf/x data from Denver show that there is both less drag and less lift on pitched baseballs (i.e., less Magnus force, meaning less movement)than at other MLB venues. As a result, fastballs are a little faster (lose less speed due to less drag) and breaking pitches do not move as much (less lift). These are all very well documented facts.
Ok. I am not a pilot nor an aeronotical engineer. I need to revise my thinking. I must agree with the effects that Alan Nathan below has described. I thank you Alan for the clarity you have brought to this (small) point I was completely wrong about.
Your description of the Magnus Force and behavior of a struck ball in the very high altitude of Denver is convincing. The ball carries farther in the dry air of Denver. Documented. I’m convinced. Thank you for setting me straight. It pays to keep an open mind. We are all fallible. Thank you again Alan Nathan.
Speak for yourself on the NE thing. I live in Pittsburgh and while the summer was dry, I think we were owed that since we received 11 inches of rain in June alone, and May was not too dry either. This was still a very ‘warm’ season.
What if they are trying new, softer balls?
Then offense would be down. Softer balls aren’t as elastic, and they’d come off the bat slower.
No no, I’m pretty sure that softer balls improve offensive output. Or is that only in football season in New England?
In Green Bay they overinflate the balls, and offense is good there as well. Maybe it’s not the ball and the talent is just better in those cities.
Jeff, if you keep analyzing this season before this season is ever over, what the hell are we all going to read when the season DOES finally end? I may get so desperate that I’ll start following football. Football, Jeff. Don’t make me do it.
Just be glad you’re not a Philadelphia fan. Football offers no reprieve for the poor sports fans of our city.
Welcome to the Team of Chip’s Dreams!
Maybe it’s just an actual influx of power?
Rookies in 2015 have hit 619 HR vs 471 HR in 2014, and 41.0 PA/HR in 2015 vs 50.9 PA/HR in 2014. Even with two weeks left to go in the season this is the highest total number of HR hit by rookies in any season in the last 15 years.
It’s not just a symptom of an overall increase in offense: Rookies in 2015 have a 90 wRC+ which is the highest for rookies in any of the last 15 years (keep in mind wRC+ is era-adjusted meaning this is merely measuring how good rookies were compared to the rest of the league’s hitters). This year’s rookies are simply a stellar class of hitters, especially power hitters.
There’s no easy way to search players who have left MLB between 2014 and 2015 but I’d expect if you measured the gap between the new players and the outgoing players whose PAs are being replaced, you’d see a more dramatic difference than in a typical year.
This even helps explain the increase as the year has gone on as some of the best rookie hitters were called up relatively late in the season (Schwarber, Correa, Sano).
Someone smarter than me should figure out to what extent the unusually good performance of rookies accounts for the increased offense this season.
Ding ding ding.
I think we have (at least a partial) answer.
And here’s Mike discussing it just today: http://www.fangraphs.com/fantasy/the-year-of-the-rookie-power/
Thank god someone mentioned the rookies!
Watch this R/G mark tumble when the “dangerous slide” policy is put in place
Fine by me.
There would probably be a lot more scoring if runners could knock the ball from defenders’ gloves, too. But no one is advocating that. Just enforce the rules that exist.
Dangerous slides have a negligible overall effect on scoring. How often do they succeed in breaking up a DP?
Because injuring good hitters like Kang is sure to hurt offense.
it’s nice to be back, Jeff
have you been working out?
I was looking for HRs, and there it is, in the last graph. Just from fiddling with the stats the past few weeks, a few things stand out to me:
great pitchers are having all-time great seasons
batters still K a lot, still have low BAs, but power is up
a couple of things that seem true, but I haven’t looked it up:
fewer triples, or fewer players with a lot of triples
inordinate number of batters with a reverse platoon split (RHB hitting better vs RHP than LHP, as is the norm).
Wrt fewer triples:
It’s only one player, who’s probably changed his approach at the plate, but I thought this was an interesting development for Trout, showing his HR/3B ratio, and HR as % of all XBH:
2012 3.75 46%
2013 3.0 36%
2014 4.0 43%
2015 9.8 56%
After being fairly constant his first three years, they went way up this year. It’s even more dramatic if you look at his numbers at the end of July:
16.0 63%
Then he slumped—at least partly the result, some of us still think, of a wrist injury–and went a month without a HR. During that month, he had three 3B, more than doubling his season’s total during the previous three months.
I think Bryce Harper single-handedly raised it.
We single-handedly raised it.
I single-handedly raised it
I did not
This is pretty much our fault.
Dude, don’t forget us!
This is clearly a result of Matt Harvey’s innings limit
The balls are being threaded tighter or just juiced in general. I have seen more balls bounce a couple of feet in front of home plate and then not come down until they are in the outfield than in previous years.
It happened because the Mariners hired finally hired me as hitting coach, at which point I promptly turned the worst-hitting team in MLB into the best-hitting team!
maybe 1 thing is the number of pitchers. With 2 weeks to go, 730 different pitchers have pitched in MLB this season- already a MLB record. Last year for perspective, there were 692 pitchers used in MLB. 389 already used in the NL(an NL record, even with the years of 16 NL teams).
A whopping 95 more pitchers used this year already compared to 2010.
Has any grad student been measuring ball coefficient of restitution for five years? Odds love to see that scatterplotted with BABIP and HR%
You are all welcome