Oh well, one man's flyball is ... another man's something else. At b-r, FBs last year generated a 225 BA, 204 this year, essentially the same gap which is the key point here. I'll guess the cited numbers don't include pop-ups which is probably a good idea but not one adopted by b-r (or their split source) yet. Another possibility is different LD definitions.

If we pro-rate this year's FBs to last, it's 1000 HRs difference which is an awful lot, about a 18% decline. In percentage terms, the effect on LDs is even greater, down from 602 to 362, a 40% drop. If these sorts of effects hold up over the season, that's about 42 fewer HRs per team.

Seems like they could just adjust the humidity in the humidor and make it a little drier.

That's a good start. Now enforce the pitch clock and limit teams to 11 pitchers on the active roster. Also a player has to stay down 30 days if optioned to the minors. Watch the strikeouts plummet, and maybe we'll get something approximating fun baseball.

Seems like they could just adjust the humidity in the humidor and make it a little drier.

The number of HRs is fine, the problem is the league is hitting .239 and pitchers are k-ing 8.5/9 IP. That's what needs to be fixed. Adding half a HR per game to this brand of baseball doesn't really make it more appealing.

The number of HRs is fine, the problem is the league is hitting .239 and pitchers are k-ing 8.5/9 IP. That's what needs to be fixed. Adding half a HR per game to this brand of baseball doesn't really make it more appealing.

Right. We'll see if MLB is willing to stick out the criticism and see if the players adjust. The adjustment isn't going to happen overnight. Or in a month, or probably even in a year. The hitters need a reason to adjust and MLB needs to be patient enough to give them that reason.

I don't think the number of homers is all that significant as long as teams are scoring runs. While run scoring is down from the last couple of years, both runs per game and homers per game are exactly where they were in 2015 - and more broadly, where they were in the post-steroid, pre-launch angle era from 2010-2015.

In other words, everything's fine. Now just speed up the games a bit, and we'll be all set.

In other words, everything's fine. Now just speed up the games a bit, and we'll be all set.

There are two issues - the game is too slow, and the batting averages (the proxy for "action") are too low. Those things are correlated. I presume the number of runs/game is the same as 2015 because batters are trading hits for walks and strikeouts. Some of that is because, as noted, the pitchers are throwing harder than ever, so the batters' recourse is to try to work a walk, hence pitches/batter is up compared with 2015.

Something else has to be done, I think. Everything is not fine. But, as noted, from a runs/game standpoint, the game is "balanced". So whatever changes are done to speed up the game and increase "action" have to be balanced. Pitch clock would help (probably reduce the average velocity somewhat) but that might have to be paired with an even deader ball, and perhaps other measures like slower infields to force fielders to play less deep and increase the batting average on ground balls.

I presume the number of runs/game is the same as 2015 because batters are trading hits for walks and strikeouts.

You can see all the year-to-year variations here. You are correct that hits are down and walks and strikeouts are up since 2015, but both walks (a little) and strikeouts (more than a little) are down since last season.

Hits are also down since last season, but it's all homers. Total hits have dropped by 0.14 per game, but total homers have dropped by 0.21 per game. Singles have increased this season.

The other thing I noticed about that chart that I hadn't before was this: Runs/game are the same in 2022 (4.26) as they were in 2015 (4.25), but batters are hitting .239/.310/.386 this year, while they hit .254/.317/.405 in 2015. How could that be?

The other thing I noticed about that chart that I hadn't before was this: Runs/game are the same in 2022 (4.26) as they were in 2015 (4.25), but batters are hitting .239/.310/.386 this year, while they hit .254/.317/.405 in 2015. How could that be?

OBP is only slightly lower (317 vs. 310), ISO is the almost the same (151 vs. 147), so the on base and extra base averages are very similar (I doubt a difference of 7 for OBP or 4 for ISO is going to be super meaningful, especially taking into account the effect of zombie runners). BAVG is dramatically lower, but means the least of the three stats as far as RPG goes.

The other thing I noticed about that chart that I hadn't before was this: Runs/game are the same in 2022 (4.26) as they were in 2015 (4.25), but batters are hitting .239/.310/.386 this year, while they hit .254/.317/.405 in 2015. How could that be?

Are errors up?

I thought zombie runners might have some effect that partially offset the drop in OPS, but I'm not sure. League earned runs per game (slightly different from ERA which is per 9 innings) is 3.86 in 2022, against 3.93 in 2015. That does not seem like much of a difference, but it might be a factor, all zombies being unearned runs and sometimes scoring without anything in the way of hits behind them (groundout/SF seems popular among zombies).

That does not seem like much of a difference, but it might be a factor, all zombies being unearned runs and sometimes scoring without anything in the way of hits behind them (groundout/SF seems popular among zombies).

In order to end a game before zombie runners, you had to actually put someone on base. Now you do not. There is automatically going to be lower OBP in the innings where the game ends, it's basically guaranteed. The only issue is that the games end sooner, so maybe it reduces the effects, but clearly the OBP has to be lower in Zombie games that end in one inning (and probably in others too, because people could walk or single in innings that didn't end games, etc.).

The OBP difference of 7 is like 7 extra baserunners per 1000 plate appearances. At an average of 40 PA per game, that is like 25 games. I could definitely see Zombie runners being responsible for part of that.

There are two issues - the game is too slow, and the batting averages (the proxy for "action") are too low. Those things are correlated. I presume the number of runs/game is the same as 2015 because batters are trading hits for walks and strikeouts. Some of that is because, as noted, the pitchers are throwing harder than ever, so the batters' recourse is to try to work a walk, hence pitches/batter is up compared with 2015.

Will the reduction in HR cause hitters to adjust and look for more contact? There will likely be some lag there. I still think there need to be adjustments, but perhaps stricter roster limits for pitchers could help too.

You can see all the year-to-year variations here.

Thanks for the link.

Looking at 2022, 2015, and 1990, all ~4.25 runs/game

OPS: 0.696, 0.721, 0.710
BIP/game: 24.33, 25.60, 27.75
SO/game: 8.33/7.71/5.67
HR/game: 1.01/1.01/0.79
W/game: 3.16/2.90/3.29
H/game: 7.99/8.67/8.75
SB/game: 0.50/0.52/0.78
CS/game: 0.16/0.22/0.36
SH/game: 0.08/0.25/0.37


So looking at that, the offense is staying the same between 1990 and 2022 because the homers have gone up, the hits have gone down, but most importantly folks are not bunting or getting caught stealing. At the same time the number of balls-in-play is down by more than 3/game and the number of strikeouts is up by almost 3/game.

So basically, analytics has recognized that steals and bunts are useless, and they have disappeared, and that has compensated for the otherwise reduced offense.

The other thing I noticed about that chart that I hadn't before was this: Runs/game are the same in 2022 (4.26) as they were in 2015 (4.25), but batters are hitting .239/.310/.386 this year, while they hit .254/.317/.405 in 2015. How could that be?

Do those 2015 numbers include pitchers batting, or is it just hitters stats? If the latter, then the lack of pitchers making added outs could explain it.

There's a lot more HBP these days - isn't there ? That adds to the base runners.

There's a lot more HBP these days - isn't there ? That adds to the base runne

HBP: 2022/2015/1990

0.38 / 0.31 / 0.20

So yes, about 0.18 more per game compared to 1990, but 1990 has 0.76 more hits/game. So it's an effect but small, comparable to the 0.20 fewer CS/game. The 0.30 fewer SH/game is about the same as the sum of the 0.18 more HBP and the 0.20 fewer CS. So a total of 0.60 fewer dumb outs and more runners on base through HBP.

It's an extreme example but Sunday's Cubs-Sox game illustrates the unearned runner effect. 1-1 after 9, the teams added 7 runs in 3 extra innings, along with some errors along the way. Final tally was 2 earned runs and 7 unearned runs, 5 of the 6 unearned runners scored.

But the big news is that right here in the 21st century, Cub pitchers in 11.1 innings somehow managed to strike out only 4 batters.

BAVG is dramatically lower, but means the least of the three stats as far as RPG goes.

No, it's the most important, especially at the moment. I'm not going to go into all of it gain but suffice to say that:

OPS = (1 + AB/PA)*BA + (BB+HPB+...)/PA + ISO

The reason BA doesn't matter once you control for OBP and ISO is because OBP and ISO both contain BA. This is a well-known issue in causal modelling.

Now AB/PA is slightly lower this year than 2015 (.895 vs .901) so you can't quite directly compare but close enough -- that 15 point drop in BA has a big impact. (Note also that the ISO is a bit lower this year than 2015.) Now it is true that singles don't really contribute a lot to run-scoring in a K-heavy environment so that drop in BA probably has less impact than in other eras. And of course the relationship between OPS and run-scoring is hardly exact.

The following years all have a 755 OPS

1995 4.85 R/G
1998 4.79
2003 4.73

Now that's also in order of OBP (small differences) but 1997 had almost the exact same BA/OBP/SLG as 1995 but 4.77 R/G. The difference between 2009 and 2003 is just 4 points of SLG but 0.12 R/G. 2022 is very similar to 1992 and 2014 but has 0.14 and 0.19 more R/G. The range on R/G for a given OPS (with close OBP/SLG and error rates and maybe K rates) seems to be at least 0.12 R/G and maybe a good bit more.

Or the other way ... 1933 had 4.55 R/G on a 707 OPS; 1951 4.55 722; 2005 4.59 on 749; 2021 4.53 on 729. Then you get the fun years like 1925 (288/350/407, 5.24 R/G).

One thing to remember about 2015 was that it was a season of two halves. In the first half, 4.1 R/G and 0.95 HR/G (per team); in the 2nd half, 4.4 R/G and 1.09 HR/G. I'm not sure that has any impact on the comparison but 2015 was a funny year.

The Manfred Men ... in extra innings so far, there have been 51 earned runs and 96 unearned runs. Obviously not all of those unearned runs were the zombies but most of them would be. There have been 102 extra-inning "games" ... I assume that's 51 games, suggesting teams are bringing home those runners the vast majority of the time which would seem to defeat the purpose. 17% of all unearned runs are in extras despite only 1% of PAs -- it's an even bigger joke than I thought it would be.

Speaking of the zombies ... have we had a 2 PA inning yet?

There have been 102 extra-inning "games" ... I assume that's 51 games, suggesting teams are bringing home those runners the vast majority of the time .

Not necessarily. 102 extra-inning games doesn't equal 102 extra innings. The Rays have played seven games but 11 innings, for instance.

suggesting teams are bringing home those runners the vast majority of the time which would seem to defeat the purpose.

Of course, the real purpose of the rule is to limit the number of extra innings played, so whether it's working would depend on how many extra innings are being lopped off compared to the pre Zombie era.

It's still awful, regardless the answer to that question.

without actually counting them, with 100 extra inn games, that should produce about 200 total innings (there's an exponential decay in the numbers 50% go to 11, 25% go to 12 etc). If most of those 96 unearned runs are on zombies thats probably a bit under half of them scoring.

Well wait a second that decay by half is before zombie baseball existed. So Im gonna guess maybe only 150 extra inn. have been played. that still means 300 zombie runners an less than half have scored?


HIstorically a man on second w/ no outs should score about 61-62% but the current season has a lowered offensive environment but that rate of zombie runners scoring seems low. I must be messing up the math somewhere.

there's no way the "vast majority" of zombies are scoring, but that's about the only thing I can comfortably say. Hmm

errors are just about the same: .56/game. .56/game in 2021/2020 (I didnt prorate the shortended season just average of total)

HBP are actually down this year: .38 vs: .445 in 2021/2020. Did Duke say they were up?

We seem to have had this discussion or similar at some pt last season and it turned out that zombie runners were the culprit but the current discussion has raised some more issues. I feel zombies are almost certainly part of why scoring seems more efficient even w/ lower off. stats, but perhaps not all of it.

This reminds me of Apollo 13 when they were returning home having orbited the moon and their trajectory to intercept the earth was off by some fraction and everyone at NASA was confused for a while. Turned out they had forgotten that when the made initial calculations for the return trip they had counted on some 95 pounds of moon rocks being on board. Having not landed on the moon due to the explosion the missing moon rocks accounted for the discrepancy. Lovell just had to fire some retro rockets to correct the trajectory.

EDIT: that story might be apocryphal. It seems there was also an issue with steam venting from the lunar module during the return trip. The LM was to intended to be the crew compartment for that long extended voyage home.

wp/pb are slightly down about .05 less per game.

oh did Walt say 102 extra inn "games" means 51 actual games?

OK so that's probably what 75 total inn.? so 150 zombie runners.

Multiply by 0.60 and there's your 90 unearned runs from zombies. So again the 61% rate of scoring w/ man at second no out remains constant.

I think that's right.

And I would assume that a 61 percent rate of scoring is actually pretty good at delivering the desired result (a shorter game). A condition that falls somewhere close to 50/50 in terms of likelihood of scoring seems like it would be highly preferable to one that is on either extreme.

Nope ... sorry, I missed some columns in the table.

102 extra inning "games" ... that's a team-game though so 51 games have gone to extras.

That's produced 110 "innings" (i.e. half-innings) of pitching. So that's 96 unearned runs in 110 innings. They aren't all zombies but there are only 6 RoE but of course there might be a lot of runs after the RoE. (Errors that give up an extra base might lead to an unearned run but rarely more than that.)

Part of what's going on is that batters are tearing it up in extras ... or teams are down to their 6th best reliever by the time they get there. The line is 280/389/464 in extras, a 146 sOPS+. Nearly 200 points better than the 9th inning OPS. That's just weird.

My point about it not working, which admittedly is speculative, is that if both teams nearly always score their zombies then you're just turning your 4-4 game into a 5-5 game. If they're ending earlier than they used to, it's not clear that's the zombie.

So last year 432 games, 535.1 IP, 372 UER, 255 ER and a 254/366/418 line, 117 sOPS+, 95 points higher than 9th inning. Of course there are a lot of IBB in extras.

2015 for comparison ... 424 games, 870.2 IP, 33 UER, 352 ER, 259/346/390, 105 sOPS+, 73 points higher than 9th. So obviously a lot more IP ... but also just 97 more ER in 335 more innings. I suppose these days when the visitors don't score, the home team wins before getting an earned run they'd have gotten so that would explain some of it.

So it's working in terms of innings but I'm not sure I understand wny. Putting a runner on second doesn't have a dramatic effect on the defensive alignment, maybe it throws the pitchers off their game.

Wow, I just realized ... this year it's 147 runs in just 110 innings pitched. Last year it was 627 runs in 535 innings. That ain't baseball.

Meredith Wills has dissected a bunch of baseballs from this season and found that we're still on the same baseballs that we saw in the second half of last season.

The humidors seem to be an issue and not just the humidors themselves. She found in testing that the way they were using the balls, storing them in the humidors, then at some point the mud rubbing, and then back into the humidor, was causing unexpected results to the balls.

I recently had her on a podcast to talk about this, but I will not link it since we were always strict about self-publicity. I'm sure an enterprising person can find it if they'd like though!

My point about it not working, which admittedly is speculative, is that if both teams nearly always score their zombies then you're just turning your 4-4 game into a 5-5 game. If they're ending earlier than they used to, it's not clear that's the zombie.

If a lot of those 4-4 games are turning into 6-4 games, you would have the same number of unearned runs being scored but you would still be getting the desired effect. And I would guess the defense does start aligning differently once a single run is scored (such as playing the infield in at some point), which makes additional runs more likely.

On second thought, that second run would only be unearned if an error is involved, I assume.

However, I think you would see occasions where the visiting team scored multiple runs in the top half of the frame, and would then be indifferent to yielding the unearned run in the bottom, leading to more of the UE runs scoring without affecting the outcome.

#24 What you're running into really is the limitations of modeling run scoring strictly with OBP and SLG. The standard error of that model is in the range of .12 runs per game.

Mind you, the best models we can make still have standard errors in the range of .09 RPG which is why we can say with some confidence that offense is mostly about OBP and SLG (and OBP is generally the more important factor)

Doubles and triples are way down as well, as the well-hit ball just does not travel as far. They sure seem to have, in not ruined the game, impaired its appeal.

Part of what's going on is that batters are tearing it up in extras ... or teams are down to their 6th best reliever by the time they get there. The line is 280/389/464 in extras, a 146 sOPS+. Nearly 200 points better than the 9th inning OPS. That's just weird.

In terms of the extra innings explosion of runs scored, I am interested in the "depth chart" factor. Is it possible that most teams look at the "zombie" extra innings era, and conclude that the outcome of an extra inning game is more of a crapshoot than ever before?

Are we finding that teams are doing a better job of using their best relievers at the point of maximum leverage, even if that is earlier in the game? If so, this would reduce the opportunities to use them in the bottom of the 10th inning with a one-run lead.

The continued decrease in the number of innings thrown by starters results in managers digging deeper into their bullpens, in general. What I'd like to know is where does the average bullpen drop off the table, in terms of quality? I suspect that most bullpens have four or five guys that are foundation pieces of their bullpen for that season, and the remaining two or three slots are going to guys that are either AAAA guys, or are dealing with injuries, or just don't have the confidence of the manager. I'd imagine that there is an average amount of depth a team has where the marginal difference in getting to that inflection point makes a pretty big difference.

I've coached a lot of basketball, and in youth sports through high school, figuring out where that point is on their bench - and then aggressively trying to get that part of their depth chart in foul trouble early - is a huge part of game planning. To me, this is sort of the same thing for bullpens. The value of a starting pitcher who can give you 6+ quality innings is largely that it allows the team to substitute an inning of its 4th-best available reliever with an inning from a solid starting pitcher. That's a lot of value - but by the 10th inning, you may be down to your 5th-best relief option.

SH/game: 0.08/0.25/0.37

With the DH in both leagues now it looks like the sac bunt is well on its way to extinction.

Doubles and triples are way down as well, as the well-hit ball just does not travel as far. They sure seem to have, in not ruined the game, impaired its appeal.

This is why they need to increase drag on the ball, not deaden it.

They sure seem to have, in not ruined the game, impaired its appeal.

It's not that MLB is losing its popularity under ManfredBall. I just think that its appeal is becoming more selective.

The Mets have only one player with more than 8 HR, and their 47 total HR is below the league average of 50.

but they have six every-day players with OPS+s of: 123 127 127 138 142 155 plus a 162 from part-timer Guillorme.
so they have a league-leading 254 RBI, vs an NL average of 207.
they also top the NL with a 116 OPS+ thanks to all those good numbers above (their Cs are horrible).
1st in AVG at .268. tops in triples with 13.
383 whiffs, vs league average of 417.

it's pretty entertaining, actually.

This is why they need to increase drag on the ball, not deaden it

You want to:

A) Increase "action", e.g. singles, doubles, triples, decrease strikeouts.
B) Limit home runs, so folks are forced to go for BIP

If you increase drag on the ball, you will also probably increase movement (e.g. curvier curveballs), the drag may decrease velocity, but not much.

So you will want to decrease the pitch velocity to compensate for the increased movement, or the strikeouts will go through the roof.

Lower the mound, incorporate the pitch clock, and make the pitchers throw more than 1-2 innings at a time.

Though as some have pointed out, with the advancements in conditioning and kinesthesiology, one may never realistically get back to 1970's average velocities.

It's not that MLB is losing its popularity under ManfredBall. I just think that its appeal is becoming more selective.

Boston's not a big [baseball] town anyway.

Wow, I just realized ... this year it's 147 runs in just 110 innings pitched. Last year it was 627 runs in 535 innings.

I think you're onto something. Since zombie runners increase scoring that much. We could reduce games to 7 innings if we start each inning with a zombie runner. More scoring, games take less time. Win/win.