I'd like to see this focus on 2 strike counts only (which is the only time you can strike out); my guess is that a lot more players would be better off putting it in play w/ 2 strikes than continuing to swing from the heels.

Isn't there also a tautology at work here too? Yeah, someone like Baez is more productive on balls in play-when he is striking out a ton. If he cuts down on his strikeouts then the advantage swings back to the Three True Outcomes, doesn't it?

Traditionally speaking, it's been swing with intention for strikes one and two, and then hedge for contact with two strikes and behind in the count. Nowadays, guys swing with intention no matter what. Swinging for contact is a skill, and if a guy doesn't do it often, he might be better off just not doing it at all, ever. Thus, Javy Baez. Maybe this is just his comfort zone.

Doesn't this study assume that the lesser swing produces the same result when the ball is struck as the harder swing, with more strikeouts and more home runs? That logically seems wrong to me.

The harder, home run swing, when producing a struck fair ball but not a home run, will produce a harder hit ball than the lesser swing will. I do not think that is refutable. My contention is that the harder hit ball is more likely to get through the infield, if it is a grounder, and more likely to fall into a gap and be an extra base hit (less time in the air) if it is a line drive in the air. As to grounders, I think fielder ranges have gotten much better over the years as the skill of range has gotten measurable and teams are creating line-ups with better range on the field. As to fly balls, the same applies; I just do not see the fat, slow left fielder any more; no one playing regularly is Greg Luzinski or Adam Dunn. With better ranging fielders, it became harder to sneak a ball through the infield or get it in the gap in the outfield, and that means you have to hit it harder, to get a grounder through or a ball into the gap. Thus, the harder swing aids the ground ball singles and extra base hits which are not home runs.

#1. I did a simple study with what little data we had years ago. Simply assumed that Adam Dunn could become Ozzie Guillen with two strikes. Why Guillen? Because at that particular time he had the second lowest swing and miss rate. And assuming that anybody could become Tony Gwynn at will didn't seem like a very realistic assumption.

Thing is that Guillen wasn't much of a hitter with two strikes on him. Almost no walks or extra base hits (career numbers of .209/.227/.260 -- though I didn't use career numbers because back when I did the study we didn't have them)

Dunn wasn't very good with two strikes, but would still be giving back runs ( .142/.263/.290 -- he was better in the time frame of my study )

The real key is that to avoid striking out with two strikes you have to swing at a lot more pitches and weak swings at balls you can't drive tend to produce a lot of groundballs to short and nobody hits a lick on those. Further, you're giving up a lot of two strike walks. Dunn had a career .476 OBP on full count PAs. Guillen's was .338 (in many fewer PAs)

Hitters might be better off swinging more earlier in the count but ... well basically every hitter will swing at any pitch he thinks he can drive. Still, I think a lot of people would be surprised that prime Mark McGwire swung at the first pitch more frequently than Ichiro did.

I'd like to see this focus on 2 strike counts only (which is the only time you can strike out);

I was curious about that too awhile back.

MLB, with two strikes
1988 .184/.248/.266
1991 .184/.253/.269
1996 .190/.265/.293
2001 .185/.258/.291
2006 .194/.264/.300
2011 .180/.247/.272
2016 .176/.246/.276
2021 .166/.243/.272

The Mark Belangers and Mario Mendozas and Mallex Smiths and Yolmer Sanchezes of the world ... But the Ted Williamses and Babe Ruths and Mickey Mantles and Barry Bondses and Mike Trouts...

ISTR in one of the Abstracts or maybe Baseball Books, James ranted about this kind of lazy hackneyed writing. MLB does not have Mark "Belangers" and Ted "Williamses" and Mike "Trouts" at their disposal. Each one is a single player with his own unique set of strengths and weaknesses. Mark Belanger. Ted Williams. Mike Trout. Etc. Whether he did or not, it's disappointing to see a writer as good as him fall into using this crutch.

To Pat Rappers' Delight:

Thank you. I thought I was the only one who hated this expression. It's not much more difficult to write "players like Mike Trout".

My other pet peeve is the term "skill position". So therefore everyone else is playing an unskilled position.

"Whether he did or not, it's disappointing to see a writer as good as him fall into using this crutch."

He hasn't been that writer in many years now. But I do recall him doing that pluralization thing back in his heyday. This kind of lazy half-assed study tho is certainly way beneath his peak.

The really lazy assumptions is that the choice is between the current mix of HR/BB/SO and the average result of BIP. The whole point of altering your approach would be to change it when Ks are most likely and HR/BB least likely. No one is saying swing for contact when the count is 3-0, 2-0, or 3-1, when you're looking for a specific pitch to crush. In unfavorable counts, the K vs HR/BB mix is going to swing heavily towards Ks, so you don't next great BIP results to be better than a bunch of Ks and an occasional dinger.

I thought I was the only one who hated this expression.

You're definitely not the only one.

It's not much more difficult to write "players like Mike Trout".

Since there really are no players like Mike Trout other than Mike Trout, you're better off with "such as."

Even Mike Trout ain't Mike Trout no more.

If "Mark Belangers" is good enough for the Bill Jameses of the world, it's good enough for me! Hopefully I've never gone so far as "Belangers, Mendozas, Smiths and Sanchezes" too often though.

Doesn't this study assume that the lesser swing produces the same result when the ball is struck as the harder swing, with more strikeouts and more home runs?

Sorta. It assumes that you can swap a TTO outcome for a BIP and that the TTO outcome you surrendered is a random (or average) one and that the new BIP has the same average result as your "regular" BIPs.

But despite all this (supposedly) harder swinging, league-average results on BIP had been incredibly stable. For 1996-2021, the median OPS on a BIP is about 682 with 21 of 26 seasons between 674 and 688. If the harder swings are leading to harder-hit, tougher-to-handle BIP then the defenses have managed to keep up. Obviously there is player-to-player variation and a fast player like Baez will usually have better BIP outcomes than a slow-footed C. And it's also true that walk rates haven't changed a lot.

So it pretty much has been about trading Ks for HRs. But even there, we've got a paradox of sorts. OPS not-in-play was much higher in the 2010s than the 2020s (but is still 150-200 points higher than OPS in-play). We also haven't seen extreme individual HR totals since the early 2000s. The increase in HRs (and HR/PA) is on the order of 500+ per year but it's more of the "every starter hits at least 15" as opposed to "sure that guy has 250 Ks but he did hit 80 HRs."

Missing from this is the notion of an optimum or an equilibrium. Javy has basically found it for us. If you K 33% of the time while walking 4% of the time and have a 333 BABIP but don't hit 40 HRs then you have finally reached the break even point of BIP vs BnIP. If you can walk more, you're Joey Gallo. (By the way, with the Mets, Javy's walk rate so far is nearly 8% with a corresponding drop in K-rate.)

Career OPSip, OPSnip, BABIP (AD is Adam Duvall, who knew he was that Gallo-esque)

JB 776 807 335
JG 635 991 265
AD 633 988 267

Javy walks in about 12% of his nIPs and homers in about 11.5%. Gallo walks in about 25% and homers in another 11%. Duvall walks about 14.5% and HRs about 13.5%. It doesn't seem to me the Javy-Duvall OPSnip gap should be that big but it is. (maybe bad math by me)

Back to the top point though. James is assuming Javy would be swapping a 807 OPS PA for a 776 OPS one and, even more relevant, gaining 72 points of OBP while surrendering just 102 points of SLG which is pretty much an even trade in production terms.

So ... it's not surprising Javy is where he is. Everybody agrees that swinging at pitches a foot outside and ankle high is totally counter-productive so he'd be a better player if he could just stop that. But, yet another paradox, changing that behavior doesn't lead directly to a BIP. The last thing you want to do is to keep Javy fron airing it out on pitches in the zone. You want to stop him swinging at crap so that either they walk him more or they throw him more pitches in the zone. If they choose the latter, then you'll get more BIP and more HRs.

Swinging balls out no matter the situation is the essential, definitional core of modern oafball. The zenith, which will probably be surpassed at some point, was the Reds in that first extra innings playoff game last year where they just kept oafishly and incompetently flailing with runners in scoring position in multiple tied extra innings only to connect with nothing but the thick and stuffy Atlanta (*) air. Just laughably inept, the kind of thing we could imagine one of the featured players of Greek mythology being driven to gouge their eyes out after viewing.

(*) Or Cobb County. Whatever.

Further on #5 ... I've added K and BB per PA and BABIP

MLB, with two strikes
1988 .184/.248/.266, 34.7, 7.4, 278
1991 .184/.253/.269, 35.0, 8.0, 281
1996 .190/.265/.293, 37.0, 8.6, 296
2001 .185/.258/.291, 37.4, 8.1, 289
2006 .194/.264/.300, 36.1, 8.0, 296
2011 .180/.247/.272, 38.1, 7.5, 287
2016 .176/.246/.276, 41.1, 7.9, 293
2021 .166/.243/.272, 43.3, 8.3, 287

So indeed there seem to be a couple of obvious break points. Sometime between 1991 and 1996 -- probably 1993-1994 with the start of sillyball -- the K-rate went up a bit but so did BABIP. Then sometimee between 2011 and 2016 there was a pretty substantial jump in K-rate with no obvious improvement in BABIP. However that also hasn't resulted in a better ISO, just a lower BA (and maybe higher walk rate). If this was a change by hitters, it hasn't helped; if it was due to pitchers doing better in 2-strike counts, then batters haven't adjusted. More likely a bit of both -- pitchers doing a better job of exploting hitter aggressiveness with 2 strikes.

There's been a big change in the number of PAs too of course. It's gone from a bit under 88,000 in 1998 to nearly 100,000 in 2019 (over 91,000 and counting this year, should be down a bit on 2019). That reflects rake and take. So it's been a 14-15% increase in the likelihood of reaching two strikes and a 15-16% increase in the probability of striking out once you get there which gives you an overall increase of about 32% in Ks which you may have noticed. Point being about half being a change in overall approach, half a change in 2-strike approach. It would be good to see if the distribution of (got to two strikes at) 0-2, 1-2, etc. has changed as well. It's possible that the 2-strike approach hasn't changed but they are now more often getting there at 0-2 and 1-2 than they used to.

Walt, your analyses are fantastic.

In the very early days, striking out was utter failure. The rudimentary gloves and poorly maintained playing surfaces made it worthwhile to put it in play. Also, there were no mounds - pitchers were on flat ground.

If we want less TTO ball, we've got to get back to the modern equivalents of those conditions.

If we want less TTO ball, we've got to get back to the modern equivalents of those conditions.

Mandating incompetence in the grounds crews might be difficult to implement.

If we want less TTO ball, we've got to get back to the modern equivalents of those conditions.

Smaller gloves, push back the mound a bit, and make the ball bouncier, but with a lot of drag, so grounders and liners are hot, but fly balls die.

OPS not-in-play was much higher in the 2010s than the 2020s (but is still 150-200 points higher than OPS in-play).

But even the Adam Dunns are Adam Dunn for 50 percent of less of their ABs. (If BB are part of OPS not in play, I don't think they should be; they're not an effect of swinging.)

#15 It'd be interesting if we could check whether the percentage of called strike 3 is rising on various counts. It's my impression that this is up a fair bit but I've never been able to examine that systematically except by going into Retrosheet myself and I just don't feel like doing that.