OFF-DAY: RUMBLINGS ABOUT BASEBALL'S POST-NEO MALAISE...

Yes, they did have off-days in 1962...for the Giants and Dodgers, April 20th was one of them. We'll return to 1962 tomorrow. (Which allows us to sharpen our pen-knives on those present-day folk who would mislead us with their overly glib--and, in some cases, downright inaccurate--use of statistics.)

So let's call this post a tale of two Joes--Posnanski (we pronounce it "Pose-nan-ski") and Sheehan (pronunced "Sheet-pan"). And they are both finally worried that their ongoing bromance with "launch angle" has turned into what we've been saying about things for almost a decade--an impending, disastrous sinkhole for offense that makes the offensive downturn in 2014 look like a walk in the park.

Both men have become skilled (if that's the right word for it...) at tossing out intriguing, dramatic-sounding stats that are always either strangely bereft of context, or simply mangled (either deliberately or through incompetence) in order to make a claim that is either exaggerated or totally inaccurate. They've done it again this week, as the early returns on baseball's offense in '22 have put panic into their brainstems.

Let's dispense with Joe S. first, as he's simply mangling stats. Professing that the downturn we've seen in offense (particularly in HRs, where we're seeing "only about 1.01 HR/G across MLB thus far) is due to a new baseball, Mr. Sheet-pan has his Chicken Little act setting off alarm bells in 5.1 surround sound. 

In order to make that point, of course, Joe has mangled the stats. He claims that hitters had a SLG exceeding .900 (!!) on fly balls in 2021, when that number is really their OPS (SLG + OBP). Now even the "mediots" that Joe and his brethren at Baseball Prospectus used to mercilessly disparage during their "formative" years will know that OPS and SLG are not synonymous. A look at hitters' performance when they hit fly balls, as retrieved from the Play Index at BB-ref.com shows that hitters slugged .694 when they hit fly balls in 2021; thus far in 2022, that figure is .559. 

Unfortunately, BB-ref doesn't make it possible to carve out SLG on fly balls into subsets of full seasons, but we strongly suspect that we're seeing a cold weather effect at play. If we could capture data that showed us the first two weeks of the 2021 season, we'd probably see that SLG on fly balls for that period was somewhere in the low-mid .600s. So there is almost certainly a downward turn in 2022--but it's not nearly as alarmingly pronounced as Joe S. claims. By positing that it's not just a cold-weather effect and doing a Chicken Little act on the baseballs, of course, Joe S. is side-stepping the necessary research to actually anatomize the cold-weather effect. (Things like that tend to get in the way of Chicken Little-like pronouncement...)

(Of course, the most interesting thing one can discover in the BB-Ref data is that, despite the craze for "launch angle" over the past 5-6 years, hitters are actually hitting fewer fly balls now than they did during the offensive explosion. For example, in 2003--MLB batters hit nearly 48,000 fly balls. In 2021, however, that figure was just over 39,000 fly balls. 

Take a guess where those 9,000 missing fly balls have migrated. That's right: they've become strikeouts. And that's why a solid analyst such as Brock Hanke recently made the forceful point that the game won't be able to solve the mess that "analytics" and its neo-sabe enablers have made until they address how to reverse the inexorably rising trend in strikeouts. A word to the wise: don't hold your breath.)

We'll see if MLB's apparent use of league-wide humidors is going to have any effect on offense, but for right now we're here to tell you that the downturn in HRs is due to a combination of factors, with the most significant one in play right now being the cold-weather effect. We can get hold of the overall stats for performance in cold weather (once again thanks to BB-Ref), and we can compare the numbers for 2021 and 2022 for the first two weeks of each year--the point in time with the highest occurrence of playing conditions where the temperature is less than 60 degrees.

When we do that, we can see that in the 126 games thus far in '22 played at < 60 degrees, the overall offensive numbers are .220/.303/.342 with a HR/game average of 0.74. 

Then, looking at the games thus far in '22 which were played at 60+ degrees, the offensive numbers are .237/.313/.391 with a HR/game average of 1.06.

Note that cold-weather games account for about 38% of the games played thus far in '22. That percentage will drop precipitously in a few weeks, as it always does--and at which point all of the numbers we've just displayed will rise. There were only a total of 458 "cold-weather" games in 2021, about 9% of the total number played (274 in April, 156 in May, 4 in June, 2 in July, and 22 in September).

We can make the same breakout for the first weeks of 2021 that we just did for 2022, of course. When we do that we see that the "cold-weather games" (< 60 degrees) in the first 13 days of the 2021 season (a total of 90 games) produced the following offensive numbers: .233/.316/.370 with a HR/game average of 0.96.

In the games in the same time frame played in 60+ degree weather (a total of 232 games) in 2021, we see these offensive numbers: .236/.313/.405, with a HR/game average of 1.27.

Note here that the "cold-weather percentage" for the first two weeks of 2021 was a good bit lower than has been the case in '22--just 28%. So conditions last year were more favorable for the "launch angle" approach, and are reflected in the HR/G average for those warmer early-season games--an average that's pretty close to what MLB achieved in 2021 as a whole (1.22 HR/game).

Also note that the "cold-weather" offensive performance is the portion that is seriously down from last year; the only thing that's declined in the warmer games, really, is the HR/G average. That tells us that things will snap back a good bit when warm weather is pervasive, and that the effect of dominant relief pitching--something that Joe P. has caviled about (fear not, he's not going to escape without getting at least some of what he deserves...)--is a temporary one. 

As you'll see in a minute, there are some nuances we can trace in the relief pitcher data that Joe P. (like the other Joe) doesn't bother to consider as he searches (as always) for how to get on both sides of the same issue. BB-Ref can let us break out appearance-by-appearance reliever data by temperature: it's a little messy to work with, but it can be done. When we do so, we see that relievers in '22 have been pitching lights out in cold weather. 

They have a 3.05 ERA in games where the temperature is < 60 degrees; in the warmer games, that ERA is 3.71. The slash line comparisons are: colder weather .210/.301/.319; warmer weather .227/.302/.380.

Note the ISO difference in those slash lines: colder, the ISO for relievers is .109; warmer, it's .153. In terms of HRs per 650 plate appearances, that's: colder, just under 12 HRs per 650 PAs; warmer, just under 20 HRs per 650 PAs. (That's a 60% increase in HRs allowed in the warmer games.)

So what's really happening this spring is that relievers have enjoyed inordinate success in avoiding the long ball during their cold-weather outings. But those cold-weather outings will be going away soon, and it will produce an inevitable decline in reliever effectiveness. 

So Joe P.'s yodeling in a recent blog post about how starting pitches are the weak link (when, in fact, starters are pitching better this year than last) is predicated on a hasty and incomplete analysis of the data. (Which is why we like to say that Joe P. puts the "pose" in Posnanski...actually, we don't like to say it--we wish like hell he'd stop doing it, but it's clearly become congenital.)

One final example of Joe P. tearing a page from the Joe S.'s Chicken Little playbook: he found a startling stat--and whatever else we may say about Joe, he's not dumb (you can't be a world-class poser if you're dumb)--about the percentage of games in 2022 thus far where the starter has gone at least six innings. Right now the stat is eye-poppingly low--around 15% (Joe had it at 14%, but it's starting to rise for reasons that we'll make clear in a moment).

It is startling, but it's also clear that it's an artifact of Manfred's folly (aka the lockout), which bequeathed us a too-abbreviated spring training. Joe actually knows all this, and does his usual sneaky backpedaling while simultaneously trying to make hay with readers due to the "shock value" of the stat--which also means not putting it into context...such as, what was the same percentage in the first couple weeks of 2021? You won't find it as part of Joe's blog post. (It was just over 30% for the first two weeks of '21, and wound up just under 40% for the entire year; you can expect the current figure to make a steady rise in that direction as the season moves out of its extended spring training mode.)

Of course, contextualizing these numbers allows "the Pose" to ring an alarm bell--and unring it simultaneously. He can point to stats that suggest that the sky is falling for starting pitchers even as the sky is falling for hitters--when the truth is that the cold weather has helped relievers go on a hot streak. It allows him to get on both sides of the "three true outcomes" miasma in the game, and defend the so-called intellectual efficacy of the "launch angle" catastrophe that subverted "analytics" and has infected the creation of new "stats" that threaten to institutionalize the very precepts that have strangled the game into two-dimensionality.

It's clear that these two guys are not going to lead baseball out of its wilderness; they've become too enamored with the sound of their own voices to be of any real use. The most ominous aspect of it is that they selectively use and loudly misapply stats in order to make it appear that they are somehow part of a "cutting edge" that has opened larger and more gaping wounds in a game that, like the world as a whole, deserves a better fate. 

Our best educated guess is that MLB will wind up at around a .240 BA in 2022, with about 1.05-1.12 HR/game. The game desperately needs to add hitters like Rod Carew, Tony Gwynn, and Wade Boggs back into its mix again--the notion that pitchers are now so highly evolved that such players can no longer exist is part of the cul-de-sac that both Joes are circling in their Hoverrounds along with all the rest of the "best and brightest." We need a sea-change in "analytics" that looks for ways to re-integrate hitters like Willie Keeler, Paul Waner, and Luke Appling, and that celebrates and sustains a diversity of excellence. We might get there, or we might not...but one thing seems certain: we won't get there by listening to the natterings of the two Joes.