Three wild ways baseball in 2018 is unlike ever before

Even three weeks into a baseball season, there are a lot of loud changes for us to appropriately react to: Launch angles are up (from 11.1 degrees last year to 11.7), home runs are down (even accounting for the weather), the weather is killing attendance and Shohei Ohtani is real.

More quietly, though, the game is always shifting and sloping. It can sometimes be hard to know what's permanent, what's portentous and what's nothing more than an April fluke. Here are a few of the under-the-radar ways that baseball in 2018 has been unlike any baseball played before it:

1. The bottom of lineups have been historically bad.

Of course, the bottom of lineups should be bad. Otherwise, as with a snowout in a domed stadium, something ain't right. But both leagues have so far seen unprecedentedly bad offense from the final spot in the lineup -- the eighth spot in the National League, the ninth spot in the American -- and, more broadly, from the bottom three spots.

The seventh-, eighth- and ninth-place hitters have hit .210/.287/.322 this year, excluding pitchers. That's only 73 percent of the leaguewide OPS, as measured by a stat called tOPS+.

Last year's No. 7/8/9 hitters, by contrast, were 90 percent as good as the rest of the lineup, with a .245/.312/.398 slash line. This isn't something that tends to vary much year by year: Over the past decade, the bottom third of the lineup's tOPS+ has never been lower than 86 or higher than 90, until this year's collapse.

If it holds up, baseball's bottom-of-the-order batters will be worse than in any season in history. The second-worst group had a tOPS+ of 79, and that was way back in 1918. No other season's No. 7/8/9 batters have ever been worse than 82 percent of the league overall. And the falloff at the very bottom of the order is even more stark:

NL No. 8 hitters: .199/.294/.296, down from .248/.324/.391 last year

AL No. 9 hitters: .207/.272/.309, down from .233/.293/.366

So what's going on here? One hypothesis is teams are getting better at putting their worst hitters at the bottom of the order -- they're more correctly identifying their worst hitters, and they're optimizing their lineups better to give those guys the fewest plate appearances. That'd be a fun story about baseball intelligence in action, but this isn't some 20-year trend that tracks the increase of data-smarts in MLB front offices. It's an abrupt shift from last year, when bottom-feeders hit pretty well, all things considered.

Another hypothesis is these weaker, inferior hitters are particularly hurt by this April's low-homer environment. What made the apparently juiced baseball of 2015-17 so interesting is it benefited almost everybody, even little dudes and dudes who aren't even very good. This wasn't the era of a handful of changed players banging 60- and 70-homer seasons, but of guys who had been hitting two, seven or 13 homers suddenly hitting five, 10 or 20. Scores of hitters responded to the apparently juiced ball by hitting more fly balls. Rational players were doing rational things.

But the ball isn't carrying this month, and all those hitters who retooled their swings are getting all the fly balls without all the homers. That might not matter much for sluggers who are really good at hitting home runs 85 feet over the wall, but it could matter for that subset of hitters who'd briefly found themselves credible home run threats. The juiced ball might have been a trap, targeting especially these hitters.

But the problem with this hypothesis is the decline in offense from No. 7/8/9 hitters isn't restricted to a drop in home runs or power numbers. The bottom of the order is hitting worse, relative to their peers, on grounders, too. They're hitting more grounders, relative to their peers, than they did in previous years. And they're striking out more, relative to their peers, than in previous years. They're just bad. You'll want to see it continue a few more months before you necessarily believe it, but for now, it's a thing.

2. Games are moving faster, yet still taking longer.

It's complicated! The length of a game is the result of factors both in and out of the players' control, in and out of the league's control, in and out of nature's control. So, yes, the average game has taken three hours and eight minutes, matching last year's record length.

But ... that's in part because there have been more extra innings than in a typical season. In (most likely) a weird fluke, 11 percent of games have gone long so far, which would be the most since 1965 if it held up (which it won't). This is on track to be the first season since 1981 in which the average game is at least 54 outs long. Games are taking longer because games are going longer; control for that, and the average nine innings have taken only three hours and one minute. That's still the third slowest in history, but it's four minutes per day faster than last year. Clearly, the MLB pace-of-play initiatives are working!

BUT -- that's not proof that players are moving faster or pace is improved. Rather, it's proof that offense is down, which means fewer batters are hitting in each game and each inning. Leaguewide, on-base percentage is .316, eight points lower than last year and the second-lowest mark since 1972. It will go up -- offense always does as weather improves -- but it's a primary reason that games are taking less time so far. A lower OBP means fewer baseball events, and less time spent with runners on base, when pace of play slows down. Simple.

EXCEPT -- this has been counteracted by the continued upward trajectory of pitches per plate appearance. The average batter has seen 3.93 pitches per plate appearance this year, up from 3.89 last year, which was itself up from 3.88 the year before, which was up from 3.83 the year before, and so on. If that 3.93 figure holds up, it would mean seven or eight extra pitches per game since 2014, each of them taking 25 or 30 seconds on average. Those seven or eight extra pitches per game are also part of why we're seeing a record 3.44 relievers per game, per team -- almost half an extra pitching change per game.

So strip all this away, and let's focus on how long each plate appearance is taking, and how long each pitch is taking:

In 2018, a pitch is thrown every 37.2 seconds, including all the time between batters and between innings. That's more than eight-tenths of a second faster than last year. It's comparable to the pace in 2012. That's good. Major League Baseball has, it appears, partly arrested the increase in general dawdliness, and sent us all back to a faster time.

But in 2018, a plate appearance takes, on average, two minutes and 26 seconds. That's two seconds faster than last year, but slower than every other year on record except for one -- 2014, which inspired the previous last round of pace-of-play initiatives.

So: Baseball in 2018 is slightly faster, thanks to the things the league can control. And baseball in 2018 is also slightly slower, thanks to the things the league can't control. I'm skeptical this latter trend will stop anytime soon. Home runs, walks and strikeouts all act as companion stats, with an increase in one usually leading to increases in the others -- and, logically, increases in pitches per plate appearance.

3. Strikeout rates, wild pitches and hit batters are up. Waaaaaay up.

Yes, strikeouts are up again. This will be the 14th year in a row strikeouts per game have gone up, and the 12th in a row a new record is set. I'm confident in these statements, because April strikeout percentages track quite closely with full-season rates, usually going up a tiny bit as the season gets deeper. Three things make this year's edition a bit interesting, though:

First, if it holds up, this would be the biggest single-season increase (in strikeout percentage) since 1986. Last year, pitchers struck out 8.2 batters per game; this year, it's 8.9. Sixteen teams right now are striking batters out more frequently than Roger Clemens did in his career. If you had any hope that strikeout rates might reach a ceiling and level off, there's no evidence we're getting close to it.

Second, besides strikeout rates going up, we see two other things: Wild pitches are way up, from last year's record 0.37 per game to this year's 0.40 per game; and batters hit by pitches are way up, from last year's 0.36 per game to this year's 0.42, which would be a record. Both are natural cousins to the strikeout. For one thing, pitches in the dirt and pitches inside are good places to get swinging strikes; but, for another, they result from the same phenomena that cause strikeouts in the first place: Pitches are fast, nasty and hard to deal with. That's true if you're trying to put the bat on it, but it's also true if you're trying to catch it or dodge it.

Finally, strikeout rates for starting pitchers have gone up more than for relievers. Last year, we wondered why increased reliever usage wasn't shrinking the gap between starters and their relievers. (More relief innings means more innings going to lesser relievers; fewer starter innings means starters can put more effort and energy into each pitch, and avoid the dangerous third and fourth trips through the order.) We might finally be seeing the gap start to close:

• Starters: 22.1 K%, up 1.6 percentage points from last year

• Relievers: 24.1 K%, up 0.8 percentage points from last year

Relievers still strike out 9 percent more batters than starters do, but that's the lowest gap since 1986, before modern bullpen usage had really kicked in. Now, that's just strikeouts. If you look at ERA, or runners allowed, or OPS allowed, there's no real change; relievers are still better than starters by about as much as they've recently been. But there's some indication that starters, increasingly conditioned for shorter roles, are pitching for strikeouts more, and getting them.

Last April, we looked at three quiet changes that had shown up early in that season, and wondered whether those changes were real. So, were they?

Sacrifice flies were, at 0.22 per game, on pace to set a record low. We speculated that the gap between 2017 and recent years would close, but not all the way -- and that's what happened. There were 0.24 sacrifice flies per game in 2017, the lowest since 1972. (This year, they're back up to 0.25.)

Wild pitches were, at 0.38 per game, on pace to set a record. As noted above, they ended up at 0.37, matching the previous high, which is in even more danger of being toppled this year.

Double plays were, at 0.81 per game, higher than they'd been in a decade. The rate came back down to 0.78 per game, higher than any season since 2009 but more or less consistent with 2015 and 2016 rates. Double plays are way down this year, incidentally. With fewer runners on first this year, and more strikeouts, and more fly balls than ever, that seems real, though three weeks into a season one can never really tell.