2017 Shift Slowdown

September 11, 2017

Back in April, we took a look at the state of shifting in Major League Baseball and predicted that teams were on pace to total 31,677 shifts during the 2017 season. While this projection was a 13 percent increase over the total number of shifts in 2016, the projected year-over-year increase was actually the smallest increase in league-wide shifts since they began their record-breaking trend between 2011 and 2012. As the 2017 season comes to a close, it turns out that our April projection for a slowdown in shifts was too conservative. For the first time since the 2011 season, defensive shifts are on pace to decrease from the prior year. Note that all shift totals referenced in this article are shifts on balls in play.

MLB Shifts by Season
2011 2012 2013 2014 2015 2016 2017
2,350 4,577 6,882 13,299 17,826 28,130 27,187*
*Prorated from 23,998 shifts thus far

With a combined 23,998 shifts so far in 2017, this season's total has far exceeded the shift totals for any season prior to 2016. However, the projected 2017 shift total does not look like it will surpass the approximately 28,000 total shifts tracked in 2016. So what’s causing this stagnation? To best understand this question it’s important to consider that each team came to its own conclusion on when to make shifting a priority. A number of teams took the shifting plunge prior to the 2013 season and thus became "first wave" shift adopters. This group includes teams such as the Astros, Pirates, Rays, Orioles. These teams continue to remain at the top of the shifting leaderboard in 2017, albeit with some teams slightly off of the last year's astronomic pace.

MLB Team Shift Prorated Leaders, 2017
Team 2016 2017 2017 (Prorated)
Brewers 1,489 1,454 1,647
Astros 1,869 1,349 1,528
Rays 1,588 1,336 1,503
White Sox 783 1,298 1,481
Mariners 1,482 1,295 1,467
Pirates 1,489 1,245 1,401
Yankees 1,380 1,117 1,274
Angels 1,479 936 1,060
Reds 877 911 1,025
Orioles 787 850 963
Note: through games of September 10

Conversely, there are a number of teams that began shifting more recently. These teams could be categorized as "second wave" shift adopters. This group includes teams like the Rockies, Cardinals, and Angels. It is the pullback by these second wave shift adopters that has caused the stagnation of total shifts in 2017.


Top Six Shift Decliners, 2017
Team 2014 2015 2016 2017(Prorated) Difference: 2016 vs. 2017 (Prorated)
Rockies 114 1,014 1,355 655 700
Cardinals 367 311 812 361 451
Angels 357 439 1,479 1,060 419
Astros 1,341 1,421 1,869 1,528 341
Braves 213 211 1,026 734 292
Giants 361 553 858 704 154
Note: through games of September 10

With the exception of the Astros, the remaining top six shift decliners in 2017 all had a sudden shift increase in 2015 or 2016, followed by a pullback in 2017. In particular, the 2017 Rockies are projected to shift less than half as many times as they did in 2016.


Of course, it's hard to draw definitive conclusions about how this year's stagnating shift trends can affect future years. However, it’s pretty clear that in 2017 some of the late defensive shift adopters seem to have had buyer's remorse with their shifting tendencies.


COMMENTS (12 Comments, most recent shown first)

It's a plausible argument, Maris. I wonder: has the historical context changed in recent years, when--partially in response to the increased frequency of shifting--everybody and his dog is uppercutting the ball?
6:42 AM Sep 16th
MarisFan61 what does for "balls in play" (in case anyone's still looking here......btw if you aren't now, people will eventually -- these articles are for posterity, and they show up prominently in google searches).....

I think I see what they're doing -- why they include sacrifice flies but not sacrifice bunts. It's not clear to me that it's the best way to do it, but I think I can see why. I wonder what other people think of it.

It gets into the 'philosophy' of sac bunts and sac flies; not mainly what the players or what the teams think of them, but what an analyst thinks of them, and what we think of them. And I suspect that is seeing this differently than players and teams do.


In trying to understand what is doing, we have to remember the context: The main reason they're deriving "balls in play" is to calculate batting average on balls in play. So, I guess they're just looking for what's most relevant for that.

I guess the reason they use sac flies but not sac bunts as an 'addend' to get "balls in play" is, they view sac bunting as "giving yourself up" -- you're not trying to get a hit -- but they don't view sac flies as giving yourself up; they consider it as though you were really trying to get a hit but you just didn't. Or maybe it's a little different than that: Maybe they're not viewing it in terms of what's in the hitters' heads but just in terms of what they think is the likelihood of getting a hit if you're bunting vs. if you're hitting fly balls. Heck, there's all kinds of ways of viewing how they're viewing it: maybe they hardly believe at all in "trying to hit a fly ball"; they figure that at least for the most part, the guy is trying to get a hit, as usual.

My own take on all those things is, sure there's a difference, and there's some tendency 'in favor of' bunts in all those respects. But it's not absolute, and I don't think we know that it's large.
Except on this one thing: Of course every time a player does a sac bunt, he was trying to bunt. :-)
That's not the case on sac flies.

But I think it's clear that a hitter's approach in a sac fly situation often is adapted for the purpose of maximizing the chance of a fly ball, just as the pitcher's approach often is adapted for the purpose of minimizing it. Also it's obvious that when a batter is sac bunting, he sometimes gets a hit; I imagine it's not as frequent as getting a hit when you're 'trying' to hit a fly ball, but it happens a fair amount.


We might decide to either include or not include sac bunts and/or sac flies in "balls in play" according to the purpose of the moment. For something like looking at shifting tendencies, I think it makes no sense not to include both of them; on the other hand, when we're using balls-in-play for a purpose like calculating BABIP, we could see it differently, as does. I agree with them in the sense that I see it differently too -- but I not only wouldn't include sac bunts in the divisor for BABIP; I wouldn't include sac flies either.
2:06 AM Sep 16th
About what Rob just asked: He's right about a decrease in balls-in-play, so I did the math on the whole thing, adjusting for it.

This still leaves a decrease in shifts compared to last year, although it cuts down the difference from a 3.35% decrease to a 2.10% decrease.

BTW, unlike what some sources seem to do (for some reason) when totaling "balls in play," I did include sac bunts. (It appears doesn't.)
If anyone wants to know anything else about what I did, including the data, let me know. (I'll retain it all for a while.)
10:57 PM Sep 13th
Upon reading this sentence, "Note that all shift totals referenced in this article are shifts on balls in play", my first thought was hey, aren't there fewer balls in play this year than last year? And if so, wouldn't that mean that shifts are actually NOT down this year? Or at least not down by as much?​
2:30 PM Sep 13th
I think the nature of the guys most shifted against doesn't feature bunting skills.

So screw them.

If it was me, and it is unlikely to be, I'd square around at least until he got a strike on me.

Think about it...a pitcher must have a freaking heart attack when he winds up and looks into a bunt realizes he's got one pal on the side of field.
11:37 AM Sep 13th
I don't think the OBP would be anywhere near .800 for batter who bunt against the shift.

About halfway down this article -​hamilton - there's some data that says during 2010-16 the batting average when bunting against the shift was .521.

That figure though omits the cost of an added strike when a batter attempts a bunt against the shift but doesn't get the bunt down in fair terrirtory. An earlier article using sketchier data - - indicates that about 62% of bunt attempts against the shift produced a strike, not a fair bunt attempt.

7:40 AM Sep 13th
One of my favorite plays now is the baserunner taking third on a napping shortstop in the shift, either rounding second or a delayed steal or whatever. Every time that happens a puppy somewhere rolls around in autumn leaves.
5:13 AM Sep 13th
.....but when they move to that other position, that other position becomes what they work on. It becomes their "groove."

I mentioned the "selection" for the position as an additional factor, not the only.
4:14 PM Sep 12th
I don't know. Seems to me most 3b men can also play 2B, and many do eventually (or vice versa). You're not talking about the skill sets of Wide Receiver and an Offensive Tackle here.
11:44 AM Sep 12th
If indeed there is a slowdown, rather than this being just a 1-year bump, I wonder if a significant factor might be something besides hitters doing better at beating the shift. If I were running a team, it sure would be.

How many times have we seen plays-not-made even though the ball was fielded by a 'shifted' fielder, seemingly because he wasn't in his familiar spot? I mean, plays that would be pretty routine for the guy who's usually there. The most obvious kind of thing, and maybe the most common, is double-plays-not-turned, but there are also others. And what about this other thing -- related, but fuzzier and harder to identify but I can't believe it isn't so: What about the extent to which shifting fielders around takes them somewhat out of their 'groove,' and makes them less good in their usual spots than they'd normally be?

Look: A third baseman has groomed himself according the kinds of plays and throws and third baseman usually makes; a second baseman, likewise. In fact it's more than that: it's not just that they've groomed themselves over the years according to those things, it's also that to a large extent they've been selected according to those things.

When they have to be shifting, for one thing they need to practice being in those other spots, and -- here's that phrase again :-) -- "I can't believe" that this doesn't at least somewhat take away from their practicing of their usual stuff, at least for most players. And besides that: I would think there's some tendency to lose some of your physical "groove" when you're doing more and different physical things.

If these things are so, it wouldn't be surprising that it would take a few years for teams to start realizing it and to be taking them into account more and more.
1:38 AM Sep 12th
There must be an ego element to a player not bunting against a shift. I would think some players would end up with an .800 on base ave if they bunted, but they just can't bring themselves to do it. I see it in slow pitch softball - no matter how deep an outfielder is playing, some batters still try to hit it over their head, only to fly out.

Then again, with a big hitter up, maybe the defense would take an .800 OB with every hit an infield single.
12:45 AM Sep 12th
This is interesting. What I'd really like to know is, how often are players bunting to beat shifts, and how successful are they? I feel like I still rarely see anti-shift bunts.
6:13 PM Sep 11th
©2017 Be Jolly, Inc. All Rights Reserved.|Web site design and development by|Terms & Conditions|Privacy Policy