Where Defensive Opportunities Have Declined Most
Major-league batters struck out in 21.7% of their plate appearances last season, an MLB record. That rate broke the previous record set in 2016 (21.1%), which broke the previous top rate of (20.4%) set in 2014 and matched in 2015. Major-league batters struck out last season at a rate five percentage points greater than in 2003. You’re probably well aware of this trend — a trend of more swing and miss, of fewer batted balls in play.
Reversing this trend seems difficult. It would require a change in incentives or, perhaps, the ball’s seam height. Batters would have to trade in power for more contact; pitchers would have to throw fewer breaking balls and with less velocity. The game keeps moving toward more power, more velo, more breaking stuff. It seems, at least to this author, that this strikeout level is pretty sticky and might continue increasing for the foreseeable future.
This phenomenon means a number of things. Most relevant to this post, it means that defenders have fewer opportunities, thousands fewer over the collective course of a season, which erodes the value of defense.
Some teams have been more aggressive with defensive placement, more willing to gamble to find a place for a bat, to trade defense for offense. The Indians stuck Jason Kipnis in center field late last summer, a position he had not played at all since the low levels of the minor leagues and not regularly since he attended Arizona State. (Kipnis posted a -3 DRS in 71 innings center). The Indians were also aggressive in slotting Lonnie Chisenhall in center earlier in the season despite his eight innings of previous experience there. Chisenhall, a college shortstop and professional third baseman turned outfielder, posted a 0 DRS in 164 innings in center, converting 34 of 36 Balls in Zone into outs.
Craig Edwards wrote about the curious case of Reds third baseman Nick Senzel trying out shortstop this spring. Craig wondered if we might see more players move up the spectrum, from third to short, in a similar way. After all, not only are fewer balls being put in play, but fewer balls are in play in the infield as ground-ball rate has declined in each of the last two seasons — and more and more batters are looking to launch.
While the makeup of a pitching staff and ballpark environment come into play when planning a defensive alignment, this decline of balls in play suggests teams could and should perhaps become more aggressive in finding ways to add more offense to their lineups. And maybe we’re already seeing that in this golden era of shortstops.
While players typically move down the defensive spectrum as they advance through professional baseball and age at the major-league level, perhaps batted-ball trends could cause more players to move up the defensive spectrum.
Overall, there were 60,249 “plays” by defenders in 2007, according to our data. Last season, there were just 49,809 — or roughly 10,000 fewer. A gradual year-to-year decline is evident over the last decade.
What I was really curious to learn is if certain positions were losing a greater share of opportunities or if the decline was spread relatively uniformly across the defensive spectrum. To simplify this study, there are issues to consider like defensive placement and alignment. I used the “plays” metric to capture opportunities handled by each position in the following chart over the last 11 seasons:
| Season | RF Plays | CF Plays | LF Plays | 3B Plays | SS Plays | 2B Plays | 1B Plays |
|---|---|---|---|---|---|---|---|
| 2007 | 8418 | 10886 | 8014 | 7221 | 10625 | 10120 | 4965 |
| 2008 | 7892 | 9684 | 7322 | 7403 | 10567 | 10525 | 4860 |
| 2009 | 7563 | 9609 | 7273 | 6943 | 10476 | 10149 | 4766 |
| 2010 | 7430 | 10086 | 6912 | 7220 | 10354 | 10380 | 4602 |
| 2011 | 6945 | 9116 | 6539 | 7366 | 10087 | 10395 | 5299 |
| 2012 | 6804 | 9129 | 6280 | 7238 | 9850 | 10181 | 5166 |
| 2013 | 6753 | 9210 | 6084 | 7197 | 9843 | 10049 | 5312 |
| 2014 | 6988 | 9466 | 6420 | 7138 | 9544 | 9699 | 5360 |
| 2015 | 6722 | 9053 | 5957 | 7037 | 9685 | 9504 | 5139 |
| 2016 | 6490 | 8555 | 5914 | 7161 | 8858 | 8615 | 4854 |
| 2017 | 6470 | 8552 | 6049 | 7008 | 8702 | 8140 | 4888 |
| % Change | -23.1 | -27.2 | -24.5 | -2.9 | -18.9 | -19.6 | -1.6 |
Not including catcher, five of the defensive positions that align themselves in fair territory have seen their play numbers drop by at least 19% compared to 2007 levels. Strikeouts have jumped by 25% over the last decade, and home runs are at record levels, too.
Interestingly, third base and first base have maintained relatively stable opportunity, whereas all three outfield positions — plus shortstop and second base — have had their opportunities drop significantly. Third basemen generally have been aided by defensive shifts, becoming responsible for covering a larger swath of territory.
If you prefer a visual presentation consider the following charts. One for most of the infielders…

And another chart of outfielder …

What does all this mean?
If a club has a ground-ball pitcher on the mound, perhaps it can be more aggressive in sneaking some additional offense into the outfield. If a team has an offensive-minded but large-framed shortstop, like a Corey Seager Carlos Correa, perhaps they could become less concerned with having to move him off the position. Instead of players generally always moving down the defensive spectrum, teams ought to consider candidates who can move up the spectrum — like Senzel — and more often trade offense for defense.
With teams more interested in roster flexibility than ever, versatile gloves that can hit become more valuable. Maybe we see more infield-outfield versatility among position players tied to their pitchers’ ground-ball rates. (While framing numbers have been erratic, catcher defense perhaps becomes even more important: its value is largely unaffected as it’s the lone defensive position in foul territory.)
Defense still matters a great deal. There’s always going to be a place and an interest in rostering elite gloves. There are still relative advantages, position to position, to be had defensively in this strikeout environment. Playing shortstop and center field still requires great athleticism.
Defenders just have fewer opportunities and those opportunities, to date, have not decreased uniformly.
A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.
Interesting that of chances declined more than infield chances. Sure hr per fb is up but there are also more fb.
That corner balls are up is logical. Uppercut guys hit less grounders but tend to pull their grounders since they top the ball if they are early and then roll over. I think that is also one of the reasons that many extreme flyball guys have lowish babips ( that and that fb hits are usually homers and don’t count for babip) as pulled grounders get killed by the shift.
Very interesting research. The selection of endpoints appears to be a significant factor in assessing the magnitude of any changes, too.
For example, these graphs make it look like 2008 produced a severe uptick in groundballs, so the percentage changes from ’08-’17 would be quite different from the numbers calculated for the full decade.
OTOH, 2014 looks like an extreme “launch angle” year, where flyballs jumped while grounders fell. The 2016-17 seasons might be more examples like ’14, but OF chances might have been held relatively stagnant in these more recent years by a large increase in strikeouts (with groundball chances showing enormous drop-offs).
Thanks for your work, Travis. The trends seem clear. What opportunities arise from these trends is probably worth looking into further.
Great article. Some of those OF chances used to be hits that are now outs (thanks to better positioning) or sometimes home runs I would imagine. The infield chances presumably include foul outs (propping up the corner infielders some) and pop ups, so infield defense may be losing “at risk” chances even faster than it appears.
One random thought about catcher defense – since hitters virtually never sacrifice stolen bases are the lone “strategic” means of advancing a runner so there may be more situational value in having a catcher that deters steals.
In the outfield particularly, wouldn’t you have to consider the weight (value) of those chances made, rather than just the absolute number?
Moral of the story: Keep signing those third basemen! (or maybe: Start signing those third basemen!)
Might it be time to take a look at changing positional adjustments for WAR then?
SS still the hardest to play and saw the most balls, just by a smaller margin, so its probably not that significant yet.
You’d need to get to the point where different positions either had completely new demands or actually saw more balls than the others
A 20% swing in opportunities over ten years is significant. The question is how often positional adjustments get recalculated.
Here’s a quick idea with numbers I threw together: with less defensive opportunities, shouldn’t we expect the gaps in positional adjustment to narrow?
From the Fangraphs Positional Adjustment page, we use the following seasonal adjustment as a guideline:
C: +12.5
1B: -12.5
2B: +2.5
SS: +7.5
3B: +2.5
LF: -7.5
CF: +2.5
RF: -7.5
Now, adjusting each closer to zero (less opportunity is bad for good defenders and good for bad defenders) according to the percentage of opportunities lost (found in the table above) yields:
C: +12.5
1B: -12.297
2B: +1.958
SS: +6.008
3B: +2.438
LF: -5.408
CF: +1.900
RF: -5.535
Is this the correct line of thinking?? Maybe, maybe not… tell me whatI’m missing!
Two simplifications in the above warrant refining.
The chances are not all the same – some are worth multiple outs, and some are worth multiple bases.
The other is the difficulty of the chances (and the risk of a failing to get an out or allowing an extra base advance) are not uniform. Shortstops make a lot more close plays and tough plays than others do. Center fielders and right fielders have longer and sometimes more hurried throws than left fielders.
This is a good first pass, but it needs some more detailed analysis to reflect these issues. If the original positional adjustment numbers are not adjusted for these factors either (I do not know off the top of my head), then it is probably better to simply start from scratch. I may do that anyway for my own amusement.
Good thinking! But my guess is that the positional numbers I pulled from fangraphs take these factors into account already. IF this is the case (big IF), then relative difficulty between positions is preserved when subtracting out rates of opportunities lost.
Cut only SS opportunities by half — SS still has the highest positional adjustment by a run (sans C).
Of course, in this hypothetical situation, 2B should lose double play opportunities. This just shows how nuanced this problem is…
No, that is the issue – 20% fewer chances is a big change, and it may mean the distribution of their difficulty has also meaningfully changed.
Completely wrong. Corner Ofers are not bad defenders. Less chances simply means they are less valuable.
Should be more like this
C-12.5
1B- (-10)
2B-2
SS-6
3B- 3
LF- (-9)
CF -1
RF (-9)
DH (-12.5)
Total (-16)
Pre adjusted Total (-16)
With all of the emphasis on launch angle / hitting the ball in the air, I would’ve bet money that infield opportunities had declined far more precipitously than those in the OF. Which is why I’m grateful Mandalay Bay refused to give me a reasonable line.
These may be a couple of dumb questions (at least around these parts), but has the positional adjustment remained fairly constant for the past 10 years? Wouldn’t the fact that 2B saw more plays for several years move it up on the spectrum (or, at least, move it closer to the next-up position)?
It is not the plays, but the aggregate difficulty and importance of the plays.
Is this why people think Matt Kemp has a chance to make the Dodgers roster?
Interestingly, what this suggests is that players that get good defensive rating nowadays such as Simmons, Betts, Heyward, etc., are actually even more better than their contemporary peers compared to how much better the elite defensive players 10 years ago were. (Utley, Beltre, Rolan, etc.)
Not more valuable, but better in terms of underlying skill, relatively speaking.
Of course, it also applies to the players with shoddy defense as well.
Imagine the kind of numbers JDM or current Kemp could have put up 10 years ago.
Yet somehow the position adjustments dont change. It seems 1B and 3B should be getting a bump here
Logically the corner infield adjustments should stay the same (because the percentage changes hardly changes). It’s the other positions that should regress towards zero.
There’s a whole bunch of fascinating questions here. Does it matter if the stadium is a hitter’s park or a pitcher’s park? Do you favor a good-glove 3B on the grounds that they’ll be more useful in shifts? Equally, do you favor a good-glove CF on the grounds that you’re prepared to weaken the corners, so you need back-up?
And then there’s the fact that not all fielding opportunities are created equal. I would prefer to see middle infielders with good gloves simply on the basis that I’m greedy when it comes to double plays. And I don’t think these numbers degrade the value of an outfielder with a strong arm. (We can all point to CFs with a noodle arm.)
One particular thing that stands out it that the traditional utility IF (good glove, weak bat) doesn’t seem to have nearly the same value as ten years ago. A strong bat (even platoon) that isn’t an utter disgrace down the middle might be more valuable.
Oh, and if the Boston FO is reading this and is encouraged to try “that” experiment again — no. No more Hanley Ramirez in left field, please. Not even if these numbers support the idea.
Very interesting study. Thanks for doing this Travis.
Like others I am a bit surprised that outfield chances have declined more than those in the infield, given recent emphasis of elevating batted balls.
I’m absolutely gobsmacked by the notion that plays by defenders are down 17% in the past ten years! I knew that walks, Ks, and HRs were up, but it just never dawned on me by how much that would reduce what we have come to know as “baseball”. No wonder it’s more difficult to sit through a nine-inning game these days. The game takes an additional half-hour, but there is less “action” than in the shorter version.
Overall, there were 60,249 “plays” by defenders in 2007, according to our data. Last season, there were just 49,809 — or roughly 10,000 fewer. A gradual year-to-year decline is evident over the last decade. Can we get the “plays” definition?
~ 17 fielded outs * 162 * 30 / 2 = not “plays”
Yes, I’m sort of worried about this. Nobody yet has asked where all those lost fielding chances go … which is at least a point of interest, and possibly the subject of a further post.
There are still basically 27 outs per team per game (ignoring extra innings, which I doubt have made a difference). And there are still 81*30 games per year (ignoring the post season, which I doubt makes a difference). That (roughly) 20% of fielding plays must have gone somewhere.
Who knows, it might be a consequence of the move towards “three true outcomes.” None of the three involve fielding chances, obviously. It can’t simply be “pitching efficiency” in general, because inducing ground balls or infield pop-ups is efficient but still counts as fielding chances.
I hate to suggest that the criterion for a “fielding chance” might have changed, or that possibly some of those “fielding chances” are assigned to “unknown” rather than to a specific position, but then again, 20% or so is a rather large lapse in the things. Where did the other outs go?
With this Machado becomes especially interesting. He is an elite defensive 3B, and while he came up as a SS and had a solid reputation there, will he be more valuable defensively as a SS?
Does this account for shifts?
Corey Seager Carlos Correa is my favorite player