It’s common knowledge that James Shields is one of the most consistent pitchers in baseball. Over nine seasons as a pro, he has pitched at least 200 innings eight times, and produced at least 2.0 WARP eight times—without ever topping 3.5. His excellent FIPs from his sophomore season on have never strayed from a .8-run band.

Yet one of the keys to Shields’ success in recent seasons has been his inconsistency—specifically, his ability to pitch significantly better with runners on base, and especially when they reach scoring position. Since joining the Royals, he has been worse than the average pitcher with bases empty, and a true ace with runners on:

  • Empty: .264/.313/.401
  • Men on: .240/.293/.364
  • RISP: .227/.287/.342

The average MLB pitcher sees his performance suffer a bit when runners are in scoring position, for reasons both intuitive (a pitcher surrounded by baserunners might be scuffling) and a bit more nuanced (the defense is less free to play extreme shifts). Not so with Shields. The natural inclination might be to chalk the difference up to statistical noise, but an even better inclination is to dive in and see what's happening.

Let’s start by looking at how Shields approaches opposing hitter when nobody is on base. Shields falls into some patterns:

Overally, he distributes his five pitches almost totally evenly against left-handed hitters. Righties get a bit more emphasis on the four-seam, cutter and changeup, with the sinker and curveball taking a backseat—though each appears with enough frequency to remain in the hitter's mind. With that distribution of pitches, a hitter wouldn't be able to sit on anything; he might as well be going up against a pitch randomizer.

But those are just the overall usages. In specific counts, Shields becomes more predictable. Against lefties he is very likely to start off a PA with a four-seam fastball, then abandon the pitch once he gets ahead or has two strikes on a hitter. (Red boxes indicate usage more than 10 percent above average. Blue is the inverse.) His change is rarely used early in the count versus lefties, but the usage jumps up to 33 and 36 percent when he is ahead in the count or has two strikes. He begins to lean heavily on his curveball when he gets ahead 0-1—but pockets it when he has two strikes and is going for the putaway. Shields knows what his out pitch is: The changeup. A hitter who is paying attention does, too.

When facing righties, the pitch selection is a bit more nuanced, but the interplay between the fastball and the changeup—the two dominant pitches in his arsenal—is simplistic depending on the count. Four-seam fastballs are the dominant offering early; the change is little threat until two strikes. Shields' broad repertoire will never be too predictable, but a hitter can at least sketch out a plan.

So how does that compare to what opposing hitters see when there are runners in scoring position?

Here, he becomes the unpredictable James Shields. All of his pitches are in play in almost any count. He uses his fastball more regularly, with minimal variation between counts. He follows a similar pattern with his changeup, leaning on it whether he's way ahead or way behind in the count. He uses a curveball more frequently as an ambush to grab the early edge in the at-bat–and less frequently in the more predictable pitcher's counts.

Much of this is true against righties, too, though the biggest difference here is that he unleashes the cutter as a weapon in all cases, giving him three co-equal putaway pitches (along with a more aggressive curveball usage).

Comparing the two pitch usage charts above can be tedious, so I created a simple comparison chart that shows the change in Shields’ pitch usage depending on whether the bases are empty or runners are in scoring position:

Negative numbers indicate that Shields’ usage drops with RISP, while positive numbers denote increases in pitch usage. As mentioned before, Shields has some distinct changes in pitch usage when runners get in scoring position. Against lefties Shields starts to pitch backward, using his fastball later in the count as opposed to early. His changeup usage increases across the board, and he minimizes his usage of his two “worst” pitches, especially later in the count.

Why might Shields abandon his fastball, especially early in the count? It’s the only pitch he throws to LHH that doesn’t generate ground balls at least 50 percent of the time. (It’s perhaps worth noting that the 36 percent LD/BIP rate from his fastball is far and away the worst for any of his pitches this year, though using line drive numbers from an individual pitch within one season brings up some serious sample size concerns.) That said, it seems obvious that Shields eschews his four-seam fastball to go with pitches that have a smaller chance of being hit hard. Minimizing the damage done with runners in scoring position seems to be Shields’ game here.

The one time Shields does elect to use his fastball is when he’s ahead or has two strikes on the opposing hitter. Interestingly, in this situation Shields throws his fastball inside to lefties, a stark contrast to his typical habit of pounding the outside of the zone with fastballs:

Against right-handed batters, Shields utilizes his cut fastball, as opposed to his four-seam fastball. Why? One reason is that he throws his cutter for strikes more often. His four-seam fastball goes for a ball against RHH 34 percent of the time, while his cutter is called a ball just 29 percent of the time. Another reason is that Shields gets more whiffs on his cutter: 10.8 percent compared to just 6.5 percent for his four-seamer.

This of course brings up the question of why Shields would throw his cutter with runners in scoring position, but not all the time. The answer is, in a word, complicated. One potential factor is that Shields' four-seam fastball has averaged 94 mph this season, while his cut fastball comes in at just 88 mph. Shields’ four-seamer, then, helps keep opposing hitters honest. If he were to replace all of his four-seam fastballs with cutters, to give an extreme hypothetical, then two-thirds of the pitches he delivered would come in between 86 and 88 mph. Shields’ changeup averages just over 86 mph, so a repertoire consisting primarily of changeup and cut fastball wouldn’t provide much variety in terms of velocity.

Another big component to this is that Shields performs best when he maintains a balanced pitch mix. Kings of Kauffman noted that Shields struggled at times in 2014 when his cutter usage crept up and opposing hitters could sit on the pitch. By maintaining a good mix of pitches Shields is able to utilize his cutter or four-seam fastball in situations where he feels they will give him an edge over the hitter, without having to worry that the hitter knows it’s coming.

The result of all of this is that Shields is highly effective with runners in scoring position; he appears to be able to “buckle down," not necessarily due to his moral fiber but due to a well reasoned plan. Does this mean he's living up to his clutch-themed nickname? He executes better when the opponent is threatening to score. He consciously changes his pitch mix, featuring different pitches and using familiar ones in new ways. James Shields is a clutch pitcher, because he's a very, very smart pitcher.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
If you listen to sports talk radio in NY, you know that many of the on air personalities laugh at the "Big Game James" moniker, specifically due to his career numbers versus the Red Sox and Yankees, which they deem not good enough. He has an ERA over 4 and a losing record against both teams. (insert obligatory comment about how W-L records and ERA are not good indicators)
His career line of 4.98/1.40 in the postseason argues against any "Big Game" hosannas.
2 questions I would have:

1) How do all pitchers change their approach with runners in scoring position? For all we know, Shields' changes are similar to all pitchers' changes.

2) If he is more effective with a different mix of pitches with RISP, why not throw those same mixes all the time?
I believe the answer to #2 lies in game theory - if he always used the same mix, then that mix would become more detectable and therefore less effective.
It makes no difference whether he has one detectable mix of pitches or two detectable mixes of pitches (RISP and no RISP), as the hitter is just as aware of the base situation as the pitcher, and can adjust accordingly.

Thus, the "game theory" hypothesis is not convincing.
That is 100% correct. Game theory for pitching only applies within the AB anyway.
Great questions!

The first one is difficult to parse out, but I'm working on that. Sam and I discussed this in depth before this article went up, but what separates Shields from other SPs isn't necessarily that he changes his approach, it's that it is more effective. We know for a fact the average MLB pitcher gets worse with RISP vs. non, likely for a variety of reasons.

As for your second question, I think there are a couple of factors. One reason is that it seems his cutter is more effective when he uses it selectively. The Kings of Kaufman article linked above goes into that to some degree. Another interesting note is that this season RHH have hit his cutter better than his fourseamer. So it seems that early in the count Shields is willing to trade the potential for giving up hits for strikes (better command of his CUT vs. FA). We opted not to include it in the final piece because we felt that batted ball numbers for 1 pitch in 1 season were too volatile to note. That said, it's interesting to note. Here's a Brooks link to pitch outcomes vs. RHH in 2014:
You completely lost me.
Yes, I realize that he has been more effective, otherwise you would not have written the article, but you are attributing that success to a change in approach. If all pitchers use a similar change in approach with RASP then it is unlikely that that change is a causative factor for the success (i.e., he probably just got lucky).

That's like saying, "Pitcher A has had amazing success pitching with runners on first base and 1 out, as compared to the average pitcher in that situation. Let's see why. Wow, it turns out that he is somehow able to induce 20% more GB than he normally does, hence the success."

Well, if ALL pitchers induce 20% more GB with a runner on first and 1 out, then that can't be the reason for his success!
I see what you mean. I asked our research team to pull the relevant data here so I'll look into it shortly.