Good days at the plate are pretty easy to identify. If you’re looking for the best game any hitter had in April, you can look at total bases (as in Ryan Braun’s three-homer game) or at hits (as in Charlie Blackmon’s 6-for-6 game) or at win probability added (as when Kyle Seager hit two homers, including a walk-off, for a one-game .906 WPA); or, simply RE24, which would lead you back to Blackmon, who produced more than five runs all by himself. Similarly, for pitchers, pretty easy: Andrew Cashner’s 9/1/0/0/2/11 was the month’s best game score, though you might opt for Jose Fernandez’s 8/3/0/0/0/14 for dominance or Julio Teheran’s 1-0 shutout for value.

Defense is trickier. How would you go about identifying the best defensive game of the year? Does a guy who makes 10 routine plays top a guy who makes a couple web gems in an otherwise quiet night? The busy fielder didn’t necessarily have much to do with those 10 balls coming to him, but he did turn them all into outs—and, as we are smart enough to know by now, what seem like routine plays aren’t always routine (as opposed to the residue of good positioning, good first steps, or an Andreltonesque ability to make difficult plays look easy and impossible plays look merely difficult). We also know that Web Gems aren’t always the most difficult plays—leaving one’s feet being noticeable but not always the sign of excellent defense.

This comes up because this year we wanted to highlight the best defensive game of each month (or so) using statistical means. We turned to Inside Edge, which provides advanced data to major-league clubs (and elsewhere), to help us out. We’re starting simple, but we might get a bit more complex as the year goes on, demonstrating ever-broader expressions of defensive excellence, separating our standouts by position and getting more precise with what we consider valuable. Mostly, of course, this is an excuse for me to make GIFs of pretty baseball things.

So, without further introduction, the best game of the 2014 season so far, according to Inside Edge. It will come as no surprise to you that it was produced by a 24-year-old National League shortstop. Of course, it’s Andrelton Sim—

Wait. Actually, it’s not. It’s Starlin Castro, a shortstop whose defense has long been questioned, who has finished first or second in the league in errors every year of his career, whose move off the position has been considered inevitable by many, whose lapses in concentration have led to tension with his club, and who has only once received any votes for a Fielding Bible award (finishing 15th among shortstops in 2012). But if this seems at all odd to you, remember: Charlie Blackmon, first paragraph.

Inside Edge puts each chance in one of six categories:

  • Routine: Greater than 90 percent chance that it will be fielded
  • Likely: 60 to 90 percent chance
  • Even: 40 to 60 percent chance
  • Unlikely: 10 to 40 percent chance
  • Remote: 1 to 10 percent chance
  • Impossible: 0 percent chance

Castro had eight balls hit to him. He fielded them all. They were categorized thusly:

Routine plays

1. Tony Campana groundout

Difficult because: Low, skipping ball; 80-grade runner, 3.8 seconds home to first.
Easy because: It was hit directly at him; hard enough that he didn’t have to hurry too much.

2. Tuffy Gosewisch double play

Difficult because: Had to hurry and deliver a clean feed to 2B, with baserunner closing in; forced him to take a little bit of an in-between hop.

Easy because: Routine grounder hit close to right at him; slow baserunner at the plate.

3. Brandon McCarthy groundout

Difficult because: Shards of exploding bat flying toward him; throwing a little against his momentum.

Easy because: Pitcher running (4.8 seconds to first); wasn’t hit hard; first baseman bailed him out after gratuitously low throw.

4. Paul Goldschmidt lineout

Difficult because: Line drive, hit hard; actually had to be positioned in the right place, rather than double-play depth that the situation might have suggested.

Easy because: Pretty much directly at him; no rush on throw.

5. Paul Goldschmidt groundout

Difficult because: Required reasonably good footwork to both get in front of it and set for the throw.

Easy because: Hit close to him on his glove side; playing deep for the power-hitter; average runner from right side, so had time to set and throw.

So those are the easy plays. All pretty easy, all balls that you’d expect to see Derek Jeter field. Also, though, with the exception of Brandon McCarthy’s groundout, balls that did required some technique—a good first step, a well-positioned player, competent footwork to square up the ball or a strong throw. You can see, for instance, how easily Bad Castro can get one step too casual on a ball like Goldschmidt’s

or a bit too nonchalant on a play like Campana’s

or a bit too quick on the feed to second base on a ball like Gosewisch’s

and end up with an error.

(Of course, had Rizzo not scooped one of his two low throws, we’re not talking about this performance at all.)

Even plays

1. Chris Owings fielder’s choice

This is actually the play I’d most expect to see on a highlight show. It’s not visible from this angle, but Castro actually flips the ball from his glove hand to his throwing hand for the quick transfer, then has to make a very unnatural pitch to second base while running in the opposite direction. Further, the second baseman Luis Valbuena isn’t at the bag yet, so he’s flipping to a moving target. One might consider docking the difficulty of the play some because Castro might plausibly have gone to first base (there were two outs in the inning), but the batter (Chris Owings) gets down the line pretty well; on a play we’ll discuss in a second, I clocked him at 4.05 to first. Note: Castro attempted an almost identical play last year and botched it.

Unlikely plays

1. Chris Owings groundout

2. Tuffy Gosewisch groundout

These plays both make Castro travel, but they’re primarily throwing plays. In the first, the final hop skips up on him a little, so that (it appears) he isn’t quite able to get a clean transition from glove to throwing hand. Once he does get it, then, it has to be a very quick draw and shoot to get Owings at first. In the second,

he lays out and, in one controlled motion after getting to his feet, pivots his body into throwing position while shuffling the ball out of his glove into his throwing hand. Both throws were strong and accurate. None of these eight plays, for what it's worth, was one of that day's Web Gems. Watching them, it's actually not that surprising. But that's not the same as saying they weren't difficult, and valuable.


If we treat each of these plays as landing right in the middle of Inside Edge’s categories—that the routine plays were all 95 percent likely to be made, that the unlikely plays were 25 percent—then Castro got eight opportunities and made eight outs, when he should have been expected to make 5.75 plays. If we assume all of those plays would have otherwise been singles, he saved a little bit more than one run, though that’s ignoring that two of the plays turned into double plays.

Did we get the right game? Is this truly the best defensive game of April 2014? I’m very happy with the choice, but it’s impossible to be definitive. Not only are our assessments of defensive plays subjective, but putting them in context brings in further subjectivity: different players at different positions behind different pitchers are asked to do different things at different rates; an outfielder who stole a home run and otherwise stood unused in left field all game would have produced similar defensive “value,” but was he actually better than Castro, who stole a bunch of singles? I mean, gosh. Who knows. Is this a statistical question or a philosophical one? I do know Castro had a heck of a game. I can buy this.

Is it merely a fluke that Castro, a player not noted for his defense, had such a performance? Maybe not. Since last summer, local writers have been quoting Cubs coaches and managers talking up his increased dedication to defense. In September, Castro talked about paying more attention to who was batting, so that his pace on each play matched the speed of the runners. In the same article, Dale Sveum praised his work getting rid of the ball:

“Getting rid of the ball when he has to now I think is a big obstacle he’s overcome,” he explained. “You don’t see him tapping the glove and taking the time, not meaning to take his time—but that was the only way he ever fielded a ground ball. … “The ball is touching his glove and he’s getting rid of it.”

Castro has talked in multiple interviews about arriving early for extra defensive work “every day” this year, and catcher John Baker noted it, too: “Watching him do his work—he wants to improve as a defensive player. I see him out there all the time, before day games taking groundballs. That, to me, is more impressive than him hitting two home runs in a game. He really is striving to be a complete, frontline shortstop.”

And, Castro has said, he has clicked with Cubs coach Gary Jones in a way he hadn’t with other coaches.

The daily sessions with Jones have focused on footwork as the new Cubs infield coach tries to help Castro reduce the mistakes. "With Jonesy, we work a little more," Castro said. "We did footwork with other [coaches], too, but it was different work. [Jones] wants the left foot in front [when I throw], and you can stride to first base."

Said Jones: “He understands situations. He understands runners. He has put all the work in and recommitted himself to wanting to be the best player he possibly can be, and it's showing.'' And Castro’s boss: “As for the idea of moving him from shortstop, Hoyer says the Cubs are ‘taking it off the table’ and sounded tired of hearing that question.”

I don’t dare cite small-sample defensive stats based on just a single month of data—unless, that is, it fits my narrative, in which case I cautiously nudge you toward Castro’s player page, where you’ll note that, at 2.9 FRAA this year, he’s already topped his career best. That said, we won't get carried away here. Overall, Inside Edge ranks Starlin Castro 16th best among the 30 busiest shortstops this year. So he's average. That's better, frankly, than I expected Castro would be at this point in his career. And for one game, he was even better.

Huge thanks to Nick Wheatley-Schaller for technical assistance, and Kenny Kendrena at Inside Edge.

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
For the curious, this is the April 22 game. Cubs won 9-2. You'd think it was the 5-1 victory on April 21 (seeing as it's 5-1 in the 8th inning). As far as I can tell, Valbuena didn't play in that game. Also, Hammel is clearly on the mound. The Cubs scored 4 in the bottom of the 8th in that game, and the DBacks scored 1 in the top of the 9th.

Castro went 2-4 with a double at the plate in this game as well.
Great stuff. I look forward to keeping up with these the rest of the year.
love how you guys chose castro. after watching his play last year compared to this year, it has been a drastic change
This article brings a lot of questions to me, but here are the two predominant ones:

1. The comment, "(Of course, had Rizzo not scooped one of his two low throws, we’re not talking about this performance at all.)" highlights the fact that this can become really subjective. In essence, the "quality" of the infielder's defense will be largely dependent on the gloves of the rest of the infield, for example. I'll be curious to see how this impacts future selections.

2. The defensive game, Castro in this case, can't only be based on ground balls. If he botches a relay through from the outfield that costs a run, that needs to play into account. As you can imagine, if a LF is not in position to back up second base on a throw from right field, and an errant throw ensues allowing a runner to take an extra base, that also plays into account a player's "defensive game". We could go on and on with these examples, but it might be interesting to rewatch the Castro game to see if we can identify any of these types of flaws that would downgrade his quality of defense.
Inside Edge does actually track secondary defense -- the plays that happen after the initial play, as you're talking about in (2). It just happened that in this case no secondary defense came into play on Castro in this game. Hopefully next month!
The timing of my reading this article was impeccable, as I just watched Castro make an inexcusable throwing error on a routine grounder with 2 outs costing a run...while I was reading this.

Regardless, he does seem to be a bit more focused this year, at least so far, and hopefully it continues throughout the season. Loved the article!
Shouldn't "his two low throws" count against him? In other words, would another player with pretty much all the same play opportunities, but with zero or one low throw have had a better defensive day?

And so similarly, I'd think Rizzo scooping low throws should get him extra credit ... I guess the mechanism for that is that from his perspective those plays, rather than being "routine" were "even" ... perhaps that's the way to decide the value to dock Castro (ie. dock Castro the value added to Rizzo)?
He fielded a two-hopper today and threw it away, costing a run and extending the inning. Just sayin'.
I love how the gifs don't start until your mouse rolls over them.
Sam, I bought a year's subscription yesterday so I could read this article. This one article is worth every penny as it helps give meaning to defensive metrics. Puts some skin on the skeleton. Defensive metrics are simple enough to understand but seeing a play-by-play breakdown of how a few groundballs add and subtract to a players value over the course of a game is very instructive.

I hope BP does something similar on a regular basis, like how you guys handled This Week In Catcher Framing last year. It would be great to turn this into a weekly feature, maybe first week of May it is "best middle infield performance of April", 2nd week is "best OF performance of April", 3rd is "best corner infield performance", and 4th is a miscellaneous topic - such as team defense (how Rizzo's clean scoops cover a multitude of throwing sins, relay throws, double plays, backing up a play), or analyzing shifts, or catcher's defense, or worst/most volatile/"most average" defensive performances, etc. These types of features are what gets me telling all my friends how much they need a subscription.

That said, even if this is all we get, thank you, and I'm glad to be back for a year.
I'd like to see a similar analysis on Nolan Arenado. He has 20 more assists than any other major league third baseman and is in the top 10 among all players at all positions in assists in the league which is rare for a third baseman.