I'm not a big fan of measuring defense by errors because I think the stat is inherently flawed. It's missing too much context, and it's a loose judgment call that can be subject to the whims of different scorers. Citing how many errors a player has made, especially an outfielder, tells me next to nothing about his defensive ability. However, there are extremes to any example, and at some point when a shortstop like Marcus Semien makes 15 errors in 40* games there is no question that he has been brutal with the glove. In this case, the errors are sending a clear message.
What do 15 errors really mean, though? What kind of effect does that have on a team, and how much of a club's struggles can be chalked up to those miscues? There are all sorts of defensive metrics out there, and I believe in most of them, but I also recognize that they are not universally accepted, that they aren't reliable in 40-game samples, and that even if they do accurately perform their desired functions they will only be measuring theoretical wins and not the specifics of how the timing of the plays actually tied into the results of the 40 games at hand.
So let's try something different. We have numbers but don't necessarily know what they mean. Let's go through Semien's game logs and see how each of his errors actually affected the team and its results. This will not be a full picture of Semien's overall defense, because there are also plays he's failed to make that weren't scored as errors, but at least this will be a starting point. Also, it's important to note that the rest of the team has made 28 errors as well, a total that would still rank fifth among all MLB teams. Semien is not the sole offender here. Anyway, here are his 15 errors (or, scroll to the end if you just want to skip to the conclusions).
Marcus Semien's errors
1. April 8: A's lead 9-0 in the 7th. Semien error, leadoff runner reaches base. Kazmir retires next two batters, then has to spend 7 more pitches to finish the inning. It was his last frame anyway, though, and the bullpen finishes it off. Verdict: Harmless, just 7 extra meaningless pitches
2. April 11: A's lead 2-1 in 8th, one on, one out. Force play at second, but Semien rushes and misses second base by maybe an inch. Runner is called out, but Mariners challenge and it is reversed. I honestly didn't even think you were allowed to challenge this call, but the ump's explanation that it wasn't technically a "neighborhood play" was somewhat acceptable. Next batter grounds out, but that's only the second out now. Then Nelson Cruz homers on an utter meatball pitch from Dan Otero; Seattle takes 4-2 lead. A's come back to tie, Abad blows it in the 11th. Verdict: Definitely harmful, cost 6 pitches and a pitching change, and contributed to 3 runs, but also far from the only cause for defeat and Otero is my main goat ... also, kind of a B.S. call by the ump and I don't fully blame Semien for it
3. April 12: A's lead 3-0 in 5th, one out. Runner reaches on Semien's throwing error. Next batter hits GIDP, inning over. Literally didn't even cost the pitcher any extra pitches. Verdict: Harmless, like it never even happened
4. April 20: A's lead 4-2 in 3rd, one out. Semien's error allows Mike Trout to reach. D'oh. Trout steals second, then scores on a single. That run is fully on Semien. On the other hand, Semien then leads off the next inning with a homer to get that run right back, and the A's go on to win 6-3. Kendall Graveman has to face one extra batter, but only one, albeit it costs 9 pitches. But he's pulled in the 4th anyway at only 73 pitches and it's not like he was cruising before the error. Verdict: Cost a run and 9 pitches, but as a silver lining he immediately earned the run back and the A's won anyway
5. April 21: A's trail 3-0 in 2nd. Pomeranz has already given up a 3-run homer to Johnny freaking Giavotella, and there are two on and one out. This is already looking dire. Pujols hits a liner just to Semien's right, but the ball plays him and he can't snare it (link). He really should have caught it because it was so close to him, and it wasn't hit that hard, but I can't say for sure that it didn't knuckle on him or something. Anyway, a run scores and Pomeranz throws one more pitch to get the last out. Buuut R.J. Alvarez enters a 4-1 game later and gets torched for 6 runs, and the A's lose 14-1. Sure, Semien cost a run, but did it really matter? Verdict: Cost a run and 1 pitch, but this game wasn't lost by Semien
6. April 21 (continued): Same game, A's trail 9-1 in 6th. Alvarez is utterly melting down, and now he's just a sacrificial lamb on the altar of saving the rest of the pen. With one on and two out, Semien makes a throwing error. A 10th run scores. Horror! Alvarez throws 6 more pitches to get the final out. Verdict: Cost a run and 6 pitches, but both the run and the extra pitches were completely irrelevant
7. May 9: A's trail 1-0 in 1st, two on, two out. Semien flubs a grounder, then makes the throw anyway and throws it away. He gets two errors on the play -- one for clanking the grounder enough to allow Logan Morrison to beat out his throw, and a second for allowing him to advance to 2nd on the overthrow. A run scores, and that run is on Semien. Jesse Hahn throws 3 more pitches to get out of the inning. The A's go on to lose 7-2 and were never really in this one anyway even without Semien's misplay(s). All hail J.A. Happ. For the purposes of this entry, I am only judging the initial clanking of the grounder. Verdict: Cost a run and 3 pitches, but the whole team combined to lose this game
8. May 9 (continued): Same game. This is the aforementioned throwing error from the previous entry. It only allowed the two runners to move up to 2nd and 3rd, where they were immediately stranded. Verdict: Harmless, like it never even happened
9. May 10: Tied 0-0 in 2nd, none on, two outs. Semien flubs a grounder, runner reaches 1st. Jesse Chavez throws 5 more pitches to end the inning and strand the runner. Semien then homers off Felix Hernandez in the next inning, and homers off Felix again later in the game. I didn't even know you were allowed to homer off Felix twice in a game. The A's lose 4-3 because Semien is the only guy who showed up (rest of team: 3-for-30). Verdict: Harmless, just 5 extra pitches, and then he homered twice off freaking Felix
10. May 11: A's lead 1-0 in 4th, one out. Dustin Pedroia is on 1st, and he's running on the pitch. David Ortiz "singles" to the gap in left-center, but it would be a double for most other players. Pedroia is already at 2nd when the ball lands, and already passing 3rd when the outfielder picks it up. Semien makes an excellent relay throw home and honestly miiiight have had him if Vogt had caught it cleanly. Can't tell from the replay if he short-hopped it, but the throw was there; honestly, I would have given Vogt the error for the missed catch. Anyway, the run is properly scored as an earned run, and the error merely allows Ortiz to go to 2nd, where he should have really been anyway. This error didn't cost the team an out so it didn't cost the pitcher any extra pitches, and Ortiz was stranded after a K, a BB, and a groundout. As for the run that scored, well, the A's simply got school by a former MVP on that one. Verdict: Harmless, like it never even happened, and I think the error should have been on Vogt
11. May 12: A's lead 8-1 in 7th. Pedroia beats out infield single to lead off, but Semien's throw gets away and Pedroia advances to 2nd. I actually have this game recorded still, so I went back and watched the play. It was a slow chopper and Pedroia is fast; there are a few elite shortstops who might have gotten him, but this was a pretty clean infield single. It was definitely worth making the throw, though, just in case, and while he airmailed it a little bit I really think Max Muncy should have knocked it down. So, this error did not cost an out, and it was such a team effort that it was initially charged to Muncy and then later changed (I would have kept it on Muncy, personally).
The next batter K's, and then Hanley Ramirez chops a grounder in the infield grass on the right side. Hanley is thrown out, but the next question is whether that might have been a double play if Pedroia had been on 1st. Well, to start, it was a 3-2 count and I think Pedroia would have been running anyway; heck, the ball might have gone through for a hit with the defense positioned for the steal attempt. And even if he hadn't been stealing, there was virtually no chance of a GIDP anyway; it wasn't hit hard enough and they barely got Hanley as it was. Mike Napoli then hits a little flair down the RF line to score Pedroia; he would have scored easily even from 2nd, and the run was properly recorded as earned. Next batter grounds out to end frame. So, the error didn't cost an out, it didn't cost any extra pitches, it wasn't even fully Semien's fault, and the runner would have scored anyway. Verdict: Harmless, like it never happened, and I think the error should have been on Muncy
12. May 13: A's trail 1-0 in 8th, two on, two out. (This is the Wade Miley game.) Pedroia hits a grounder to the hole in the left side, Semien ranges to his right and picks it. However, he airmails the throw badly, all the way into the seats. One run scores, and that run is fully on Semien. As a result, the inning continues, and Evan Scribner gives way to Fernando Abad to face Big Papi. Ortiz is retired on 4 pitches. The A's lose 2-0, but it's because they go 0-for-14 with runners in scoring position; at one point, Semien hits a leadoff triple but is stranded by his teammates. Blame that unearned run on Semien, but blame this loss on the entire rest of the team -- Semien never even batted with a runner in scoring position, and he didn't bat again after his error while his teammates went down quietly, so there's little he could have done to help. Verdict: Cost a run, a pitching change, and 4 pitches, but it didn't lose the game
13. May 17: A's lead 2-1 in 5th. I was at this game. Adam Eaton reaches on Semien's error to lead off. A single and a groundout make it corners with one out. Jose Abreu singles down the line on a ball that Brett Lawrie absolutely positively should have fielded, scoring Eaton. The next batter singles, and it was Scott Kazmir's own fault for not covering the bag. Kazmir walks the next batter with the bases loaded, then deflects what could maybe have been an inning-ending GIDP and knocks it into left field for two more runs. He's finally replaced by Otero, who needs only 4 pitches to record the GIDP to end it. Semien definitely started this inning off on the right foot, but it was an absolute shitshow of horrid team defense between him and Lawrie and Kazmir. By my count, there were six batters in this inning who should have been retired, and only one of them reached because of Semien -- sure, it was the leadoff guy, but being spotted a runner on 1st doesn't excuse the display I saw out there from the rest of the team. He played a role, but I think you'd be crazy to blame him for all of this. A's never really got back in it and went on to lose 7-3. Verdict: Cost part of a run amidst a larger team-wide circus act
14. May 17 (continued): Same game, A's now trail 5-3 in 8th, none on, two out. Semien flubs grounder, batter reaches 1st. Otero is pulled, and Abad throws one pitch to get the final out. Verdict: Cost a pitching change and 1 pitch, otherwise harmless
15. May 18: Tied 1-1 in 3rd, one on, one out. Sogard fields a grounder and glove-flips to Semien for what should have been a fairly routine GIDP from Semien's perspective. However, Semien misses the feed and fails to get either runner out. It looks like maybe he wasn't expecting the glove-flip or just didn't get his glove up quickly enough, or maybe he was rushing and forgot to focus on the task at hand; regardless of excuses, that's a play he needs to make and it costs two outs. The next batter walks, but then Chris Carter hits into a GIDP that Semien fields, takes to the bag himself, and finishes with a strong throw. Drew Pomeranz needs only 7 extra pitches to clean up Semien's mess, and no runs score. Pom ends up leaving the game early with an injury anyway, and the A's go on to win. Verdict: Cost 7 pitches, but ultimately harmless because his pitcher picked him up
Those are the details of each of Semien's 15 errors. Let's tally them up!
|Harmless, but cost a few pitches
|Partially responsible for 1 run
|Fully responsible for 1 run that didn't end up mattering
|Partially responsible for 3 runs that blew a save in a loss
Semien has 15 errors, but I think two of them were wrongly attributed to him. He should only have 13 errors. None of his errors have caused his pitcher to throw more than nine extra pitches or face more than two extra batters, and on average his 15 miscues have cost only 3.27 extra pitches. It appears that none of his errors have led to seriously extended innings.
Only seven of the errors resulted in a run scoring (9 total runs), and only five of those nine runs were completely Semien's fault. Only one of the run-scoring instances actually factored meaningfully into a game, and it was the least egregious of all the errors -- simply missing 2nd base with his foot is just a dumb short-hop and it required replay to even call it, and that would never even have been noticed in all of baseball history before 2014. And anyway, Otero was the real goat for the meatball homer. Even if you want to give Semien credit for that whole blown save, which I think would be unfair, it still means his errors have led to only one actual loss.
Add it all up, and here's what we get. Semien's 15 errors have cost the A's 49 extra pitches, 3 pitching changes, 5 full runs, partial share of 4 more runs, and a tiny share of one blown save, all over the course of 40 games. And he's done all that while posting a 141 OPS+ with the second-most homers, the most extra-base hits, the most steals, and the fourth-most RBI on the team.
Before I go any further, let me make something clear. I'm not trying to convince you that Semien has not been a bad defender. I have eyeballs too, and he has most definitely been inadequate. Furthermore, beyond these errors there are other plays that weren't made but were scored hits, and double plays that weren't made -- especially the latter, since by rule you can't get an error just for failing to get the second out on a GIDP, no matter how routine. You can see those lesser mistakes in his wretched UZR, which is by far the worst in the league, small sample be damned. These 15 errors aren't his only offenses, they're just his worst, his most unambiguous, and the ones that are most commonly cited. They're the difference between a weak shortstop who can't make a lot of difficult plays, and a negative influence who misses routine plays and/or actively causes harm.
What I'm trying to do is play a bit of devil's advocate, and to give us a better look at what a player's error total actually means. In terms of real results, Semien's errors have cost the team, at most, one win, despite costing individual runs in several other games. Semien has had 172 fielding chances, which means he's still completed 157 plays successfully -- he really does usually make the plays. He's also been involved in 25 double plays, which is among the MLB leaders, so again it's not like he blows it every time and it just doesn't show up in the box score. Nor are his errors all in key late-game situations or anything, where they would be even more glaring.
He's in the top-5 in total chances, so he's making as many outs as anybody. Given that he's doing that while playing behind a pitching staff that induces the sixth-smallest ground ball rate in MLB on the 10th-lowest contact rate and ranks only 19th in hits allowed, maybe there aren't that many extra "error-ruled-a-hit" screwups mixed in there after all, at least not enough to drastically change these conclusions. Note that Defensive Runs Saved rates Semien as exactly neutral, with a value of 0 (zero). That's only one dissenting statistic amid a forest of uniformly negative opinions, but the point is that there does actually exist one real measure that doesn't hate Semien's glove at all.
And remember, shortstop is a high-error position to begin with -- five others around the league have at least eight miscues, and only a handful don't have at least four. The MLB runner-up is Ian Desmond, the guy in Washington who was good enough to push Yunel Escobar off the position, and Desmond's got 11. Semien has 15 errors, but it should really be 13, and that's really only 5-8 more than an average replacement (who, by the way, won't hit for anything remotely close to a 141 OPS+).
Marcus Semien has been a terrible defender. Marcus Semien's defense has not cost the A's much of anything in terms of wins and losses. Both of those statements can be true, just like a team can get a bunch of hits but fail to string them together into runs or a pitcher can throw nine frames and lose 1-0. It's all about timing, and the vast majority of the errors came at times when they just didn't matter, which makes sense since most innings don't produce runs anyway and most baserunners don't end up scoring. Again, I am not excusing the errors or saying he has not been a poor defender. The point isn't that Semien hasn't been bad, it's that making a handful of extra fielding errors isn't necessarily the guaranteed kiss of death it's made out to be, and also that it shouldn't overshadow a legitimate bat in a golden age of pitching. Not every error in baseball allows a crucial game-winning run to score, or even allows a run to score at all.
If you're looking for culprits behind the A's 14-27 record, then absolutely cite Semien's defense as a thing that has been poor and needs to improve. But please don't cite it as the cause of the losses, because that is just not true. If you're torn between bullpen and defense for your scapegoat, I assure you the bullpen is the place to focus your attention.
* Stats in this post do not include Tuesday's games, except for the updated W-L record in the final paragraph.
Separate note: This post does not consider the timeliness of errors made by other Athletics. There may well be some that cost Oakland some games. But they weren't made by Semien.