By In Stuff

BR Questions of the Week

Brilliant Reader JRoth has a question about my overwhelming dislike of errors in baseball:

“How many homeruns does Pedro Alvarez have? 32, which is the number that stubborn ignoramuses who believe in errors would claim, or 33, the *real* number, based on how many times he hit a ball and crossed home plate on the same pitch?”

Well, first of all, I do not think people who believe in errors are stubborn anything — and I’m pretty sure I never said that. Errors are a part of baseball. Many people like that, and that’s great. I’m sure I’ve written this before: I remember once asking Bill James if he would get rid of errors if given the chance. I can’t find the email just now but I remember he basically said that even though he was one of the first strong voices to point out the absurdity of errors, he would not get rid of them. They are part of the game now.

So, I don’t want to overstate this. I have no illusion that the error is going anywhere I don’t know that I would change that if I was named king of baseball. I just think it’s fun to point out the absurdity of errors every chance I get so people will not take them so seriously.

But to get to JRoth’s specific point — earlier this year, Pedro Alvarez singled to right field and the ball slipped under the glove of Chris Denorfia. Alvarez ran all the way around the bases to score. It was filled (correctly, I think) as a single and three base error. So, if my errorless world existed, how would we score that? As a home run?

OK, two thoughts. One, the Alvarez play is not the sort of error that troubles me. The reason: Alvarez was not credited with an out. He was credited with a single and then a three-base error followed. That may or may not be sloppy scorekeeping, but it has nothing to do with my beef with the error. My beef revolves around errors that punishes a hitter (or rewards a pitcher) by recording an out that NEVER HAPPENED. That just drives me insane.

I’m sure I’ve written the story about the probably apocryphal quarterback in the mountains of North Carolina many, many years ago — I remember hearing this story when I first started at The Charlotte Observer. The kid apparently was putting up insane passing statistics, so much so that the Observer sent someone all the way out to this little town to watch the kid play.

On the first play of the game, the quarterback threw a pass to a receiver who dropped it. The reporter saw the statistician credit the quarterback with a 25-yard pass.

“You can’t do that,” the reporter said.

“Why not?” he said. “It wasn’t his fault the guy dropped it.”

That is the error stat keeping that I vandalizes our statistics. As far as how many bases Alvarez should be credited for, well, now we’re getting into a different kind of error, one that has its own vagaries and peculiarities and we can discuss at the another time.

But my second thought is this: While I don’t know if Alvarez should get home run credit for Denorfia’s mistake, I do know that if Denorfia had lost the ball in the lights or the sun and Alvarez ran around the bases that WOULD be an inside-the-park home run. I do know that if two fielders went after the ball and collided and the ball bounced away and Alvarez ran around the bases that also WOULD be an inside-the-park home run. If Denorfia mistakenly took a step in off the crack of the bat, and it sailed over his head and just out of his reach and Alvarez rand around the bases, that too WOULD be an inside-the-park home run.

This is why it drives me nuts when people say, “You shouldn’t reward a player for another person’s mistake.” We do it all the time, in every one of our sports. You could argue that every good thing that happens for one person in sports is due, at least slightly, to a mistake by another.

Many people like that baseball places judgment on its statistics. They want a game where a guy who should have been out is called out in his own personal stats. I get it. I just disagree. I’d rather count stuff.

* * *

Brilliant Reader Ross asks: “Should MVP equal highest WAR in all cases?’

The clear answer here is no. WAR is not close to a perfect statistic — there are no perfect statistics. There are different versions of WAR, different ideas about how to to measure a player or pitcher’s value, wildly different opinions about things as basic as how much a player actually contributes with his defense and how to value clutch performance.

That said, as an MVP voter I would like to give the award to the best player. I think WAR is helpful in identifying the best player. It’s not the only statistics. There are others like RE24, which I wrote about here, and various fielding independent pitching stats and cool defensive stats and, absolutely, the basics like strikeouts and walks, home runs and ERA, on-base percentage and slugging. So is watching a lot of baseball. So it talking to a lot of people around baseball. I don’t think voting for an MVP should be a one-step process.

There’s something else here too, something that often gets overlooked: There is a point where a difference in statistics is basically meaningless. If one player has a 7.2 WAR, and another has a 7.0 WAR, choosing the higher based on a 0.2 difference is, in my view, a copout. I think even a half a win — even a whole win, even more — certainly can be within the margin of error.

What has made the Mike Trout-Miguel Cabrera thing so compelling is that Trout has had a HUGE advantage in WAR while Cabrera has had a HUGE advantage in the Triple Crown numbers. That’s an anomaly, I think. In the National League right now, Andrew McCutchen has a 7.5 WAR, Clayton Kershaw has a 7.2 WAR. I think picking McCutchen based solely on that is a disservice to the award and to what WAR attempts to do.

In the end, WAR is a tool. It’s flawed and controversial, but it’s also fascinating and, as of right now, among the the best tools we fans have for measuring the all-around contributions of a player. It’s a pretty good starting point for, what I hope, is always a much longer thought process.

* * *

Brilliant Reader Sean writes in not so much with a question but to talk a bit about Robinson Cano’s amazing numbers at Camden Yards.

Up to the minute:

Cano at Camden: .360/.405/.578 with 27 doubles and 13 homers in 73 games.

Not bad. I thought it might be fun to look at some great numbers in every ballpark. Because, well, I’m crazy.

Angel Stadium (Los Angeles Angels): Mike Piazza: .438 average, .734 slugging in 15 games.

Ballpark at Arlington (Texas): Let’s go with Carlos Delgado: .316 average might not wow you, but his 15 doubles and 19 homers in 41 games get the job done.

Turner Field (Atlanta): Barry Bonds, of course. .352/.482/.800 with 14 homers in 34 games.

Camden Yards (Baltimore): Cano is impressive, but I’d go with Josh Hamilton — .362/,408/.723 with 10 homers in 23 games. He might have considered going East instead of West as a free agent.

Fenway Park: Written about it before — Dave Kingman might have been a Hall of Famer had he played his whole career at Fenway . In 18 games there, he hit 13 home runs.

Wrigley Field: Mike Schmidt hit 50 home runs in just 138 games. The wind always seemed to be blowing out when Schmidt came to town.

U.S. Cellular Field (White Sox): How do you explain Karim Garcia? .410/.422/.951 with 10 home runs in 16 games. He played for seven different teams, but never the White Sox — no idea why.

Great American Ballpark (Cincinnati): Troy Tulowitzki is hitting .413 with seven homers in 19 games. Lance Berkman always mashed in Cincinnati — he has hit 23 homers in his 61 games.

Progressive Field (Cleveland): I was just on a podcast with Will Leitch, and I was trying to explain to him just how popular Albert Belle was in Cleveland. It’s kind of a hard thing to explain. I’m sure there were many, many people in Cleveland who did not like Belle, but I have to be honest: I NEVER met any of them. Belle — more than MannyBManny, more than Jim Thome, more than Kenny Lofton, more than Sandy Alomar, more than any other player — seemed to represent to the people I knew the resurgance of Cleveland. He was tough, angry, under appreciated, ripped by the press, and he just kept hammering baseballs — even as the Browns bolted out of town. He was a sure thing. He was especially good at home, where every game sold out. He hit .325/.417/.681 in more than 200 games at what was then Jacobs Field. He banged 77 doubles and 69 homers. And everyone I knew in Cleveland loved him and wished the media would shut up and leave him alone.

Coors Field (Colorado): Hey, how about Eric Karros? In 46 games, he hit .370/.419/.740 with 21 homers.

Comerica Park (Detroit): Albert Pujols has only played 14 games in Detroit. In those 14, he has hit .464/.559/.768 with four homers.

Minute Maid Park (Houston): Carlos Delgado is at or near the top of numerous ballpark lists. In his 14 games in Houston, Delgado hit .442 and slugged .769.

Kauffman Stadium (Kansas City): Adrian Beltre has always blasted the Royals. He’s hitting .374/.408/.640 with eight home runs in 36 games in Kansas City.

Dodger Stadium: Miggy Cabrera has eight home runs in 18 games and a .742 slugging percentage — he’s about the only guy out there who could hit with that kind of power in that cavernous place. Ryan Braun is hitting .365/,444/.667 in his 16 games.

Marlins Park (MIami): Too early to get much data yet, but Giancarlo Stanton is hitting .307 with 30 homers in 116 games.

Miller Park (Milwaukee): Shawn Green had 11 homers in just 19 games.

Target Field (Minnesota): Again, like Miami, it’s too early, but Robbie Cano is hitting .382 in Minnesota. Kansas City’s Salvador Perez is hitting .411 with five homers in 19 games.

Stad Olympique (Montreal): Well, I’m including it to honor Jonah Keri’s upcoming book on the Expos. Sean Casey hit .470 in Montreal in 18 games. Barry Bonds blasted 30 homers in 96 games.

Citi Field (New York Mets): Joey Votto is hitting .350 with six homers in 17 games.

Yankee stadium: Miggy again, no question: .417/.493/.967 with 10 homers in 16 games. Yikes.

Real Yankee Stadium: Here’s a fun one for you — Geronimo Berroa (.729) and Jose Guillen (.707) both had higher slugging percentages at Yankee Stadium than a guy named Babe Ruth (.697). Of course, it was a different Yankee Stadium. Still, Berroa’s 10 homers in 31 games — as well as his .383 batting average — is impressive.

Oakland Coliseum: Remember Larry Parrish? He hit 14 homers in 36 games. It pleases me to now end that the highest batting average at Oakland Coliseum, 50 at-bats or more, is the one and only Neifi Perez, who hit .394 there.

Citizen’s Bank (Philadelphia): Give it up to Luis Gonzalez, who hit .433 in 19 games. Pujols has hit 10 home runs in 26 games.

Chase Field (Arizona): MannyBManny hit .471 in 17 games — Richie Sexson hit 12 home runs in 27.

PNC Park (Pittsburgh): Pujols. Again. He’s hitting .376/.452/.718 with an amazing 29 homers in 89 games.

Petco Park (San Diego): Andruw Jones loved the big dimensions of Petco. He hit .319 with nine homers in 19 games.

Safeco Field (Seattle): Got to tell you, Carlos Delgado’s name seems to come up at virtually every stadium. Delgado was an excellent road hitter — his home/road split are very, very close — and in Seattle he hit .386/.476/.792 with nine doubles, a triple and six homers in 19 games.

AT&T Park (San Francisco): Usually the players with the crazy impressive stats are road players for obvious reasons — small sample size. But at AT&T Park only one player — Ryan Braun in 16 games — has a higher slugging percentage than Barry Bonds. In almost 500 games, Bonds hit .335/.526/.763 with 160 home runs and a ridiculous 552 walks.

Busch Stadium: Ryan Howard was born in St. Louis, grew up in St. Louis — he’s hitting .368/.512/.695 with nine homers in 26 games in St. Louis.

Tropicana Field (Tampa Bay): Chris Richard played for four teams in 280 big league games — he sure loved Tropicana Field. He hit .463 with 13 extra bases hits in 18 games (13 starts) in Tampa Bay. He slugged .907 in Tampa Bay, .415 everywhere else.

Rogers Center (Toronto): Adrian Gonzalez is hitting .356 with 17 extra base hits in 18 games.

Nationals Park (Washington): Give it up to Giancarlo Stanton — .343/.426/.818 with 11 doubles and 12 homers in 26 games. Yeah, Washington could use a guy like him.

Print Friendly, PDF & Email

33 Responses to BR Questions of the Week

  1. Dana King says:

    I sort of agree on errors, but I also look at it from the fielder’s perspective: how do we account for plays he should reasonably have made, but did not?

    Revered Pirates announcer Bob Prince used to advocate for the concept of a team error to cover some of the examples you cited. An infield pop-up that falls between three fielders? Clearly not a hit, but who is charged the error? The team. Same with a collision.

    It might have been Bill James in an Early Abstract who talked about including times a batter reached base via error as part of his OBP, on the rationale Rickey Henderson provokes more errors than Willie Mays Aikens, if only because the fielder knows he has to hurry. I could get behind that, acknowledging the fielder didn’t do his job, but also understanding the hitter may have had something to do with that.

    • Grant says:

      Well, an ‘error’ is not the only way a player fails to make a play he should. Much of the criticism of Jeter is that he was absolutely terrible going to his left, costing many hits because he couldn’t get to grounders up the middle (ish). Those don’t show up as errors, but they matter.

    • JRoth says:

      Publicly available FIELDf/x data would revolutionize the game as much as everything that’s happened since the first Abstract, IMO. Sadly, it seems that it will be kept proprietary,

    • invitro says:

      “Publicly available FIELDf/x data would revolutionize the game”

      What would change? Curious.

    • FranT says:

      Manny Machado made a bone-headed play against the Yankees the other night when he thought he had a force at third but didn’t. The Orioles ended up with no outs, but the hitter lost points on his batting average due to the “fielder’s choice”. Alas, Machado was not charged with an error.

  2. Tim says:

    An example from cricket might be helpful in the case of errors. In cricket, if the bowler bowls a wide (bowls the ball so poorly that the batsman cannot possibly hit it), 1 run is credited to the batting team, but not to the batsman. The run is put in the “Extras” column for that match. In some cases, the wide will go into the field of play and become a live ball, and more than one run is scored, and occasionally, the ball travels all the way to the boundary of the pitch, and five runs are recorded in the “Extras” category. Bases reached that are the result of errors could go into a newly created “Extras” category. The batter should be credited with an at-bat, plate appearance and an “Extra” (multiple “Extras” for each base reached), which would be included in his Batting Average, On Base Percentage and Slugging Percentage. It wouldn’t be a hit, so Pete Rose would stay happy, but it would be included into all averages, which would make Mike Trout happy (and Miggy Cabrera a little less so). Official scorers would still get to decide what was a hit or an error, and all the stupidity with earned runs that goes along with it, so they would stay happy. I’m thinking this is the best answer.

    • Tim says:

      Actually, you would have an “Extras” column for each reach-on-error occurrence and an “Extra Bases” column for each base reached on error. That will make the OBP and SLG work out right.

  3. I think errors are good to track, but it should be up to statisticians and analysts to decide how to give credit for those plays. If a player hits a ground ball to SS and the fielder throws it into the stands, so the batter reaches 2nd base, give the SS an error but call it a double. An analyst can decide if he wants to exclude that play, count it as an out, or give the batter partial credit if, say, he’s trying to evaluate the batter’s true talent. It’s not any different than when an analysts exclude IBB’s from a pitcher’s stat line, really. Or, to use an example from football, when Brian Burke at looks at teams’ run/pass balance on first down, he writes: “I looked at plays where the score was within 10 points and the game is either in the 1st or 3rd quarter. Ultimately, the goal is to look at situations where neither team is desperate or hurried, and both teams are still playing a game of maximizing their net point advantage.”

    The bottom line is, statistics should record what happened; drawing conclusions from those statistics is the job of analysts–especially since not every statistician is going to ask the same questions.

    • JRoth says:

      Yeah, that’s good stuff.

      I appreciate Joe’s response, of course. Part of my issue with his position on errors is that ISTM that it’s in a sense the opposite of DIPS theory – instead of trying to correct a batter’s line for the fielders he’s faced, it wipes out any evidence at all that, in fact, he got lucky to the tune of a dozen points of AVG or SLG. I certainly agree that errors are very clumsy (given my druthers, I’d be a lot more aggressive about calling them – guy falls down on dry turf and misses an easy fly, it’s an error even if he never gloves it), but I don’t think pretending they don’t happen gets us any closer to what is generally considered the goal of most sabermetric analysis – to determine the underlying talent. Getting on base more due to errors has a talent component, of course, but it’s not identical to overall hitting talent (a comparable stat is probably HBP: some guys top the leader boards every year, and it’s obviously a skill/talent, but other guys appear in the top 10 one year and don’t get a single HBP the next; if you’re analyzing a player’s OBP, you need to know which kind he is).

    • Robert says:

      The answer is somewhere in between. Some errors are totally chance (medium grounder right at somebody takes a funny hop) and some are talent driven (you have to scoop and throw quicker on a grounder from Jarrod Dyson than Billy Butler). I think it’s obviously incorrect to count all errors as outs, but I think it’s equally incorrect to count them all as hits. But I think it depends on what you want your stat to do. The answer is obviously somewhere in between for predictive talent purposes. But from the perspective of counting production or MVP voting, a hitter should be rewarded for what he actually does, all the time, to help his team win. Reaching on a error helps his team win just as much as popping one up in the lights and reaching, yet we (insanely, IMO) score those plays differently.

    • Robert says:

      This comment has been removed by the author.

    • Ian R. says:

      @JRoth – I don’t think the underlying goal of sabermetric analysis is to determine a player’s underlying talent. That’s a big one, of course, but it’s only relevant to one category of analysis: predicting future performance. The other traditional branch of sabermetrics focuses on analyzing past performance, and the idea there is to get an accurate view of what actually happened.

      When it comes to predicting future performance, sure, we’d want to know whether a player is getting a lot of hits because he’s hitting the ball hard or because a lot of his ground balls happen to sneak past infielders. Component stats like line drive percentage will help us figure that out. When we’re just looking at past performance, though, I don’t see the point in distinguishing between a line-drive single and an error.

  4. Jose Bautista owns Target Field. He’s batting .362/.423/1.043. Bautista is tied with Joe Mauer for 10th all-time in home runs hit at Target Field with 10. Bautista has the most HRs at TF hit by an opponent.

  5. Scott says:

    This from tonight’s box scores (unofficial):

    1st M Cabrera safe at second on throwing error by third baseman C Gillaspie, A Dirks scored.


  6. Ross Holden says:

    Thanks for answering, Joe! Good analysis

  7. Wilbur says:

    This comment has been removed by the author.

  8. Wilbur says:

    Joe and readers, how would you deal with (categorize statistically) an error where no one reaches base or advances? An example would be a dropped pop foul. And would it matter what the batter ended up doing in the at bat?

  9. invitro says:

    Why can’t we both give the batter a hit instead of an out that did not happen, and -also- give the fielder an error? I agree with Joe that taking hits away on ROE is a serious problem; 9 hits in a season is not small potatoes. I want to have hope that fixing this is so obvious that it will be done soon, but there are too many fans whining about rewarding a batter for a fielder’s mistake for me to feel truly hopeful. Someone can come along later and judge the “quality” of the hit to see how repeatable it is, to use in WAR or whatever.

    Taking away hits is the biggest problem. The second biggest problem is not issuing errors to fielders unless the fielder gloved the ball. I don’t understand why people think it’s logical to require touching the ball for a play to be an error. Like hits stolen by ROE calls, I think this happens lots of times in a season. I wish there were reliable stats on it. Labeling these errors retroactively (the day after) would be fine.

    I think the “team error” situations really are small potatoes, though I’d appreciate stats to confirm/deny here also.

    I would give Alvarez a single and three Xtra Base Advances (XBA) that I believe should be a regular stat anyway, and also give the fielder an error. But as Joe said, Alvarez was not robbed of a hit in this situation, so it’s not an example of the worst scoring offense due to error calls.

  10. ethegolfman says:

    As a DBacks fan, let me point out that 12 of Richie Sexson’s games at Chase Field were as a member of the DBacks, 6 HRs including one of the longest HRs in the history of CHase Field. A 480 ft bomb off the scoreboard which was featuring his face at the time so he knocked one off his own face there.

  11. Phil says:

    As much as I would love to find a way to “fix” errors (ha ha), the logical extension is that we ignore the fielders entirely. After all, look at how many times a batter faces the shift — and then singles through the vacated shortstop hole. The announcers say “Well, without the shift, that’s an out.” So perhaps the positioning of the fielders shouldn’t matter, and we just look at a spray chart with speeds and spins and determine what “should” be hits, as though we’re playing wiffleball (“off the porch railing is a double, hitting the tree leaves is a homer”). . . . I see this as too slippery a slope to navigate safely.

  12. PDS Math Guy says:

    You know, I understand this wasn’t the purpose of the anecdote, but I’m a baseball fan and not a football fan, and your story about crediting the quarterback with a completed pass makes perfect sense to me. I don’t understand why that isn’t done all the time!

    And isn’t there some unofficial stat in tennis called “unforced errors”? Who decides whether it’s forced or unforced? Serious question…

    • Grover Jones says:

      The Wall Street Journal did a study on “unforced errors” in tennis within the past month. They said they were wildly under-reported. They did their own counting of random matches and saw twice as many “unforced” as the official scorer (or whatever they call him in tennis).

  13. PDS Math Guy says:

    This comment has been removed by the author.

  14. Ben Wildner says:

    I think this article managed to bring back a little jolt of terror to every fanbase in the bigs. Certainly brought back a little suppressed Shawn Green PTSD for this Brewers fan.

  15. wow9gamer says:

    I have no illusion that the error is going anywhere I don’t know that I would change that if I was named king of baseball.
    World of Warcraft Gold Billig
    World of Warcraft Gold Kaufen

  16. mckingford says:

    Of course Miggy hit a “home run” this year that was an even more egregious error than the Alvarez play, even if it wasn’t ruled an “Error”. May 22 against the Indians he hit one to deep centre…to the warning track, but not deep enough to clear the wall. Bourn arrived just in time to come to a stop under the ball, which hit the heel of his glove and bounced over his head, and then over the wall for a “home run”. Without Bourn’s intervention the ball never had a chance to clear the wall (except perhaps on a bounce, which would have resulted in a double).

    There was never any discussion of this being anything other than a home run. Yet it was clearly an error. I say this as a Tigers’ and Miggy fan (and as a Miggy fantasy owner), but the only proper ruling would have seen this ruled a 4 base error. But it wasn’t an out, so I certainly agree with Joe that Miggy shouldn’t be credited with one either. (Where’s JRoth’s outrage over this play?)

    • JRoth says:

      Unlike a lot of seam heads, I don’t get outraged over the current rules about scorekeeping; my whole point is that Joe does, but, in his recent posts on the issue, wasn’t acknowledging the absurd results from his anti-error stance (nowhere in the previous 2 posts on the topic did he pay any mind to extra bases on errors).

      Grant Brisbee at Baseball Nation recently wrote a post asking readers what baseball rules they’d change if they could, and I think that literally every single response was about scorekeeping. For some reason, this matter is very dear to people, and I just couldn’t care less. I was needling Joe because he was acting as if one scorekeeping outcome is unconscionably absurd, while another, equally absurd one, would be completely fine. The mistake is imagining that there’s some set of scorekeeping rules that will avoid absurdity. Baseball is an absurd game, so there isn’t.

  17. Unknown says:

    I think this statement from Joe sums up the argument against errors as they are currently defined: “You could argue that every good thing that happens for one person in sports is due, at least slightly, to a mistake by another.”

    Is there really that much of a difference between a pitcher hanging a curve ball and a groundball going through the SS’s legs?

  18. Tonus says:

    So why not add “ROE” (Reached On Error) as a statistic, and add it to on-base percentage? A batter who is walked intentionally gets similar treatment, although the number is also folded into his overall walk total.

  19. Richard S. says:

    “There’s something else here too, something that often gets overlooked: There is a point where a difference in statistics is basically meaningless. If one player has a 7.2 WAR, and another has a 7.0 WAR, choosing the higher based on a 0.2 difference is, in my view, a copout. I think even a half a win — even a whole win, even more — certainly can be within the margin of error.”

    Agreed. When you think of all the things that can affect the outcome of a game, that are not dependent on the actual play…

    * An umpire blows a call, but the play is so close anyway that no one challenges it or even notices.
    * A rainout shifts a team’s pitching rotation, giving them favorable matchups over the next few weeks.
    * A swarm of insects descends on the playing field, discombobulating the pitcher into giving up a key run.

    … you realize that there is a significant amount of pure chance that goes in to a team’s record. You hope it evens out over the course of a one hundred and sixty-two game season.

    Now you can’t really call it a “tie” when one player bats .332 and another bats .331 – the numbers are clearly different. But with something as contrived as WAR, it’s entirely possible that a difference of a full run might not be a statistically significant difference.

  20. clashfan says:

    Chipper Jones, at old Shea Stadium:


    88 377 101 20 1 19 55 .313 .407 .557 .964

Leave a Reply

Your email address will not be published. Required fields are marked *