clock menu more-arrow no yes

Filed under:

RBI - Really Bad Indicator

New, 858 comments
Is BFI getting enough RBI? Or do the A's need a better ROI on their OBP to avoid runners being LOB? Please LMK, ASAP. KThx.
Is BFI getting enough RBI? Or do the A's need a better ROI on their OBP to avoid runners being LOB? Please LMK, ASAP. KThx.

Ah, the RBI. Runs batted in. The ol' ribeye steak. To the old-world baseball mind, it is the quintessential measure of a run producer. The more runs you drive in, the better you are. You can't argue, it's science! Juan Gonzalez agrees; he used RBI to win two MVP awards, each over a long list of much better players who enjoyed much better seasons.

To the more enlightened modern baseball mind, however, the RBI is an incomplete statistic. It is a raw total in a world which looks increasingly toward rates and percentages. Telling me that you have 80 hits isn't impressive until you tell me that you did it in 240 at-bats. That's a great batting average, rook! Ditto for your 100 strikeouts, Mr. Pitcher; did you rack those up in 100 innings, or 200? That's going to make a big difference when I analyze your performance. Are you a good basestealer if you swipe 50 bags? Probably, but I'm a lot more interested in whether your success rate is above or below 80%. If it took you 75 attempts, then, no, you're a crappy basestealer who just runs a lot. Hopefully you can see a trend forming: A player's raw numbers are only informative when put into the proper context.

And yet, the RBI is still there. It's still there in the box score, it still shows up on the TV screen when a player comes up to bat, it's still listed on everyone's Baseball-Reference page, and it's still right there on the back of the baseball card. Well, I assume it is. I haven't actually seen a baseball card in at least 10 years. They still make those things, right?

What does the RBI really mean, though? Is 100 a lot? 80? 120? How many does it take until you're good? How many until you're officially a "run producer?" Can you be a bad hitter even though you drive in lots of runs? More specifically, who is the best run producer on the 2012 A's? Brandon Inge and Josh Reddick have 48 and 49 RBI, respectively; are they evenly matched when it comes to driving in runners? Raw RBI totals can begin to answer these questions, but there is a much better statistic which is as not-new as it is completely ignored by everybody: RBI Percentage. Follow me after the jump to hear all about it!

RBI Percentage was created (to the best of my knowledge) by David Pinto at Baseball Musings. The formula is simple:

(RBI - HR) / Runners On

(Pinto also multiplies by 100, so that the result looks like a percentage, 22.5%, rather than a batting average, .225)

Take the player's RBI total, subtract his home run total, and divide the difference by the number of runners who were on base during his plate appearances. The purpose of subtracting the home runs is simple: This is a measure of how efficiently you drive in the baserunners you are given to work with. When you hit a solo home run, you're just driving in yourself. That is still good, and it's still technically run production, but it misses the spirit of this metric. A 2-run homer is a different story, though, because at least you've still driven in the only baserunner you were given. A grand slam is awesome, but it's no more awesome than a bases-clearing double in terms of your effectiveness at driving in other runners.

Before we look at any numbers, there are a couple of obvious flaws in this metric. If I was going to tweak the formula, I would remove the plate appearances (and the corresponding baserunners) in which a player was intentionally walked or hit by a pitch. As Jack Cust showed, an unintentional walk with runners on base can be the sign of a poor run producer; on the flip side, Barry Bonds showed us that even the best hitter in history can't drive in those runners when he is intentionally walked and, therefore, unable to take a swing. It is literally impossible to drive in a run on an intentional walk if the bases aren't loaded, so it shouldn't count against the hitter. Same goes for a HBP. I would also like to see a separate version which only accounts for runners in scoring position (since driving in a runner from 1st is much more difficult, and leaving one at 1st is much more excusable).

With that, it's time to get nerdy! Let's warm up with something simple. Here are Oakland's RBI leaders this year (through today; Inge's numbers include his time in Detroit):

Rank Player RBI
1 Yoenis Cespedes 50
2 Josh Reddick 49
3 Brandon Inge 48
4 Seth Smith 33
5 Jonny Gomes 29
6 Coco Crisp 22
7 Brandon Moss 21
8 Kurt Suzuki 17
9 Cliff Pennington 16
10 Jemile Weeks 15
11 Kila Ka'aihue 14
12 Chris Carter 12
13 Derek Norris 10

Here are those same 13 players, but re-ranked according to RBI% (not counting today's game, but still counting Inge's time in Detroit):

Rank Player RBI%
1 Yoenis Cespedes 20.90%
2 Brandon Inge 19.25%
3 Derek Norris 17.95%
4 Chris Carter 15.00%
5 Jonny Gomes 14.29%
6 Coco Crisp 13.82%
7 Kila Ka'aihue 12.20%
8 Seth Smith 12.09%
9 Brandon Moss 11.90%
10 Kurt Suzuki 11.11%
11 Josh Reddick 10.83%
12 Cliff Pennington 8.84%
13 Jemile Weeks 6.81%

Well, that certainly shook things up. A few notes about these numbers:

  • League average (or at least, median) appears to be around 14.50%
  • Cespedes's mark ranks 6th in all of MLB (minimum: 100 baserunners)
  • Inge's ranks 19th
  • Norris and Carter can be partly ignored due to tiny sample sizes
  • There is a substantial difference between Gomes and Smith in RBI situations
  • Coco gets it done far more often than I give him credit for
  • Weeks ranks 5th worst in MLB (281st out of 285)
  • To put Weeks' futility in context, Gio Gonzalez has a mark of 7.14% right now
  • To put that into the proper context, Gio only has 2 total RBI, so it's not really a fair comparison. But still. Step it up, Jemile.
What can we learn from this different way of looking at run production? First off, Inge has been every bit as valuable as his inflated RBI total suggests. That still doesn't mean that he's a good overall hitter, because he's pretty clearly not. What it does mean is that the skills he does have, like power and solid situational hitting, have made him about as efficient at driving in the baserunners he is given as Prince Fielder (19.35%) and David Wright (19.12%). Between the grand slams and the sac flies (3 of the latter, which leads the team), Inge has done a great job of capitalizing when there are ducks on the pond. He may not get on base a lot, but given his spot in the lineup (5th-7th, finding himself in RBI situations often), he gets his job done at an incredible rate.

The same cannot be said for Josh Reddick, though. Nearly half of his 49 RBI have come from driving himself in on home runs. When runners are on base, Reddick barely drives in 10% of them (or, just over half as many as Cespedes). He is actually in some pretty good company that low on the list (Curtis Granderson and A-Rod are each below 11.5%), but he is also in some really, really bad company. Suzuki has a higher percentage, and Alcides Escobar isn't far behind. Justin Smoak has a nearly identical percentage (10.73%), and he was recently demoted due to sucking at hitting. By the Mariners.

What does the rest of the league look like? Here are your current Top 20 in RBI (Top 21, really):

Rank Player RBI
1 Miguel Cabrera 82
2 Josh Hamilton 81
T-3 Josh Willingham 72
T-3 Jason Kubel 72
T-5 Carlos Beltran 71
T-5 Adam Dunn 71
7 Edwin Encarnacion 70
T-8 Prince Fielder 69
T-8 Ryan Braun 69
10 Mark Trumbo 68
T-11 Mark Teixeira 67
T-11 David Wright 67
T-11 Carlos Gonzalez 67
14 Andrew McCutchen 66
15 Jose Bautista 65
16 Matt Holliday 64
T-17 Andre Ethier 61
T-17 Adrian Beltre 61
T-17 Albert Pujols 61
T-17 Billy Butler 61
T-17 Jay Bruce 61

...And here are your current Top 20 in RBI%, minimum 100 baserunners (copied directly from Baseball Musings because OMG so much typing):

Player Runners On Runs Batted In Home Runs RBI Pct.
Josh Hamilton 240 81 28 22.08
Kyle Seager 229 60 11 21.40
Hector Sanchez 104 24 2 21.15
J.P. Arencibia 161 50 16 21.12
Jason Kubel 237 72 22 21.10
Yoenis Cespedes 177 50 13 20.90
Michael Trout 160 49 16 20.62
Miguel Cabrera 284 82 24 20.42
Yuniesky Betancourt 138 35 7 20.29
Andrew McCutchen 217 66 22 20.28
Miguel Montero 228 56 10 20.18
Carlos Gonzalez 234 67 20 20.09
Andre Ethier 251 61 11 19.92
Carlos Ruiz 213 56 14 19.72
Tyler Colvin 143 41 13 19.58
Allen Craig 179 49 14 19.55
Wilin Rosario 123 40 16 19.51
Prince Fielder 279 69 15 19.35
Brandon Inge 187 46 10 19.25
William Middlebrooks 162 43 12 19.14

(Note: David Wright would be #21; and Buster Posey, #23, just missed being on both lists, at 60 RBI and 18.80%)

As constructed, only 7 of the top 21 RBI men also appear in the top 20 for RBI%. If I'd raised the bar to 150 baserunners, then Wright would have made that 8 of the top 21. Among those top RBI men, the worst percentages belong to a pair of former Athletics: Holliday (15.72%), and Willingham (15.77%).

I don't have any grand conclusions today, just a different way of looking at a concept which you've been thinking about for decades. You thought you knew what RBI were. You still know what they are, in the literal sense. Now, however, you have a new way of analyzing their meaning. A high RBI total will always be good, because it'll mean that you banked a certain number of runs for your team. The runs you drive in don't go away just because you left a bunch of other guys on base. On its own, though, a higher RBI total doesn't necessarily make you better than the next guy, if the next guy accrued his more efficiently. You always knew that. Now you can measure it, and compare players' efficiency rates. Now you can truly find out which players are most talented at driving home runners.

Ryan Howard is writing an angry letter about this as we speak.

(Author's Note: Despite ripping on Juan Gonzalez in the intro, it should be noted that in his two MVP seasons, he did in fact lead the AL in RBI%, minimum 250 baserunners. So, while he was still undeserving of the award overall, at least he truly was good at the one thing everyone thought he was good at. Howard has also ranked highly a couple of times, but not in his MVP season, or lately).