The Death and Resurrection of the RBI

Historically, mainstream sports media has dumbed down baseball statistics. Whether it's displaying a player's stats on the TV screen or providing support for why a hitter should be considered for the Most Valuable Player (or any other offensive) award, we've been spoon fed three main statistics:

Batting Average
Home Runs
Runs Batted In

There are other statistics, but this is what we are supposed to care about most. In fact, we've been told that if you are a truly dominant hitter and lead the league in all three, you should be awarded the Triple Crown. It's the epitome of offensive performance in baseball. Or so we're told.

Lately, the Sabermetrics community has continued to chip away at these long held assumptions. It seems that chicks will forever dig the long ball, but more and more fans are questioning the value of batting average and RBI, in particular.

While batting average may not be completely discarded (instead, seen as a statistic of merit, but inferior to on base percentage), the Run Batted In is viewed by many stat heads as having little or no value at all.

Why the RBI is a Flawed Statistic

It was his work as an RBI machine that netted him a spot in the Hall of Fame... Because he played in the shadow of Johnny Bench, Joe Morgan, and Pete Rose, Perez was underappreciated during his career... -- on Tony Perez, from Cooperstown: Hall of Fame Players

Each night, I read a page or two from this wonderful book to my eight-year-old son, giving him a nice history of the players who are in the Hall of Fame. I couldn't help but cringe when I read this passage.

It's not that I don't think that Tony Perez was a great hitter. He had a long and distinguished career. He may even be deserving of Hall of Fame status. But insinuating that his inclusion as a baseball great falls entirely on this flawed statistic makes a very bad argument.

The problem is that the act of driving in runs is not an individual statistic. It is reliant not only on Perez getting hits (and timely hits), but having teammates who are on base. It is additionally helpful if those teammates are average to above average base runners to maximize runs scored (and therefore maximize runs batted in).

Perez played on some terrific teams. Beyond the Hall of Famers listed above, he also played with Lee May, George Foster, Ken Griffey and Dave Concepcion. Was he overshadowed? Or did Perez benefit as a result?

Part of the reason Perez drove in so many runs was that the Reds were regularly on base. His team scored the most runs in the National League four times from 1968 through 1976 and was second or third three other times during that span.

An example of a potential disparity of riches is represented in the 1975 season, one of the years the Reds scored the most runs in the National League. In nearly 43% of his plate appearances that year, Tony Perez came up with runners in scoring position. He batted fourth or fifth 87% of the time. By comparison, Dick Allen of the Philadelphia Phillies (a distant second in runs scored), batted fourth or fifth 91% of the time. Yet, he came up with runners in scoring position 32% of the time (which was actually the second highest rate of his career).

Perez needed to come through in clutch situations, and he did so frequently. He had a .512 slugging percentage with runners on base that year compared to .403 with the bases empty. But make no mistake, he was granted more opportunities to drive in runs than the typical player. Even had he fared poorly in clutch situations, the vast opportunities assured him of a nice RBI total (he finished with 109 in only 137 games, which was good for third in the NL and short of Greg Luzinski's mark of 120 despite playing in 24 fewer games).

And of course, part of the reason Perez was often up with runners in scoring position was that he hit either clean-up or fifth in the lineup 61% of the time during his career. This may seem obvious, but batters hitting third, fourth or fifth will have the most opportunities to drive in runs.

To again use 1975 as a comparison, the league leader in hits that season was Dave Cash, with 213. He played all 162 games that season as the Phillies lead-off hitter. As a result, he came up with a runner in scoring position only 22% of the time, resulting in a mere 57 RBI.

Of course, part of the reason Perez drove in more runs that year was also that he had more extra base hits than a guy like Cash, who was a singles hitter. Yet, you can't discount the fact that Perez came up in prime position to drive in runs nearly double the time.

Moreover, we are led to believe that singles hitters don't drive in runs because they are rarely in position to post high RBI totals. The reality is that a guy like Cash who piles on singles throughout the season would put up comparable RBI totals to a clean-up hitter with many more home runs and a much lower batting average -- only if, of course, both hitters were to come up to the plate with identical opportunities.

No two players have the same opportunities to drive in runs, and in fact the disparity when comparing different teams and lineup positions is significant. As a result, the RBI statistic is greatly flawed.

The Solution

Now, I'm not the first to propose some sort of solution to this (note: of course, I hoped I was but a Google search brought me back to reality). Tom Ruane of Retrosheet wrote a piece about Joe Carter a while back. Carter was a guy who would regularly hit for a low average but drive in a large number of runs.

Tom Tango, who developed the wOBA statistic, provided some inspiration for this study. Tango's stat is seen by many as being one of the most important measures of offensive worth. Among other things, wOBA assigns a runs created value to each offensive result. Runs created, however, combined both runs scored and runs driven in as a result of that outcome.

Luckily, Tom's a very accessible guy, and he provided a general RBI value for a single, double, triple, home run, strikeout (which is always zero), other out (including sacrifices), walks and hit by pitch, both with and without the DH. How? Lots of work, I'm sure. Taking all identical situations and averaging how many runs are batted in over several years of data.

First, the values (which have been shortened to be more manageable):

Single = .22
Double = .43
Triple = .64
Home Run = 1.6 (including driving one's self in)
Strikeout = 0
Other Out = .0275
Walk/HBP = .02

What does this mean? It means, given the average situation -- runners on base, base running ability of those on base, etc. -- a single will result in .218 runs driven in. It should be remembered that a batter does not need to get a hit to drive in a run, which is why outs, walks and hit by pitch are also included.

Suddenly, we can take the offensive output of any two players and determine which would have been the better run producer given identical circumstances. It's a wonderful stat, though of course it still is not perfect. Such a stat would not consider if a player were more "clutch" in situations where runners are in scoring position.

Yet, the variation (which I call XRBI) is still a vast improvement over the current RBI statistic. Below is a collection of the top 100 players in career XRBI, also including their total RBI and respective ranks. While there aren't a lot of major differences at the top (Hank Aaron is first either way), you'll note that several players you do not normally consider run producers (singles hitters like Pete Rose and Lou Brock) move way up in run production given a level playing field of run producing opportunities. Think that's weird? Look outside the box.

Oh, and Tony Perez, though he did indeed benefit from playing on offensive minded teams, was still a great run producer in terms of XRBI -- still 41st overall.

Player Name	RBI	RBI Rk	XRBI	XRBI Rk	DIFF
Hank Aaron	2297	1	2275	1	0
Barry Bonds	1996	4	2059	2	2
Willie Mays	1903	10	2002	3	7
Babe Ruth	2217	2	1941	4	-2
Stan Musial	1951	6	1899	5	1
Rafael Palmeiro	1835	14	1789	6	8
Frank Robinson	1812	18	1785	7	11
Ken Griffey	1829	16	1776	8	8
Carl Yastrzemski	1844	12	1775	9	3
Eddie Murray	1917	9	1755	10	-1
Dave Winfield	1833	15	1675	11	4
Mel Ott	1860	11	1665	12	-1
Cal Ripken	1695	24	1662	13	11
Pete Rose	1314	91	1646	14	77
Reggie Jackson	1702	23	1624	15	8
Lou Gehrig	1995	5	1617	16	-11
Jimmie Foxx	1922	8	1614	17	-9
Ted Williams	1839	13	1604	18	-5
Alex Rodriguez	1706	21	1604	19	2
Sammy Sosa	1667	26	1604	20	6
Gary Sheffield	1676	25	1586	21	4
Ty Cobb	1937	7	1583	22	-15
Ernie Banks	1636	28	1573	23	5
Manny Ramirez	1788	19	1555	24	-5
Al Kaline	1583	37	1546	25	12
Andre Dawson	1591	34	1535	26	8
Frank Thomas	1704	22	1529	27	-5
George Brett	1595	32	1528	28	4
Mickey Mantle	1509	50	1519	29	21
Mike Schmidt	1595	32	1510	30	2
Billy Williams	1475	52	1490	31	21
Eddie Mathews	1453	55	1483	32	23
Fred McGriff	1550	41	1480	33	8
Harmon Killebrew	1584	35	1475	34	1
Jim Thome	1565	39	1470	35	4
Harold Baines	1628	29	1462	36	-7
Willie McCovey	1555	40	1451	37	3
Craig Biggio	1175	150	1439	38	112
Rickey Henderson	1115	183	1436	39	144
Paul Molitor	1307	97	1428	40	57
Tony Perez	1652	27	1428	41	-14
Robin Yount	1406	68	1419	42	26
Tris Speaker	1529	45	1410	43	2
Al Simmons	1827	17	1399	44	-27
Rogers Hornsby	1584	35	1398	45	-10
Luis Gonzalez	1439	59	1388	46	13
Jeff Bagwell	1529	45	1386	47	-2
Willie Stargell	1540	42	1384	48	-6
Chipper Jones	1445	57	1377	49	8
Dwight Evans	1384	71	1367	50	21
Dave Parker	1493	51	1364	51	0
Jeff Kent	1518	48	1345	52	-4
Honus Wagner	1732	20	1343	53	-33
Brooks Robinson	1357	77	1338	54	23
Carlos Delgado	1512	49	1335	55	-6
Mark McGwire	1414	66	1329	56	10
Darrell Evans	1354	78	1328	57	21
Rusty Staub	1466	53	1325	58	-5
Jim Rice	1451	56	1315	59	-3
Roberto Clemente	1305	98	1313	60	38
Carlton Fisk	1330	85	1306	61	24
Ivan Rodriguez	1264	114	1304	62	52
Steve Finley	1167	156	1302	63	93
Goose Goslin	1609	30	1293	64	-34
Andres Galarraga	1425	63	1291	65	-2
Joe Carter	1445	57	1289	66	-9
Vada Pinson	1170	155	1288	67	88
Vladimir Guerrero	1318	90	1287	68	22
Graig Nettles	1314	91	1287	69	22
Orlando Cepeda	1365	76	1276	70	6
Cap Anson	2076	3	1269	71	-68
Gary Gaetti	1341	80	1266	72	8
Duke Snider	1333	83	1265	73	10
Chili Davis	1372	74	1262	74	0
Joe Morgan	1133	172	1258	75	97
Joe DiMaggio	1537	44	1256	76	-32
Mike Piazza	1335	82	1251	77	5
Larry Walker	1311	93	1248	78	15
Paul Waner	1309	95	1247	79	16
Charlie Gehringer	1427	62	1244	80	-18
Jose Canseco	1407	67	1239	81	-14
Dale Murphy	1266	113	1238	82	31
Nap Lajoie	1599	31	1225	83	-52
Ron Santo	1331	84	1224	84	0
Juan Gonzalez	1404	69	1221	85	-16
Johnny Bench	1376	73	1218	86	-13
Al Oliver	1326	87	1217	87	0
Steve Garvey	1308	96	1215	88	8
Sam Crawford	1525	47	1212	89	-42
Garret Anderson	1353	79	1210	90	-11
Yogi Berra	1430	61	1209	91	-30
Lou Brock	900	338	1204	92	246
Tony Gwynn	1138	169	1204	93	76
Roberto Alomar	1134	171	1200	94	77
Don Baylor	1276	107	1187	95	12
Eddie Collins	1300	101	1185	96	5
Ellis Burks	1206	129	1184	97	32
Jason Giambi	1330	85	1182	98	-13