Joe Carter - Clutch Hitter or RBI Vulture?

Dave Paisley

The Joe Carter debate rages long and hard. He is traditionally a clutch RBI guy. He doesn't have any pretty numbers except those 100 RBI every year. When he retires, there will be heated argument about his possible election to the Hall of Fame. He will finish his career with over 400 home runs and about 1500 RBI. Throw in a World Series winning home run, and he seems a popular shoo-in.

Mowever, more hard-headed baseball analysts look at his almost pathetic batting average and on-base percentage and wonder whether he is anything more than an average Joe. A career OPS of .770 to date is hardly outstanding. Right now, that's about league average, hardly the kind of numbers you want to see from an outfielder, or especially a DH.

A counter-counter-argument to his raw OPS numbers, though, is the fact that he still has an uncanny ability to get those RBIs, despite the horrendously low OPS.

I've been fooling around with some simple predictive models for generating runs from OPS, and sure enough, Carter is way off the normal curve. I decided to see how he compared with a moderate sample of regular American League hitters over the past five years.

I selected thirty-seven regulars, from Frank Thomas to Ozzie Guillen. Most of them have at least 2000 at-bats over the last five years, and almost all have 1500. I then extracted their total at-bats, runs scored, RBI and OPS. I've been using a very simple analysis for runs generated (RG), which is simply the average of runs and RBI. The data is shown in this table...

Name	AB	R	RBI	OPS	RG/AB	RG/ABpr	DRG
Buhner	2495	462	534	.900	.200	.178	.021
Carter	2833	392	509	.763	.159	.142	.017
Canseco	1804	306	373	.893	.188	.177	.012
Gonzalez	2384	395	560	.939	.200	.189	.011
Palmer	2104	336	372	.826	.168	.159	.009
Fielder	2444	342	467	.821	.166	.158	.008
Sorrento	2036	303	379	.832	.167	.160	.007
Leyritz	1416	208	252	.814	.162	.156	.007
Fryman	2849	423	465	.791	.156	.150	.006
Raines	1773	337	244	.827	.164	.159	.005
Griffey	2428	509	528	1.016	.214	.209	.004
Berroa	2067	334	349	.835	.165	.161	.004
Martinez, T.	2445	360	490	.873	.174	.171	.002
Guillen	2226	261	227	.633	.110	.108	.002
Palmeiro	2827	500	537	.912	.183	.182	.002
Ventura	2216	344	396	.850	.167	.165	.002
Ripken	2890	402	439	.769	.146	.144	.002
Vizquel	2495	381	233	.692	.123	.123	.000
Surhoff	2166	312	344	.801	.151	.152	-.001
Thomas	2498	534	599	1.086	.227	.228	-.001
Phillips	2644	538	299	.830	.158	.160	-.002
Rodriguez, A.	1384	260	228	.900	.176	.178	-.002
Salmon	2588	456	497	.931	.184	.187	-.003
Williams, B.	2598	455	409	.865	.166	.169	-.003
McGwire	1499	307	375	1.098	.227	.231	-.003
Hoiles	1830	287	315	.862	.164	.168	-.004
Sprague	2567	316	340	.728	.128	.133	-.005
Ramirez	1938	334	372	.939	.182	.189	-.007
O'Neill	2425	399	462	.922	.178	.184	-.007
Vaughn, M.	2645	458	548	.970	.190	.197	-.007
Martinez, E.	2013	413	388	1.006	.199	.207	-.008
Wilson	1756	187	243	.718	.122	.130	-.008
Rodriguez, I.	2564	382	353	.803	.143	.153	-.009
Bordick	2399	245	229	.638	.099	.109	-.010
Valentin	2391	390	353	.860	.155	.168	-.013
Lofton	2664	536	267	.843	.151	.163	-.013
Knoblauch	2774	531	285	.842	.147	.163	-.016

Stats Glossary

AB - At-Bats
R - Runs Scored
RBI - Runs Batted In
OPS - On Base Percentage plus Slugging Average
RG - ((R+RBI)/2)/AB
RG/AB - Actual RG
RG/ABpr - Predicted RG from regression
DRG - Actual RG - Predicted RG

I then performed a regression of RG/AB against OPS. The correlation coefficient was .95, about as good as you will ever get with this kind of data. The results can be seen in the chart. The red line is the regression line, and it's pretty evident that the data hugs the line quite closely except for a few points. One of those points is Joe Carter, but there is one player even farther above the line than he is, and that's Jay Buhner. Remember, these are five-year numbers, and so any major deviation here has been going on for a long time.

What is it that makes Joe Carter and Jay Buhner so much better at generating runs than their peers? It can't be speed on the basepaths. These guys are as slow as molasses. Besides, look at poor Kenny Lofton - almost as far below the line as Carter is above it.

Maybe some of these guys really do hit better "in the clutch", i.e., with runners in scoring position. Some do, some don't.

I can hear some of you screaming "It's the lineup position, stupid. We can see leadoff guys like Knoblauch and Lofton languishing way down there, while cleanup hitters like Carter hang out on the high side. They obviously get more chances to drive in runners, don't they?"

Well, you'll just have to come back for part 2 to see if all that's true.

Why yes, Dave Paisley was the first one in line at CompUSA the day Excel 98 came out. Share your favorite macro tips at drdjp@strikethree.com.