Ahead in the Count: So How Good are MVPs Really?

December 1, 2010

With the Most Valuable Player and Cy Young awards announced in the last two weeks, we saw a first-time MVP in each league, a first-time American League Cy Young winner and a National League Cy Young winner who had won the American League Cy Young Award seven years prior. Winning consecutive MVP or Cy Young awards is a rarity, though we have seen recent repeats by Albert Pujols and Tim Lincecum. In the last 18 years (1993-2010, which encompasses the last two rounds of expansion), we have seen just six of 36 MVP awards go to the previous year’s winner, and just nine Cy Young Awards to the previous recipient. But the best hitter or best pitcher in the league is usually not a different person every year.

Anybody who wins an MVP or Cy Young was almost definitely both good and lucky. No one is going to argue that it is possible for a bad player to win such an award, but it may be less obvious to some people why award winners are also lucky. First of all, the fact that one year is still a pretty small sample, and that plate appearances or innings pitched must be a finite number for any hitter or pitcher means that there will be some luck in play. However, someone would need to be a tremendous baseball player to outplay everyone else in the league while getting very bad luck himself or even moderately bad luck, because there are plenty of great players who have good luck in any given year.

But just how lucky are the player who played the best and the pitcher who pitched best each year? To move closer to the answer, I looked at all 72 MVPs or Cy Young Award winners during the last 18 years, and examined how well they played in the year before and the year following their award-winning performance. Of course this will not exactly estimate their true talent level, but it is closer to their true talent level than their award-winning performance. I decided to look at WARP, TAv, and PA for hitters, and WARP, ERA, and IP for pitchers, to get a sense of whether players were contributing more, and whether that was because they were playing more or playing better.

Below, I list the MVPs from 1993 through 2010 in each league:

National League

Year	Name	Prev Yr WARP3	WARP3	Next Yr WARP3	Prev Yr TAv	TAv	Next Yr TAv	Prev Yr PA	PA	Next Yr PA
1993	Barry Bonds	12.3	12.4	8.6	.384	.373	.346	612	674	474
1994	Jeff Bagwell	6.2	11.7	5.4	.321	.386	.315	609	479	539
1995	Barry Larkin	6.6	6.8	7.7	.283	.312	.325	501	567	627
1996	Ken Caminiti	7.1	10.4	6.6	.307	.341	.317	602	638	576
1997	Larry Walker	1.8	5.1	4.6	.270	.327	.311	304	664	524
1998	Sammy Sosa	0.8	5.9	5.1	.255	.319	.306	694	722	712
1999	Chipper Jones	6.7	5.9	5.2	.316	.341	.311	707	701	686
2000	Jeff Kent	5.5	9.3	7.6	.295	.336	.303	585	695	696
2001	Barry Bonds	8.4	13.0	12.5	.358	.428	.451	607	664	612
2002	Barry Bonds	13.0	12.5	11.0	.428	.451	.403	664	612	550
2003	Barry Bonds	12.5	11.0	10.8	.451	.403	.438	612	550	617
2004	Barry Bonds	11.0	10.8	0.6	.403	.438	.339	550	617	52
2005	Albert Pujols	8.4	8.5	9.5	.339	.339	.351	692	700	634
2006	Ryan Howard	2.0	7.1	5.1	.297	.340	.317	348	704	648
2007	Jimmy Rollins	4.3	7.1	5.7	.278	.297	.284	758	778	625
2008	Albert Pujols	10.2	11.1	12.0	.335	.367	.361	679	641	700
2009	Albert Pujols	11.1	12.0	8.9	.367	.361	.344	641	700	700
2010	Joey Votto	4.6	7.7		.331	.350		544	648
	Average	7.4	9.4	7.5	.332	.362	.342	595	653	587

American League

Year	Name	Prev Yr WARP3	WARP3	Next Yr WARP3	Prev Yr TAv	TAv	Next Yr TAv	Prev Yr PA	PA	Next Yr PA
1993	Frank Thomas	8.9	7.1	10.4	.358	.346	.381	711	676	517
1994	Frank Thomas	7.1	10.4	7.8	.346	.381	.349	676	517	647
1995	Mo Vaughn	3.7	4.0	4.5	.305	.314	.320	463	636	752
1996	Juan Gonzalez	2.0	2.1	2.2	.287	.301	.290	374	592	574
1997	Ken Griffey Jr.	7.4	8.7	8.1	.320	.325	.313	704	720	706
1998	Juan Gonzalez	2.2	3.2	3.1	.290	.308	.299	579	669	629
1999	Ivan Rodriguez	8.7	7.4	5.8	.286	.284	.311	617	630	389
2000	Jason Giambi	5.1	7.8	8.8	.320	.359	.366	695	664	671
2001	Ichiro Suzuki		6.3	5.0		.294	.284		738	728
2002	Miguel Tejada	3.6	6.0	3.5	.267	.287	.273	683	715	703
2003	Alex Rodriguez	8.3	7.1	5.9	.315	.314	.300	725	715	698
2004	Vladimir Guerrero	4.0	6.9	6.4	.321	.322	.320	467	680	594
2005	Alex Rodriguez	5.9	8.5	4.7	.300	.340	.305	698	715	674
2006	Justin Morneau	0.5	4.4	3.9	.251	.306	.284	543	661	668
2007	Alex Rodriguez	4.7	9.0	7.1	.305	.345	.322	674	704	594
2008	Dustin Pedroia	4.6	8.4	4.7	.283	.299	.281	581	726	714
2009	Joe Mauer	7.4	9.0	6.1	.311	.346	.312	633	606	584
2010	Josh Hamilton	1.0	8.2		.254	.346		365	571
	Average	5.0	6.9	5.8	.301	.323	.312	599	663	638

Josh Hamilton and Joey Votto were very valuable players in 2010, each worth about eight wins, but how many people think that those two players are the best in their respective leagues going forward? The only players who have won consecutive MVPs in the current era are Barry Bonds, Frank Thomas, and Pujols, and chances are that Pujols would be a better bet going forward than Votto, regardless of which one you would rather have had in 2010. Votto is an amazing player who played above his head in 2010. But how far above his head?

Looking at just 1997-2010 winners (to avoid issues related to the 1994 strike), we see that the average MVP in both leagues was worth 8.1 wins, which was 26 percent higher than the 6.1 wins that they put up the previous season and 19 percent higher than the 6.6 wins that were worth the following year. Their TAvs went from .317 to .342 in their MVP season then back down to .327 the following year, while their PA went from 606 to 670 to 620. Thus, the 26 percent increase in production in MVP seasons is partly a result of a 10 percent increase in PA, and the 19 percent production decline the following year is partly a result of the 9 percent decrease in PA.

It seems safe to say that MVPs are probably worth a little over 6-7 wins in terms of skill level, and since MVPs are not necessarily the most skilled players in the league, we can round up and say that the very best players in the league are probably worth seven wins in true skill level, though seasonal MVPs tend to be worth about eight wins in their award-winning seasons. These players tend to get more than four plate appearances per game in their MVP campaigns, but tend to get less than that the year before and the year after.

Moving onto Cy Young winners, there are certainly more repeat winners, but we still see a distinct spike in production and playing time during their winning seasons.

National League

Year	Name	Prev Yr WARP3	WARP3	Next Yr WARP3	Prev Yr ERA	ERA	Next Yr ERA	Prev Yr IP	IP	Next Yr IP
1993	Greg Maddux	9.4	6.4	10.5	2.18	2.36	1.56	268.0	267.0	202.0
1994	Greg Maddux	6.4	10.5	10.8	2.36	1.56	1.63	267.0	202.0	209.7
1995	Greg Maddux	10.5	10.8	7.3	1.56	1.63	2.72	202.0	209.7	245.0
1996	John Smoltz	3.6	6.9	4.9	3.18	2.94	3.02	256.0	253.7	167.7
1997	Pedro Martinez	2.2	8.1	5.4	3.70	1.90	2.89	216.7	241.3	233.7
1998	Tom Glavine	4.8	7.0	3.1	2.96	2.47	4.12	240.0	229.3	234.0
1999	Randy Johnson	6.4	6.3	6.5	3.28	2.48	2.64	244.3	271.7	248.7
2000	Randy Johnson	6.3	6.5	7.0	2.48	2.64	2.49	271.7	248.7	249.7
2001	Randy Johnson	6.5	7.0	8.0	2.64	2.49	2.32	248.7	249.7	260.0
2002	Randy Johnson	7.0	8.0	1.2	2.49	2.32	4.26	249.7	260.0	114.0
2003	Eric Gagne	5.5	8.4	4.2	1.97	1.20	2.19	82.3	82.3	82.3
2004	Roger Clemens	4.3	5.7	7.7	3.91	3.28	1.87	211.7	214.3	211.3
2005	Chris Carpenter	1.7	5.2	5.1	3.46	2.83	3.09	182.0	241.7	221.7
2006	Brandon Webb	5.6	7.2	6.3	3.54	3.10	3.01	208.0	229.0	235.0
2007	Jake Peavy	3.4	8.8	6.3	4.09	2.54	2.85	202.3	223.3	173.7
2008	Tim Lincecum	1.6	7.5	6.8	4.00	2.62	2.48	146.3	227.0	225.3
2009	Tim Lincecum	7.5	6.8	4.6	2.62	2.48	3.43	227.0	225.3	212.3
2010	Roy Halladay	7.8	9.0		2.79	2.44		239.0	250.7
	Average	5.6	7.6	6.2	2.96	2.39	2.74	220.1	229.3	207.4

American League

Year	Name	Prev Yr WARP3	WARP3	Next Yr WARP3	Prev Yr ERA	ERA	Next Yr ERA	Prev Yr IP	IP	Next Yr IP
1993	Jack McDowell	6.1	5.7	3.4	3.18	3.37	3.73	253.7	260.7	256.7
1994	David Cone	5.1	6.2	1.8	3.33	2.94	3.82	254.0	171.7	229.3
1995	Randy Johnson	7.2	8.6	1.3	3.19	2.48	3.67	172.0	214.3	61.3
1996	Pat Hentgen	1.3	6.7	5.4	5.11	3.22	3.68	200.7	265.7	264.0
1997	Roger Clemens	7.0	11.3	7.3	3.63	2.05	2.65	242.7	264.0	234.7
1998	Roger Clemens	11.3	7.3	1.4	2.05	2.65	4.60	264.0	234.7	187.7
1999	Pedro Martinez	5.4	8.4	10.1	2.89	2.07	1.74	233.7	213.3	217.0
2000	Pedro Martinez	8.4	10.1	4.0	2.07	1.74	2.39	213.3	217.0	116.7
2001	Roger Clemens	3.5	5.0	1.9	3.70	3.51	4.35	204.3	220.3	180.0
2002	Barry Zito	3.1	5.8	3.5	3.49	2.75	3.30	214.3	229.3	231.7
2003	Roy Halladay	6.2	6.9	2.0	2.93	3.25	4.20	239.3	266.0	133.0
2004	Johan Santana	5.9	6.5	5.6	2.87	2.77	3.33	231.7	233.7	219.0
2005	Bartolo Colon	0.5	3.3	-0.5	5.01	3.48	5.11	208.3	222.7	56.3
2006	Johan Santana	5.9	6.5	5.6	2.87	2.77	3.33	231.7	233.7	219.0
2007	CC Sabathia	3.8	5.6	7.3	3.22	3.21	2.70	192.7	241.0	253.0
2008	Cliff Lee	-1.2	7.0	5.0	6.29	2.54	3.14	97.3	223.3	231.7
2009	Zack Greinke	5.2	9.9	3.6	3.47	2.16	4.17	202.3	229.3	220.0
2010	Felix Hernandez	6.1	7.9		2.49	2.27		238.7	249.7
	Average	4.9	7.3	4.1	3.44	2.73	3.50	212.3	232.5	195.5

Roger Clemens, Randy Johnson, Pedro Martinez, Greg Maddux, and Lincecum won consecutive Cy Youngs, but Roy Halladay and Johan Santana both won two non-consecutive awards. Winning the Cy Young is hard, but the best pitchers seem to be able to do it more easily than the best hitters seem to be able to win MVPs.

Again looking at only 1997-2010 (so that we can focus on players who had the potential to participate in 162-game seasons the year before and the year after their awards), the average Cy Young winner across both leagues jumped from 5.0 WARP the year before their award to 7.3 WARP the year of to 5.0 WARP the year after. Their ERAs went from 3.25 to 2.56 to 3.16, while their innings pitched went from 211.3 to 231.0 to 200.5.

Unlike the MVPs who performed better the year after their award than they did the year before their awards, the Cy Young winners had the same overall production on average. In fact, they had ERAs that were lower the year after their award, but they had innings pitched numbers that were lower, too. This is probably due to health consequences of pitching more during their Cy Young seasons, since they failed to pitch as many innings but still pitched better. I believe this indicates that the Cy Young winners’ true skill level was a little higher during their award-winning season than the following season, because we know that the ability to stay on the mound is a skill. This means that 5.0 WARP is lower than their true skill level by a small margin.

This suggests that the skill level of pitchers during their Cy Young season was around six wins (again rounding up from 5-6 wins of skill level for Cy Young winners, because the Cy Young winner might not be the best pitcher), despite the fact that Cy Young winners’ performances are worth 7-8 wins. Again, this shows how much luck can factor into pitching. However, the repeat winners also suggest that the best pitcher is often so much better than the second-best pitcher in the league that he can still win despite not having amazing luck.

Back in June, when I wrote an article explaining the luck that had played into the 1.16 ERA that Ubaldo Jimenez had through his first 13 starts, I received a tweet directed at me that read: “Magic seasons happen. Deal with it.” While the tweet was erased soon after, and the ERA was erased with a 4.14 performance by Jimenez during the 20 starts he had after my article was posted, it should be very clear that magic half-seasons happen. Of course, they are not really magic half-seasons, but they are lucky half-seasons that happen to very talented players. To a lesser extent, luck exists in full seasons as well, meaning that “magic” seasons do happen; the tweeter was right. This small study should be a good clue as to just how magical they are. The answer is that there are about 1-2 wins of luck for MVPs and about 2-3 wins of luck for Cy Young winners that factor into the average winner’s award-winning season.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now

Matt Swartz

Latest Articles

You need to be logged in to comment. Login or Subscribe

jessehoffins

12/01

This is really cool, suprsing how more reproducible the pitchers results were. It'd be cool to compare mvps scores under nate silver's old "best player in baseball" metric and compare that to who nate's system would have picked.

Reply to jessehoffins

BillJohnson

12/01

Curiously missing from this article is any mention of the word "team." Certainly for the MVP, and probably also for the CY, one of the bits of "luck" possessed by most winners is that they played on successful teams -- teams that, in many ways, rode their own waves of luck to an extent even greater than the one that carried the MVP himself. Any chance of expanding this article to look at that?

Reply to BillJohnson

swartzm

12/01

I'm not sure I understand what I would change. Is your belief that the luck is correlated between players on a team? If so, I wouldn't call that luck. Maybe you mean that the things that tend to cause unsustainable jumps in performance are often common among players on a team? Maybe so, but it would be tough to quantify that I think, but I'm open to suggestions. Thanks.

Reply to swartzm

BurrRutledge

12/01

I think he's simply saying that the voters take into account the success of the MVP/Cy Young candidates. Unclear whether this would actually impact any of your findings.

You could validate by studying the award winner and separately analyze the runner up. Is there more variance in performance for one over the other?

Reply to BurrRutledge

BillJohnson

12/01

Well, to pick an obvious example: Just how lucky were the 2010 Cincinnati Reds? Because the fact that they made the playoffs, while St. Louis did not, is the main reason why Joey Votto won the MVP in a landslide over a guy (we all know who that is, right?) who had an otherwise very comparable, maybe slightly superior, year. Obviously, we won't know until after 2011 just how much of the Reds' success this year was because of luck (Pythagenpat says something about that, but is hardly the whole story), and how much was a real improvement by the Reds that is sustainable. For similar cases in the past, however, we should be able to answer that question. Try something like this: for two players with comparable, MVP-worthy years (or at least years worthy of MVP consideration), see whether there's any correlation between the votes they received and the fraction of their teammates playing well above their 50th-percentile projections. For those same two guys, repeat the exercise the following year, if necessary finding other peers to compare their performance to. Repeat.

Luck certainly *is* correlated between members of a lucky team, if such a thing exists (and I think the evidence is clear that it does). Correlation does not imply causality.

Reply to BillJohnson

swartzm

12/01

Okay, so firstly, I'm 100% sure that you are right that the better the performance of a player's teammates, the more likely a similar performance will lead to an MVP Award. I'm also 100% sure that the luckier your teammates were, the better there numbers were. However, I'm not asking that question. What we do know is that luck between teammates should not be correlated at all. If we define luck as we generally do-- randomness-- then we should not expect a correlation between teammates' luck. If Albert Pujols flips a coin that comes up heads and then gives the coin to Matt Holliday, his coin should still have a 50/50 shot of coming up heads. If Pujols' true talent level is a .400 OBP guy and Holliday is a .380 OBP guy, then Pujols getting on base more than 40% of the time should not affect Holliday's likelihood of getting on base more than 38% of the time (excluding things like IBBs).

Reply to swartzm

BurrRutledge

12/01

"Luck certainly *is* correlated between members of a lucky team."

Are you simply saying that a) teams have better W-L records when more of their players outperform their projections?

Or are you saying that b) lucky players who are good enough for MVP consideration are more likely to win the MVP because they had more lucky teammates, thereby getting the team into the playoffs?

Reply to BurrRutledge

BillJohnson

12/02

A is certainly true, and obvious. The question is: is B? Substituting "if" for "because" in your otherwise pithy summary.

Reply to BillJohnson

BurrRutledge

12/02

Fair 'nuff.

Reply to BurrRutledge

dkeenan80

12/01

The 2004 stats for Santana are the same as his 2006 stats. I think the numbers listed are his 2006 stats. Don't think this affects your overall conclusion however. Good article.

Reply to dkeenan80

swartzm

12/01

D'oh. Fortunately, that mistake was made copying from Excel into Word, so the averages were done using Santana's correct statistics. His 2003-05 numbers were: 3.4, 8.6, 5.9 WARP; 3.07, 2.61, 2.87 ERA; 110.3, 228, 231.7 IP. Thanks.

Reply to swartzm

iddscoper

12/01

I would be curious to see the difference in the MVP analysis if you exclude Bonds from '01-'04. He was just playing on a completely different level, and I think that has to skew the results a bit, maybe enough to account for the difference in MVP and CY luck.

Reply to iddscoper

swartzm

12/01

I'm actually inclined to include Bonds, because it's not that surprising that one player every couple decades will exceed their competition by such a large margin. I think you want to include outliers here. Though for the curious, the WARP change would now be from 5.0, 7.5, 6.1 WARP, and .299, .327, .310 TAv. So it would still be about the same with a little more of a pre-MVP to MVP jump and a softer MVP to post-MVP decline. Bonds' 2005 injuries cause a 10.2-win drop in production that was atypical but his strong 2000 performance made the average improvement in MVP season look smaller. Still, I think we should include Bonds. Sometimes MVPs are won by Ruths and Bonds and sometimes by Vottos and Hamiltons.

Reply to swartzm

cdmyers

12/01

Another factor is that there are simply more batters vying for the MVP than there are pitchers. If we assume that only the starting rotation and the 8 starting position players from each team are realistically able to compete for the award, that's almost half again as many position players. I'd think that that greatly increases the chance that someone strikes gold and wins off of a very lucky season. This in turn would make repeat wins harder, even if amazing hitting seasons are no less repeatable than amazing pitching seasons.

Reply to cdmyers

ScottBehson

12/02

Most WS winning teams are both REALLY GOOD and a bit lucky.

It seems as if this pattern holds or most MVPs and CYs, too

Reply to ScottBehson

Ahead in the Count: So How Good are MVPs Really?

Thank you for reading

Latest Articles

A Candle for Rickey

Early PECOTA is Available Now!

Earworms for Drafting $

The Top-500 Dynasty Prospects, December 2025 $

The Grim World of Children Asking Santa for Baseballs $

Matt Swartz

Latest Articles

A Candle for Rickey

Early PECOTA is Available Now!

Earworms for Drafting $