Designated HitterApril 23, 2009
WAR and Remembrance
By John Walsh

Baseball fans love to argue. Did Dustin Pedroia really deserve the MVP award last year? (After all, he was only 18th in the AL in OPS.) Sure, Manny can hit (can he ever!), but he gives it all back with the glove, right? On the flip side, is Adam Everett, with his fabulous defense, a valuable player? We older folks like to argue about the players of our youth: For example, who had the better career, George Brett or Wade Boggs? In the end, it usually comes down to putting a value on a player, a total value that includes hitting, defense, baserunning and everything else.

Well, Sean Smith -- you know, the guy who does the CHONE player projections -- is putting an end to some of these arguments. What Sean has done, bless his soul, is evaluate players on just about every aspect in which a player contributes to winning. And he's done this for all players going all the way back to the middle of the last century. Bravo, Sean!

So, what are these different aspects of baseball, the important contributions a player can make towards winning? Here's the list:

o batting
o baserunning
o avoidance of grounding into double plays
o defensive range
o catcher defense
o defensive arm for outfielders
o double-play proficiency for infielders

Sean has analyzed over 50 seasons of play-by-play data available at Retrosheet and determined each player's value in the above categories, expressed in runs above or below that of an average player. For the defensive categories, players are compared to the average for that position. I won't go into the methodology for all these categories, you can refer to Sean's explanations here. I do want to mention Sean's Total Zone system, which he uses to measure defensive range. After hitting, defensive range (and catcher defense) is the biggest contribution to a player's value. Total Zone uses Retrosheet play-by-play data to evaluate defensive range for all players of the last 55 years or so. It's a clever system that squeezes just about every bit of information from the play-by-play data, data that is not as complete as modern play-by-play data from professional statistics providers like Baseball Info Solutions or STATS, Inc. See here for more details on Total Zone.

Of, course there's a lot more here than just defense, as you can see in the list above. Now, we've known how to measure baserunning and outfield arm proficiency for a while and the other categories, given the Retrosheet data are treated in a similarl way. The important thing that Sean has done is to 1) put in the dirty work to make all these different evaluations and 2) put them altogether to allow us to get a total picture of player value. Oh, and 3) he's posted it all on the web for all to use (at no charge).

Do you realize how great this all is? I recently wrote an article for the Hardball Times that did an in-depth comparison of Carl Yastrzemski and Manny Ramirez. I got the hitting from baseball-reference.com, defensive range from Sean's own Total Zone system and the outfield arm ratings came from my own work at THT. I couldn't locate comprehensive baserunning information, so I had to work that out (a less complete analysis) on my own. Now, to write that article, I would could do all my "shopping" at Baseball Projection.

Sean then goes a couple of steps further with the data he has compiled. He translates "runs above average" to "runs above replacement", since a player's true value is best measured against a replacement level player. Along the way he gives each player a "position adjustment". Remember when I wrote that range is measured against the average defender at the same position? Well, the position adjustment accounts for the fact that the value of an average fielder is not the same for each position.

The last step is translating runs into wins and, since we are now relative to replacement, these are Wins Above Replacement, or WAR. I've been very brief in describing the system, if you want more info about determining overall player value, I heartily recommend a series of posts at FanGraphs, which goes through the process step-by-step, starting here.

Speaking of FanGraphs, those good folks have been doing similar work. They also produce WAR values for all players, using a different fielding system (known as UZR) and play-by-play data purchased from Baseball Info Solutions. Their data set goes back only a few years, though, so you need to use Sean's WAR database, if you want to look at, I dunno, who really should have won the MVP awards in 1974...

-------------------------------------------------------------

Jeff Burroughs is the guy who, when reciting the names of MVP winners, you always leave off the list. Well, him and Zoilo Versalles, I guess.* It's not that he was underserving of the award, although, he was, as we shall see shortly. It's just that looking back, he doesn't seem like much of a star. He actually was a very good hitter for a few seasons and I'm sure he's not the MVP-winner with the worst career.

*What? You mean, you don't find yourself reciting the names of AL MVP winners? That's strange, I do it all the time. Pennant winners and World Series champs, too. Just don't ask me who the 13th President of the United States was.

Jeff Burroughs in 1974 was probably the best hitter in the American League. The 23-year-old Texas Ranger hit .301/.397/.504, which is even better than it looks, since offensive levels were quite a bit lower 35 years ago. Burroughs finished third in on-base average and slugging percentage and finished among the top ten in just about every important offensive category. He only led the league in one category, but it was the right one for garnering MVP votes: RBI.

We can get an overall measure of Burroughs' hitting by considering the Batting Runs part of the WAR database. Here are the AL leaders for 1974:

 ------------------ ------ --------- 
| Name             | Team | BatRuns |
 ------------------ ------ --------- 
| Jackson_Reggie   | OAK  |      49 | 
| Burroughs_Jeff   | TEX  |      48 | 
| Carew_Rod        | MIN  |      35 | 
| Allen_Dick       | CHA  |      34 | 
| Rudi_Joe         | OAK  |      34 | 
| Yastrzemski_Carl | BOS  |      33 | 
| Bando_Sal        | OAK  |      27 | 
| Tenace_Gene      | OAK  |      27 | 
| Gamble_Oscar     | CLE  |      27 | 
| Grich_Bobby      | BAL  |      27 | 
 ------------------ ------ --------- 

Burroughs is right there with Reggie Jackson at the top of the list. Jackson finished fourth in the MVP balloting, which may be explained by Burroughs' advantage in RBI, 118 to 93. In any case, from a hitting standpoint, Burroughs was certainly not a bad choice for MVP.

But, baseball is more than hitting, of course — how did Burroughs do in the non-hitting categories? Burroughs was not a fast player, at all, so we don't expect him to excel at baserunning, defensive range and avoiding the GDP. But did he at least hold his own? Did the 1974 American League MVP at least approach the average players in the "extra" categories? I'm sorry to report that he did not.

Here's how Burroughs fared in the non-hitting categories:

o Defensive range - Burroughs was 17 runs worse than an average right-fielder. That's the worst range mark of any AL player in 1974.

o Outfield arm - sometimes slow guys have good arms. Not in this case. Burroughs cost his team an additional five runs with an ineffectual throwing arm.

o Baserunning - Two stolen bases and three caught stealings give you an idea of Burroughs' speed. He was also below average in advancing on the basepaths, giving him a net baserunning value of -3 runs.

o GDP - Burroughs grounded into 17 double plays in 1974, a few more than the average batter would have, given the same opportunities. Good for -2 runs.

o Position - it's not his fault, of course, but Burroughs played right field in his MVP year, which is an offense-first position. The adjustment for right fielders is -8 runs.

The 1974 AL MVP was below average in every single non-hitting category for a grand total of -35 runs. Yikes, that negates a good chunk of his batting runs (which was +48, you'll recall). In fact, without considering hitting, Burroughs was the very worst player in all of baseball in 1974 and he was one of only four players who was below average in each of the non-hitting categories. This dude was seriously one-dimensional.

So, who should have won that 1974 AL MVP? Well, if you don't require your MVP to play on a playoff team (Burroughs's Rangers did not make the playoffs), then you could rank MVP candidates according to their overall win value, or WAR:

 ----------------- ------ --------- ------- ----- ------ ------ ---------- ------ 
| Name            | Team | Batting | Range | Arm | BsRn | GIDP | Position | WAR  |
 ----------------- ------ --------- ------- ----- ------ ------ ---------- ------ 
| Grich_Bobby     | BAL  |      27 |     5 |   3 |    5 |   -2 |        4 |  6.9 | 
| Jackson_Reggie  | OAK  |      49 |     0 |  -2 |    0 |    2 |       -8 |  6.7 | 
| Carew_Rod       | MIN  |      35 |    -9 |   2 |    5 |    2 |        4 |  6.6 | 
| Rudi_Joe        | OAK  |      34 |     0 |   3 |    1 |    1 |       -8 |  5.6 | 
| Campaneris_Bert | OAK  |      13 |     6 |   1 |    4 |    1 |        8 |  5.4 | 
| Money_Don       | MIL  |      19 |     0 |   2 |    3 |    0 |        4 |  5.4 | 
| Maddox_Elliott  | NYA  |      19 |     4 |   6 |    4 |   -1 |       -2 |  5.1 | 
| Bando_Sal       | OAK  |      27 |    -4 |   0 |    1 |    0 |        3 |  5.0 | 
| Tenace_Gene     | OAK  |      27 |     4 |   0 |   -5 |   -1 |       -2 |  4.6 | 
| Robinson_Brooks | BAL  |       5 |    14 |   1 |    0 |   -1 |        4 |  4.4 | 
 ----------------- ------ --------- ------- ----- ------ ------ ---------- ------ 
BsRn - baserunning runs
Range - includes catcher defense 
Arm  - includes infield DP rating

For me, it comes down to Bobby Grich, Jackson and Rod Carew. Pay no attention to the 0.3 wins separating these three — no system is accurate enough to distinguish players this close. Grich played a prime defensive position and played it exceptionally well. He won a Gold Glove at second base in '74, and was excellent with the bat and on the basepaths. Reggie, we already saw, was one of the top two hitters in the league, and he hangs on to those batting runs by coming out average in the other categories (except for position adjustment). Carew was top notch in everything except defensive range (he was still playing second base at this point).

In the actual vote, Grich finished ninth and Carew seventh. You might notice the absence of somebody from the above list: Jeff Burroughs, who totaled 4.0 wins over replacement for the season.

-------------------------------------------------------------------

Over in the National League, the voters did not fare much better: they elected Dodger first basement Steve Garvey over several more valuable players. The problem in this case was not neglecting the other categories (although I suspect many writers did so), but rather not doing a good job of evaluating offensive value.

Sean Smith's WAR database rates Garvey as the NL's ninth most productive hitter in 1974:

 --------------------- ------ --------- 
| Name                | Team | Batting |
 --------------------- ------ --------- 
| Schmidt_Mike        | PHI  |      49 | 
| Wynn_Jimmy          | LAN  |      47 | 
| Morgan_Joe          | CIN  |      46 | 
| Stargell_Willie     | PIT  |      46 | 
| Smith_Reggie        | SLN  |      40 | 
| Zisk_Richie         | PIT  |      33 | 
| Bench_Johnny        | CIN  |      32 | 
| Garr_Ralph          | ATL  |      31 | 
| Garvey_Steve        | LAN  |      29 | 
| McCovey_Willie      | SDN  |      28 | 
 --------------------- ------ --------- 

Why did the voters elect Garvey over these other superior hitters? Well, some of these guys were on non-contending teams, including Mike Schmidt, but that doesn't explain why Garvey's teammate Jimmy Wynn finished fifth in the voting (not to mention the Pirates, Reds and Cardinals in the above list).

Garvey batted .312/.342/.469 on the year, with 21 homers and 111 runs driven home. He did not lead the league in any category, though he was Top 10 in several. Here's my take on how he won the MVP: he batted over .300, knocked out 200 hits and had the highest RBI total of players on an NL playoff team (the other being the Pirates). That and the great hair, of course.

Did Garvey do anything in the non-hitting categories to boost his case and vault him over the better hitters in 1974? No, not really. Here are the numbers:

 --------------------- ------ --------- ------- ----- ------ ------ ---------- ------ 
| Name                | Team | Batting | Range | Arm | BsRn | GIDP | Position | WAR  |
 --------------------- ------ --------- ------- ----- ------ ------ ---------- ------ 
| Garvey_Steve        | LAN  |      29 |     0 |   0 |    3 |    2 |      -10 |  4.8 | 
 --------------------- ------ --------- ------- ----- ------ ------ ---------- ------ 

I don't think of Garvey as a speedster, but he was above average in the speed categories of baserunning and avoiding double plays. He was average in defensive range and arm (although he was famous for having a very weak arm), but he takes a -10 run hit for playing first base. An overall WAR value of 5 is nothing to be ashamed of, but Garvey was not among the ten most valuable National League players in 1974:

 --------------------- ------ --------- ------- ----- ------ ------ ---------- ------ 
| Name                | Team | Batting | Range | Arm | BsRn | GIDP | Position | WAR  |
 --------------------- ------ --------- ------- ----- ------ ------ ---------- ------ 
| Schmidt_Mike        | PHI  |      49 |    17 |   1 |    1 |    2 |        4 | 10.0 | 
| Morgan_Joe          | CIN  |      46 |     3 |   1 |    8 |    1 |        4 |  8.8 | 
| Wynn_Jimmy          | LAN  |      47 |    12 |   2 |   -1 |    2 |       -2 |  8.4 | 
| Bench_Johnny        | CIN  |      32 |    11 |  -1 |   -1 |    0 |        9 |  7.5 | 
| Evans_Darrell       | ATL  |      18 |    18 |   2 |    2 |    1 |        4 |  6.8 | 
| Stargell_Willie     | PIT  |      46 |     1 |   1 |   -2 |    0 |       -7 |  6.2 | 
| Rose_Pete           | CIN  |      18 |    15 |   5 |    4 |    1 |       -9 |  6.0 | 
| Smith_Reggie        | SLN  |      40 |     8 |   0 |   -2 |   -3 |       -7 |  5.7 | 
| Cedeno_Cesar        | HOU  |      20 |     4 |   4 |    7 |    1 |       -2 |  5.7 | 
| Oliver_Al           | PIT  |      28 |     6 |  -4 |    6 |   -3 |       -4 |  5.2 | 
 --------------------- ------ --------- ------- ----- ------ ------ ---------- ------ 

Wow, look at the fabulous season that Mike Schmidt had. Best hitter in the league, one of the best defensive players and above average in all the other categories. Achieving a WAR of 10 is no small feat: it has only been done 36 times since 1955.

The fantastic thing about having this WAR database (did I thank Sean for this yet?) is it makes clear just how some very good players end up getting underrated, because a lot of their value comes in the non-hitting categories. Jimmy Wynn, Darrell Evans and arguably Cesar Cedeno fall into this group. Wow, just noticed that Pete Rose had a great year with the glove in 1974.

In case you were wondering, Steve Garvey ranked 14th in WAR in the NL in 1974.

-------------------------------------------------------------------

So, I hope I have given you a flavor for just how useful Sean's WAR database really is. You could use it to answer many, many questions, of course. Which players are underrated because much of their value is in the non-hitting categories? Which players were the most well-rounded or one-dimensional? Who had value because of speed and who despite of a lack of it? Or let's talk about teams: The 1985 Cardinals stole 314 bases — how much impact did their baserunning have on their offense? Were they the best baserunning team of the last half-century? Who were the best defensive teams and the worst?

Oh, the mind reels at the possibilities. All the numbers are there, waiting to be looked at. Thank you, Sean.

John Walsh is a regular contributor to the Hardball Times. He welcomes comments via email.

Comments

I have to take the defensive value adjustments for first base.

A first basemen with a good glove can save more errors and runs than a shortstop with great range can as the first basemen are taking balls thrown to them from 5 positions.

They are involved in a huge percentage of the plays in any game and if they are not able to pick the ball out of the dirt or take it on a bounce or stretch to get that ball that is offline, then the rest of your team is going to be getting errors and the opposing teams will have many more chances to score runs.

The first basemen don't get many errors by and large because they are throwing the ball to other bases less than other positions, but they do affect the errors other infielders have.

The difficulty comes in measuring how this glove skill or lack thereof effects the game.

It would be interesting to see how TEAM defense was affected when someone like Wally Joyner, who was known for having a great glove at 1B, was out or was replaced.

websoulsurfer,

You are right that a good first baseman can help his team by scooping up bad throws. However, you overestimate the impact of a good scooper. Check out MGL's research on this exact topic at fangraphs.com.

MGL concludes:

So before you start opining about how your favorite first baseman is so great defensively because he “saves so many errors,” consider that scooping ability is probably worth less than a ¼ of total defensive ability or value at first base. Fielding grounders is at least 75% of the package and “scooping” is the rest. But every little bit helps.

I looked at the top 300 listing on retrosheet and was disappointed that only years after 1955 were used so there was no way to see where the Babe stood and where one of my favs, Stan the Man would rank since he only played 8 yrs after 1955 and those were decline years mostly, would love to see something about pre 1955, Mays/Mantle/Williams also get shortchanged a bit.

bill,

Yes, the play-by-play data collected by retrosheet only goes back around 55 years. This WAR analysis is dependent on play-by-play data, hence the impossibility of evaluating the guys that played before then.

A lot of Mays' and Mantle's careers are covered, though, and in fact, I was thinking about looking at those two more closely.