This is Part II of Boy of Summer's MVP Thoughts. You can read Part I here.
As I mentioned, some sportswriters are really not happy about this, perhaps most notably, Jayson Stark. As Rob Neyer confirms, Stark is a prince of a guy, so I'll be as kind as possible. However, it should be noted again that I think he's dead wrong on this issue, and that I think he's not much of a statistician, when you get right down to it. Most baseball writers would take that as a complement.
Examining Jayson Stark's MVP criteria is not as easy as it sounds, becasue he seems to keep grasping at different things, anything, to get people to believe him instead of the Rules.
If you've read Stark's columns for any length of time, you've learned that he's a pretty good journalist. He gets good stories, and he writes them well. He's interesting and creative and seems like a decent fellow. But he seems to have something less than a firm grasp on how statistical analysis ought to be utilized. One of the worst things you can do with stats, and one of the things that Stark does fairly often, is to pick an arbitrary number that seems to support your point, and don't bother to give any other information that might make your point look anything less than salient.
In an article he wrote in September about why A-Rod shouldn't be the MVP, he said, "Since 1994, when baseball broke into six divisions and created twice as many pennant races, no player from a losing team has finished within 100 points of the winner. That includes A-Rod last year. And it undermines his candidacy this year."
OK, he picked 1994 because of the break between the two and three division formats, meaning that there are more contenders and therefore (theoretically) more potential MVPs. But the 100 points could have been anything. It sounds like a nice, round number, but it's arbitrary.
When players have MVP voting clauses written into their contracts, they're written by placement, not point totals. And while what Stark says is true, it's misleading. Players from losing teams have been as high as second or third in the voting on several occasions in that timespan. A-Rod was second in 2002, Griffey was second in 1994, and Frank Thomas finished third to Griffey in 1997, all playing on losing teams. Players from non-playoff teams have been up pretty high in the tally as well, including Carlos Delgado (4th in 2000), Griffey (4th in 1998), and Mo Vaughn (5th in 1996).
That same article made three "key" points in the argument against Rodriguez's candidacy for the AL MVP:
1) WHAT PART OF "VALUABLE" DON'T YOU UNDERSTAND? - Stark argues that the BBWAA has almost always chosen a player from a winning team for the MVP, so why should we change that tradition now?
2) WHERE WAS HE WHEN THEY NEEDED HIM? Stark argues that A-Rod faded when his team sank from contention in June, and that he didn't consistently produce the way an MVP should.
3) THIS ISN'T 1991 - Stark argues that an MVP should only come from a losing team if there are no true/close pennant races, and therefore no clear player who makes the differene between his team making the playoffs or not.
Let's take these on one at a time...
1) WHAT PART OF "VALUABLE" DON'T YOU UNDERSTAND? The meaning of the word "valuable" has nothing to do with the context of the object in question. The Hope Diamond would be no less valuable in the bottom of a landfill in Staten Island than it is in the Smithsonian, it just wouldn't be as useful. Once the diamond's recovered, the landfill would admittedly be somewhat less valuable without it.
The fact that baseball writers have traditionally used winners, especially playoff contenders, to determine the MVP does NOT mean that it should be this way. History held many 'traditions' to be perfectly acceptable right up until someone showed that there was a better way. Ever read a short story called 'The Lottery'?
2) WHERE WAS HE WHEN THEY NEEDED HIM? This has to be Stark's weakest argument. He cites the following as evidence that Rodriguez was not the MVP:
So as the ship sank, he sank with it. It really wasn't until his team's season was essentially over that he began compiling many of these alleged MVP numbers.
He hit the second-most home runs in the league in April (8). But he was 10th in May (6), 20th in June (5) and 21st in July (5), before leading the league in August (15) and September (7).[Ed. NOTE: A-Rod was 5th in the AL in Sept.]
He was fifth in the AL in RBI in April (22). Then he fell to 44th in May (13), 34th in June (15) and 17th in July, before leading the league in August (31) and ranking third in September (15).
OK, how about slugging percentage -- a stat that doesn't depend on the contributions of anyone around him? He was fourth in the league in April (.667). But then he was 54th in May (.462), 31st in June (.540) and 37th in July (.505), before an .849 August (first) and .594 September (fifth).
Does that sound like an MVP season to you? It doesn't to us.
Well, Jayson, maybe it does and maybe it doesn't but since you don't give us anyone else's numbers to compare with A-Rod's, how are we to know whether it does or not? He could have picked a lot of other stats (runs scored, on-base%, etc.) but I suspect that he chose these because they supported his point. I'm going to use OPS (on-base% + Slugging%) to try to rebut his point, because it's a pretty good rough measure of a player's offensive contributions, apart from his teammates contributions, and also because I didn't feel like charting all of the possible stats.
Here are the monthly ranks for for each of the top ten MVP points-getters:
April May June July Aug Sept Avg (T) Avg (5) Season
Delgado 1 5 14 12 16 10 9.7 8.4 1
Ramirez 16 40 6 4 20 2 14.7 9.6 2
A-Rod 3 49 22 26 1 13 19.0 13.0 3
D. Ortiz * 45 18 11 3 5 16.4 16.4 5
Posada 18 36 40 25 11 6 22.7 19.2 10
Beltran ** 16 43 17 9 14 19.8 19.8 12
V. Wells 58 19 9 23 31 12 25.3 18.8 13
B. Boone 23 10 17 18 68 25 26.8 18.6 15
No-mah 49 11 10 44 28 92 39.0 28.4 20
Stewart 54 38 *** 24 44 51 42.2 42.2 32
First, a few explanations:
Only the top 100 players, in terms of at-bats, were rated each month. I had to cut it off somewhere.
*David Ortiz wasn't the everyday DH in Boston until June, but he amassed essentially one month worth of at-bats between April and May, which got him an OPS around .800, which would rank him around 45th in the league in any given month, hence the 45 in May and nothing in April.
**Carlos Beltran was injured the first month of the season and only got 38 at-bats in April.
***Shannon Stewart got only 33 at-bats in June.
The Avg(Tot) column is the average rank of the totals for all six months of the season. The Avg(5) column is the average of the player's five best months, which helps to normalize for players who didn't get enough at-bats in a particular month to qualify due to injury (Beltran and Stewart) or platooning (Ortiz). It also gives some grace to players who may have had one really off month that skewed their average. The Season column is the overall rank of that player for (you guessed it) the season.
OK, with that said, what can we learn from this?
A) Carlos Delgado is the only player who meets Stark's criteria of "consistently" producing. He's the only player in the top 20 in every month, never straying higher than 16th (Aug), and therefore ranked 1st overall at the end of the year. However, based on Stark's insistence upon players from winners and playoff contenders, Delgado's no good either. The standard deviation in his rank (which I didn't post here) was two and a half times smaller than anyone else's on the list.
B) Manny Ramirez, A-Rod, Vernon Wells and Bret Boone are the only other players who rank in or near the top 30 in the AL five out of six months. Other players either didn't qualify for a month or had more than one month in which they ranked over 30th.
C) Nomar Garciaparra was particularly flaky, ranking as high as 11th in May and as low as 92nd(!) in September. Ouch. Still, he managed to finish a solid 20th at the end of the year.
D) Jayson Stark's pet candidate, Shannon Stewart, was so valuable that he never ranked higher than 24th, and easily averaged the lowest among the lot. Even when he was supposedly somehow turning the Twins around and leading them into the postseason, he ranked as the 44th and 51st best hitter in the AL down the stretch. And Stark wants you to vote for him?!?
Regarding the fate of the Rangers with relation to Rodriguez's play, the fact that the team was as close to the .500 mark as they were (25-27 on May 29) was essentially a fluke, a mirage created by the ability of the early season to skew our views of reality (remember when we thought the Royals were contenders?).
The team's Pythagorean record at the end of May (the wins and losses you'd expect them to have based on the total runs scored and allowed) was a couple of games worse than what they actually had, which means that they'd been lucky to do as well as they did to that point. Certainly, there's no denying that June was Rodriguez's (and the Rangers') worst month, but there aren't many players who avoid a 3-week swoon all season, and the man couldn't do anything about the Rangers' pitching staff's 6.63 June ERA. And besides, A-Rod's worst month (49th in OPS) is about as good as Shannon Stewart's average month (42).
3) THIS ISN'T 1991 Stark contended that MVPs only come from losing teams whn there are no pennant races. This simply isn't true. He uses the AL's only example (Cal Ripken in 1991) because it supports his point. But if you dig just a little deeper, and choose another league to analyze, I don't know, let's say...the National League, you find that this isn't the case at all. Andre Dawson won the NL MVP in 1987 on a Cubs team that finished last, even though both the Mets and the Expos finished within four games of fisrt place in the NL East.
Also, in 1959, when Ernie Banks won the NL MVP on a losing Cubs team, both the Braves and Giants finished within four games of the Dodgers, who won their first pennant in LA. So we see that the Senior Circuit doesn't necessarily discriminate against a truly great and valuable player just because there are pennant races and good players on the teams vying for them.
Stark's post-mortem article on the MVP debate revisits the issue of tradition, which we've already addressed, so I won't get into that again. However, he also asks "Where would the team have finished without him?" and answers it accurately: Last. Again, though, his answer is uninforming and misleading.
In any other division in all of MLB, the Rangers (71-91) would not have been last. They were, admittedly, the worst team in a relatively strong division, but taking A-Rod away and replacing him with Joe Average Shortstop costs the Rangers something like six to ten wins in the standings, which is a lot. I'm sure that as bad as the Rangers were last year (and the year before that, and...) a lot of Texans are really glad that they didn't lose 100 games in 2003, and A-Rod is a major reason for that.
If that's not Value, then value doesn't exist.