Thursday, October 13, 2005

Worst award selections

Another long-range project of mine involves working my way back through history analyzing and recording each season's individual achievements. Having now made it back through 1985, I thought I'd offer up my opinion on the worst award selections of the past 20 years (keeping in mind that a few might be made in this seasons awards):
Honorable Mention: Larry Walker (1997 NL MVP), Miguel Tejada (2002 AL MVP), Dennis Eckersley (1992 AL Cy Young), Kerry Wood (1998 NL Rookie of the Year)

#10: Mark Davis, 1989 NL Cy Young
Baseball people just don't really know how to measure a relief pitcher against starting pitchers. I will grant you that Davis was the best closer in the NL. He had a 1.85 ERA with 44 saves and 92 K in 92.2 IP. He won the award because, as often happens, sportswriters gradually convinced themselves that Davis's season was "special" and that he was carrying the NL Champion Giants.
Davis' season was darn good for a reliever, but I think he got the award because the voters didn't see a viable starting pitcher as a candidate. Mike Scott led the league with 20 wins (and 10 losses), but his 3.10 ERA wasn't near the top ten, and the Astros finished 3rd. Scott finished 2nd in the voting. Greg Maddux came in 3rd, which wasn't such a bad choice. Maddux went 19-12 with a 2.95 ERA for the division-winning Cubs. He's actually a better candidate than Davis.
But the best NL pitcher in 1989 was, in my opinion, Orel Hershiser. Hershiser's 2.31 ERA was 2nd in the league. He threw 256.2 IP, allowing only 9 HR and 77 BB against 178 K. Not only that, but Hershiser threw 256.2 IP, which completely obliterates Davis' advantage in ERA, as far as I see it.
But Hershiser didn't win because his record was 15-15. Even today, in the era of Moneyball, a pitcher couldn't win the Cy Young with a 15-15 record. So it was laughable to suggest such a thing in 1989. Hershiser got exactly 7 votes (compared to 107 for Davis).
#9: Ivan Rodriguez, 1999 AL MVP
This was another case of the voters convincing themselves (in spite of any actual evidence) that a player's contribution went above and beyond. Also, given that Pudge hadn't gotten an MVP award yet, he was due. He "led" the Rangers to an AL West title with a 356/558/332 line. This, of course, was good but certainly not great for a hitter in the AL's premiere hitting park during the hitting renaissance ofthe '90s. Rodriguez was, of course, a fine defensive catcher. But does that outweight the gulf in hitting between him and the other candidates? Pedro Martinez finished a close second to Pudge and would have made a much better choice for MVP. Roberto Alomar finished 3rd and Manny Ramirez finished 4th, all fairly representative of their contribution. But the AL's best player in 1999 finished 6th in the voting, which is a major surprise considering that he is a media darling and a perpetually overrated player. Yes folks, the real AL MVP in 1999 was none other than Derek Jeter.
In the much less hitter-friendly Yankee Stadium, Jeter posted a 438/552/349 line, a huge step above Pudge. Rodriguez does have the clear edge in defense, but it would take a gargantuan defensive difference to make up for Jeter's edge in hitting. And would the sportswriters actually argue that Jeter's leadership was less than valuable than Pudge's in the year where Jeter's Yankees swept the Rangers in the ALDS on the way to an easy World Championship?
#8: Mo Vaughn, 1995 AL MVP
Let's start out by saying that Mo was at least the 3rd-most valuable 1st baseman in the AL. Frank Thomas hit 454/606/308 and Rafael Palmeiro hit 380/583/310. Vaughn hit 388/575/300. It doesn't even touch Thomas. You might be able to argue him past Palmeiro, except that Raffy played better defense in a tougher park for hitters. There is the steroids question, but that's a hazy question for 1995. Even if you can get him past Palmeiro (and Mark McGwire, who hit 441/685/274, albeit in limited work), you've still got a problem. You cannot get Vaughn past the massive obstacle that is Frank Thomas, effectively blocking his MVP. Vaughn had 15 more RBI than Thomas, and 1 more double. Of course, he hit one less home run and struck out 76 more times. So how in the world can you picture Vaughn leaping the gap between his 388/575/300 hitting line and Frank's 454/606/308? You could say that Frank had already won 2 MVPs, and it was Mo's turn. But the MVP award is not an Elementary School Participation ribbon. If someone deserves multiple MVPs, they get them. The sportswriters, of course, just decided that Vaughn was better for no apparent reason. Vaughn's team made the playoffs, whereas Frank's didn't. But that's an extremely silly argument that I'm surprised people still make (well, not that surprised).
But it gets better -- Frank Thomas probably wasn't even the best player in the AL. Was it Edgar Martinez, who hit 479/628/356, albeit as a DH? Or maybe it was Albert Belle, whose 401/690/317 line would have won him the award, were he not such a hothead. Or perhaps it was Tim Salmon, who hit 429/594/330. It's a tough question (my vote's for Edgar), with many options. But I'll tell you one thing for sure. It was not Mo Vaughn.
#7: Juan Gonzalez, 1998 AL MVP
Conventional wisdom in baseball is often contradictory. Most "baseball men" distrust 1-dimensional players. But these same people also LOVE guys who rack up lots of RBIs. So what happens when a 1-dimensional player racks up a lot of RBIs? His name is Juan Gonzalez, and he won 2 MVP Awards he didn't deserve.
Giving a 1-dimensional player the MVP isn't such a bad idea, so long as that one dimension is pretty damn good. Juan Gonzalez was lucky to be on a team that gave him a lot of RBI opportunities, and he took advantage of them. He had a MLB-leading 157 RBIs. But other than that? He hit 366/630/318, a good mark, but not great for someone in a hitter's paradise during a hitter's era. Gonzalez was below-average defensively and offered nothing apart from his slugging (50 doubles, 45 HR, .630 SLG). But his team won the division, and the voters went ga-ga.
Compare his numbers to Albert Belle's 399/655/328 mark in a less-friendly ballpark. Albert even drove in 152 runs of his own, making it odd how the voters gave Gonzalez any clear edge. Although his Chicago days were much less volatile, Albert was never a supreme joy. So you argue that Albert's "chemistry" (without even knowing whether it was particularly bad that year, or even what Juan-Gone was like) makes him unfit for the MVP? Okay. What about Ken Griffey, Jr.? Can't argue against his chemistry. He led baseball with 56 HR and hit 365/611/284 in a tougher park for hitters with good defense and 20 SB. Even if you can't accept Albert as better than Juan, you must accept Griffey. Either way, you've got two guys who are demonstrably better than Gonzalez, and that doesn't even take into account Edgar Martinez, Mo Vaughn, Alex Rodriguez, Bernie Williams, Carlos Delgado, Nomar Garciaparra, and even Roger Clemens, all of whom probably had a better season than Juan.
But would you believe that Juan won another MVP award that he deserved even less? See below.
#6: Bob Welch, 1990 AL Cy Young
Cy Young voters are absolutely mesmerized by Wins. It doesn't matter that a 5th-grader could debunk the myth that Wins are the best indicator of a pitcher's quality. It doesn't matter that intelligent baseball fans have distrusted wins for over 50 years. The sportswriters love them, and they have misled voters toward awarding a bad Cy Young more often than any other single factor. Just as RBIs have the power to make MVP voters suspend reality, Wins have the power to make otherwise intelligent men argue not just against specific evidence, but against the very concept of evidence.
Having said this, my case can be made that Bob Welch was a dud. Here is Welch compared to Pitcher X:
Welch: (27-6, 2.95 ERA, 77 BB:127 K, 28 HR, 238 IP)
Pitcher X: (21-6, 1.93 ERA, 54 BB:209 K, 7 HR, 228.1 IP)
Pitcher X is Roger Clemens, and he makes Bob Welch look like Rick Vaughn. Bob Welch was, in my estimation, the 9th-best pitcher in the AL in 1990. Clemens finished second in the voting, but it wasn't that close. Dave Stewart finished 3rd, followed by Bobby Thigpen (with his all-time record 57 saves) and Dennis Eckersley (with his 0.61 ERA). Any of the 4 would have been a much better choice than Welch. Oddly enough, the voters wouldn't reward Eckersley's fantastic 0.61 year with a Cy Young Award. They'd wait until he had a worse year and give him the Cy Young and the MVP as well (see below).
#5: George Bell, 1987 AL MVP
At least this award has gained the reputation it deserves as being silly. It is, of course, RBIs that once again led the BBWAA astray. Bell led the league with 134, and his team made the postseason. He hit 352/605/308 and was about average on defense. Wade Boggs hit 461/588/363, although he wasn't fabulous on defense either. So I have evidence that Wade Boggs was a far, far better hitter than George Bell. What is the argument that makes Bell a more valuable player? RBIs? We've already been there, and that won't cut it. Or maybe you favor Alan Trammell, who finished 2nd in the MVP race. He hit "only" 402/551/343 in a tougher hitter's park. In my opinion, Bell is 14th, behind Boggs, Trammell, Mark McGwire, Kirby Puckett, Roger Clemens, Paul Molitor, and many more.
#4: Steve Bedrosian, 1987 NL Cy Young
Again, the BBWAA decided to pick (apparently at random) a relief pitcher to win the Cy Young. They again convinced themselves by the "Bullshit Snowball Effect" (whereby a bad idea gathers steam if enough people repeat it) that Bedrosian had something "special" beyond his numbers. Did he really?
Well, he did lead the league in saves with 40. We don't know how many save opportunities he had, but I'd speculate it was enough to make his raw number of saves of less importance. His ERA was 2.83, light-years behind Montreal closer Tim Burke, whose ERA was 1.19. Burke only had 18 saves, but perhaps that's because he was pitching for the Expos. There's no reasonable way (beyond saves) to argue Bedrosian past Burke . . . or Todd Worrell, or Dave Smith. Or maybe even John Franco. Bedrosian had a worse ERA than all of them. He allowed 11 HR, much more than any of them. His 28:74 BB:K ratio was good, but not significantly better than his compatriots. So we can't demonstrate with any certainty that "Bedrock" was the most valuable relief pitcher in the NL. He was probably the 3rd-most valuable closer (4th, if you count John Franco), making him a puzzling choice for the Cy. And, of course, this doesn't take into account good starting pitchers like Orel Hershiser (3.06 ERA, 264.2 IP, 74 BB: 190 K), Nolan Ryan (2.76 ERA, 211.2 IP, 87 BB:270 K) or Mike Scott (3.23 ERA, 247.2 IP, 79 BB: 233 K). Or Rick Reuschel, Bob Welch, Rick Sutcliffe, Dwight Gooden ...
#3: Andre Dawson, 1987 NL MVP
People continue to insist that Andre Dawson was an elite player. The growing insistence upon on-base percentage has threatened this perception. Of course, when confronted with evidence that may disprove their belief, people will reject the evidence out of hand. So Dawson's career OBP (.323) is dismissed.
Dawson's 1987 season wasn't any better than any of his other good seasons . . . except that he led the league in RBI (137) on a last-place team. And he hit 49 HR, which is impressive . . . unless it's in Wrigley Field in 1987 (an aberrantly hitter-friendly year). Dawson's defense wasn't bad, but his knees weren't what they used to be. He stole 11 bases, too. All this to go with a 329/568/287 hitting line. You'd have to be Ozzie Smith to even be considered for the MVP with a hitting performance like that. And considering that it took place in a hitter's park in a hitter-friendly year, it becomes a pretty disappointing line.
But no one looked past the RBI when making Dawson's case for MVP. And for good reason, since every other number argued against his candidacy. Dawson was, in my opinion, clearly the 4th-best right fielder in the NL in 1987. Tony Gwynn (447/511/370), Dale Murphy (417/580/295), and Darryl Strawberry (398/583/287) all had far superior seasons. You might even be able to argue Mitch Webster (361/435/281) past Andre, honest to God. Where does Andre rank among all NL players? I don't really know; I stop ranking at 20th place.
It's funny how often people use past voting in the MVP race to justify future voting. Which is a very effective way to keep making the same mistakes over and over again. Some years people say that an MVP has to come from a contending team. Then some years they'll completely change their minds and reward guys like Dawson or Alex Rodriguez for being on bad teams. There is absolutely no good sense involved in any of this, and anyone making such arguments should be avoided at all costs (that means you, Joe Morgan).
#2: Juan Gonzalez, 1996 AL MVP
I mentioned earlier that Gonzalez's 1998 MVP was an excellent idea compared to his 1996 honor. It all comes back to RBIs again. But Juan didn't even lead the league; Albert Belle did (of course he wasn't getting the MVP). But Juan hit 47 HR and 144 RBI the same year the Rangers won their first-ever division title. So the voters gave him the MVP.
So often in these discussions, people try to quote statistics and argue about how good a player is. But that's not really the question. The question isn't whether Juan Gonzalez, or Andre Dawson, is a good player; is he better than these other guys? It's not a question of simple quoting of statistics -- it's entirely comparative. Statistics are meaningless unless placed in context; Gonzalez's 47 HR and 144 RBI look much less impressive when you consider the context of his lineup, his ballpark, his league, and his era. It's also an infamous example of cherry-picking evidence; using the bits that serve your purpose while ignoring the other parts of the whole. If someone tries to sell you a house, you want to see the whole house. You don't just want a tour of the outside while the realtor tells you what the inside looks like. Because the part doesn't always accurately represent the whole. But the reality is that the "baseball men" sold the voters (in Juan Gonzalez) the house from The Money Pit; it looked really great at first, but once you actually get to the root an reality of the entire subject, it all crumbles into dust. Of course, the baseball men won't admit the house has fallen down around them. They'll just move on to the next huckster who makes them look silly.
And after that pained analogy, I present, in order, a list of the best right fielders in the AL in 1996:
1. Manny Ramirez (399/582/309), neutral park, no defense
2. Paul O'Neill (411/474/302), neutral park, above-average defense
3. Tim Salmon (386/501/286), neutral park, average defense
4. Jay Buhner (369/557/271), neutral park, no defense
5. Juan Gonzalez (368/643/314), hitter's park, no defense

Gonzalez doesn't make my list of the 20 best players in the league that year, and he's probably orbiting 30. Where's the argument?
#1: Dennis Eckersley, 1992 AL MVP
Dennis Eckersley: 1.91 ERA, 51 saves, 80 IP, 11:93 BB:K ratio - wins MVP & Cy Young
Roger Clemens: 2.41 ERA, 246.2 IP, 62:208 BB:K ratio -- doesn't even win Cy Young
So Eckersley has a slight edge in ERA, considering that Clemens pitched in a much tougher park. But Clemens has . . . 3 TIMES AS MANY INNINGS PITCHED. But wait, friends, it gets worse.

Dennis Eckersley: 1.91 ERA, 51 saves, 80 IP, 11:93 BB:K ratio - wins MVP & Cy Young
Frank Thomas: 439/536/323, league-leading 46 doubles, 122 walks
Roberto Alomar: 405/427/310, excellent defense, league-leading 34 Win Shares

This is the height of folly. And if I thought the voters had learned anything from this odious dereliction of duty, then I might be able to laugh at this award. But folks, the next colossal blunder like this is just waiting around the corner. This is what happens when you throw evidence, the scientific method, and every modicum of common sense straight out the window. Arrggh.

With that said, I'd like to post a list of the awards I agree with completely. Just to give you a sense that the BBWAA and I do see eye-to-eye every once in a while. All awards are since 1985.

AL MVPS I agree with:
Vladimir Guerrero (2004), Jason Giambi (2000), Frank Thomas (1994), Rickey Henderson (1990), Robin Yount (1989), Jose Canseco (1988).
NL MVPS I agree with:
Barry Bonds (2001, 2002, 2003, 2004), Jeff Kent (2000), Jeff Bagwell (1994), Barry Bonds (1990, 1992, 1993).
AL Cy Youngs I agree with:
Johan Santana (2004), Pedro Martinez (1999, 2000), Roger Clemens (1997, 1998), Pat Hentgen (1996), Randy Johnson (1995), David Cone (1994), Roger Clemens (1991), Bret Saberhagen (1989), Roger Clemens (1986, 1987).
NL Cy Young I agree with:
Eric Gagne (2003), Randy Johnson (1999, 2000, 2001, 2002), Pedro Martinez (1997), Greg Maddux (1992, 1993, 1994, 1995), Tom Glavine (1991), Orel Hershiser (1988), Mike Scott (1986), Dwight Gooden (1985)

Having gone through all of this, I must admit that no way of voting, including my own tabulations, are perfect. I'm sure if I made a close examination of my calculations and analysis (some of which is 2 years old), I'd find some conclusions that I've changed my mind about already. The point is to learn from past mistakes and progress. And any system that fails nearly 50% of the time is not being run by the most effective voting machine.
Peace out ...

No comments: