The Most Valuable Player debate always starts, and usually ends, with how well the candidates hit. But a run saved is a run earned -- as Ben Franklin would have said had he lived a century or so later -- and the awards won by Ivan Rodriguez and Ichiro Suzuki in recent years suggest that the voters do pay attention to performance in the field.

Baseball Prospectus rates fielders every year with a system that still doesn't have a name of its own: "BP Fielding" will do. It starts by rating the defense of the team as a whole, breaking down the contributions of the team's pitchers and fielders. It then sorts the team totals into contributions from each position, and each position is then broken down into contributions by each player who played there. In all of this, the system rates and adjusts based on various splits in the data which define the characteristics of the team: strikeout rates; groundball/flyball and left/right tendencies of the pitching staff; how many baserunners are allowed. It's not a perfect system, and using it in-season is complicated by a slightly reduced data set and, of course, a smaller sample size than a full season would produce.

Let's see how the American League MVP candidates have done in the field this year, and whether their defense helps or hurts their MVP case:

The longest shots
These players are having good enough seasons that they would have been MVP candidates on a winning team, but not so good to remain candidates when their team loses. Mike Sweeney, for instance. He's leading the league in batting average for the last-place Royals. The system says he is having a good year defensively at first base -- 10 runs better than average, which would be a personal best. That helps, but just a little; his history will count against him.

Jim Thome is second in the league in home runs, but eight runs below average at first base. That's no surprise; he was -14 runs last year. That hurts. Magglio Ordonez is also in this category; he's basically average defensively, coming in at +2 runs. No change.

The Yankees
I really don't think a Yankee will win the MVP this year, as they have at least four legitimate candidates who may splinter the votes amongst themselves.

Jason Giambi is a former MVP, and again having a monstrous offensive season. But he is a first baseman, arguably the least important defensive position, and has a terrible defensive reputation. The BP rating does not support that: he rates at +4 runs so far this season, after a +3 last year and -1 in 2000. He's basically average, even if he does look bad. (Nick Johnson does rate substantially better, though, so the team might be better off with Giambi at DH.) His defense should make no difference in his candidacy, but his reputation probably hurts him.

Bernie Williams is a multiple Gold Glove winner, but the numbers have not been kind to him. His rating has slid from a +3 in 1999, to a -8 in 2000, to -18 last year, and has already reached -19 this year. If he keeps it up, he'll finish the year at -28, which would be his third consecutive 10-run drop. Bernie turns 34 next month, and it looks like the legs are going.

The keystone combo of Alfonso Soriano and Derek Jeter are both legitimate candidates, although the talk seems a lot louder for Soriano than any other Yankee. Sorry, New York, but from a strictly statistical point of view they are both horrible defensively. Soriano rates 13 runs below average so far this season, after going -18 last year. Jeter is even worse: -20, following seasons of -17, -24, and -16 (whatever flaws the system may have, it is consistent).

A big part of why the two don't score well is that they don't record many double plays. The Yankees are on pace to turn 70 double plays at shortstop, which would tie the 1991 Blue Jays (and the immortal Manuel Lee) for the fewest in the AL in the past 20 years. (The NL, with more basestealing and bunting, has fewer double-play opportunities as a rule, and has had half a dozen teams with fewer than 70 shortstop double plays in that span.) Yes, the Yankees have good pitchers, which limits the number of opportunities, and yes, they're not a groundball staff, but that's already been taken into consideration: they're not turning as many as you'd expect, and somebody has to take the blame for that. The middle infielders should definitely be hurt by their fielding abilities in the MVP voting.

The other contenders
Seattle's best candidate this year is last year's MVP, Ichiro Suzuki. Defense was a big part of his winning, and the BP system concurred, rating him an impressive +20 runs last year. This year he appears to be a little off that, but he's still solidly above average at +8. It helps, but you generally have to play even better to get a second MVP in a row.

Oakland's best candidate was looking like Barry Zito, but Miguel Tejada has certainly moved into the outskirts of consideration. He's been a very average fielder, with a +2 this year, and a -1 and a 0 in the two years prior; but an average fielder at shortstop is a valuable thing. I think his glove helps his case.

The Red Sox also offer up their shortstop, Nomar Garciaparra, although Derek Lowe and Pedro Martinez deserve consideration. Like Tejada, Garciaparra is a basically average shortstop in the field (+1 this year, 0 last year, +3 in 2000), which again is a help.

The Twins' best bet is Torii Hunter. He's having an impressive year at the plate, he's widely acclaimed as the best center fielder in the AL, and Baseball Prospectus 2002 backs that up -- he's had the best center-field rating in the majors two years in a row, at +34 and +20.

Except that this season, he's rating a dismal -10. Look at the team putouts for Minnesota:

Left  Center  Right
275    296     275

That's about as close to a perfectly even split as you'll see; the center fielder has a shade under 35 percent of the team's putouts, less than any team I can find; a normal share would be 40 percent. Last year, the Twins' numbers were 352/504/366, which is typical. The left fielder is the same (Jacque Jones), the center fielder is the same (Hunter); only right field is different (Matt Lawton replaced primarily by Dustan Mohr and Bobby Kielty).

Between July 1 and August 8, we have this oddity. In 297.1 innings,

Twins LF: 87 putouts, 2.63 per game
Twins CF: 78 putouts, 2.36
Twins RF: 84 putouts, 2.54

Hunter's share, over that time, is just 2.32 putouts per game. Something is wrong with him, and I don't think it's just the hip flexor that sidelined him last week. While his defense this year should be hurting him in the MVP race, I don't think this downturn is public knowledge, and his reputation remains sky-high.

That leaves one more strong contender: Alex Rodriguez. He's possibly the best hitter in the league; currently the AL's home-run and RBI leader. Regrettably, playing for a team with no pitching and a Swiss cheese outfield, he's also in last place in the AL West.

Jeter rated well below average on defense, Tejada and Garciaparra as average. Rodriguez is a better hitter than any of them, and he also rates as a better fielder:

2000 +17
2001 +6
2002 +14

So it turns out that the best hitter in the league also benefits the most when you take fielding into consideration. Alex Rodriguez is clearly the best player in the league, but whether he's the Most Valuable will be in the hands of the voters.

You can check out more work from the team of writers of the Baseball Prospectus at baseballprospectus.com. Clay Davenport can be reached at clayd@baseballprospectus.com. Baseball Prospectus is a registered trademark of Prospectus Entertainment Ventures, LLC.