Since the candidates are all hitters, we can start with the batting stats. Helton and Bonds were the league's two toughest outs (.463 On Base Percentage to .440) and 1-2 in Slugging as well (.698 to .688). But while Coors did its usual number (scoring was up 65% there vs. Rockies road games), Pac Bell (for some unknown reason) was the third best pitcher's park in the league at -16%.

Nobody knows exactly why this happened, given the drooling occasioned by Pac Bell's 307 foot right field porch. Sometimes weather and other factors cause an odd year at a park; homer-happy Wrigley was the top pitchers' haven this year at -19%. Perhaps it's the massive power alleys, or the batting visibility. Still, park effects are real and usually turn out to be steady enough over time that they can't be ignored. Clearly, Bonds was a better hitter than Helton. For the mathematically inclined, the Rockies were outscored by the Giants 5.3 to 3.9 runs/game in road games, but still put up prettier numbers by scoring 7 runs/game to the Giants' 5 at home. Helton's slim advantages of 1.5% in slugging and 5.2% in OBP over Bonds are obviously the work of the same thin air effect.

How about Bonds and the other 1B/OF? Edmonds, Guerrero, Sosa, Andruw Jones and Sheffield were all great, but to my mind the top competitor was Bagwell. Bagwell had more than 100 more plate appearances than Barry Bonds, and was not so far off in production with the bat (.310 batting/.615 slugging/.424 OBP). Put another way, Bagwell's numbers minus Bonds' would be:
PA SLG OBP R RBI
112 .300 .339 23 26
In other words, Bagwell's extra at bats were just filler; the Giants were better off with Bonds and with a bench player doing 100 plate appearances. Yeah, he had great Runs and RBI totals, but Bagwell also had some teammates having big years. And the writers are justifiably suspicious of guys who put up big offensive numbers in small ballparks for bad teams.

I ran the Runs Produced numbers last week for the AL; while it's not a true measure of offensive productivity, Runs Produced shows how many actual runs a player participated in actually putting on the board, so it's a good bottom-line sanity check on some of the more esoteric stats. The NL leader board:
Todd Helton 243
Jeff Bagwell 237
Jeff Cirillo 215
Jeff Kent 206
Brian Giles 199
Richard Hidalgo 196
Jim Edmonds 195
Sammy Sosa 194
Chipper Jones 193
Andruw Jones 190
Quick, hurry out and name your kid "Jeff." If you're wondering, Bonds was 12th, at 186; Piazza was off the chart at 165. The top three guys here all had a lot of help from the home cookin', which is ironic for Bagwell because the Astrodome killed him last year. But Enron Field was second only to Coors this season, pumping up scoring by nearly 20%, and the surge in Bagwell's stats (becoming the first player since Ted Williams to score 150 runs in a season) was more the effect of new scenery than any change in his performance.

I would gladly have honored Piazza, who has been repeatedly screwed in the MVP voting and had the lead going into September. But the double whammy of a late-season fade and a gap in offensive numbers puts Piazza at a big burden to get an edge for being a catcher. His defense is now poor enough that you can't give him much credit
for catching, so the Mets wait still another year for an MVP.

That leaves the quiet Giants. But how good was Kent, really? Kent, too, batted a lot more than Bonds; let's run the same "marginal production" comparison as with Bagwell:
PA SLG OBP R RBI
88 .187 .318 (-15) 19
Yes, Kent scored 15 fewer runs than Bonds despite nearly 100 more plate appearances. As you can see, Kent's superior offensive numbers were also the result of bulk rather than quality; even at a lower level of playing time, Bonds did more good to the Giants. I won't go through the whole argument here, either, but Kent's RBI count was influenced by Bonds' "runs scored" more than the other way around.

But wait; we're forgetting something. Outs are not only made at the plate. Jeff Kent was caught stealing 9 times and hit into 17 double plays; Bonds' totals, even at age 35, were 3 and 6. Put differently, if you lumped in the outs on the basepaths (at least the ones the major stat services keep track of), Bonds' 16-point edge in OBP
swells to 24 points and Kent's "marginal" OBP drops to .262. Those numbers tell a simple story: Kent made so many more outs than Bonds that he just can't be viewed as a comparable hitter.

Bonds was the best hitter in the National league this season; Kent finished in the top ten, but would have to rank behind Bagwell and a few others including Sheffield, Sosa, Giles, and Guerrero.

Defensively, both Bonds and Kent are about average at their positions at this stage; Both were around the league average (Bonds a bit above, Kent a bit below) in range factor, and both above average in fielding percentage. For Bonds, that's a sign of decline; he used to be the best defensive left fielder alive. For Kent, adequacy is real progress.

What about timing? Kent partisans point to his early-season heroics that kept the team afloat while Bonds was injured. But if you recall, the Giants were still running fairly close with the Diamondbacks in the NL West, and facing 8 games with Arizona in the last 11 days of the season, until Bonds went berserk in September (11 homers and 27 RBI in 24 games between August 30 and Sept. 28) and the team blew the race open. That has to be at least a draw for Bonds, possibly a slight edge.

Leadership? Like I said, their teammates hate Bonds and ignore Kent, so you can take your pick. Thus, the entire issue between Bonds and Kent comes down to how much you value the difference between a left fielder and a second baseman.

Here are Bonds and Kent compared to (1) the average player at their position and (2) the next best hitter at their position:
PLAYER PA Avg Slg OBP R RBI ParkEff SB (CS+DP)
Bonds 607 .306 .688 .440 129 106 -16% 11 9
Avg NL LF 702* .284 .490 .364 106 99 None** 12 14**
Sheffield 612 .325 .643 .438 105 109 -15% 4 19

Kent 695 .334 .596 .424 114 125 -16% 12 26
AvgNL2B 715* .284 .418 .359 100 71 None*** 20 20**
Alfonzo 650 .324 .542 .425 109 94 -11% 3 14
*This represents the batting line of all of each team's second basemen and left fielders, not the starters.

** The DP averages were unavailable, so I calculated an approximate average among starting second basemen and left fielders.

***The league totals average out park effects, at least to any significant degree.

Alfonzo and Sheffield were actually fairly close comparisons to Kent and Bonds, the major differences being that Sheffield is an awful fielder and Alfonzo batted behind Jay Payton and Derek Bell instead of Bonds. If cumulative stats like Runs Produced are your thing, Kent stands out at +49 over the league average at his position, while Bonds is at just +10. But if you prefer OPS, Bonds comes out at +274 to Kent's +243. When you add in baserunning outs, Bonds (-5) stretches his lead over Kent (+6).

I might have voted for Kent too, since it was so close and Bonds is a jerk who has plenty of hardware. I certainly can't fault sportswriters for parsing the numbers a little differently than I would have. But on a strictly objective analysis, here's my ballot:

1. Barry Bonds
2. Jeff Kent
3. Mike Piazza
4. Jeff Bagwell
5. Edgardo Alfonzo
6. Vladimir Guerrero
7. Gary Sheffield
8. Sammy Sosa
9. Brian Giles
10. Todd Helton

TRIVIA QUIZ: Before he was drafted by the Pirates, what team originally selected Barry Bonds in the amateur draft?

In 1998, I surveyed the field and found that the numbers pointed to Nomar – even though I felt that the player with the greatest overall impact was probably Pedro. Last season, I was firmly in the Pedro column. As we turn to 2000, guess what? At the risk of sounding unoriginal or (gasp!) pro-Red Sox, there simply was not a more valuable player in the American League than Pedro Martinez. If someone tells you otherwise, ask them how many presidential candidates they voted for.

Here's the case for Pedro:

1. The Facts
Twenty-five American League pitchers threw at least 190 innings in 2000. Twenty-two of them allowed at least 100 runs; 24 allowed at least 96 runs. The second best pitcher in the league was Roger Clemens, tagged for 96 runs in 204.1 innings, although you can quibble about Bartolo Colon, who let in 86 runs but worked just 188.2 innings. Pedro allowed 44. In 217 innings. That’s the same number Jaime Navarro allowed in 33.2 innings. To find the next AL pitcher who allowed less than 80 runs you have to go down to the number 50 man in innings, Jose Mercedes, who allowed 71 runs in 145.2 innings.

Compared to the next best full-time starter in the AL, Pedro saved his team somewhere in the neighborhood of 50-55 runs. Compared to the average of the other top 24 starters (111 runs), the margin is 67. And that’s without adjusting for Pedro throwing more than the average number of innings. Compared to the average AL starter’s ERA (5.09), projected over 217 innings, Pedro comes out 81 EARNED runs ahead. And we still haven’t factored in Fenway; over the past 2 seasons, Pedro is 19-7 with a 2.04 ERA at home, 22-3 with a 1.76 ERA on the road.

In other words, it’s safe to say that Pedro Martinez saved the Red Sox at least 50 runs compared to any other AL pitcher, close to 70 runs compared to the average number 1 or 2 starter in the league, and well more than 80 runs compared to an average AL starter.

2. The Hitters
So that’s the baseline: at least 50 runs ahead. Was any non-pitcher even close to 50 runs better than any other, combining offense and defense? Because the margin is so large, I figured I could just tinker with the big numbers to see if I found something; forgive me if there are a few imprecisions here.

Who were the best in the league? I started by asking the experts. I just got my copy of the STATS, Inc. 2001 Major League Handbook in the mail (order it NOW if you like to follow baseball by the numbers). The Bill James Runs Created formula (explained on page 393 of the book) rates Frank Thomas as the biggest stick in the league, by a margin of TWO runs over Jason Giambi and THREE runs better than Carlos Delgado. To be fair, those three guys were about 20 runs ahead of the rest of the pack, and Runs Created is just one of many statistical estimates of productivity (based on stuff like total bases, walks, etc.), plus it ignores defense, position, and park effects. The per-out measures of productivity put Giambi ahead. Either way, it’s hard to see one guy towering over the pack. How far are the top guys above the average? As a baseline, the average AL team scored 857 runs, 95 per lineup slot. Runs Created, of course, means nothing if it doesn’t add up to that: how many runs each player was responsible for. The top guys are in the 160s, and thus about 70 runs above average. Good, but no Pedro.

How else to slice the pie? Well, one simple stat that tells us how many runs a player actually put on the board – not what he was worth but just how many times his actions sent a run across the plate – is Runs Produced, which is the total of Runs and RBI, with homers subtracted (because driving yourself in is still just one run). Luck and teammates have a big impact here; at the extremes, a player who bats 1.000 and hits 40 homers but always bats with the bases empty will “produce” only 40 runs, while a player who bats 500 times with the bases loaded and hits .150, all singles, will produce close to 150 runs. The totals will be higher than the Runs Created method, because most of the top guys play in good offenses and thus feed off each other.

Still, if somebody’s putting an extra 50 runs on the board, over and above everyone else, it’s gotta show up here. Seven AL players cleared 200 Runs Produced in 2000:
Alex Rodriguez 225
Mike Sweeney 220
Frank Thomas 215
Carlos Delgado 211
Johnny Damon 208
Edgar Martinez 208
Jason Giambi 202
Hmmmm . . . don’t see a 50-run advantage here, or anything like it. As I noted above, we’re further above the 95-run average here because all these guys hit in the top/middle of the order and all except Delgado come from the league’s top 5 offenses.

Still, we have five guys on this list who DH or play first base like a DH, and the top guy is a shortstop in a pitchers park... so we have a suspect. A-Rod was 7th in the league in OBP (.420), and 5th in SLG (.606), and while four guys (Thomas, Giambi, Delgado, and Manny Ramirez) were ahead of him in both, none of those guys is much distinguishable from the others, none has any defensive value, and Rodriguez slugged .702 on the road, the best in the AL. Only Manny Ramirez, barely, was within 50 points; Ramirez, Thomas, Giambi and Delgado all hit better at home, although in Giambi’s case that was against the tide of a pitcher’s park. Despite playing two more games at home, 135 of A-Rod’s 225 Runs Produced came away from Safeco, which depressed scoring by a whopping 18% compared to Mariner Road games. If you double his road total to 270 you get closer to the 50-run advantage mark, but a fairer adjustment would be to increase his overall total by 9%, which puts us at around +20 compared to Giambi (rather than +5 compared to Sweeney).

If you’re wondering, after park adjustments A-Rod is still about 20 back in the Runs Created column, so this is hardly the final word. But we are looking for ANY evidence of a Pedro-like impact.

That leaves Rodriguez with at least 30 runs to make up in the field against Giambi to stand out the way Pedro does. It’s a difficult comparison to make, but A-Rod made in the neighborhood of 20 more plays than the average major league shortstop, and Giambi made only six errors, so it would be rather ridiculous to say that Rodriguez saved an extra 30 runs in the field if each man is compared to an average fielder (or even a poor one) at his position. On the other hand, as I plan to discuss in more detail in a later column, the A’s, for reasons that can not entirely be rationally explained, surrendered a staggering proportion of unearned runs this season. I don’t know how many of those resulted from throwing errors, but only two major league teams approached Oakland’s penchant for throwing errors, so some of the blame for this probably belongs to the first baseman.

Thus, for Rodriguez to approach the impact of Pedro, the answer must reside in his being further above-average at his position, offensively, than Giambi. That’s a tall order, given that three of A-Rod’s thirteen compadres are Jeter, Nomar and Tejada. As it turns out, measuring by Runs Produced, the average AL 1B rates 18 runs ahead of the average SS. Looking at the percentages, a clearer picture comes into view: the first basemen were 37 points ahead in On Base Percentage and 80 points in slugging. So the real productivity difference may be larger than 18 runs.

What if we compare Rodriguez to another shortstop? Tejada, for all his gaudy totals, posted league average batting and on base percentages and wasn’t that much higher in slugging. Jeter’s defensive stats are horrendous - he made nearly 100 fewer outs in the field than Rodriguez, and if you add those 100 outs to Jeter’s batting record you would see that his net contribution to the Yankees, while still impressive, isn’t the stuff of MVPs. The real competition is Nomar. Before adjusting for the parks and the difference in playing time, Nomar is arguably Rodriguez’ equal at the plate, posting a higher OBP (.434 to .420) and close in SLG (.599 to .606). I’ll spare you the numbers here but I would wager that the real offensive difference between these two isn’t more than 40 runs and was probably closer to 10-15.

It’s also difficult to find a comparison to Nomar in the field that gives Rodriguez much daylight. Rodriguez made about 70 more outs and about half as many errors, but most of the difference in outs is due to an extra 100 innings in the field; the range factors are more comparable. Again, I can’t see giving Rodriguez more than another 10-15 runs (charitably) with the glove. No way can he be worth an extra 50 runs a year, across the board, than Garciaparra.

To conclude, even a charitable reading of the statistical measures leaves A-Rod about 10 runs short of Pedro, when contrasted to his peers, and probably more than that. The real numbers probably put A-Rod much closer to Giambi, who really was a better hitter. Nobody here can really be said to approach Pedro’s dominance, even with an extra 130 games to make that impact felt.

3. The Standings
It’s risky sometimes to start the "Where-would-the-team-be-without-this-guy?" analysis, which tends to discriminate against players on awful teams (as Branch Rickey once told Ralph Kiner, “We can finish last without you”) and players on great teams (the 1998 Yankees and 1986 Mets were both MVP-less). But take an objective look at the Red Sox, really: the team finished TWELFTH in the AL in runs scored. In Fenway. The defense was nothing special, and the numbers reflect that: the Sox matched the league average in fielding percentage and turned a major-league low number of double plays. Jose Offerman (!) was the only regular in the top five at his position in range factor. Five men started 15 or more games for this year’s Red Sox, and the ERAs of the other four were 6.13, 4.78, 5.48, and 5.11. Pedro was the team’s only pitcher to qualify for the ERA title, and except for Ramon (with 10) no other Sox pitcher won more than 8 games.

Yet, this team – with a bad offense, four bad starting pitchers, mediocre fielding, rotten team chemistry, and chronic instability – was mathematically eliminated by the Yankees (on the way to their third straight World Championship) with ONE game left in the regular season. Derek Lowe and the bullpen deserve some of the credit for that, but the rest is Pedro’s. The Sox were 64-69 when anyone else took the mound, a .481 team (projects to 78-84 overall); with Pedro, they were 21-8, a .724 team (projects to 117-45 overall). And they scored 1.6 runs per game in his losses, and two-thirds of the wins were by three runs or less. In other words, the deciding factor in the Red Sox’ improvement with Pedro on the mound wasn’t luck or run support; it was Pedro.

4. Outside the Box Score
The MVP voters, as expected, tabbed Giambi over the other two offensive monsters (Delgado and Thomas) based largely on his clubhouse impact. Giambi also got hot down the stretch, with he and Rodriguez having their best days of the year on the season’s penultimate day.

Many voters believe that a pitcher can’t have that impact. That’s why pitchers rarely manage, you see; they aren’t respected as the same sort of workhorse, carrying the load every day. But as clearly as the numbers show Pedro’s impact, the intangible factors lean his way as well. Yes, the Red Sox took the field 133 times without him this summer. But think of the electricity in the air when the man pitches; think of the intimidation of other teams.

Most of all, think of how easily this Sox team could have quit this season; what kept them alive? It wasn’t a loose, happy clubhouse, or a beloved manager. It was knowing that every fifth day, a man would take the mound and turn the Lansings and the Spragues and the Gilkeys of the world into a better team than the 1927 Yankees. It was knowing that any bunch of bozos might win a playoff series with Pedro in the rotation. Big slugging first basemen, even hard-hitting shortstops, aren’t nearly as rare.

SUMMING UP
Here’s this man’s AL MVP ballot, if they would have given me one:
1. Pedro Martinez
2. Alex Rodriguez
3. Jason Giambi
4. Carlos Delgado
5. Frank Thomas
6. Nomar Garciaparra
7. Darin Erstad
8. Edgar Martinez
9. Troy Glaus
10. Mike Sweeney