The Stats That Matter Most (2003 season finals)

Many moons ago, trendsetting sabermetrician Bill James demonstrated a relationship between runs scored, runs allowed and wins which, because of the necessity of squaring the elements, he dubbed the Pythagorean Theorem (of baseball).  The formula is that the square of runs scored divided by the sum of the square of runs scored and the square of runs allowed equals a team's expected winning percentage.  From there, you simply multiply the winning percentage and games played and you have a projected win total.  The measure works remarkably well, and, obviously, the more games a team plays, the better the relationship ordinarily is.

Here are the numbers through the end of the 2003 regular season:

Orioles Comments:  The acknowledged Orioles late season meltdown can be laid almost entirely at the hands of the offense.  On the morning of August 21, the Orioles were -5 in RDIFF and were scoring 4.97 runs per game.  From that point until the end of the season--a span of 37 games--they were outscored 194-122 (that's 3.30 R/G for those of you keeping track).  They allowed 5.24 runs per game during the meltdown, compared to 5.01 beforehand.  When the dust settled, the Orioles ranked 10th in the league in R/G, barely ahead of the 11th place Angels; the Birds had been fifth in the league at the All-Star break.  In OR/G, the Orioles finished 9th, essentially where they'd been all season long.  When it was all said and done, there was no meaningful difference--in a composite sense--between the performance of the Orioles and the Cleveland Indians this past season.

AL Notable Numbers:  The Yankees finished with the league's best DIFF/G mark, but the Mariners had the highest projected Pythagorean win total (a function of a DIFF/G mark almost as good as New York's and playing, overall, lower scoring games where runs were more precious).  Astute observers will note that the Mariners, despite the highest projected win total in all of baseball, missed the postseason.  Seattle's -6 WDIFF mark was the largest discrepancy in the AL.  The Pythagorean Theorem implies that the M's should have beaten the Oakland A's out by four games in the AL West.  Both teams used pitching/defense (A's 1st, M's 2nd in OR/G) as a key to success, overcoming mediocre offensive performances (M's 7th, A's 9th in R/G).  The difference between the two, at least in part, was a function of one-run and extra inning game records (Oakland:  25-20 1-run, 10-4 extras; Seattle:  16-15 1-run, 6-6 extras).

The Theorem also implies that the Chicago White Sox should have beaten the Minnesota Twins out for the Central Division crown this year.  But Chicago finished -3 in WDIFF while Minnesota was +5.  Again, one-run game records may have had something to do with it (the Twins were 22-20 in one run contests, the White Sox 18-22).  As a side note, the Kansas City Royals also finished +5 in WDIFF; the team was outscored this year, so no one in the organization should have any illusions about the club's performance in 2003.  Still, the Royals improved by 123 runs in the RDIFF department in comparison to last year.

The Boston Red Sox overcame subpar pitching (8th in the AL) by hitting the cover off the ball to take the AL wildcard slot.  Boston finished with a ML best 5.93 runs per game, more than 0.4 runs per game better than the AL runner-up (Toronto).  The Jays were no group of slouches either, scoring better than 5.5 runs per game and finishing a solid +68 in RDIFF (that's close to a full run per game better than the Orioles).  The Devil Rays had another lousy year, losing 99 games and finishing 13th in R/G and 12th in OR/G, but the Theorem implies that they should have won 67 games rather than 63 and the team's -137 RDIFF was a 108-run improvement over 2002.  Tampa Bay didn't field a good team in 2003, but it was a dramatically better club this year than last. 

The Texas Rangers were awful, even worse than their record (+3 WDIFF).  The Rangers allowed more runs than the Red Sox scored, and were easily the most prolific run-yielding team in all of baseball.  The team's RDIFF was worse than Tampa Bay's.  While Texas' record actually got worse this year (71 wins this year, 72 in 2002), the club's RDIFF mark worsened by 104 runs.

As awful as the Rangers may have been, they weren't in the same time zone as the Detroit Tigers.  Everyone knows that the Tigers narrowly avoided the ignominy of matching or exceeding the 1962 New York Mets' 120 losses, but...Detroit scored 124 runs fewer than the next worst scoring AL team (Tampa Bay); the Tigers were outscored by an average of more than two runs per game; after being outscored by 289 runs in 2002, the Tigers actually got worse in 2003 (-337); Detroit's raw run differential was worse than that of the '62 Mets (-337, -2.08/G to -331, -2.07/G).  I can't say for certain, but Detroit's two-year run differential mark of -626 is probably the worst in modern major league history.

NL Notable Numbers:  The Atlanta Braves had the National League's best RDIFF and it's best record, but still outpaced expectations by four wins.  The Braves really had a remarkable offensive year; despite playing in what has historically been a good pitcher's park, Atlanta outscored every other club in the NL, and all but the Boston Red Sox in the AL.  They managed 54 more runs than the Coors Field-based Colorado Rockies.  In a switch from traditional accomplishments, the Braves finished with a very average OR/G mark (9th in NL).  The Philadelphia Phillies finished -5 in WDIFF despite a 10-3 record in extra inning games.  They lost the wildcard race to the Florida Marlins, despite a superior Philly RDIFF.  The Marlins finished +3 in WDIFF and the club's 30-23 record in one-run games may have been a factor.  The Montreal Expos managed a winning record despite a slightly negative RDIFF and the New York Mets were obviously the East's worst team, most a function of a poor offense (3.99 R/G, 15th).

The Central was simply upside down.  The Chicago Cubs won the Central despite having the division's third best RDIFF.  In fact, they were so far behind the Houston Astros (+128 RDIFF) that it's shocking.  The Astros finished a remarkable -8 in WDIFF, and it cost them the division.  The Cubs parleyed a 27-17 mark in one-run games to a +2 WDIFF mark.  Houston was fourth in the NL in R/G and third in OR/G--both marks better than that of the Cubs (fourth in OR/G, ninth in R/G) and missed the playoffs.  (By comparison, the New York Yankees finished third in both categories and finished with the AL's best record.)  Not only does the Theorem imply that the Astros should have finished comfortably ahead of Chicago (nine games), it also implies that the second place team should have been the St. Louis Cardinals, not the Cubs.  The Cardinals project to 89 wins, three more than Chicago.  St. Louis had the league's second most prolific offense, but was done in by its 11th ranking in OR/G and a pathetic 14-25 record in one-run games.  The Cardinals finished with a -4 WDIFF.  As for the rest of the division, the Pittsburgh Pirates really did belong in the mid-70s in wins (the team improved offensively from last season, by nearly 2/3 of a run per game), but the Milwaukee Brewers were actually three games better than expected, despite a 94-loss season; only two NL teams allowed more runs per game than the Brewers.  Then we have the Cincinnati Reds, who were miserable this year, but were actually worse than their record.  The Reds--who weren't fielding anything close to a major league team by the end of the season--had been overachieving all year and, despite a 69-93 record, were the worst team in the National League in a statistical sense.  They projected to 100 losses, a function of a lousy (13th) offense and a miserable OR/G mark (15th).  The only NL team to allow more runs than the Reds was the Colorado Rockies, the Coors Field tenants.  The Reds overachieved in the win column by finishing with a 30-21 record in one-run games and a remarkable 15-5 mark in extra inning contests.

Out West, the Giants were the best team, but still finished +6 in WDIFF.  San Francisco finished 6th in R/G, but second in OR/G (PacBell was at least partly responsible for this).  The Los Angeles Dodgers allowed only 3.43 runs per game (more than a half-run better than the second ranked Giants and the best mark in MLB); the problem is, the Dodgers scored only 3.54 runs, the worst mark in either league...worse than Detroit (for crying out loud)!  The Arizona Diamondbacks were a less extreme version of the Dodgers--they prevented opponents from scoring pretty well, but didn't do much offensively themselves (5th in OR/G, 10th in R/G).  Both Los Angeles and Arizona finished within a game of expected win totals.  The Colorado Rockies demonstrated their typical Coors Field effect (3rd in R/G, 16th in OR/G) and were marginally better than their 74 win total.  The San Diego Padres were essentially as bad as their 64-98 record would suggest and were among the league's bottom feeders in terms of offense (14th R/G) and pitching/defense (13th OR/G).

The Stats That Matter Most (thru games of August 20, 2003)

The Stats That Matter Most (thru games of July 13, 2003)

The Stats That Matter Most (thru games of June 24, 2003)

The Stats That Matter Most (thru games of May 20, 2003)

Back to Stats and Numbers

Back to Birds in the Belfry Home Page