One of the more perplexing conundrums this season has been the uneven start of the Pittsburgh Pirates. They don't look like a championship team, and yet they have been posting championship results—until recently at least.
Depending on the day, that isn't necessarily true of the won-loss record (which can be skewed earlier in the season by lucky wins or unlucky losses), but has generally been true of a league-leading ERA, and from time to time, impressive hitting statistics.
First, a look at the pitching. Taken as a group, Paul Maholm, Zach Duke, and Ian Snell are not pitching better than last year, based on the number of walks, strikeouts, and home runs achieved (these are measures of "sabermetric," rather than actual ERAs, meaning that these pitchers have been "luckier" so far this year).
Nonetheless, the Pirates' rotation is decidedly better, because of the "backenders," Ross Ohlendorf and Jeff Karstens, acquired from the Yankees last year in the trade of outfielder Xavier Nady and reliever Damasco Marte.
Ohlendorf is actually a middle, not back of rotation type, and while Karstens is no great shakes, he is a legitimate fifth starter. That is more than can be said of the Pirates' two backenders in 2008, Tom Gorzelanny, and the tag team of Matt Morris, Phil Dumatrait, and assorted minor leaguers.
On the whole, the Pirates have at least an average rotation, but with a number of "lucky" (non)-runs having created the appearance of a barn-burner.
A similar situation exists on offense, where the Pirates appear to be about "league average," with a statistical "mean" of about 4.5 runs per game. But there are two ways to achieve this. One is to have a team that consistently scores three to six runs per game.
The other is the Pirates' way; eight or 10-run "runaways" in several games, and zero runs in a number of others. Because of a statistical property called "variance," this is the worst way to achieve a given average. The one and two-run losses that have become a Pirates' staple recently hurt more than extra runs in a blow out help. So the Pirates' run distribution isn't nearly as useful as one that hugs the statistical averages.
So far in the season, the Pirates have scored 109 runs and allowed only 90. The key is a (19) "run differential" inferior only to those of the Toronto Blue Jays, St. Louis Cardinals, and Los Angeles Dodgers.
It also suggests a win-loss percentage of around .600, except for the following: As alluded to above, Pirates' pitchers have been "lucky" for 10-15 runs, meaning that they should have allowed 100-105 instead of 90, based on raw statistics. And because of the bad distribution, the 109 runs scored may only be worth 100-105. Making these adjustments fully explains the Pirates' .500 record.
In the end, that record probably tells the story: of a team that has come a long way from last year, but still has a long way to go before becoming a contender.