FanGraphs contributor Marc Hulet wrote a post today involving, in part, discussion of the Astros' offensive woes (what else is new?), and the subject brought to mind the fact that I hadn't really taken the time to look at our overall runs scored versus runs allowed, other than the general feeling that the pitching has been okay and the offense has been terrible.
For those who don't know, Pythagorean Expectation is, per Wikipedia:
...a formula invented by Bill James to estimate how many games a baseball team "should" have won based on the number of runs they scored and allowed.
The formula is as follows:
With 10% of the season done, it seemed like a good time for a first look utilizing this method of analysis; the sample size is still small, but I will revisit this subject in less depth throughout the season as we see our stats stabilize, and every sixteen games seems like a reasonable frequency at which to do so.
So let's dig in. The bad news? The Astros have been far, far worse than their 6-10 record would suggest. In fact, based on the above formula, we're on pace to be historically bad, among the worst teams in MLB history.
The good news is, there are a number of reasons to believe that the Astros should finish the season well above their current Pythagorean Expectation. Here are the whys and wherefores of both ends of the spectrum. WARNING: Lots of numbers ahead! Proceed with caution!
To date, the Astros have scored 45 runs and allowed 74 opposing runs. According to the formula above, that projects out to a .265 winning percentage, which ties the eighth-worst season in Major League history, the Detroit Tigers' 43-win season in 2003. If we continue to play like we have been over the course of the entire season, we will likely wind up below 50 wins. The crazy thing is, the Orioles have been even worse than we have; They've scored four more runs than us, but have allowed thirteen more to score.
Our Pythagorean Expectation of 43 wins is 23 less than our actual winning percentage (.375, or a 66-win pace) would suggest. Since that is pretty close to rock bottom, there isn't really much more I can say on the pessimistic side of things; it is what it is.
So since there's nowhere to go but up, it's time to discuss the myriad of explanations for this poor start, as well as the team's current hot streak, which both point to improvement (if not of our win/loss percentage, at very least of our Pythagorean Expectation).
The first and arguably the most important explanation has been the stiff competition the Astros have faced. 15 of our 16 games have been against teams who were above .500 both in win/loss percentage and Pythagorean Expectation in 2009. Two of those teams (the Phillies and Cardinals) were FAR above .500 in both categories and were among the best teams in the NL. We also faced an array of very difficult starting pitching matchups, including three of the best pitchers in baseball in Tim Lincecum, Roy Halladay, and Adam Wainwright. Five of the sixteen pitchers we faced had earned run averages below 3.00 in 2009; for perspective, only nine pitchers (minimum 100 IP) in the entire National League managed that feat last season.
The second explanation is that Lance Berkman was out of the lineup during this span. It's hard to measure his exact impact in terms of runs scored due to the way his presence interacts with the rest of the lineup, but suffice it to say, he is by far our best hitter. In a "down" season last year, he still had the thirteenth highest OPS in the National League.
Finally, there was a fair share of poor luck involved in our poor start. Our batter BABIP (batting average on balls in play) was the lowest in the NL in our first 16 games at .267. Compare this to 2009, when our BABIP was .293, and the lowest-BABIP team was the Reds, at .282. Our pitcher BABIP has also been unusually high in the first 16 games, at .321. Even though pitchers and especially batters do have some amount of control over this stat, these are still extreme outliers and are very unlikely to be sustained over the rest of the season.
So with that out of the way, we come to the subject of our current hot streak. In the last ten games, the Astros are 6-4. The level of competition in these games has still been somewhat higher than we should expect over the course of the rest of the season, but it's a much more reasonable spread of opponents than looking at the entire 16-game record; two of the teams we faced were around average or a little above last season (the Cubs and the Marlins), one was very good (the Cardinals), and last night's game was, finally, against a bad team, the Pirates.
The bad news is, our Pythagorean Expectation in this ten-game span was still lower than the 6-4 record suggests, and was, in fact, well below .500. We scored 32 runs during this stretch while allowing 37 runs to score against us. Our Pythagorean expected win/loss percentage was .428, which is a 69-win pace. Interestingly, this is not far from our current pace based on our overall win/loss percentage (66 wins). However, one must keep in mind that Lance Berkman only played in four out of those ten games, and in those games, he was fresh off the DL and likely a little rusty (not to mention still slow because of his knee injury). His contribution should improve significantly.
If you assume that Berkman is going to add three additional wins, about what he contributed last season (as compared to his injury replacement Geoff Blum), that aligns very close to what CHONE projects for our 2010 season: 73 wins, or the second worst in MLB behind the Toronto Blue Jays.
So the sample size is still too small to make firm predictions, but if you place any stock in the basic measurement of runs scored versus runs allowed, things are not looking good for your Astros. We will not, fortunately, be historically bad, as our current overall Pythagorean Expectation indicates, but we are still facing the possibility of being among the worst teams in baseball, perhaps the very worst.
I'll revisit this subject more succinctly after we've played another tenth of the season, in, oh, about two and a half weeks.
