Now that we've had a few weeks to get over the euphoria after watching the best A's season in at least a decade, it's time to run some numbers. The 2012 A's won far more than anybody predicted they would. They obliterated preseason projections; Dan Szymborski, for example, stated that they had "good odds on losing 100." Here I'll look at some aspects of the underlying performance to get an idea of whether we are dealing with the 2008 Rays or the 2012 Orioles. Without getting too far into individual performances or offseason rosterbation, let's find out what really happened in 2012 and put together some reasonable expectations heading into 2013.
The A's won a lot in 2012, this we know. But did they earn that division title, or did they steal it? There are definitely some ominous surface signs: with 14 walkoff wins, a 25-18 record in one-run games, and an 11-5 record in extra innings, one might be tempted to suggest the club was a beneficiary of a generous sprinkling of luck from the baseball gods. However, this assertion would be only partially correct. First, let's look at the team's pythagorean expectation:
As you might remember, pythagorean expectation was created by Bill James. The intent is to figure out how often a team should be expected to win just based on runs scored and runs allowed. Plugging in the A's 713 runs scored and 614 runs allowed, we get 93 wins. Using this method, it looks like the team wasn't so lucky after all. So far, so good!
Wins above replacement (WAR) gives us another method for figuring out how many games the team should have won. FanGraphs and Baseball-Reference publish the most well-known versions of this stat. FanGraphs had the A's down for 23.7 position player WAR and 18.1 pitcher WAR in 2012. They awarded 669.4 total position player WAR and 460.3 total pitcher WAR across all of MLB. Thus:
81-(669.4+460.3)/30 = 43.3
gives us the FanGraphs replacement level. What I've done here is taken the average number of wins (half the games, so 81) and subtracted the total number of WAR divided by the total number of teams. In other words, based on the amount of WAR FanGraphs awarded, we can conclude that they expected a team of replacement-level players to win about 43 games in 2012, or a .268 winning percentage. Add in the A's WAR and you get a measly 85 games. FanGraphs thought the A's got really lucky this year.
How about Baseball-Reference? B-R gave the A's 18.6 wins above replacement on offense, 0.9 on defense, and 23.0 for pitchers. They awarded 517.4 total position player WAR and 356.9 total pitcher WAR. Using the same formula as above, B-R comes up with a replacement level of about 52 wins, or a .320 winning percentage. Add in what the A's got and you get 94 wins.
Clearly, there is a huge gap between the FanGraphs and B-R assessments of the A's overall performance this season. While there is also a big gap in their replacement levels, they are both reasonably close to the .300 winning percentage many online analysts cite. And if the team's performance relative to the league were equal in both models, the final win totals should have been close. One area that really stands out is the A's share of pitcher WAR. In the FanGraphs version, Oakland got 3.9% of all pitching wins, while the team picked up 6.5% of B-R pitching wins. The main culprit appears to be the bullpen, with Balfour, Cook, Blevins, and Neshek all getting an extra win or more from B-R. Straily and Griffin also fared far better with B-R than with FanGraphs. In any case, simply average the three methods and you get about 91 wins.
Before I get to conclusions, there's one more area I want to look at: bang for the buck. This article uses Marginal Payroll (the amount of money spent beyond what's required to field a minimum-salary team) and Marginal Wins (the number of wins above what a replacement-level team would get) to calculate a Cost Per Marginal Win for each team. And the A's blew away the field this season as the only team that spent less than $1 million per Marginal Win. The worst? The Boston Red Sox, who spent nearly $8 million per MW. Improving the outlook for next year is the following table (numbers from the above-linked Ballpark Digest post):
Among 2013 division-mates, only the Rangers finished in the top half in efficiency, and just barely. The rest of the division appeared to have relative difficulty in translating payroll into wins, which bodes well for a team trying to keep its competitive window open on one of the smallest budgets in baseball.
Overall, while Oakland's 2012 record seems to be cautiously, though not fully, supported by the underlying performances, many of those performances probably weren't sustainable due to regression to the mean. Gomes just had his best year—by far—since 2005. Crisp and Colon are almost guaranteed to both regress and have their underlying skill-sets get worse. Moss, Carter, and Reddick are all strong regression candidates. However, there is reason for optimism as well. The team should be able to count on Anderson for more than 6 starts this season. They added Chris Young, a great bounceback candidate. There is a long list of previously inexperienced (at the MLB level) players who will go into 2013 with a year on the books (Norris, Donaldson, Carter, Cespedes, Milone, Parker, Griffin, Straily, Doolittle, Cook); some of them will surely improve.
With a few tweaks (in particular: get a shortstop) they can probably put a high-80s true talent team on the field in 2013. If you believe player development will equal or outpace the overall regression, then they could be even better than that. But in all likelihood it will take some luck for the team to reach into the 90s in wins and challenge for another division title. However, the outlook is very strong and clearly far better than it has been in several years. I'm optimistic that Oakland will give Texas and Anaheim another good run in 2013.