The pre-season narratives for the Orioles and Nationals were as different from one another as they were similar in the national media and the online stats community. No matter where you looked, the 2012 Orioles were a fluke and the 2012 Nationals were the beginning of a dynasty, even though both teams progressed to their respective Division Series and got bounced at the same time.
The national media narrative had the Nationals as a sexy World Series pick and the Orioles returning to cellar-dweller status in the tough AL East. PECOTA, the sabermetric darling of a system, projected (not predicted, but still, projected) the Nationals to win 87 games and the Orioles to win a mere 75.
On paper, there was plenty of logic behind this. The Nationals had what looked to be the best young rotation in baseball, combined with a solid offense that included budding superstar Bryce Harper. The Orioles had a middling rotation and an unbalanced offense, and had won 93 games in 2012 largely on the strength of an superhuman bullpen. I didn't buy it as far as the Orioles went, but I absolutely believed the Nationals narrative (in fact, I secretly dreamed of a DC-Baltimore World Series to turn the regional rivalry into a credible one).
With 117 games in the books as of Sunday, though, the teams' fates are almost exactly reversed from the projections and predictions of writers and statisticians alike. The Orioles are winning at a .556 clip (65-52), while the Nationals are losing to the tune of .487 (57-60). Neither team is far out of line with its Pythagorean record, so their seasons haven't been fluky from the runs scored/runs allowed viewpoint, either. So what happened?
First, some numbers (through Sunday's play):
|Awful/Amazing 3B play ratio *||1/50||1,000/2|
|Games lost to wall collisions **||0||31|
* Non-factual, but meaningful
** Factual, but mean-spirited and non-meaningful
These statistics tell a cumulative story that's been told many times about both teams since the season started to take shape. The Nationals are pitching well -- but not quite as well as expected -- and struggling to score runs. The Orioles are getting middling pitching, excellent defense, and piling on the runs. That's the what, but really it's the why that's worth exploring.
In the link above (or, if you're lazy, right here), I rattled off a pre-season list of reasons the Orioles were going to outperform expectations. In short, the Orioles were mostly lucky in the first half of 2012, not the second half -- and the 2013 club looks a lot more like the second-half team than the first one. That's mostly borne out, with the added bonuses of surprise MVP candidate Chris Davis, Brooks Robinson reborn as Manny Machado, and Chris Tillman refusing to regress to the mean. Not to mention that almost the entire roster was in or entering its prime. Put it all together, and you have a team that scores plenty of runs, allows few enough runs with pitching most days, and prevents tons of runs with its defense.
In short, even if the 2012 Orioles had their share of luck in winning 93 games, there was no real reason to believe that the 2013 club was destined to be a 75-win disappointment.
And then you have the Nationals. Winning an MLB-best 98 games in 2012, they brought back a largely unchanged roster, tinkering around the edges by trading away Michael Morse, and obtaining Denard Span, Rafael Soriano and Dan Haren. That's the position the team thought it was in -- they could sign Dan Haren as their #5 starter just because he was there.
And now the pre-season darlings find themselves below .500 and out of the playoff race in all but the most strict mathematical sense. What happened? It's easy to point the finger at some injuries -- Harper, Stephen Strasburg, Jayson Werth and Ross Detwiler have all spent time on the DL -- but every team deals with injuries, and none of those injuries were for a significant portion of the season.
The simple explanation is that the offense hasn't hit (Span, Adam LaRoche and Danny Espinosa have all been wildly below expectations), and that the pitching has been good but not as elite as it was billed (sixth-best ERA in the NL). But belying that point is something a little less obvious. The number of individuals on the Nationals' roster who obviously had seasons above their normal capabilities -- if not out-and-out career years -- in 2012 is obvious in hindsight. LaRoche was never going to hit like he did in 2012 this year, his age 33 season. Span was brought in as a surefire leadoff hitter, when in fact 2012 was his first above-average offensive season (by OPS+) since 2009. Espinosa had never really adjusted to the league and proven he was a major-league hitter, especially against righties. In reality, Ian Desmond is the only National who blew up in 2012 and sustained his performance into 2013.
One of the chief flaws of PECOTA and subjective sportswriting alike is that they overemphasize the recent past. PECOTA primarily looks at the past three seasons' worth of data, not factoring in injury history or typical aging trends nearly as much as it probably should. As such, a ballclub full of players just entering their prime, or successful reclamation projects (like the Orioles) might come out looking a lot worse than their actual value, while a ballclub full of veterans just exiting their prime, or frequent injury risks (like the Nationals) might look a lot better. It's easy for a casual analyst to fall into the same trap, whether they're actively relying on a statistical projection system or not. Teams like the Nationals and Giants that had sustainable-looking success, and brought back mostly the same clubs, seem like surefire winners, when in fact they might just be a slightly older club with some vulnerabilities that didn't come to the forefront the year before.
One of the reasons baseball is the greatest professional sport is the fact that anything can happen in an at-bat, a game, or a season. I'm not trying to say that the seasons of the Orioles and Nationals are nails in PECOTA's coffin, or that they prove all sportswriter predictions to be idiotic. They do, however, demonstrate the folly of saying anything about a baseball season that hasn't started yet with any kind of certitude, and they definitely seem to demonstrate that the conclusions that seem the most obvious based on the recent past may not tell a casual observer as much as they seem to. The unpredictability of the 162-game grind of a baseball season will always surmount even the sharpest minds, and there's really nothing wrong about that at all.