We all like to look into the future. The problem is that the future doesn't like to be looked at. Things change, often quickly, and factors we never consider end up making a hash of our best predictions.
Clay Davenport doesn't pretend to look at the future, instead using his advanced math skills to create a simulation (run one million times) that helps us assess the most likely scenarios. Davenport's statistics once made up the heart of the Baseball Prospectus statistical engine, back when Nate Silver's PECOTA was among the most accurate simulations around. Without the Davenport Translations, none of it worked.
Davenport continues to do the hard math work that allows us to look forward to the season. One of his coolest tools is what he calls the Playoff Odds Predictor. Since the POP uses actual results, it gains accuracy as more games are played. Watching the swings of playoff odds is absolutely fascinating in September, and none more so than the last few seasons.
Using his hand-crafted simulator, Davenport recently published his 2013 projections, some of which will surprise you. Let's take a look through some of the findings:
The six division champions as predicted aren't going to surprise anyone. Davenport has Toronto (87 wins), Detroit (92) and Los Angeles (91) winning in the AL. In the NL, the projected winners are Washington (86), Cincinnati (87) and San Francisco (92.)
HOW ABOUT THE WILD CARD?
OK, this gets complicated. In the AL, Davenport has four teams all winning 85 games. A four-way tie would create havoc for baseball's schedulers with Oakland, Texas, New York and Tampa requiring a playoff before the playoffs.
Suffice it to say, a bit of luck—good or bad—is enough to push this from reality. It's going to be very close, however, and small things like a trade or an injury could swing two divisions. In fact, the Astros could end up being a big factor. If they're as bad as many think, those wins shifted into the AL West could push the A's and Rangers into the postseason.
It's easier in the NL, where the Dodgers and Braves take the wild-card slots and face each other in the projected play-in game.
HOW BAD ARE THE ASTROS?
Davenport's system projects the Astros to win 75 games. That's still last in the AL West, but it is significantly higher than other projections. While many have the Astros losing 100 games or more (including myself), Davenport doesn't even have them as the worst team in the league!
That honor and the first pick that goes with it is expected to go to Miami. The latest fire sale in South Florida has brought a team expected to win just 67 games. That's definitely not going to help sell seats in that new park.
WHO IS HIGHER THAN EXPECTED?
Some may be surprised to see the Rays as an 85-win wild card, or the Red Sox improve to 83 wins this season. Neither is truly surprising given the talent and the tendency of records to even out over the course of seasons absent of any major change.
More will be surprised that the Pirates are projected to finish in second place in the NL Central, breaking a 21-year run of losing seasons. While 81 wins isn't a winning record, for the downtrodden fans of Pittsburgh, not having a losing season could be a difference-maker.
WHO IS LOWER THAN EXPECTED?
The Phillies aren't going to be happy to see themselves projected not only in third place in the NL East, but a game under 500 (80-82). Davenport is also calling for Baltimore to take a step back from last year, going 78-84. Part of that is that no projection system is going to think that the Orioles' record in one-run games last season is sustainable.
The Cardinals' drop to 500 might be surprising, but that's a tough division and the injury risks on the team seem to be weighing them down. The Brewers also project lower in Davenport's system than many others, though it admittedly does not take the recent Kyle Lohse signing into account.
NO 100 WIN OR LOSS TEAMS?
Not according to Davenport. In fact, the compression that we've seen increasing over the past couple seasons appears to be even more exaggerated. The spread between the best teams and the worst is just 25 games.
If you throw a blanket over the teams stuck in the middle, just plus/minus five from a 500 record, you'll end up capturing 19 of the 30 teams. That means that a lot of things that look like luck could end up swinging almost all the races, and that competition will be tight all summer.
CAN I USE THESE FOR VEGAS OVER/UNDER WIN TOTALS?
"Entertainment purposes only," people.