02 October 2014

MLB Postseason Projections 2014: Giants Win Should Aid Favored Nationals

This post is part of a series tracking the postseason series win probabilities throughout the 2014 MLB Playoffs. Click here to enjoy the full series of posts.

AT&T Park • Photo credit: BspangenbergCreative Commons 3.0

Unlike Tuesday Night's game, the story of last night's San Francisco triumph is one of steadiness and domination. That's good for the Washington Nationals (and, to a lesser extent, for the Los Angeles Dodgers).

According to RP's Power Scores, the Pittsburgh Pirates were a much better team than the Giants. Had the Bucs come through against Bumgarner, the Nats' WS% would have been hovering around 17%. Instead, the Nationals will have home field advantage against San Francisco, and are accordingly poised to win it all with a 19% probability. That puts them fractionally ahead of the Los Angeles Angels, who would enjoy home field advantage in the Fall Classic.



It's saying something about my system's view of the Washington Nationals that they are still expected to win it all despite the fact that the AL has World Series home field. In an alternate universe where the National League won the 2014 All Star Game, the Nationals would have a 21% chance to win it all vs. the Angels' 18% shot. Home field or not, the NL has the slight edge overall as the LDS begins.



Today I'll be focusing on the AL match-ups, as the NLDS does not begin until Friday.

My numbers tell me that the Angels should beat the Kansas City Royals 61% of the time. Kansas City enters the first round as the weakest AL hopeful with a Power Score of only .519. They will starting on the road against the squad with MLB's best record and a .572 Power Score that ranks the Angels third overall.

The most likely outcome is an Angels win at home in Game 5, and every Angels victory scenario is more likely than every Royals' winning scenario. Kansas City's best bet at present is to put the Angels away at home in Game 4.


The second ALDS matchup is the closest of the first round. The Baltimore Orioles, with their home field advantage, should beat the Detroit Tigers about 55% of the time. The most likely outcome is an Orioles win in five, while Detroit's best bet is to advance at home in four.


Above is a table I will be updating throughout the postseason. It lists the most likely World Series outcomes disaggregated as much as possible under my methodology. Though the Nationals and Angels are in a dead heat to win it all, the four most likely scenarios have the Angels taking home the commissioner's trophy.

At the end of it all, we'll see where the final outcome ranked on the list of the most likely. This answer two questions: 1) how accurate was my model? and 2) how unlikely was the final outcome?



Though I never claim to be able to "beat Vegas," I think it's important to put my model up against the world's most active sports prediction markets: Vegas betting parlors. The table above re-presents the RP models' numbers as odds, compares them to Vegas odds (as per Bovada), and then re-scales the Vegas odds so that they add up to 100%.

There is some agreement, some disagreement. Vegas loves the Dodgers more than my model does, the Angels less, and gives more credit to underdogs like St. Louis, Kansas City and Detroit. The only serious disagreements are Vegas' expectations that Detroit and St. Louis have better hopes than Baltimore and San Francisco, respectively.

Interestingly enough, when converted into percentages, Vegas and I reach the same conclusion about the AL-NL balance: Vegas expects the NL to win 51% of the time.


Finally, I am debuting a new chart this year: Interesting* World Series Match-ups. This table presents the likelihood of World Series match-ups that I find particularly interesting, such as the possibility of an October "Battle of the Beltways" between Washington and Baltimore (10%), or an All California Series between the Angels and either the Dodgers or the Giants (18%).

*Your interest may vary.

In contrast, an All-Missouri series seems rather unlikely (2%). Among possible World Series rematches, the most likely would give the Dodgers a chance to avenge their being swept by the 1966 Baltimore Orioles (9%). The Tigers have a 4% shot at facing the Giants again, in what would be the most recent World Series rematch. Finally, there is a 4% chance of replaying the only all-St. Louis series in MLB history, in which the war-depleted Cardinals defeated Pete Gray and the St. Louis Browns in six games. The St. Louis Browns would later move to Baltimore and change their name to the Orioles.

Stop by tomorrow for updated LDS, LCS and WS probability numbers plus deeper analysis of the NLDS.

1 comment:

Kenneth Matinale said...

Radical Baseball: Tournament odds from another blogger: JD Mathewson.

http://radicalbaseball.blogspot.com/2014/10/tournament-odds-from-another-blogger-jd.html