Analyzing Predictions

As I was writing my predictions last week, particularly the segments about teams with big shake-ups over the off-season, I kept thinking about how these teams would be over- or underestimated. Two things got me thinking about this in particular. One was Sports Illustrated’s prediction on the Marlins, which put them second in the NL East. As I pointed out last time, they were 17 games behind the Braves and 8.5 games behind the Nationals in 2011, and although the Marlins went out and made a lot of big moves, they have just as many what-ifs to worry about. They’ll likely improve, but it probably won’t be enough for them to leap into second place.

The other thing was ESPN’s predictions from last year. If you’ve forgotten, ESPN’s 48 experts predicted the Red Sox or Phillies to win the 2011 World Series 40 times and (lose it 29 times). Relatedly, the Red Sox added Carl Crawford and Adrian Gonzalez, while the Phillies added Cliff Lee. The Red Sox, as you may recall, didn’t even make the playoffs, while the Phillies lost in the first round. For an event as random as the MLB playoffs, there seemed to be a unanimity in picking teams that had made big offseason moves. Making the playoffs is very close to an even playing field for the entrants, and yet, the results didn’t seem to match that at all.

So, I decided to look at ESPN’s predictions for this season to see if I could pick up on this pattern again. This year, they had 50 experts predict, and the answers were much more spread out.

The picks to make the playoffs weren’t particularly notable. The picks do seem to line up with which teams are best.

Team-Times predicted to make the playoffs
Tigers-50
Yankees-49
Rangers-49
Angels-46
Phillies-44
Dbacks-43
Rays-38
Marlins-29
Giants-28
Reds-27
Cardinals-26
Brewers-17
Red Sox-16
Nationals-15
Braves-14
Dodgers-5
Rockies-2
Blue Jays-2

I really don’t see any problems with that, to be honest. The best teams seem to be near the top (also, as a note, 18 teams got at least one playoff appearance vote). The World Series picks are different though, I think. (Note: 15 teams were picked to appear in a World Series.)

Team-Predicted World Series-Predicted Wins-Win%
Angels-21-18-.857
Diamondbacks-16-0-.000
Phillies-14-5-.357
Rangers-9-8-.889
Rays-8-7-.875
Tigers-7-6-.857
Cardinals-5-2-.400
Yankees-4-3-.750
Reds-4-0-.000
Giants-3-0-.000
Braves-3-0-.000
Brewers-2-0-.000
Nationals-2-0-.000
Red Sox-1-1-1.000
Marlins-1-0-.000

With the Diamondbacks and Phillies near the top, the appearances list doesn’t seem to lean to teams that had the splashiest off seasons. But, and maybe this is just me, it looks like the teams with higher winning percentages were the ones that made the biggest off season moves. In theory, a seven game series is pretty close to a toss-up, and yet, the Angels (Albert Pujols, C.J. Wilson), Rangers (Yu Darvish), Tigers (Prince Fielder), and Yankees (Hiroki Kuroda and Michael Pineda) all appear near the top. Also, those four combine for 35 of the picks. The Rays are the only outliers, carrying an .875 winning percentage despite their big acquisitions being Carlos Pena and Luke Scott.

Conversely, the Diamondbacks, Cardinals, Phillies, Reds, Giants, Braves, and Brewers all appear under .500. Between these seven teams, the big acquisitions were Trevor Cahill, Carlos Beltran, and Aramis Ramirez (and the latter two came at the expense of Albert Pujols and Prince Fielder, so they probably register as net losses). The Nationals, Red Sox, and Marlins are also exceptions. However, both the Nationals and Marlins would qualify as huge improvements by just making the World Series, and the three teams only appear in four World Series between them anyway, so the results may be not as telling.

I’m not sure how much these results actually mean; it could be that I’m missing something obvious (and please, if I am, feel free to point it out). However, it does seem like writers (or, at the very least, the writers at ESPN) seem to pick teams that added a lot of parts in the offseason to win it all more frequently, despite the facts that the playoffs are fairly random. Maybe this just means that the writers think these are the best teams in the league. But, for example, do the Angels’ additions really make up the, say, 16 game difference that existed between them and the Phillies in 2011? Even if the Angels are better than the Phillies now (which is definitely a possibility), is it enough for there to be a .500 point swing in their predicted winning percentages once they reach the World Series. I don’t know for sure; it’s interesting, though, to say the least.