Filed under:

# MLB 2015 over/under win totals: Vegas vs. regression

The first win totals for the 2015 MLB season were posted by the Atlantis Casino in Reno. We look at what regression to the mean predicts for each team this year, and compare those win totals to the lines.

February is the time when baseball news starts peeping through the snow like seedlings, giving little reminders that warm weather is just around the corner*. The sports book at the Atlantis offered one such sign Friday when they posted the first over/unders of the 2015 season.

* - Sorry for the bad metaphor. We've gotten an Altuve and a half of snow over the last three weeks and I'm going a little stir-crazy.

The numbers are below. The Nationals and Dodgers lead the way with win totals over 90, while the Twins and Phillies are bringing up the rear with win totals under 70.

Going just by last year's records, there are a number of surprises. But the market, of course, is incorporating teams' offseason transactions, as well as public interest (slightly inflating the numbers of popular teams like the Cubs and Red Sox). Keep in mind, these numbers aren't projections; they're designed to get people to put money on one side or the other.

A few years ago, Rany Jazayerli developed a linear regression to make rest-of-season predictions from a small number of early-season games. One component was a team's projected winning percentage given their performance over the last three seasons:

P = .1557 + (.4517 * X1) + (.1401 * X2) + (.0968 * X3),

where X1 is a team's winning percentage from last season, X2 is a team's winning percentage from two years ago, and X3 is a team's winning percentage from three years ago. Let's compare these values of P to the over/unders. The chart below includes two estimates, one based on each team's actual record since 2012, and the other based on their Pythagorean records over that time.

Both methods seem to agree with the over/unders placed on the Brewers, Mets, Tigers, Royals, and Rangers. But there are some large discrepancies based on this winter's fire sales and shopping sprees: the Cubs, Padres, Red Sox, and White Sox have win totals far above their past results, while the Athletics, Braves, Phillies, and Rays all have records far below what their past few seasons would suggest.

It's probably a bad idea to use the Plexiglas Principle to guide your gambling, but this chart (presented in table form below) does highlight the big discrepancies between public perception and each team's past performance. And don't forget, these lines aren't set in stone, and will change as bets start coming in. By the time the first pitch of a spring training game is thrown, this chart will look very different.

Team O/U Exp Wins Exp Wins (Pyth) WP Delta Pyth Delta
Nationals 93 90.2 89.8 2.8 3.2
Dodgers 91 88.9 87.6 2.1 3.4
Angels 87.5 89.0 88.5 -1.5 -1.0
Cardinals 87.5 88.0 85.9 -0.5 1.6
Tigers 86.5 87.5 86.5 -1.0 0.0
Red Sox 86 77.5 79.0 8.5 7.0
Pirates 85.5 85.8 84.3 -0.3 1.2
Mariners 85 81.7 83.0 3.3 2.0
Giants 85 84.7 83.4 0.3 1.6
Orioles 84.5 89.5 87.3 -5.0 -2.8
Padres 84 78.0 76.7 6.0 7.3
Blue Jays 83.5 80.1 81.4 3.4 2.1
Royals 83 84.4 82.5 -1.4 0.5
A's 82.5 87.5 92.1 -5.0 -9.6
White Sox 82 75.3 75.1 6.7 6.9
Marlins 81.5 75.4 76.0 6.1 5.5
Cubs 81.5 73.4 73.4 8.1 8.1
Indians 81 83.1 81.3 -2.1 -0.3
Mets 81 78.5 79.9 2.5 1.1
Yankees 80 84.3 80.4 -4.3 -0.4
Brewers 80 80.7 80.3 -0.7 -0.3
Reds 79 81.6 82.7 -2.6 -3.7
Rays 77.5 81.5 82.5 -4.0 -5.0
Rangers 76.5 77.2 76.9 -0.7 -0.4
Astros 73.5 69.3 70.7 4.2 2.8
Braves 73.5 83.5 83.1 -10.0 -9.6
Diamondbacks 72.5 73.3 75.0 -0.8 -2.5
Rockies 70.5 71.6 76.2 -1.1 -5.7
Twins 68.5 72.5 74.3 -4.0 -5.8
Phillies 67 76.3 75.5 -9.3 -8.5

. . .

All statistics courtesy of Retrosheet.

Bryan Cole is a featured writer for Beyond the Box Score who only knows when to fold 'em. You can follow him on Twitter at @Doctor_Bryan.