In previous posts, we’ve described various models and
algorithms we use to handicap football and basketball. In baseball we don’t use models. We start with the basic premise that the
worst team is going to win 60 games and the best team is going to lose 60
games, it’s the other 42 games that determines its season. Therefore we use a criterion to find when the
underdogs are going to win their 60 games and the favorite is going to lose
their 60 games at the right price.
Before we get into discussing the specifics around the
criterion one MUST understand prices when betting a money line sport. Obviously, each money line converts to a
different win percentage to be profitable.
For example a bet at +145 has to win at a much lower percentage to be
profitable than +105. An easy way to
illustrate why price is important is to maybe compare a 5 cent difference on
each bet (say +145 vs. +140) when played in 600 games over a baseball
season. If you hit at 50% you would earn
an extra +1500 (300 wins x 5 cents). At
$100 a game, that would be an extra $1,500 in your pocket, and an extra 2.5%
return on your bets. In a business where
a 10% return is considered outstanding (and maybe even unsustainable) getting a
few extra cents each day can make an enormous difference. If you don’t understand this it won’t matter
what criteria or approach you use, it will be extremely difficult to make a
profit betting on baseball.
The following is the specific measures we use in our criteria
and how we use them. Our back testing
has shown that each of these increases profitability. We’ve described what basic concept we use
and some details around it.
Underdogs: We only play underdogs. If you’re a favorite, you’re not on our
card. To start most people don’t like
playing underdogs. They don’t win as
much. They are not as good as their
opponent, and people simply don’t understand the pricing of underdogs. This leads to mispriced underdogs across the
board. There has been a multitude of
articles and even an academic study showing that betting on underdogs can be profitable.
Run Differential: After
compiling the list of all the days’ underdogs, the next thing you want to do is
throw out the super shitty teams. Even
though underdogs are profitable, not all are.
This is what we use to screen out betting on teams that are going to
lose more than their fair share (maybe even more that they don’t win 60). Run differential (“run dif”) is the amount of
runs scored – runs allowed. Run dif is widely
considered to be a good predictor of a teams quality. The way we
screen using run differential is we throw out any team who has a negative run
differential that is a great than a certain percentage of their games
played. For example if you wanted to use
60% as the screen and a team has played 30 games, if their run differential is
worse than -18 (30 x .60) than you would throw them out.
Above Average
Pitchers: Pitching is a huge
determinant in both price and outcome.
We want to eliminate teams playing above average pitchers. There’s a multitude of stats and advanced
stats you can use. In the past we’ve
used Jeff Sagarin’s power rankings. You
can use ERA, FIP (fielding independent pitching) or whatever you want. You can throw out teams playing the best 15,
20 or 25 pitchers in MLB or a giving league based on any of the measures
above. The idea is to determine some
sort of ranking that eliminates teams facing above average pitchers.
We happen to like using FIP as a
measure and eliminate top pitchers from the previous month. Why FIP and the previous month? For one FIP is a better measure than ERA and
for 2 you can search top pitchers by month from previous years at www.fangraphs
.com. This aided with back testing and
knowing when we actually go to bet we can use the previous months ranking. We had used Sagarin’s rankings in the past,
but you can only back test against previous year’s FINAL standings. This can be difficult when you’re in the
current year because you don’t know if
that pitcher your betting will be in the top 15, 20, 25, etc. that year.
Pricing: The last
metric is around pricing. We did some
back testing to see if there was a certain minimum and maximum price where the
rest of the criteria started to see less of a return. From 2009-2011 playing underdogs greater than
+180 (i.e. +190) led to less of a return.
Playing teams at less than +110 made a difference in some years, but not
in others so we go ahead and play any dog less than +180.
Once
you’ve written down all the underdogs, scratched off the crappy teams and eliminated
any team playing an ace pitcher, you’re left with who we usually bet on. Who is that you may ask? Well it often ends up being mediocre or
better teams, against mediocre or worse pitchers at prices where you may not
even need to hit 50% to turn a profit.
What you’ve also probably ended up with is a team that is likely playing
in one of its 60 games it should win, 60 games it opponent should lose, or even
if its in one of the 42 games that’s 50/50 you’re still getting +$$$. Leave thoughts or questions in the comments or on Twitter. What other criteria makes sense to you?
MLB for May 9th
COL +115
OAK +125
TB +125
BAL +120
SEA +115
SF +120