One of the biggest hangups that I have had with the concept and model for a baseball fund is that the money line style of betting is confusing. For anyone who is used to the point spread betting that is used in football and basketball, money lines will be like learning a foreign language.
In order to build a model that I can use to assess the odds of each game, I will need to have the ability to quickly convert what the sportsbooks say about a team’s chances of winning and compare that to an expert’s assessment of those odds. The Vegas opinion on each game is represented in a money line, while the experts opinion on each team overall is represented by total wins for a season. I will have to convert both into a percentage so that I can compare apples to apples.
Converting Money Lines into Percentages
My trading model is going to be based on converting the standard money line for each game into a percentage that represents each team’s chances of actually winning the game. This percentage represents the implied odds of the team winning an individual game.
I found an article called How To Convert Betting Odds To Their Implied Probability by a dude named Andrew Brocker. In that post, Andrew breaks down the simple formulas for converting a money line number into a decimal probability.
For favorites, which are represented by a negative money line, we take the absolute value of the money line number and divide it by the same number plus 100. For example, on Monday the Red Sox were listed as favorites against the Cardinals at -130. We would simply take 130 / (130 + 100) and get .565 as their odds of winning.
For underdogs, which are represented by a positive money line, we use the exact same formula, but this time there are no absolute values to concern ourselves with. Using the same example, the Cardinals were listed at +110, so we would take 110 / (110+100) and get .523 for their odds of winning.
Note that the two winning percentages total to more than 100%. That is because of the juice that Sportsbook.com is charging to take the bets. This is a cost of doing business similar to commissions and slippage in a trading account, and it is something we will have to contend with.
Converting Win Totals into Implied Odds
The process of converting win totals into implied odds is going to be even simpler. We are just going to divide the number of games we project each team to win over the course of a season by the total number of games they will play (162). We will continue to use Baseball Prospectus‘s PECOTA projections for these numbers.
Sticking with the same example, BP projects the Red Sox to be an 88 win team this season and the Cardinals to be an 87 win team. That converts to implied odds of .543 for the Red Sox and .537 for the Cardinals. These implied odds show that the teams are a closer match than the money line gives them credit for, which means the Cardinals are probably the better bet in this instance.
Adjusting the Win Total Projections
One major factor that the money line is likely considering that our projections overlook is the impact of the starting pitcher. The best way that I can think of to work this into the model is to simply add the Wins Above Replacement Player (WARP) value that PECOTA gives each starting pitcher to the teams win totals.
In the example we have been using, Shelby Miller will be starting for the Cardinals. He is projected to have a WARP of 0.4 in 2014, which would make the Cardinals an 87.4 win team, making their implied odds .540.
John Lackey will be starting for the Red Sox. He is projected to post a WARP of 0.6 this year. Moving the Red Sox to an 88.6 win team adjusts their implied odds to .547.
Based on this assessment, Shelby Miller and the Cardinals have a slightly better chance of winning the game than Vegas is giving them credit for, and the Red Sox have a slightly lower probability of winning than they are getting credit for.
However, since this is just a spring training game, we would be crazy to consider betting because neither team’s priority is winning the game. It is entirely possible that the Cardinals could have Shelby Miller throw nothing but changeups in an attempt to develop a third pitch. It is also possible that the Red Sox pull Lackey after the second inning just as a precaution.
In the next baseball fund post, we will look to develop our projection model a bit further and then attempt to build some sort of spreadsheet that will make it possible to quickly process every game on a daily basis looking for opportunities.