Alright gang – today I’m going to talk about Closing Line Value (CLV): 1) what it is, 2) how to measure it, 3) why it’s useful and 4) ultimately why it’s flawed. This seems to be a polarizing topic so I’m expecting some disagreement and backlash.
I’ll cover 1) and 2) in today’s post and 3) and 4) tomorrow in Part II.
Closing Line Value
For those of you who are new to sports betting, the “Closing Line” is the line/odds of a game when the market for a game closes (i.e. just prior to kickoff/first pitch/tip off, etc.). Closing Line Value (CLV) is simply a comparison between 1) the line/odds that your bet was placed at and 2) the Closing Line.
The theory behind CLV is that if you’re getting a line better than what is offered at the close of the market, that’s generally a good thing. Simple example: you bet the Yankees at -125 and they closed at -150. You got positive CLV. Congrats!
Measuring CLV
Unfortunately, there is no standard approach to measuring CLV.
The Casual Approach: Casually, folks would say you got “25 cents” of CLV. Clearly this is a good thing, as a $100 bet at -125 would win $80, while a $100 bet at -150 would only win $67.
The Win Probability Approach: To get slightly more technical, we can compare the breakeven win probability of your bet at -125 vs the closing line of -150. The breakeven win probability of -150 is 60.0% while the breakeven win probability of -125 is 55.6%. The difference of 4.4% in breakeven win probability is another way to quote your CLV.
The Expected Value Approach: A third approach is to measure CLV based on the expected value of the bet. If you made a bet at a breakeven probability of 55.6% and the closing breakeven probability is 60.0%, you could say that “price” of your bet increased from 55.6% to 60.0% (increase of 4.4%). Therefore your “return” (increase in value) was 4.4% / 55.6% = 8.0%.
Removing Vig
Some people prefer to review their CLV absent the book’s vig. To make this adjustment, we simply remove the half of the vig for that bet (we assume half the vig is charged on both sides of the bet).
Assuming a standard 10-cent baseball line (+140/-150) we would have a closing vig of 1.6%. Our no-vig CLV measurements would be as follows:
The Casual Approach:With a closing line of +140/-150, we estimate that the “fair” price of the favorite is -145. Thus, a comparison of your bet at -125 and the fair price of -145 would only yield “20 cents” of CLV.
The Win Probability Approach: Subtracting half the vig from our breakeven win probability yields a no-vig CLV of 3.6% (4.4% - 0.8%).
The Expected Value Approach: The closing breakeven probability of -145 is 59.2% so the “price” of your bet increased from 55.6% to 59.2% (increase of 3.6%). Therefore your “return” (increase in value) was 3.6% / 55.6% = 6.5%.
While you’re free to measure CLV however you feel like it, theoretically the no-vig expected value approach should best estimate your long-term return based on CLV.
CLV for Point Spread and Totals
To measure CLV for points spreads or totals using the Win Probability Approach or the Expected Value Approach, you need to estimate the push probabilities of the numbers that were crossed (i.e. if you bet -2.5/-110 and the market closed at -3.5/-110, you crossed the 3). You can then compare your bet with the implied “fair” moneyline of your bet based on the closing line. Referencing our NCAAB half point price of 9 cents on the 3, we estimate -2.5/-128 to be the equivalent of -3.5/-110. You can then calculate your CLV just as you had before.