If you pull up a Formula 1 betting market online, it will look like this:
Source: https://www.oddschecker.com/us/motorsport/formula-one/singapore-grand-prix/winner
In this post we will look at how the F1 betting market works and relate this back to driver win probability.
American Betting Spreads
F1 spreads or “odds” are of the American variety (consider that horse race betting was the most popular form of gambling in the U.S. for a long time).
There are just a few things to know about American betting spreads:
(1) “Favorites” have a greater than 50% chance of winning, and have negative spreads. “Underdogs” have a lower than 50% chance of winning, and have positive spreads.
(2) A spread of -100 for Favorites or +100 for Underdogs equals a 50% win probability. Therefore, you won’t see any spreads between -100 and +100. If a Favorite had a spread of 0, they would have a win probability of 0 and would be an Underdog, and vice versa.
(3) The functions converting spreads to win probabilities (see code below) have an exponential shape. Therefore you can see very high numbers in absolute value that are associated with very small/large win probabilities.
Below are all 20 driver spreads for the upcoming Singapore Grand Prix (from the above source).
We will record these in R, which after many races will give us a nice data set for other analysis.
Below are the formulas to convert betting spreads to probabilities; these are the blue line charts above.
It’s worth it to set up a function this way, rather than just calculate it once to make a column in the dataframe.
Then any time we want to convert a spread to a probability, we can just use underdog_prob() or fave_prob()
Now we add probabilities to our dataframe.
This says if the spread is negative, calculate the win probability with fave_prob(), otherwise use underdog_prob()
The sum of the 20 driver probabilities (sum_prob) should equal 100%, but the betting market is not perfectly efficient.
(The fact that betting markets are inefficient is why it may be worth trying to make money in them. More detail on that in the future.)
The total of all 20 driver win probabilities happens to be 112%.
Let’s visualize the win probabilities for all drivers on a chart.
Despite the presence of 20 drivers on the grid, the betting market has assigned most of them no shot at winning the race in Singapore.
Only the drivers from Red Bull, Ferrari and Mercedes have a non-negligible win probability.
The spreads do make sense considering the record for this year’s 16 races so far:
Vertappen 11, Leclerc 3, Perez 1, Sainz 1
(Red Bull 12, Ferrari 4)
But it may be interesting to assume something unexpected could happen with a probability greater than 0.
So let’s make some bets!
For the below bets, I used betonline.ag – note you pay about 10% of your cash to play (grab yourself a free drink!)
Bet 1: The Field over Max Verstappen
Max’s spread of -215 implies a 68% win probability.
That happens to match his record this year, 11/16
Nonetheless, just for fun, I am betting on the Underdog which is The Field (anyone else) for +145 or 41%.
That is a little bit of a rip off, because I should be getting this for 32% (100% – Max’s 68%).
The good news is that if Max loses, my winnings will be better than 2:1, having bet on the Underdog.
Potential Winnings = ($36.25 + $25.00) / $25.00 = 2.45x
Note that the winnings ratio is the inverse of the probability, i.e. 1/41% = 2.45, and 1/2.45 = 41%
Bet 2: Charles Leclerc over The Field
Charles’ win probability of 20% also lines up with his record this year, 3/16
We don’t have a reason to think this is a “cheap” bet, but if anyone is going to beat Max, it would probably be Charles.
By the way, Charles won the last Singapore Grand Prix which was in 2019.
Also I like him.
Charles’ Underdog spread is +400 (20% win probability)
Potential winnings = ($100.00 + $25.00) / $25.00 = 5x
Here the same inverse relationship between earnings and probability applies, 1/20% = 5, and 1/5 = 20%
Bet 3: Haas over Alpha Tauri
Finally, there are betting markets for team vs. team set ups as well.
Much more to come later on Haas, the American team, but for now I’ll bet on them over AlphaTauri.
Note that this year so far, Haas is in 7th place with 34 points, and AlphaTauri is in 8th place with 33 points.
Haas is the Underdog with a spread of +120 or a win probability over AlphaTauri of 45%
Potential winnings = ($30.00 + $25.00) / $25.00 = 2.2x
1/2.2x = 45%, 1/45% = 2.2x
Please leave your questions or comments!
Enjoy the (night!) race in Singapore on Sunday October 2 @ 8:00am ET!