Building an NFL Model - Part 3 | Betting Markets

G’day, lads! In this third and final article on the NFL Model, I’ll be calculating the probabilities for the three main NFL markets: Moneyline, Spread and Total Points. All of this will be based on the data from our simulations and the distribution they’ve produced.

If you haven’t already, download the NFL Model for daily analysis and join the most active sports modelling community on YouTube!

Download the ‘Excel LADZ - 2024 NFL Model’: https://www.patreon.com/posts/excel-ladz-2024-111333965

Moneyline

Consider the matchup between the Browns and the Jaguars shown below. The Jaguars are slight favourites due to their small expected points advantage, reflected in their win percentage of 50.94%. This means they are expected to win 50.94% of the time. On the other hand, the Browns are slight underdogs with a win probability of 49.06%. As mentioned in Part 2, this probability is based on the simulations generated in the “Simulations” worksheet. Out of the 5,000 simulations run, the Jaguars won 50.94% of the time, which is 2,547 simulations, leaving the remaining 2,453 games for the Browns.

To calculate this, I’ve used the SUMPRODUCT function shown above. This formula compares the points scored by the Browns in each of the 5,000 simulations (in column B) with the points scored by the Jaguars (in column C). It checks whether the Browns’ simulated points exceed the Jaguars’ for each game and counts the number of times this occurs. Dividing by 5,000 gives the percentage of simulations in which the Browns won.

On a technical note, you may be wondering why the double negative “--” is necessary at the start of the SUMPRODUCT function. The double negative (--) in the SUMPRODUCT function converts logical TRUE and FALSE values into numerical 1’s and 0’s. When the Browns’ simulated points exceed the Jaguars’, the comparison returns TRUE, and the first negative turns it into -1. The second negative then converts -1 into 1, making it usable by SUMPRODUCT. Without this conversion, SUMPRODUCT would not be able to sum up the results correctly. This process allows the function to count how many times the Browns win in the simulations.

Spread

The calculation for the spread is similar to that of the moneyline, with one key difference: a team is given a points advantage or disadvantage based on the spread set by bookmakers. Spread betting, unlike the moneyline, focuses not just on which team wins, but by how much they win or lose. In this case, the Browns are considered the underdogs and are given a 1.5 point advantage. This means that in every simulation, 1.5 points are automatically added to the Browns’ score (in Column B of the ‘Simulations’ worksheet) when comparing it to the Jaguars’ (in Column C). As a result, the Browns’ chances of winning increase from 49.06% to 51.34%, since the additional points make them more likely to cover the spread. Spread betting allows for more balanced odds, even when one team is favoured, by adjusting the point margin.

As seen above, the probability of the Browns covering the spread is 51.34%. To convert this probability into American odds, we can use a simple formula. For probabilities greater than 50%, American odds are calculated as: -100 * (p / (1 - p)), where p is the probability. With a 51.34% probability, the corresponding American odds would be around - 106, meaning you would need to bet $106 to win $100. If the probability were less than 50%, you would use the formula: (100 / p) - 100 to find the positive American odds. By adjusting the spread in cells B10 & B11 (for example, to 10.5 & -10.5), you can significantly alter the probability, and consequently, the odds, reflecting the new chances of the Browns covering the spread.

Total Points

In total points betting (also called “Over/Under”), you’re betting on whether the combined score of both teams in a game will be over or under a set threshold determined by the bookmaker. This market focuses solely on the total number of points scored, without concern for which team wins the game.

To calculate the probability of a match finishing with more than a certain number of points, the SUMPRODUCT function sums the simulated points for both the Browns (Column B) and Jaguars (Column C) and counts the instances where the total exceeds the selected threshold. This gives us the percentage of simulations where the game went “over”, which can then be used to assess the likelihood of hitting the over in total points betting. To find the “under” probability, I’ll switch the greater than sign, “>”, to less than, “<”.

Finding Trades with Positive Expected Value

By comparing the model’s probabilities for the Moneyline, Spread or Total Points with the implied probabilities from the bookmaker’s odds, you can spot bets that are more likely to win than the odds suggest. These bets represent positive expected value, giving you an edge in the long run.

Imagine sportsbooks are offering +150 odds on the Browns to win. To determine the implied probability, given the odds are positive, I’ll use the formula: 100 / (Odds + 100) = 100 / (150 + 100) = 40%. Now, there’s a clear difference between the sportsbook’s implied probability (40%) and the model’s (49.06%). As a result, betting on the Browns at +150 is advantageous because, over time, you would expect to profit more from the model’s higher predicted probability compared to the implied odds offered by the sportsbook.

Thanks for reading lads! JOIN to download the Excel LADZ - NFL Model and become part of the greatest sports modelling community on YouTube!