Rating Gap to Win Probability Calculator
Enter a rating gap and a league to see the model's win probability.
- Edge the model sees
- +98 rating points (+50 gap + 48 home edge)
- Away win
- 38.1%
- Home spread
- −4
- Home moneyline, no vig
- −163
- The model's constants
- NFL: scale 0.00557, shift -0.059, home edge 48 rating points, 0.0403 points per rating point (fitted on 1,260 games)
The model's logistic: probability = 1 / (1 + e^−(scale × edge + shift)); soccer adds a draw band, P(home) = σ(z − w), P(away) = 1 − σ(z + w). With no fit, scale is ln(10) / 400 and shift is zero.
Educational tool. Bandwagon does not take or place wagers. Results are arithmetic on the numbers you enter, not advice.
How the rating Gap to Win Probability Calculator works
Ratings are Elo-style with a 1500 mean (docs/ODDS.md). Before a game the home side gets the league's home advantage in rating points; the edge is home − away + that. Win probability is 1 / (1 + e^−(scale × edge + shift)).
Scale starts at ln(10) / 400, the classic Elo curve, and the nightly fit replaces it and the shift with values that minimize log loss on finished games. Soccer adds a draw band of fitted half-width w: P(home) = σ(z − w), P(away) = 1 − σ(z + w), the rest is a draw.
Every constant is printed under the result, so you can reproduce it by hand. On a neutral site the home edge is dropped.
Rating Gap to Win Probability Calculator example
You enter
- League
- NFL
- Rating gap (home − away)
- 50
- Where
- Home field
You get
- Edge the model sees
- +98 rating points (+50 gap + 48 home edge)
- Away win
- 38.1%
- Home spread
- −4
- Home moneyline, no vig
- −163
- The model's constants
- NFL: scale 0.00557, shift -0.059, home edge 48 rating points, 0.0403 points per rating point (fitted on 1,260 games)
In the NFL a home team rated 50 points above its opponent has an edge of 98 points once the 48-point home advantage is added, a 63.7% win probability, and a line of −4.
Rating Gap to Win Probability Calculator questions
What is a rating gap?
The difference between two teams' Elo-style ratings, home minus away, before home advantage.
What is the home advantage in rating points?
A per-league constant, 48 in the NFL, 100 in the NBA. It is shown under each result.
Why are there constants for soccer's draw?
Soccer ends level often enough that a two-way logistic is not enough, so an ordered logistic adds a draw band.
Where do the ratings come from?
Bandwagon recomputes them every night from every final score. They are on /odds/<league>.
Related calculators
Win Probability to Spread Calculator
Turn a win probability into the point spread the model would set, per sport.
Moneyline to Spread Calculator
Convert a moneyline to the point spread it is worth, with Bandwagon's fitted constants per sport.
Spread to Moneyline Calculator
Convert a point spread to the fair moneyline and win chance, per sport.
Live on Bandwagon
Convert an Elo-style rating gap to a win probability using Bandwagon's logistic and its per-sport constants: scale, shift, home edge and points per rating point.