The Bet Better Margin Index is a daily CSV of the margin bookmakers build into their odds, by sport, market class and market shape, measured across tens of millions of pre-match markets. Here is how to load it, tidy it and plot it in R.

1. Read the file straight from the web

library(readr)
library(dplyr)
mi <- read_csv("https://betbetter.world/studies/margin-index?format=csv",
show_col_types = FALSE)
glimpse(mi)
# as_of_date, sport, market_class, shape, markets_n, books_n,
# mean_pct, median_pct, min_pct, max_pct

Each row is one day, one sport, one market class (head to head, spread, total, player prop) and one shape (two-way or three-way). mean_pct is the average margin in per cent of stake and markets_n is how many markets were measured that day.

2. The latest day, by sport

latest <- mi %>%
filter(as_of_date == max(as_of_date),
market_class == "h2h", shape == "2way",
sport != "ALL", markets_n >= 50) %>%
arrange(mean_pct) %>%
select(sport, mean_pct, markets_n, books_n)
latest

On most days the NFL and college football sit at the cheap end, about 4.4 cents in the dollar, and T20 cricket at the dear end, near 7.

3. The trend across the season

library(ggplot2)
mi %>%
filter(sport == "ALL", market_class == "h2h", shape == "2way") %>%
ggplot(aes(as_of_date, mean_pct)) +
geom_line() +
labs(x = NULL, y = "Average margin (% of stake)",
title = "What bookmakers keep on a two-way bet, every day",
caption = "Bet Better Margin Index, CC BY 4.0")

Reuse

The file is CC BY 4.0. Cite it as: Glush, E. (2026). The Bet Better Margin Index: daily measured bookmaker margins across market types. https://betbetter.world/studies/margin-index

Data licensed CC BY 4.0, free to reuse with a credit. Model estimates, not advice. 18+. If gambling is causing you harm: gamblinghelponline.org.au.

Originally published at https://betbetter.world/rstats/pull-the-bookmaker-margin-index-in-r. Data is CC BY 4.0, free to reuse with a credit to Bet Better.

Decimal odds hide two numbers: the bookmaker's probability for each outcome, and the cut they take. Ten lines of R separate them.

1. Implied probability

odds <- c(home = 1.85, away = 2.05)
implied <- 1 / odds
implied
#  home  away
# 0.541 0.488

The two implied probabilities add to 1.029, not 1. That extra 2.9 per cent is the overround, the margin built into the prices.

2. The margin in cents per dollar

overround <- sum(implied) - 1
margin_pct <- 100 * overround / sum(implied)
round(margin_pct, 2)
# 2.8   (per cent of every dollar staked, across the market)

3. Remove it to get fair probabilities

fair <- implied / sum(implied)
fair
#  home  away
# 0.526 0.474
fair_odds <- 1 / fair
round(fair_odds, 2)
# 1.90  2.11

This is the simple proportional method. It assumes the margin is spread evenly across outcomes, which it usually is not: bookmakers load more of it onto the longer price. The calibration study measures that tilt across 18,238 settled markets.

4. Do it for a whole file

A clean file of bookmaker prices, one row per priced outcome, is on the Internet Archive under CC BY 4.0:

prices <- read.csv("https://archive.org/download/betbetter-market-prices-2026/betbetter-market-prices-2026.csv")
str(prices)
library(dplyr)
margins <- prices %>%
group_by(sport, game_date, home_team, away_team, market) %>%
summarise(overround = sum(1 / bookmaker_price) - 1,
margin_pct = 100 * overround / sum(1 / bookmaker_price),
sides = n(), .groups = "drop") %>%
filter(sides == 2)
summary(margins$margin_pct)

Group by market, apply the three steps, and you have a margin per market that you can compare by sport, bookmaker or date. That is exactly how the Margin Index is built each day.

Data licensed CC BY 4.0, free to reuse with a credit. Model estimates, not advice. 18+. If gambling is causing you harm: gamblinghelponline.org.au.

Originally published at https://betbetter.world/rstats/odds-to-probabilities-in-r. Data is CC BY 4.0, free to reuse with a credit to Bet Better.

A calibration curve asks one question of a set of probabilities: when the market said 60 per cent, did the outcome happen about 60 per cent of the time? Here is the whole thing in R, using the open market calibration dataset: 36,476 outcome observations, both sides of 18,238 settled two-way markets, vig removed, no bookmaker names and no bettor selections.

1. Load

Download market_calibration.csv from the Kaggle mirror (CC BY 4.0), then:

library(dplyr); library(ggplot2)
cal <- read.csv("market_calibration.csv")
# columns: event_month, sport, market_family, market_implied_probability, outcome

2. Bin and compare

curve <- cal %>%
mutate(bin = cut(market_implied_probability, seq(0, 1, 0.1), include.lowest = TRUE)) %>%
group_by(bin) %>%
summarise(predicted = mean(market_implied_probability),
observed  = mean(outcome),
n = n())
curve

3. Plot against the diagonal

ggplot(curve, aes(predicted, observed)) +
geom_abline(slope = 1, intercept = 0, linetype = 2) +
geom_line() + geom_point(aes(size = n)) +
coord_equal(xlim = c(0, 1), ylim = c(0, 1)) +
labs(x = "Market probability (vig removed)", y = "Observed frequency",
title = "Betting markets are well calibrated",
caption = "Bet Better market calibration dataset, CC BY 4.0")

Every decile lands within about two points of the diagonal. Look closely and the low bins sit slightly above the line: outcomes priced under 50 per cent happen a touch more often than their vig-free price says, which is the favourite-longshot tilt showing through after the margin is stripped out.

4. Split by market family

cal %>% count(market_family)
# repeat the binning with group_by(market_family, bin):
# player props tilt about twice as much as game lines

Data licensed CC BY 4.0, free to reuse with a credit. Model estimates, not advice. 18+. If gambling is causing you harm: gamblinghelponline.org.au.

Originally published at https://betbetter.world/rstats/calibration-curve-in-r. Data is CC BY 4.0, free to reuse with a credit to Bet Better.