Short answer: For a two-way esports market, convert each decimal price to implied probability, add the two figures and divide each by the total to remove overround proportionally. At 1.80 and 2.10, the raw percentages are 55.56% and 47.62%, totaling 103.17%. The normalized figures are about 53.85% and 46.15%. They remain market-derived estimates, not true probabilities.
Apex Betting is an independent editorial site, not a bookmaker, casino, account service or source of paid picks. This guide is for adults 18+ and explains a method; it does not promise an outcome.
Raw and normalized percentages
The raw conversion is 1 / odds. Normalization uses each raw percentage divided by the complete total. In the example, 55.56 / 103.17 gives roughly 53.85%. Rounding each input too early can leave the normalized pair slightly above or below 100%, so keep extra precision until the final display.
Proportional normalization is transparent and easy to reproduce. It assumes the overround is distributed proportionally, which may not match how a market prices favourite-longshot bias or liability.
Multi-team lobbies need the full field
Apex event-winner markets can contain many teams. Every active outcome belongs in the denominator. Scratched teams, field entries and ‘any other’ selections need explicit treatment. If a page captures only the top five quoted teams, it cannot report the complete market margin or normalized probabilities.
Record whether a team was removed and whether prices were taken before or after that change. Comparing totals across different fields otherwise creates a false movement signal.
Probability is conditional on event definition
A probability for winning one game is not a probability for winning a six-game series or a Match Point final. The roster, stage, patch and settlement rules define the event. A clean model description states those conditions before presenting a percentage.
Historical frequencies require the same care. Qualification creates a selected population, and older patches may not represent current play. Confidence intervals or sensitivity ranges are more honest than a single unexplained decimal.
Test sensitivity instead of claiming an edge
Compare the market-derived percentage with a range of plausible model values rather than one point estimate. If a small roster or patch adjustment reverses the conclusion, the evidence is fragile. Track forecasts prospectively and include all predictions, not only the interesting ones.
Even a calibrated estimate cannot make gambling safe. Keep a fixed personal limit and treat no action as an acceptable outcome whenever the model, market field or rules are incomplete.
Two-way normalization
| Side | Odds | Raw implied | Normalized |
|---|---|---|---|
| A | 1.80 | 55.56% | 53.85% |
| B | 2.10 | 47.62% | 46.15% |
| Total | 103.17% | 100.00% |
Continue inside this guide
Sources and scope
Sources were checked on 12 August 2026. They support the definitions and consumer checks below; they do not endorse this site or any gambling decision.
- EA: ALGS season overview
- EA Help: how Apex Legends Ranked works
- UK Gambling Commission: put a limit on spending
FAQ
Is normalized probability fair probability?
It is a no-overround estimate under a chosen normalization method. It is not independently verified truth.
Can I normalize only the favourites?
No. A complete market calculation requires every mutually exclusive outcome still in the field.
Why do different sites show different percentages?
They may use different prices, timestamps, fields, rules or overround-removal methods.
Related checks on this site
These routes connect this answer with closer checks, methodology and fresh local coverage.
