Learn
Understanding odds and simulation results
A practical introduction to the mathematics used throughout Novus Odds.
American odds and implied probability
Positive American odds describe the profit returned on a 100-unit winning stake. For example, +300 represents 300 units of profit on a 100-unit stake. Negative odds describe the amount that must be risked to generate 100 units of profit.
Every price can be converted into an implied break-even probability. A +300 price implies a 25% break-even probability before considering differences between the market price and the true probability of the outcome.
Expected value
Expected value combines the probability of winning, the payoff when winning, and the loss when losing. Positive expected value means the mathematical expectation is above zero under the assumed probability model. Negative expected value means the expected loss is greater than the expected gain.
Expected value does not guarantee the result of an individual event or a limited sequence of events. Variance can produce long periods where observed outcomes differ substantially from mathematical expectation.
Why win rate is not enough
A strategy that wins 20% of the time can theoretically outperform a strategy that wins 70% of the time when the prices paid for those outcomes are sufficiently different. Betting analysis therefore requires both probability and price.
Novus Odds displays win rate alongside implied probability, simulated true probability, expected return, and observed return so results are not interpreted using win percentage alone.
Monte Carlo simulation
Monte Carlo simulation repeatedly samples uncertain outcomes to study the range and distribution of possible results. Large experiments can help demonstrate long-run tendencies, but increasing the number of simulations does not make an incorrect probability model accurate.
A simulation is only as meaningful as its assumptions. Synthetic experiments are useful for studying mathematics and behaviour under controlled assumptions, but they should not be treated as forecasts of future real-world sporting events.
Variance and streaks
Random outcomes naturally create winning streaks, losing streaks, drawdowns, and temporary deviations from expected performance. These effects can occur even when every probability used in an experiment is perfectly known.
This is why the Odds Lab reports both expected and observed results. Differences between them are a core part of studying probability rather than an error that simulations should eliminate.
Strategies and fallacies
The Strategies hub and Strategy Lab include filters such as hot-streak chase and consistent price bands. These are stress-test heuristics for studying behaviour under synthetic assumptions — not recommendations to wager.
Casino systems like Martingale are included specifically to demonstrate how house edge and finite bankrolls produce ruin risk despite short-run winning streaks.
Next
Strategies hub
Explore educational playbooks for sports, casino, bankroll, and fallacy strategies with deep links into the labs.
Open strategies hub →Help
Breakdowns guide
Read every lab panel — buckets, overlays, hot streaks, horse fields, casino house edge — with practical educational use cases.
Open breakdowns guide →