Probability Theory for Bettors: The Basics You Actually Need

Most bettors lose money because they don't understand probability. They chase short-term results, they mistake luck for skill, and they don't grasp why a seemingly solid bet can lose.

You don't need to be a mathematician to understand the basics of probability. But you do need to understand a few core concepts. This guide explains the foundations in plain language, using football examples throughout.

Probability As A Number Between 0 And 1

Probability is a measure of likelihood. It's always between 0 (impossible) and 1 (certain).

If an outcome is certain to happen, probability is 1.0 (or 100%).

If an outcome is impossible, probability is 0.0 (or 0%).

If an outcome is equally likely to happen or not happen, probability is 0.5 (or 50%).

In football:

A team will definitely either win, draw, or lose. The combined probability of all three is 1.0. Each outcome has some probability between 0 and 1.

A team at 2.0 odds has an implied probability of 0.5 (50%). The bookmaker thinks there's a 50% chance they win.

Understanding probability as a number between 0 and 1 helps you convert odds to probability. Odds = 1 / Probability. If probability is 0.5, odds are 2.0.

Independent vs Dependent Events

An independent event is one whose outcome doesn't affect another event.

A dependent event's outcome is affected by other events.

In betting:

If you're betting on two separate football matches happening at the same time, the outcomes are independent. Team A's result doesn't affect Team B's result.

If you're betting on the same team in two consecutive matches, the events are loosely dependent. A team that won last week has some momentum for this week. But they're not strongly dependent. Each match is largely determined on its own merits.

This matters because combining independent events multiplies their probabilities. If Team A has a 50% chance of winning and Team B (separate match) has a 50% chance, the probability both win is 0.5 times 0.5 = 0.25 (25%).

But if the same team has a 50% chance of winning this week and a 50% chance of winning next week, the probability they win both isn't necessarily 25%. The events are somewhat dependent. Momentum, injuries, team morale carry over.

For most betting purposes, assume events are independent unless you have specific reason to believe otherwise.

Probability vs Odds

Probability and odds are related but different.

Probability is the likelihood expressed as a decimal between 0 and 1.

Odds are the numerical representation of that probability as offered by a bookmaker.

The conversion:

Odds = 1 / Probability

A probability of 0.5 converts to odds of 1 / 0.5 = 2.0

A probability of 0.25 converts to odds of 1 / 0.25 = 4.0

A probability of 0.67 converts to odds of 1 / 0.67 = 1.49

Understanding this relationship is fundamental. When you see odds, immediately convert them to implied probability. This helps you assess whether the odds offer value.

Expected Value (EV)

Expected value is the average profit or loss per bet over time.

A positive EV bet is profitable long-term. A negative EV bet loses money long-term. An EV of zero is break-even.

The calculation:

EV = (Probability of winning × Profit if you win) minus (Probability of losing × Loss if you lose)

Example:

You back a team at 2.0 odds with a £10 stake. You assess their true probability at 52%.

Probability of winning: 0.52 Profit if you win: £10 (stake times odds minus one) Probability of losing: 0.48 Loss if you lose: £10 (your stake)

EV = (0.52 × £10) minus (0.48 × £10) EV = £5.20 minus £4.80 EV = +£0.40

This bet has a positive expected value of £0.40 per £10 stake. Over many bets like this, you profit.

If the odds were 2.2 instead of 2.0, and you assess true probability at 52%:

EV = (0.52 × £12) minus (0.48 × £10) EV = £6.24 minus £4.80 EV = +£1.44

A better edge gives better EV.

If you assess true probability at 48% (worse than the implied 50%), and odds are 2.0:

EV = (0.48 × £10) minus (0.52 × £10) EV = £4.80 minus £5.20 EV = -£0.40

Negative EV. You're taking worse odds than your assessment suggests. This is a losing bet.

This is the entire game. Find bets with positive EV. Size them proportional to the EV. Over time, positive EV bets profit.

Variance and Standard Deviation

Variance is the fluctuation of results around the expected value.

High variance means results jump around wildly. Low variance means results are consistent.

In betting, even with a positive EV, you'll experience streaks of losses due to variance.

Example:

You have a betting strategy with a +2% EV per bet (you profit 2p per £1 staked over the long-run). This is a small edge.

Over 100 bets, the +2% EV says you should profit £2 per £100 staked. But variance is huge. You might be up £20 or down £15 due to lucky or unlucky short-term results. The +2% EV is drowned out by variance.

Over 10,000 bets, variance smooths out. Your results converge toward the +2% EV. You'll be close to a +£200 profit per £10,000 staked.

Standard deviation measures variance. High standard deviation means wide swings. Low standard deviation means consistent results.

In football betting, variance is always present. You need to accept that even good bets lose sometimes, and even bad bets win sometimes.

The Law Of Large Numbers

The law of large numbers states that as the number of trials increases, the average result converges to the expected value.

This is fundamental. This is why bettors care about long-term results, not short-term results.

Example:

A bet has a +5% EV. You stake £10 each time.

Over 10 bets: You might be up £20 or down £15 due to variance. The +5% EV (expected profit £5) is invisible against the noise.

Over 100 bets: You might be up £30 or down £10. The +5% EV (expected profit £50) is becoming visible.

Over 1,000 bets: You're very likely to be close to +£500, plus or minus £100 due to remaining variance. The +5% EV is now very obvious.

Over 10,000 bets: You're almost certainly close to +£5,000, plus or minus £300. The EV dominates variance. Your result is predictable.

This is why professional bettors think in terms of thousands of bets, not tens of bets. They're letting the law of large numbers work. Over time, edge becomes obvious.

For you, this means:

• Don't judge your betting strategy on short-term results
• 100 bets is not enough to know if you're any good
• 1,000 bets might be
• 5,000+ bets clearly shows if you have edge or not
• One losing month means nothing
• One winning month means nothing
• One losing year might mean something, but could still be variance
• Five years of consistent profit means you have edge

The 95% Confidence Interval

In statistics, a 95% confidence interval means you're 95% certain the true value falls within that range.

For betting, this helps you understand variance.

If you have a +2% EV betting strategy and you place 1,000 bets with £10 stakes, your expected profit is £200. But due to variance, you might end up anywhere from +£100 to +£300 (roughly, the 95% confidence interval).

If you end up at +£150, that's within the expected range. You haven't proven your strategy doesn't work.

If you end up at -£500, that's far outside the expected range. Either you have bad luck (unlikely but possible) or your strategy doesn't work.

This is why bettors look at long-term results. More bets shrinks the confidence interval. Your results become closer to the true EV.

Why You Shouldn't Judge Short-Term Results

A bet with positive EV can lose. A bet with negative EV can win. In the short-term, luck dominates skill.

Over 100 bets:

• A skilled bettor with +3% EV might be down £20 due to bad variance
• A terrible bettor with -3% EV might be up £30 due to good variance

If you judge based on results, you'd think the terrible bettor is skilled and the skilled bettor isn't.

This is the gambler's fallacy. Judging a system by short-term results leads to abandoning winning systems and doubling down on losing ones.

The answer is patience. Stick to a system with positive EV. Let time and the law of large numbers prove it works. Don't judge based on 50 bets or 100 bets. Wait for 1,000 or more.

How To Test If You Have An Edge

The process:

1. Define your betting strategy and the bets it generates
2. Track every bet you place over months (or years)
3. Calculate the profit/loss and strike rate (win percentage)
4. Calculate the expected EV of your bets
5. Compare actual results to expected results

If your actual profit is close to expected EV over 1,000+ bets, you have demonstrated edge.

If actual profit is far from expected EV (especially negative when EV is positive), you likely don't have edge or your edge is smaller than you think.

This requires discipline and honest record-keeping. Most bettors don't do this. They remember the wins, forget the losses, and think they're better than they are.

In Summary

• Probability is a measure of likelihood between 0 and 1.
• Understanding probability helps you convert odds and assess value.
• Expected value is the average profit per bet over time.
• Positive EV bets are profitable long-term.
• Variance is the fluctuation around the expected value.
• Even positive EV bets lose sometimes.
• The law of large numbers states that over many trials, results converge to the expected value.
• This is why long-term results reveal edge.
• Short-term results tell you almost nothing due to variance.
• You need thousands of bets before edge becomes obvious.
• Understanding these concepts fundamentally changes how you think about betting.
• You stop judging systems on short-term results.
• You focus on identifying positive EV bets.
• You accept variance as a fact of betting.
• You think in terms of seasons and years, not weeks and months.
• This doesn't guarantee profit.
• But it gives you the right mental framework for sustainable betting.

FAQs on Probability and Betting

What is a good expected value for a bet?

Anything positive is profitable long-term. But practically, you want at least 3-5% EV per bet to justify the effort and variance. A +1% EV is mathematically profitable but requires so many bets and so much time that few achieve it. Professional bettors target 5-10%+ EV.

How many bets do I need to know if I have an edge?

Depends on the size of the edge. A large edge (10%+) becomes obvious in 500-1,000 bets. A small edge (2-3%) needs 5,000-10,000 bets. A tiny edge (under 1%) might need 20,000+ bets to prove statistically.

Can I lose money with a positive EV betting system?

Yes, in the short-term due to variance. But over the long-term, positive EV systems should profit. If you've been betting a +EV system for 5,000 bets and you're down money, something is wrong. Either your edge is smaller than you think, or you're misjudging probabilities.

Is it possible to have a betting edge?

Yes, but it requires better probability assessment than the bookmaker. If your model predicts 55% for an outcome and the bookmaker's odds imply 50%, you have edge. Finding these edges consistently is the hard part.

Why do professionals bet such small percentages of their bankroll per bet?

Because they're thinking probabilistically. A 2% EV bet with £1,000 stake earns £20 expected value. Over 1,000 such bets, that's £20,000 profit. But there's variance. Conservative stake sizing survives downswings without bankruptcy.

Can I use expected value to calculate how much I should win over a season?

As a rough guideline, yes. If you place 2,000 bets with an average EV of +3% per bet, you expect to profit roughly 3% of your total staked amount. But variance means actual results might be significantly different, especially early in the season.

Is there a minimum number of bets before I know if I'm any good?

At least 500, ideally 1,000+. But even 1,000 bets can still be explained by luck if your edge is small. 5,000+ bets is more definitive.