Separate Good Results from Good Decisions
This phase is the heart of the subject. It explains how a bettor can lose a good bet, win a bad bet, and still know which was which. That is the difference between gambling randomly and evaluating wagers like prices.
Chapter 13Expected Value
Expected value, usually shortened to EV, is the average amount you expect to win or lose per bet if the same opportunity could be repeated many times. It is the central concept in rational betting.
Suppose you can bet an outcome at +150. That means you profit 1.5 units for every 1 unit staked. If you believe the true probability is 45%:
A positive EV does not mean the next bet wins. It means the price is favorable relative to your probability estimate. Over time, consistently positive EV is what gives a bettor a chance to overcome vig and variance.
Human brains love outcomes. EV cares about process. If you bet +0.12 EV and lose, you did not necessarily make a mistake. If you bet -0.12 EV and win, you did not necessarily make a good decision.
Chapter 14Fair Price and True Probability
To judge whether a line has value, you need some view of the outcome’s true probability, or at least a better estimate than the market’s current price implies. Once you have that, you can convert it into a fair price.
From Probability to Fair Price
If you think an event happens 50% of the time, the fair decimal price is 2.00. If you think it happens 40% of the time, the fair decimal price is 2.50. If you think it happens 62%, the fair decimal price is roughly 1.61.
Fair price means the no-margin price at which your expected value is zero. Any better price than fair is positive EV. Any worse price is negative EV.
| Your Probability | Fair Decimal | Fair American |
|---|---|---|
| 50% | 2.00 | +100 |
| 45% | 2.22 | +122 |
| 62% | 1.61 | -163 |
Where True Probability Comes From
In practice, bettors estimate true probability using different levels of sophistication:
- Market-based — using no-vig prices from sharper books as a baseline.
- Model-based — building statistical forecasts from data.
- Context-based — adding informed judgment about injuries, matchups, rest, or weather.
- Hybrid — starting from the market, then making disciplined adjustments.
The critical point is that true probability is never known with certainty. It is estimated. The quality of that estimate determines whether your EV calculation is meaningful or fictional.
False precision is common. Saying an outcome is 54.7% likely does not make the estimate rigorous. Sometimes a rough but honest range is better than a fake exact number.
Chapter 15CLV, Variance, and Sample Size
Closing Line Value
Closing line value, or CLV, compares the price you got to the closing market price. If you consistently beat the closing line, that is often evidence that your bets were placed at favorable numbers.
For example, if you bet over 47.5 at -110 and the market closes 49.5 at -110, you likely beat the market by two points. That does not guarantee a win on the bet, but it usually indicates you captured value relative to the final consensus price.
Variance
Variance is the reason short runs can look absurd. A bettor with a small edge can still lose for weeks. A bettor with no edge can run hot for months. This is not a flaw in the concept of EV. It is what uncertainty looks like when outcomes are realized one by one.
Sample Size
A record of 14-6 tells you almost nothing by itself. A record of 540-470 starts to say more, but even then the quality of markets, prices, and record-keeping matters. Betting performance is noisy enough that tiny samples are close to useless.
CLV Is a Process Signal
It asks whether you got a better number than the market eventually settled on.
Win Rate Is an Outcome Signal
It reflects what happened, but can swing hard in small samples.
ROI Depends on Both
Long-run profitability still requires edge, discipline, and enough volume for noise to shrink.
One of the fastest ways to become worse is to overfit to a tiny recent stretch. You can learn from losses, but you should not rewrite your worldview after ten bets or even fifty if the markets are volatile.
Chapter 16A Simple Betting Process
Most effective betting processes are more boring than people expect. They rely on repeatability, not excitement.
Choose a narrow market first
It is easier to learn one league and one market type deeply than to chase every sport on the board.
Get the best available price
Line shopping is one of the most immediate edges available to any bettor, because a better number directly changes EV.
Estimate fair value honestly
Use no-vig market information, data, and context. Do not pretend certainty where you have none.
Bet only when price exceeds your threshold
Passing is part of the process. Not every opinion deserves a wager.
Log the wager and review later
Track the market, the number you got, the closing line, and your reasoning. Then review patterns instead of relying on memory.
At a high level, sports betting process is simply this: estimate a fair price, compare it to the market price, act only when the difference justifies the risk, and measure whether your process beats the market over time.