Value bets – why a good bet can lose

The first chapter of the Playbook covered bankroll management and the second covered the closing line. This chapter covers the concept both of them ultimately rest on: value. Value is the only durable idea in betting. Everything else – tips, instinct, knowing a league, statistics – is just a way of finding it.

What value means

Every price is a claim about probability. Odds of 2.00 claim the event happens about 50% of the time. Odds of 1.50 claim about 67%, odds of 4.00 about 25%. The probability is one divided by the odds: 1 / 2.00 = 0.50.

A value bet is a bet where the true probability of the event is higher than the probability the odds claim. If a team wins 55% of the time and the odds are 2.00, the bet has positive expected value: staked at 100, the average return is 0.55 × 200 − 100 = 10. The edge here is 10% of the stake.

If the same team wins only 45% of the time, the same odds of 2.00 are a bad bet, even if the team happens to win this particular match. Value is not a property of the result. It is a property of the relationship between the price and the probability, and it exists before the match starts.

Value is not the same as long odds

The most common misunderstanding is that a value bet means a long price. It does not. Odds of 1.40 can be a value bet if the true probability is 78% (the price claims 71%). Odds of 6.00 can be a bad bet if the true probability is 12% (the price claims 17%).

The second misunderstanding is that a value bet is a "safe" bet. The opposite is true: a value bet at 2.50 typically loses more often than it wins. If the true probability is 45% and the odds are 2.50 (claiming 40%), the bet has a 12.5% edge, and it still loses 55 times out of 100. That is the nature of a value bet: a single result says almost nothing about the quality of the bet.

Why prices are sometimes wrong

A bookmaker's price is the market's best guess with a margin added (chapter 4 covers the margin in detail). The market is good but not perfect. Prices are most often wrong in three situations.

The first is new information the market has not yet priced: a line-up change, a resting goalie, travel and a game on the second night of a back-to-back. The second is disagreement between bookmakers: when the same market is priced clearly differently across books, at least some of them are wrong, and the best price may hold value. The third is public pressure: favourites, home teams and big clubs attract the public's money, which pushes their prices down and the other side's prices up.

In all three cases, finding value requires your own estimate of the probability, made independently of the market. An estimate that copies the price cannot find an error in the price.

How the Sorceress looks for value

The Value picks lane works in two stages. First, the Sorceress makes a blind estimate of the match: it estimates the probability of the event without seeing any odds, using only statistics, line-ups and the match situation. Only then is the estimate compared with the market. This way the estimate cannot anchor itself to the price.

A pick reaches the Value lane only if the blind estimate gives the best available price an edge of at least four percent, and in addition a trigger is present: a move in the odds, clear disagreement between bookmakers, or a concrete reason such as an absence. A numeric edge alone is not enough, because without a reason the edge is more likely an error in the estimate than an error in the market. The lane also has an odds cap, because at long prices even a small error in the estimate looks like a large edge.

Each value pick's card shows the market probability, the Sorceress's estimate and the difference between them, the edge. This is deliberately visible: the reader can judge for themselves whether the reasoning is strong enough for the edge claimed.

How value is measured

Because value bets lose often, hit rate is a poor measure of their quality. Two better figures are published for every pick.

Level-stake return shows what a set of picks has produced when each is staked at 100. It is the final measure, but a slow one: single weeks swing a lot, and a reliable picture takes hundreds of picks.

The closing line and CLV show much sooner whether a pick had value. If the price drops between publication and kick-off, the market has moved towards the Sorceress's estimate, and the pick most likely had a real edge regardless of how the match ended. Chapter 2 covers this in detail. On the Value lane, CLV is the main quality measure, because it separates skill from luck far faster than results do.

The statistics page shows Value picks as a lane of their own: number of picks, hits, return and CLV. The figures update after every settlement run, and losing weeks are shown exactly like winning weeks.

What this means for the bettor

In value betting you are not playing a single match but a long series of decisions. Three practical rules follow.

Stake flat or as a percentage of your bankroll, as described in chapter 1. Value is expected value, and expected value only materialises if a losing run does not eat the bankroll before the sample is large enough.

Compare prices. The same pick can be a value bet at 2.10 and a bad bet at 1.90. The Sorceress names the bookmaker where the best price was found at publication, and each pick's card has an odds comparison.

Judge picks in hundreds, not tens. A run of ten value bets can lose outright even if every one of them was correct. Only a couple of hundred picks tell you whether the estimate beats the market. That is why everything is published: only a long and complete history lets the reader judge this for themselves.

Summary

Value is the gap between the price and the true probability. It exists before the match, and the result does not change it. A value bet loses often and is still profitable if the edge is real. A real edge requires an estimate made independently of the market and a reason why the market would be wrong. The Value picks lane is built on exactly this: a blind estimate, an edge of at least four percent, a trigger and public CLV tracking.

The next chapter covers odds and the margin: why a bookmaker's probabilities add up to more than 100 percent, and what that means for the bettor's return.

Betting is for adults only. Please gamble responsibly.