Den här sidan på svenska · This page in Swedish
Value betting in Sweden
Value betting means placing a bet when the odds are higher than they should be. A single bet can lose, so it is the many bets that count. The worked example below has 100 bets with 100 kronor on each, and they end 250 kronor up on average. It also shows how our system has done over 39 025 bets.
A worked example with 100 bets
This is what a value bet looked like in our system.
From the odds across the whole market, we calculated that a bet like this wins 50 times out of 100. That makes the fair odds 2.00. At ComeOn the odds were 2.05, and that makes the bet a value bet. Chelsea lost 2–0, so this particular bet lost.
Now picture 100 bets like it, with 100 kronor on each.
| 100 bets of 100 kr at odds 2.05 | Number | Amount |
|---|---|---|
| Bets that win | 50 | +5 250 kr |
| Bets that lose | 50 | −5 000 kr |
| Left on average | +250 kr |
A worked example, not a result.
250 kronor is 2.5 per cent of the stakes. The odds were the same number of per cent above the fair odds, and that is called the edge. Further down the page, the same thing is called expected value. The example is a typical value bet: in the seven days from 6 to 12 November 2026, the typical edge was 2.71 per cent when the system found the bets, and the typical odds were 2.05. By typical we mean that as many bets had more as had less.
Most of the time it is not exactly 50 out of 100 bets that win, and even 100 bets can end in a loss. That is why value betting is counted over thousands of bets.
This is how it has gone for our system. From 1 March to 12 November 2026, the system has 39 025 settled bets. ROI, the profit divided by the stakes, is 3.91 per cent, with the stake the system worked out for each bet.
We replayed the matches 20 000 times and let luck choose which ones were included: in 95 out of 100 replays, ROI is in profit, between 2.68 and 5.13 per cent. It took a little over 2 600 bets before 95 out of 100 replays ended in profit. 24 of 36 full weeks ended in profit, and the longest losing run was 5 weeks in a row.
The tool for value betting shows the bets the system finds, with the stake worked out.
What is value betting?
Odds are a price on a probability. Odds of 2.00 mean the bookmaker expects the outcome to happen about every other time. Odds of 4.00 mean about one time in four. Divide one by the odds and you get the probability the odds imply. One divided by 2.00 is 50 per cent.
Value betting rests on the price sometimes being wrong. A bookmaker sets odds on thousands of matches a week, reacts at different speeds to news, and often moves its odds according to how customers bet rather than what is most likely. That creates situations where one bookmaker's odds are higher than the probability justifies. A bet like that is called a value bet, or värdespel in Swedish.
The word says nothing about which team wins. It only says that you get a better price than the fair one. Fair odds are one divided by the probability. If the probability is 40 per cent, the fair odds are 2.50. Anything above 2.50 has value, anything below does not, and that holds however the match feels. The glossary page on value betting explains the term briefly.
The hard part is not the formula but the probability. Nobody knows it exactly. The best estimate is usually the market itself: the odds at the bookmakers and exchanges that take the largest stakes and have the lowest margin, with the margin removed. That is the starting point we use in the rest of this guide and in our tool. If you are completely new to the subject, the best place to start is our page on betting.
How does value betting work? Expected value with a worked example
100 kr returns 4.20 kr on average over many identical bets. The single match returns 345 kr or nothing.
Expected value is what a bet returns on average if you could place it over and over again. The formula is short: the probability times the odds, minus one. If the answer is above zero, the bet has positive expected value. If it is below zero, you are paying too much. The glossary page on expected value explains the term at an easy pace.
Here is an example with real odds. Nations League on 1 October 2026, Greece v the Netherlands, kick-off 20.45. Just under an hour before kick-off, ATG, Svenska Spel and Unibet offered odds of 3.45 on a home win. At that moment, the market's fair probability of a home win was 30.2 per cent by our calculation, which means fair odds of 3.31.
Do the maths: 0.302 times 3.45 is 1.042. Minus one is 0.042. The expected value is plus 4.2 per cent of the stake. With 100 kronor on the bet, the expected value is 4 kronor and 20 öre, on average, over many identical bets. The single match returns either 345 kronor or nothing.
The match ended 2–2. That particular bet lost. That does not change the fact that the price was good. A bet with a 30 per cent probability loses seven times out of ten, however good the odds are. The value does not show in one match. It only shows as an average over a thousand bets or more.
Same match, different price. Just before kick-off, the three bookmakers had cut the odds to 3.35, and the market's fair probability was then 30.0 per cent. Do the maths again: 0.300 times 3.35 is 1.005. The expected value is close to zero. Same team, same match, but the price decides whether the bet is worth placing. The value bet calculator does the maths for you when you enter the odds and the probability.
The margin: the bookmaker's price, built into the odds
Before you compare your probability with a set of odds, you need to know that the odds carry a fee. Add up the probabilities that a bookmaker's odds imply for every outcome of a match and you end up above 100 per cent. What lies above one hundred is the margin, the bookmaker's payment for taking the bet.
A real example from the same match. Just before kick-off, our odds comparison showed sixteen bookmakers with a Swedish licence. ATG, Svenska Spel and Unibet had 3.35 on a home win, 3.55 on a draw and 2.23 on an away win. One divided by 3.35 is 29.85 per cent, one divided by 3.55 is 28.17 per cent and one divided by 2.23 is 44.84 per cent. The sum is 102.86 per cent. The margin is just under 2.9 per cent.
Remove it: divide each probability by 1.0286. A home win becomes 29.02 per cent, a draw 27.39 and an away win 43.59. Now the sum is 100. Turn the numbers around and you get the bookmakers' odds without the margin: 3.45, 3.65 and 2.29. The difference from 3.35, 3.55 and 2.23 is what you pay.
At 888sport and Mr Green the odds were 3.00, 3.30 and 2.25, a margin of 8.1 per cent. That is why the same team costs different amounts at different bookmakers, and why the highest odds are always the lowest price. A single match is not typical. Measured on closing odds from 28 September to 2 October 2026, ATG and Svenska Spel averaged just under 6 per cent, and 888sport and Mr Green just under 7. The glossary page on margin goes through the maths in detail, and the page on betting sites shows how often each bookmaker has had the best odds in our measurement.
The margin also explains why most odds have no value. The market's fair odds just before kick-off were 3.33, 3.58 and 2.38 by our calculation. The best odds on the Netherlands, 2.31 at ComeOn, were therefore below the fair odds. At ATG, Svenska Spel and Unibet, almost the whole margin sat on the away win, while their 3.35 on a home win was just above the fair odds. If you remove the margin at one bookmaker, you get that bookmaker's own price, not the market's. Value betting is about finding the exceptions, where one bookmaker's odds are above the market's fair price.
Closing odds: the measure that shows whether the price was good
The result of a bet says almost nothing about whether the bet was good. It takes more than a thousand bets before wins and losses say more about the method than about luck. That is why anyone who bets systematically measures something else: the odds just before kick-off, the closing odds.
The idea is simple. Up to kick-off, information and money flow into the market, and the odds move towards what is actually likely. The last price before kick-off, with the margin removed, is therefore the market's best estimate. It is called the fair closing odds. If you got higher odds than that, you had a better price than the market. If you got lower odds, you had a worse one. The comparison is often called closing line value.
Back to Greece v the Netherlands. The odds of 3.45 were available just under an hour before kick-off, and the fair closing odds turned out to be 3.33. So the price you could take was 3.6 per cent better than the market's last price. That measure can be read the same evening, without waiting for a thousand matches.
If you get a better price than the fair closing odds time after time, it suggests you are finding prices that the market then corrects. It is no guarantee, and the account can still be in the red for a long time. If you get a worse price time after time, you have no edge, even if the last few weeks have gone well. On our odds pages, the bookmakers' closing odds stay up after every match. They still include the margin, so odds that beat them all can still be below the fair odds, as with the Netherlands above.
Variance and bankroll: how much it swings, and what Kelly says about the stake
Variance is the name for the swings, and they are large even when every bet has value. Take a bet at odds of 2.20 with three per cent expected value, which means a probability of 46.8 per cent. Place 100 such bets with the same stake and you are still behind in about 40 cases out of 100. After 300 bets, in about 32 cases out of 100. After 1 000 bets, in about 19 cases out of 100. That is what a bet with an edge looks like in reality.
Losing streaks are part of it. At the same odds, you get at least six losses in a row during 100 bets in two cases out of three. During 1 000 bets, you get at least ten in a row in more than half the cases. Anyone who does not know that in advance is easily tempted to raise the stake to win back the loss. Never do that. Pausing or stopping is always a right decision, even in the middle of a streak.
That is why value betting starts with a bankroll. It is a sum set aside for betting, which you can afford to lose and which you do not top up to chase losses. The stake is calculated as a share of that sum, not as an amount you feel like.
How large a share? The Kelly criterion answers that. The formula is the expected value divided by the net win of the odds, that is the odds minus one. In the Greece example: 0.042 divided by 2.45 is 0.017, or 1.7 per cent of the bankroll. That is full Kelly, and full Kelly assumes the probability is exactly right. It almost never is. That is why most people use half or quarter Kelly. Quarter Kelly gives 0.43 per cent, which is about 86 kronor with a bankroll of 20 000 kronor. The Kelly calculator does the maths for you, and the glossary page on the Kelly criterion explains what the fraction does to the swings.
Two more rules. Several bets at the same time call for smaller shares than the formula gives for each one, because several losses can come at once. And if Kelly shows zero or a negative number, there is no bet to place, however good the match feels.
Being limited: when the bookmaker cuts your stake
A bookmaker lives on its margin. A customer who takes odds above the market's time after time removes that margin, and the bookmakers' terms give them the right to restrict such customers. It is called being limited: the maximum stake you may place is cut, sometimes to a few tens of kronor. The account remains, but in practice it becomes useless for value bets.
How fast it happens depends on the bookmaker, on how much you bet and on how clear the pattern is. Bets that are almost always above the market's price, bets placed just before the odds are cut, and stakes in odd amounts are the signs usually mentioned. None of that is prohibited. It is price comparison, but the bookmaker does not have to accept every bet.
Two things usually delay it. Spread your bets across several bookmakers with a Swedish licence, and keep the total deposited at all bookmakers within the bankroll. Round your stakes too. Still, expect it to happen at most bookmakers sooner or later, so it does not come as a shock. Never open more than one account at the same bookmaker and never bet in someone else's name. That breaks the terms and can lead to accounts being closed. The guide on limited accounts goes through what usually triggers it and what you can do.
Value betting, arbitrage and matched betting: what sets them apart
The three methods are often mixed up. They rest on different things and ask different things of you.
Value betting takes the risk in every bet. You place a bet with positive expected value, lose often and count on the value paying off over many bets. You need no bonus and no offer, only odds that are higher than the probability justifies and enough patience to let thousands of bets decide.
Arbitrage, or sure bets, rests instead on two bookmakers offering such different odds on the same match that the stakes on all outcomes together are less than what is paid out, whichever outcome it is. That only holds as long as every bet stands. The margin is often one or a couple of per cent, the opportunities are short-lived, and the pitfalls are voided bets, odds that change before you are done and being limited quickly. The arbitrage calculator works out the stakes, and the pages on arbitrage and sure bets explain more.
Matched betting starts from a welcome offer. You meet the terms of the offer with as little risk as possible, often through opposite bets at two bookmakers. The method only works as long as there are offers left, and a bonus may only be given the first time you play with a licence holder. So the offers run out. Value betting needs no offer, but it does require that you can stand long losing streaks.
How our tool does it
Doing the maths by hand works for one match. For hundreds of matches a day, you need something that fetches the odds, removes the margin and compares the bookmakers automatically. That is what Edge Analytics does, and this is how. We compare other tools and what they cost in the guide comparing value betting tools.
The tool reads odds and boosts from 30+ bookmakers, around the clock, and compares them with a fair price that we calculate from the whole market. The bookmakers and exchanges that take the largest stakes and have the lowest margin weigh the most, because they move fastest when new information arrives. Bookmakers that share the same odds supplier count as one vote. That calculation is behind the figure of 30.2 per cent in the example above.
A bet is only shown when it meets several requirements. The odds must be at least two per cent above the fair odds. The odds must be between 1.20 and 3.50, where the swings are manageable. At least five independent sources must have priced the match, and at least one of them must be a sharp reference.
For each bet you see the bookmaker, the odds and the value in per cent. The suggested stake is calculated against the bankroll you enter yourself, and you decide what to do with it. The bets can be sent to your phone via Telegram, and under My Bets you bookmark the bets you have placed and follow the result over time.
You can try it for seven days at no cost. You add a card when you start, and after seven days Edge Analytics costs SEK 1,999 a month until you cancel. You can cancel at any time under Account. Try it free.
Gamble in moderation
Value betting is mathematics, but the money is real and the stake can always be lost. Decide on the bankroll before you start and stick to it. Every bookmaker with a Swedish licence requires you to set a deposit limit. Set it low, so that the total at all bookmakers fits within the bankroll. Gambling for money is for those aged 18 or over. If it feels like it has gone too far, Stödlinjen, the Swedish gambling helpline, answers on 020-81 91 00. Through Spelpaus you can exclude yourself from all gambling that requires registration with bookmakers licensed in Sweden.
Common questions about value betting
What is value betting, in short?
That the odds you get are higher than the outcome's probability justifies. Fair odds are one divided by the probability. At 40 per cent that is 2.50. Anything above 2.50 has value, anything below does not.
How does value betting work in practice?
You estimate the probability, usually from the market's odds with the margin removed, compare it with the odds at each bookmaker and only place bets where the probability times the odds is greater than one. Then you let many bets decide, with stakes that are a small share of a bankroll.
How do I calculate expected value?
The probability times the odds, minus one. Example: 0.302 times 3.45 is 1.042. Minus one is 0.042, or plus 4.2 per cent of the stake on average over many bets. The value bet calculator does the maths for you.
Why am I losing when the bets have value?
Variance. A bet with a 30 per cent probability loses seven times out of ten even when the price is good. At odds of 2.20 and three per cent value per bet, you are behind after 100 bets in about 40 cases out of 100. So measure against the fair closing odds as well, but stick to the bankroll you decided on in advance. If the bankroll is gone, it is time to stop.
What are closing odds?
The odds just before kick-off. With the margin removed, they are the market's last and best estimate, the fair closing odds. If you got higher odds than that, you had a better price than the market. That measure says more about your method than wins and losses do in the short term.
How much should I stake on a value bet?
A share of the bankroll, not an amount you feel like. The Kelly criterion works out the share from the expected value and the odds. Most people use half or quarter Kelly. In the Greece example, quarter Kelly is 0.43 per cent of the bankroll.
Will I be limited for value betting?
Expect it, sooner or later, at most bookmakers. Bookmakers may restrict customers who take odds above the market's time after time. Spread your bets across several bookmakers with a Swedish licence and keep the total deposited within the bankroll. Round your stakes and never open more than one account at the same bookmaker.
Is value betting the same as arbitrage?
No. Arbitrage locks in a small margin in theory through stakes on every outcome at different bookmakers, but a voided bet, changed odds or being limited can mean it does not work out. Value betting takes the risk in every bet and counts on the value paying off over many. Matched betting rests on welcome offers.
Is value betting legal in Sweden?
Yes. Betting with bookmakers that hold a Swedish licence is legal from the age of 18, and value betting is a way of choosing which odds are worth their price. The bookmakers may, however, restrict customers they do not want.
Do I need a tool for value betting?
For one match, the formula and a calculator are enough. For hundreds of matches a day, you need something that fetches and compares the odds automatically. Edge Analytics does that. You can try it for seven days at no cost. After that it costs SEK 1,999 a month until you cancel, which you can do at any time under Account.