- The number: the best odds pay between 4.4% and 6.7% more than the market average.
- The effect: a "blind" bet on the favorite improved by about 3.3 points of return.
- The limit: the theoretical best odds are not always available to you, and they do not fix a bet with no value.
What is line shopping and why does it matter?
Line shopping is comparing the price of a bet at several sportsbooks.
Then you bet at the one that pays the most.
It does not change your analysis, it changes your payout.
It is a silent edge because it does not depend on being right more often.
It depends on collecting more when you are right.
Over time, that difference adds up.
To use a comparison tool step by step, there is the guide on how to use an odds comparison tool.
Here we measure how much it is worth, with data.
How much more the best odds pay: 34,345 matches
We took every match with complete odds.
We compared the maximum posted odds against the market average and against one big reference sportsbook.
| Outcome | Best vs average | Best vs big sportsbook |
|---|---|---|
| Home win | +4.4% | +4.8% |
| Draw | +5.4% | +5.6% |
| Away win | +6.7% | +6.4% |
Source: football-data.co.uk, seasons 2019/20, 2021/22 to 2025/26 and the 2026/27 games played so far.
Sample: 34,345 matches from 16 European leagues with complete opening odds.
Leagues: first and second divisions of England, Spain, Italy, Germany and France, plus the Netherlands, Portugal, Belgium, Greece, Scotland and Turkey.
Our own calculation.
Notice the pattern.
The less likely the outcome, the larger the gap between sportsbooks.
With favorites there is less room to differ; with away teams and draws, more.
The big reference sportsbook had the maximum odds in only 10.3% of cases.
How the return of a "blind" bet changes
To measure the effect we built a bet with no criteria at all.
That means betting one unit on the market favorite in every match.
That isolates the price: the hit rate is identical, only the odds we collect change.
The favorite won in 51.3% of matches.
That does not depend on the sportsbook.
The only thing that varies is how much they pay when it wins.
| Blind bet (1 unit) | Best odds | Average | Big sportsbook |
|---|---|---|---|
| Always the favorite | −0.6% | −4.0% | −4.6% |
| Always the non-favorite | −4.2% | −10.6% | −9.8% |
| Always the home team | −2.9% | −7.0% | −7.2% |
| Always the away team | −5.0% | −10.6% | −10.0% |
Source: football-data.co.uk, seasons 2019/20, 2021/22 to 2025/26 and the 2026/27 games played so far.
Sample: 34,345 matches from 16 European leagues with complete opening odds.
Leagues: first and second divisions of England, Spain, Italy, Germany and France, plus the Netherlands, Portugal, Belgium, Greece, Scotland and Turkey.
Our own calculation.
Return: profit or loss per unit staked, on average.
Favorite: the outcome with the lowest average odds.
Look at the favorite row.
At the average, the blind bet loses 4.0% per bet.
At the best odds, it loses 0.6%.
The improvement is about 3 points.
That is line shopping: lowering the cost of betting.
A market with no margin is theoretical: read this table carefully
The draw bet at the best odds came out at +0.1%, slightly above zero.
Do not read it as a winning strategy.
It is the result of choosing, after the match, the highest price posted by any sportsbook.
In practice, those odds may not be available, may drop before you bet or may come with low limits.
Building the 1X2 with only the best odds for each outcome, the margin fell to 1.5% against 6.1% for the average.
That reduced margin is theoretical.
So take it as a ceiling on what line shopping can earn.
The real figure will land below it.
Even if the improvement is half of that, the effort is still worth it.
Is it always like this? Return by season
The average can hide very different years.
We repeated the favorite-bet calculation for each season.
| Season | Matches | Best odds | Average |
|---|---|---|---|
| 2019/20 | 5,188 | −1.9% | −5.8% |
| 2021/22 | 5,661 | −0.9% | −4.5% |
| 2022/23 | 5,595 | +2.3% | −1.1% |
| 2023/24 | 5,598 | +0.3% | −2.9% |
| 2024/25 | 5,483 | −1.0% | −3.9% |
| 2025/26 | 5,438 | −2.5% | −5.6% |
| 2026/27 (partial) | 1,382 | −2.1% | −5.1% |
Returns go up and down from one season to the next.
In two seasons the favorite came out positive at the best odds.
But in every season the best odds returned more than the average.
That is the stable part: the price improvement.
The unstable part is the hit rate.
That is why you should never draw conclusions from a few bets.
How much it changes by league
The price improvement when betting the favorite ranged from 3.0% to 4.2% depending on the league.
There are no leagues where line shopping does not help.
Second divisions are among those with the biggest difference.
| League | Matches | Odds improvement on the favorite |
|---|---|---|
| LaLiga 2 | 2,885 | +4.2% |
| Süper Lig | 2,103 | +3.9% |
| Ligue 2 | 2,112 | +3.9% |
| Serie B | 2,360 | +3.8% |
| 2. Bundesliga | 1,917 | +3.7% |
| Super League (Greece) | 1,471 | +3.6% |
| Ligue 1 | 2,029 | +3.4% |
| Championship | 3,456 | +3.3% |
| Primeira Liga | 1,915 | +3.3% |
| Serie A | 2,369 | +3.3% |
| Premiership (Scotland) | 1,385 | +3.3% |
| Pro League (Belgium) | 1,853 | +3.3% |
| Bundesliga | 1,899 | +3.2% |
| LaLiga | 2,379 | +3.2% |
| Eredivisie | 1,852 | +3.1% |
| Premier League | 2,360 | +3.0% |
Source: football-data.co.uk, seasons 2019/20, 2021/22 to 2025/26 and the 2026/27 games played so far.
Sample: 34,345 matches from 16 European leagues with complete opening odds.
Leagues: first and second divisions of England, Spain, Italy, Germany and France, plus the Netherlands, Portugal, Belgium, Greece, Scotland and Turkey.
Our own calculation.
Labeled example: 1.90 against 2.00
All the numbers below are an example.
Imagine a bet with a 52% estimated probability.
At one sportsbook it pays 1.90 and at another 2.00.
| Example | Odds 1.90 | Odds 2.00 |
|---|---|---|
| Expected value per unit | −1.2% | +4.0% |
| Expected result over 100 bets of 10 | −12 | +40 |
| Sign change | Yes |
With the low odds, the expected payout is negative.
With the high odds, it is positive.
A 0.10 difference in the odds changed the expected result.
The best odds do not rescue a bad bet, but a good bet at bad odds is wasted.
Calculate it with your own data in the expected value calculator and convert formats with the odds converter.
What are the limits of line shopping?
- Availability: the highest odds may be at a sportsbook where you have no account.
- Time: odds move; the best may disappear in minutes.
- Limits: some sportsbooks restrict people who always seek the best price.
- Hidden costs: withdrawals, fees and bonuses with fine print.
- Bias: the dataset reports the maximum odds observed; you will not always see them in time.
And one more important limit.
Line shopping reduces what you lose when you bet.
It does not give you an edge over the market if the bet has no value.
That is what value betting is for.
How to apply it without going crazy
- Decide the bet before looking at prices.
- Compare the same market at two or three sportsbooks you trust.
- Convert the odds to implied probability and compare with your estimate.
- Write down the odds you took and the best ones available.
- Every month, add up how much you gained in price: that is your real line shopping.
With a simple log you will see your real improvement, not the theoretical one.
If you want an AI to build that spreadsheet, see the prompt in the odds comparison guide.
Prompt to measure your own line shopping
I have a betting log with these columns: date, match, market, odds taken, maximum market odds, result (W or L) and stake. Calculate: 1) my actual return, 2) my return if I had always taken the maximum odds, 3) the difference in points. Show me the formula used and a table by month. Use only my data. If columns are missing, tell me and do not invent values.
Compare what you get with the figures in this article.
If your real improvement is much smaller, look for why: accounts, limits or arriving late to the price.
What to read next
If you want to get practical, start with the odds comparison guide.
For goals markets, continue with team goals.
For bets by half, see first-half bets.
The big picture is in AI for betting: the complete guide and in the best markets with AI.
Our picks are recorded in the public record and discussed in the Betting Journal.
And if you want to build your own AI, read create your AI for betting.
Responsible gambling: only for adults 18 and over and with money you can afford to lose.
Good odds are not a guarantee.
Frequently asked questions
What is line shopping?
It is comparing the odds for the same market at several sportsbooks and betting at the one that pays the most.
It does not change what you bet.
It changes how much you collect if you are right.
How much does line shopping improve returns?
In our data, always betting the favorite went from −4.0% to −0.6% per bet when using the best odds instead of the average.
That is about 3.3 points of difference.
It is our own calculation over 34,345 matches and does not guarantee future results.
Are the best odds always available?
No.
The best odds in the dataset are the highest any sportsbook posted, but they may be at a sportsbook where you have no account, with low limits or short-lived.
That is why the figures are a theoretical ceiling.
Do you need many accounts?
The more sportsbooks you use, the more often you will find better odds.
Start with two or three you trust.
Do not open accounts for a difference of hundredths.
Are good odds enough to win?
No.
The odds improve the payout when you are right, but the probability of being right does not change.
If you bet without value, the expected result is still negative.
That is why it is combined with expected value.
Is line shopping the same as a sure bet (arbitrage)?
No.
Line shopping picks the best price for one bet.
Arbitrage covers every outcome at different sportsbooks looking for a fixed profit.
They are different practices with different risks.