Draw progression: how the system works and why it loses in the end
A draw progression is a staking system: you back the same team to draw match after match, and after every loss you raise the stake so that the first draw wins back everything lost and leaves a small profit. Most cycles end quickly and in profit, which is why the system looks safe. The problem is long runs without a draw. In our data of more than 300,000 league matches from 37 leagues, a run of at least 15 matches without a draw happened 1,797 times, and in two league seasons out of three at least one team had one. Barcelona have not drawn in 42 consecutive LaLiga matches (as of 28 September 2026) Check on Google. At draw odds of 3.60, a run like that turns a £20 profit per cycle into £2,616 staked after 15 matches and £347,440 after 30.
Key takeaways
- 1
In 60% of cycles the draw comes by the third bet, but one cycle in eighty runs longer than 15 matches.
- 2
After 15 or more matches without a draw, teams drew 23.4% of the time, exactly what the market expected.
- 3
In our model of a 38-match season at odds of 3.32 on a draw fairly priced at 3.60, the typical season makes £171–180, but at every budget the average is a loss of 7.8% of turnover, which is exactly the margin.
- 4
A margin of about 8.5% (odds of 3.32 instead of a fair 3.60) almost doubles the total staked after 20 matches without a draw, from £13,396 to £25,924.
How a draw progression works
A draw progression is a version of the martingale, the system of doubling your stake after a loss. The rules are simple:
- Pick a team, usually one that draws often.
- Back the draw in every one of its league matches.
- After a loss, work out the new stake so that a win covers every loss so far in the cycle and leaves a set profit.
- When the draw comes in, the cycle ends and you go back to the starting stake.
The stake for the next bet comes from one formula: (losses so far + target profit) / (odds − 1).
This is how the stakes rise at draw odds of 3.60 with a target of £20 per cycle:
| bet in the cycle | stake | total staked if the draw still has not come |
|---|---|---|
| 1st | £7.69 | £7.69 |
| 2nd | £10.65 | £18.34 |
| 3rd | £14.75 | £33.09 |
| 5th | £28.27 | £81.78 |
| 10th | £143.89 | £497.99 |
| 15th | £732.26 | £2,616.13 |
| 20th | £3,726.58 | £13,395.69 |
| 25th | £18,965.15 | £68,254.55 |
| 30th | £96,516.61 | £347,439.81 |
Each stake is 38% bigger than the one before it (3.60 / 2.60 = 1.38). After 20 matches without a draw you have £13,396 on the line, and the 21st bet would need a stake of £5,160 to win £20.
Why a draw progression feels safe
A draw comes up in about 26% of league matches (leagues, seasons and how accurate draw odds are, we cover in our football draw statistics). On data from 28 leagues from 2015–2026, we checked after how many bets a draw progression on one team ended (52,474 cycles):
| cycle ended with a draw | share of cycles |
|---|---|
| on the 1st bet | 26.3% |
| by the 3rd bet | 60.0% |
| by the 5th bet | 78.2% |
| by the 10th bet | 95.2% |
| by the 15th bet | 98.8% |
| by the 20th bet | 99.7% |
Most of the time, then, the system "works": cycle after cycle ends in a small profit, and that builds the belief that all it takes is patience. The other side of the same table looks like this: roughly one cycle in twenty lasts longer than 10 matches, one in eighty longer than 15 and one in three hundred longer than 20. Anyone who plays a progression for a few seasons will sooner or later run into one of those cycles.
Runs without a draw are more common than you think
In our data of more than 300,000 league matches from 37 leagues:
- a run of 15+ matches without a draw happened 1,797 times,
- a run of 20+ matches 457 times,
- a run of 30+ matches 35 times.
In 90% of league seasons at least one team went 12 matches without a draw within that season, in 65% 15 matches, and in one season in five 20 matches.
Selected longest runs from our data, starting with England's top two divisions:
| team | league | matches without a draw | period | how the run ended |
|---|---|---|---|---|
| Tottenham | Premier League | 32 | Apr 2018 – Feb 2019 | 1-1 v Arsenal |
| Blackburn Rovers | Championship | 32 | Apr 2022 – Jan 2023 | 1-1 v Bristol City |
| FC Schaffhausen | Challenge League (Switzerland) | 45 | Mar 2017 – May 2018 | 3-3 v Vaduz |
| Barcelona | LaLiga | 42 | since Sep 2025 | ongoing (as of 28 Sep 2026) |
| CF Montréal | MLS | 41 | Jun 2017 – Jul 2018 | 2-2 v Portland Timbers |
| Juventus | Serie A | 38 | Feb 2016 – Feb 2017 | 1-1 v Udinese |
| Club Brugge | Pro League (Belgium) | 36 | Aug 2015 – May 2016 | 2-2 v Oostende |
| Rennes | Ligue 1 | 28 | Oct 2024 – Aug 2025 | 1-1 v Angers |
| Borussia Dortmund | Bundesliga | 27 | Dec 2019 – Nov 2020 | 1-1 v Eintracht Frankfurt |
What a run without a draw costs at real odds
For runs from 2015–2026 we know the closing draw odds of every match. They decide how much a run costs: after each lost bet the next stake grows by 1 / (odds − 1), so the lower the draw odds, the faster it climbs. These are closing odds without the bookmaker's margin, which is the most favourable case: at a bookmaker the odds are lower by the margin, so the stakes would grow faster still.
| team | matches without a draw | draw odds in the run | stake rise per lost bet |
|---|---|---|---|
| Blackburn Rovers, Apr 2022 – Jan 2023 | 32 | 3.23–4.24 | 31–45% |
| Tottenham, Apr 2018 – Feb 2019 | 32 | 3.36–9.71 | 11–42% |
| Manchester City, Oct 2018 – Aug 2019 | 31 | 3.55–19.61 | 5–39% |
| Juventus, Feb 2016 – Feb 2017 | 38 | 3.31–9.01 | 12–43% |
| Rennes, Oct 2024 – Aug 2025 | 28 | 3.13–4.77 | 27–47% |
Compare Blackburn and Manchester City: two runs of almost the same length. At Blackburn's odds every lost bet raised the next stake by between 31% and 45%. A favourite's draw pays 6, 10 or even 19, so each win covers a lot and the stakes barely move: at 19.61 the next stake is only 5% bigger. At a constant 3.60, 32 lost bets add up to £666,116; at a constant 10.00, to £562. And a progression is usually played on "drawish" teams, evenly matched sides with low draw odds. That is exactly where a long run hurts most.
Runs going on now
As of 28 September 2026:
- Barcelona: 42 league matches without a draw. Their last draw was 1-1 against Rayo Vallecano on 31 August 2025, and since then they have only won or lost. As clear favourites, their draw was priced between 4.02 and 26.32 during the run, so each lost bet raised the stake by only 4–33%, but it has kept rising for 42 matches.
- Arsenal: 16 matches without a draw; Borussia Dortmund and RB Leipzig: 15 each. At a constant 3.60, 16 lost bets add up to £3,630 and the 17th would need £1,404; 15 lost bets come to £2,616, with £1,014 needed on the 16th.
Any of these runs could end in the next round or carry on. The data cannot tell which.
Is a draw "due" after a long run?
No. This is the classic gambler's fallacy, and it does not work the other way either: a run of draws does not raise the chance of another draw. We checked how often a team drew depending on how many matches without a draw it had behind it, and compared that with the chance the market gave it in the closing odds:
| matches without a draw before the match | cases | draw came up | market expected |
|---|---|---|---|
| 0 (the previous match was a draw) | 52,638 | 25.9% | 25.9% |
| 1–4 | 104,223 | 25.9% | 25.7% |
| 5–9 | 34,224 | 25.9% | 25.2% |
| 10–14 | 7,834 | 23.5% | 24.5% |
| 15 or more | 2,701 | 23.4% | 23.4% |
After 15 matches without a draw a team draws even less often than average, because runs that long belong mostly to teams that rarely draw: favourites winning match after match, or sides with a very open style. The market knows this and prices it in. A draw is never "owed".
A progression with a budget: how often it breaks
Nobody has unlimited money, so a real progression has a budget. We worked out what happens with one: at the start of the season you pick a team, back the draw in every one of its league matches with a target of £20 per cycle, and stop when the next stake is more than what is left of the budget. The season has 38 matches, like the Premier League's. The draw's fair odds are 3.60, a 27.8% chance in every match whatever happened before (the table above shows a run does not change it), and the bookmaker pays 3.32, roughly the 8.5% margin we measured on the draw at the cheapest of three recreational bookmakers we track. A cycle still open at the end of the season counts as a loss. These are exact probabilities from the model, not a simulation of past seasons:
| budget per team and season | progression breaks | season ends in the red | typical season (median) | average season |
|---|---|---|---|---|
| £1,000 | 22.9% | 29.6% | +£171 | −£79 |
| £5,000 | 5.4% | 15.0% | +£180 | −£116 |
| £20,000 | 1.2% | 11.7% | +£180 | −£148 |
| £100,000 | 0.3% | 10.9% | +£180 | −£178 |
This is the textbook picture of a martingale:
- A typical season ends in a small profit, £171–180.
- On average a season ends in a loss, because every so often the whole budget goes. With £5,000 that happens in roughly one season in nineteen.
- A bigger budget does not help. The progression breaks less often, but the average loss grows, because more money goes through: average turnover rises from £1,014 with a £1,000 budget to £2,283 with £100,000. At every budget you lose 7.8% of turnover, exactly what the margin takes from each bet (3.32 / 3.60 = 0.922).
Without the margin the system does not win either. At fair odds of 3.60 the same model averages exactly £0 at every budget: the progression still breaks in 15.6% of seasons with £1,000 and in 2.9% with £5,000, and those losses cancel out the small wins to the penny. The classic martingale at 2.00 fares no better: on virtual football with an 85% return, a run capped at seven bets still loses 15% of what it stakes.
Why a progression cannot win in the long run
- Every single bet is worth zero at best. At fair odds a bet returns on average exactly what you stake, and at a bookmaker the odds carry a margin: 8.5–10% on the draw at the three recreational bookmakers we track, so each £1 returns 91–92p on average. A progression does not change this arithmetic. It only changes the shape of the results: many small wins and a rare, very large loss.
- The budget always has a ceiling; the run does not. The question is not whether you will hit a long run, but when. In 65% of league seasons some team went 15 matches without a draw.
- Stakes have limits. A bookmaker sets a maximum stake and a maximum payout per bet, and accounts that put large sums on single events often have their stakes limited. A progression that needs £5,160 on one bet after 20 matches without a draw, and £96,517 on the 30th, will sooner or later hit a limit. Then you cannot place the stake that wins back the losses, and the whole run stays a loss. How maximum stakes work, we explain in our guide to maximum bet limits.
- The pressure builds. A bet of more than £5,000 to win £20 is a short road to chasing losses and decisions made in the heat of the moment.
The margin also makes the stakes climb faster, because every bet pays less. Total staked on a draw whose fair odds are 3.60, with a target of £20 per cycle:
| matches without a draw | fair odds 3.60 | odds 3.32 (about 8.5% margin) | odds 3.27 (about 10% margin) |
|---|---|---|---|
| 15 | £2,616 | £4,303 | £4,754 |
| 20 | £13,396 | £25,924 | £29,592 |
| 30 | £347,440 | £934,380 | £1.14m |
After 20 matches without a draw the margin almost doubles the cost of the progression, from £13,396 to £25,924 at 3.32. With an unlimited budget and no stake limits a progression would always collect its £20 in the end. With a real budget, at a real bookmaker, the margin takes a share of every bet, the stakes run into the limits, and a long run can be neither sat out nor won back.
Where the numbers come from and what they do not say
- League matches only. The runs and cycles cover league matches only; cup and European matches are not included. If you also back a team's draw in the cups, its run looks different.
- The record runs come from more than 300,000 matches from 37 leagues, including second divisions, over more than 20 years.
- The cycles and the draw odds of the runs come from the 28 leagues for which we have closing odds (from the 2015/16 season). The amounts in pounds are calculations at constant odds, in the stakes tables and the season model; real draw odds change from match to match.
- A run ends in our data when a team drops into a league outside our database. Some runs may therefore have lasted longer than we show.
- Ongoing runs are as of 28 September 2026. They may have ended since.
- This describes the past, not a forecast. The data shows how often long runs happened, not when the next draw will come.
If you still want to back draws
A draw is an ordinary bet like any other, and there is nothing wrong with playing it. Just do it without a progression:
- Use a flat stake, set before the season, not one worked out from your previous losses.
- Look for value, not a "due" draw. Compare the bookmaker's odds with the chance from our forecasts or from the fair odds calculator. If the bookmaker's odds are not higher than the fair odds, the bet has no edge. There is more on this in the guide to value betting.
- Set a daily and a monthly loss limit, ideally straight away in your bookmaker account. Our bankroll rules will help.
- Write every bet down. A betting record shows after a few months what your draws really cost you.
If you catch yourself raising your stakes to win back losses, stop and take a break, and if it keeps happening, talk to someone about it.
FAQ
It is a system in which you back the same team to draw match after match and, after every loss, raise the stake so that the first draw wins back all the losses and leaves a small profit. It is a version of the martingale.
No. In our model of a 38-match season at odds of 3.32 on a draw fairly priced at 3.60, the typical season made £171–180, but the average result was a loss at every budget: 7.8% of turnover, exactly the margin. At fair odds the average is exactly zero. The rare long runs without a draw take back everything the successful cycles win, and with a margin more than that.
In our data of more than 300,000 league matches the record belongs to FC Schaffhausen of the Swiss second tier: 45 matches without a draw in 2017–2018. In the top leagues Barcelona have gone 42 LaLiga matches without a draw (ongoing as of 28 September 2026), and Juventus went 38 in Serie A in 2016–2017. In England, Tottenham in the Premier League (2018–2019) and Blackburn Rovers in the Championship (2022–2023) both went 32. In Ligue 1 the record is Rennes' 28 (2024–2025), in the Bundesliga Borussia Dortmund's 27 (2019–2020).
In theory, yes: with an unlimited budget and no stake limits every run ends in a draw sooner or later, and the cycle ends in profit. In practice the budget is finite and bookmakers cap stakes and payouts. At odds of 3.60, 20 matches without a draw already mean £13,396 staked, and a margin of about 8.5% (odds of 3.32) raises that to £25,924.
More often than you would think. A run of 15+ matches without a draw happened 1,797 times in our data, and in 65% of league seasons at least one team had one within the season. One draw progression cycle in eighty lasted longer than 15 matches.
No. Teams after 15 or more matches without a draw drew 23.4% of the time, exactly as often as the odds expected and less often than the average of all matches (about 26%). A run does not raise the chance of a draw.
No budget is safe. At draw odds of 3.60 and £20 per cycle, 20 matches without a draw already mean more than £13,000 staked, and 30 matches more than £347,000. A bigger budget only makes the progression break less often, and when it does break, you lose more.
In the classic martingale you double your stake after a loss, because you bet at odds close to 2.00. In a draw progression the odds are higher, usually 3–4, so the stake grows more slowly, by between a third and a half per bet (38% at 3.60). The mechanism is the same, though: every win has to cover all the earlier losses, and a long losing run ends in a loss that cannot be won back.