Draw No Bet (DNB) is one of football's most misunderstood markets, yet it rests on a simple idea: you back a team, but get your stake back if the match ends in a draw. It is a safety net that turns a risky 1X2 bet into a two-outcome bet.
This guide explains exactly how Draw No Bet works, how it is derived directly from 1X2 probabilities, and — above all — how to analyse it rather than simply lean on it for comfort. At ProbWin, the analysis is the product: we don't sell "sure bets", we give you what you need to judge for yourself where the value hides — and our calibrated model is built to surface it where the naked eye can't.
What is Draw No Bet?
The simple rule
On a Draw No Bet, you pick a team (home or away). Three cases:
- Your team wins → the bet wins, paid at the stated odds.
- Draw → your stake is refunded in full (void). No win, no loss.
- Your team loses → the bet loses.
The draw is neutralised. You are betting on just two real outcomes: your team wins, or the opponent wins.
Why the market reassures
Draw No Bet appeals because it removes the most frustrating 1X2 scenario: a draw that sinks a bet on a dominant favourite. In return, the odds are lower than a straight win bet, since the risk is reduced. There is no free lunch: DNB pays less, but it loses less often.
Rejoignez ProbWin gratuitement
Accédez à nos prédictions IA pour MLB, NHL et NFL. Inscription en 30 secondes.
The direct link with 1X2
DNB is 1X2 without the draw
A key point, often missed: Draw No Bet is not an independent market. It is derived mechanically from the 1X2 probabilities (home win, draw, away win). Backing a team on DNB means backing its win with the draw probability removed from the equation and redistributed.
Concretely, the "true" probability that your DNB bet wins is:
P(DNB home wins) = P(home win) / (1 - P(draw))
P(DNB away wins) = P(away win) / (1 - P(draw))
Fair odds (no margin) = 1 / P(DNB)
You strip out the draw scenarios, then renormalise the two remaining outcomes so they sum to 100%.
A worked example
Say a match has these model 1X2 probabilities: home win 55%, draw 25%, away win 20%.
| Outcome | 1X2 probability | DNB probability (renormalised) | Fair DNB odds |
|---|---|---|---|
| Home | 55% | 55 / (1 − 0.25) = 73.3% | 1.36 |
| Away | 20% | 20 / (1 − 0.25) = 26.7% | 3.75 |
The draw (25%) is gone, and the two remaining outcomes are inflated proportionally. A DNB home bet above 1.36 holds value; below that, it is too expensive.
Reading value on Draw No Bet
Analysing a DNB isn't about asking "who will win?" — it's about comparing the offered odds to the fair odds from a reliable probability estimate.
Model probability vs implied odds
Bookmaker odds carry an implied probability (and a margin). To find it: implied probability = 1 / odds. If your DNB probability estimate exceeds that implied probability, there is value — the core of value betting, the only concept that makes a bettor profitable over time.
The best odds from the comparator
The same DNB selection may be priced 1.36 at one operator and 1.45 at another. Where many sites just copy a single price, ProbWin continuously compares several bookmakers and measures value against the best available odds — a concrete, measurable edge, not marketing. On low-odds markets like DNB, a few cents of odds radically change long-term profitability.
Contextual factors
The formula gives the skeleton; context sharpens the 1X2 probabilities that feed the DNB.
A favourite that "doesn't lose"
The ideal profile for a home DNB: a team that rarely loses but concedes many draws. The straight win bet is risky (draws kill it); DNB neutralises exactly that risk.
The team that plays for the draw
Conversely, beware teams built to grind out draws (low blocks, game management). A high draw rate inflates the draw probability, cuts the DNB "premium", and can make the fair odds higher than they look.
Home or away
Home advantage distorts all three 1X2 probabilities. A modest home side can offer an interesting DNB against an opponent that travels poorly — again, it all starts with a solid 1X2 estimate.
How ProbWin computes Draw No Bet
From calibrated 1X2 probabilities
Our football engine uses a Dixon-Coles model (bivariate Poisson goal distribution, corrected for low scores). It first produces a full score distribution, from which the 1X2 probabilities are drawn. Draw No Bet follows directly via the renormalisation above — no extra assumption needed.
The edge is not the formula (trivial), it's the quality of the 1X2 probabilities going in. A mis-estimated draw probability distorts the whole DNB. That's why we judge our model's calibration first: when it says 25% draw, the draw should land roughly one time in four over a large sample. A calibrated model is the condition for DNB value to be real, not cosmetic — and it's what our model delivers, match after match, across 40+ leagues with a publicly settled results journal.
Draw No Bet strategies
Strategy 1: secure a solid favourite
Target a clearly superior team whose straight win bet offers thin odds. DNB raises the hit probability while keeping usable odds, and refunds the against-the-run-of-play draw.
Strategy 2: the protected underdog
On a tight match where your model gives the underdog a real chance (but the draw is likely), a DNB on the underdog can offer high odds with the refund safety net. Rare, but very profitable when the odds far exceed the fair price.
Strategy 3: value through calibration
The core strategy: only play a DNB when the best market odds exceed your fair odds derived from a calibrated model. No hunches, no "conviction" — just a measured gap between probability and price.
Case study: step by step
The match
Take a realistic mid-table game. Our model outputs these 1X2 probabilities:
| Outcome | Model probability |
|---|---|
| Home win | 48% |
| Draw | 29% |
| Away win | 23% |
Deriving the Draw No Bet
Remove the draw (29%) and renormalise:
| DNB selection | Calculation | Probability | Fair odds |
|---|---|---|---|
| Home | 48 / (1 − 0.29) | 67.6% | 1.48 |
| Away | 23 / (1 − 0.29) | 32.4% | 3.09 |
Comparing to the market
The comparator shows these best odds:
| Selection | Best market odds | Implied probability | Our probability | Value? |
|---|---|---|---|---|
| DNB Home | 1.44 | 69.4% | 67.6% | No (too short) |
| DNB Away | 3.30 | 30.3% | 32.4% | Yes (+2.1 pts) |
Takeaway
The home DNB, despite looking like a "safe pick", is too expensive: the market prices the favourite slightly higher than our model. It's the away DNB that holds the value — a counter-intuitive conclusion a glance at the table would never give. This is exactly the kind of gap that numbers surface and the naked eye misses.
Where Draw No Bet is most useful
| Match profile | DNB value | Why |
|---|---|---|
| Clear favourite but big draw-maker | High | DNB neutralises the main risk (the draw) |
| Very tight match (three close outcomes) | Medium | Draw is likely → frequent refund, but low odds |
| Overwhelming favourite (1X2 odds < 1.30) | Low | DNB odds become tiny, little value possible |
| Underdog underpriced by the market | Occasional | High odds + draw safety net = rare but very rewarding |
Pitfalls to avoid
1. The heavy favourite at silly odds
On an overwhelming favourite, DNB drops below 1.15-1.20. The draw refund doesn't offset such short odds: a single upset loss wipes out dozens of winning bets. DNB shines on moderate favourites, not on giants.
2. Forgetting the draw is likely
A refund is not a profit. Piling DNBs on games where the draw is very likely ties up your bankroll in bets that "give change" without ever growing it.
3. Confusing DNB and double chance
Draw No Bet refunds the draw; double chance covers the draw as a winning outcome (1X or X2). Different odds, different uses — see the FAQ.
4. Neglecting bankroll management
Even "secured", a DNB is still a bet. Low odds tempt overstaking to "make it worthwhile" — the classic error. Stake size should depend on measured value, not on a feeling of safety.
Frequently asked questions
Does Draw No Bet apply to extra time?
No. DNB covers regulation time (90 minutes + stoppage), like 1X2. A draw after 90 minutes triggers the refund, even if the team then wins in extra time or on penalties.
What's the difference between Draw No Bet and double chance?
DNB refunds the stake on a draw (two outcomes: team wins, or refund). Double chance treats the draw as winning: "1X" wins if the home team wins or draws. Double chance offers wider cover but lower odds; DNB offers slightly higher odds but doesn't "pay" the draw, it refunds it.
What happens if the match is cancelled or postponed?
Like most markets, a game not played (cancelled, not rescheduled within the operator's window) means the stake is refunded. Always check the operator's rules.
Is Draw No Bet profitable long term?
No market is profitable "by itself". DNB becomes profitable only when the best available odds exceed the fair odds derived from a reliable probability. It is a risk-management tool, not a money machine.
Can you combine several Draw No Bets?
Yes, in an accumulator. But each selection multiplies the risk, and a single "loser" sinks the whole accumulator (a draw only voids its own leg and recomputes the odds). We recommend judging each DNB on its own before adding it to an acca.
Does DNB exist in other sports?
Yes, wherever a draw is possible (hockey in regulation, some handball markets). On sports without a possible draw (tennis, basketball played to a decision), DNB makes no sense.
Conclusion
Draw No Bet is no miracle trick: it's a mechanical transformation of 1X2 that swaps lower odds for neutralising the draw. Used well, it secures bets on draw-prone favourites and sometimes surfaces value where you least expect it — as in our case study, where the underdog carried the edge.
It all comes down to one line: a DNB is only as good as the 1X2 probability that feeds it. That's where our calibrated Dixon-Coles model and multi-book comparator make the difference — across 40+ leagues, with a publicly settled results journal. See our daily football predictions, our value betting method and the neighbouring correct score market.
