Exact cards and fixed holdings
Enter one full hand exactly and place individual known cards with other seats when the auction or play has exposed them.
Bridge SolverContract bridge simulation
Bridge Simulation Software for Data-Driven Decisions
Generate deals that match what the auction and known cards tell you, then compare contract choices over the same field. Bridge Solver turns a partnership question into a repeatable simulation with visible assumptions.
Analyses require a Bridge Solver account.North♠ A K 8 4♥ Q T 7♦ K 9 3♣ J 8 2
South constraintsHCP 10-12♠ 4-5 cardsBalanced OR semibalanced
| Contract | Make | Avg score | IMP edge |
|---|---|---|---|
| 4♠ | 58.6% | +327 | +0.44 |
| 3♠ | 81.2% | +176 | -0.44 |
Workflow
How bridge simulation works
A useful Monte Carlo bridge simulation begins with a precise question and a defensible model of the unseen hands.
- 01
Describe the known hand
Enter one complete 13-card hand and choose which seat owns it, or make all four seats descriptive when no exact hand is available.
- 02
Model the auction information
Set HCP and suit-length ranges, assign known cards and combine shape or bidding filters for the other seats.
- 03
Generate matching deals
Bridge Solver samples complete 52-card deals and keeps only layouts that satisfy the active constraints.
- 04
Compare the outcomes
The generated deals are analyzed consistently so make rates, scores, tricks and contract differences come from the same sample.
Capabilities
Represent the hands you actually meant
The scenario builder is designed for bridge information, not just random hand generation.
HCP and suit ranges
Give every descriptive seat minimum and maximum high-card points and suit lengths rather than relying on a single generic shape.
Bidding-related filters
Use rules for balanced and unbalanced hands, notrump ranges, passed hands, weak openings, overcalls, controls and keycards.
Negative constraints
Exclude incompatible holdings with expressions such as !balanced or !weak 2 spades when an action makes those hands unlikely.
Three analysis modes
Compare two target contracts, analyze one specified contract, or ask for the best-contract ranking over the generated deals.
Saved, inspectable results
Review aggregate results and individual boards, filter the saved sample, and share a public result when discussion matters.
Example
Compare an invitation, game and slam decision
Suppose North has a known invitational hand after partner shows spades, but the partnership is debating whether to stop, bid game or investigate slam.
Build the relevant population
Enter North exactly. Describe South with the HCP range, spade length and shape promised by the auction. Give East and West broad but credible ranges, then remove holdings that conflict with their calls or passes. The result is not every possible bridge deal; it is the set of deals your assumptions say remain plausible.
Run the same accepted deals through 3 spades and 4 spades. A separate run can test 6 spades when the question is whether a slam try has value. Keeping the sample and assumptions aligned matters: otherwise a difference may come from the generated population rather than the contract decision.
Reading results
Use more than make percentage
Make rate answers how often a contract reaches its target under double-dummy play. Average score also captures overtricks, undertricks, game and slam bonuses, doubles and vulnerability. IMP difference converts the board-by-board score swing, while matchpoint comparisons show how often one target outscores the other in the saved sample.
A game can be preferable with a make rate below 50 percent when the vulnerable game bonus compensates for the failures. A safer part-score can still lead at matchpoints when the extra level converts too many plus scores into minuses. Read the result in the scoring context you are actually trying to decide.
Check the sample before trusting the headline
- Confirm the generated count and acceptance rate.
- Inspect boards on which the contracts disagree.
- Test whether a narrow HCP or shape assumption changes the answer.
- Keep vulnerability and declarer position faithful to the auction.
Methodology
What the simulation assumes
Bridge Solver first generates complete deals that satisfy the exact cards and descriptive rules. It then uses double-dummy analysis where contract outcomes are required. Double dummy knows all four hands and assumes optimal play, so it is consistent and powerful for comparing large samples, but it does not model a defender missing a difficult switch or declarer choosing a line without seeing the hidden cards.
The simulation is only as representative as its constraints. Auction inferences are partnership-dependent, and two reasonable models can produce different answers. Treat sensitivity testing as part of the analysis: widen a range, remove a filter, or compare a second interpretation before presenting the result as robust.
Questions
Bridge simulation FAQ
Is this the same as solving one bridge deal?
No. A solver answers what can happen when all four hands are known. A bridge simulation generates many complete deals consistent with the information you entered, solves each generated deal, and summarizes how a decision performs over that distribution.
Can I enter the auction?
The Simulation workflow does not currently interpret a full auction automatically. You represent auction information through HCP ranges, suit lengths, known cards, shape rules and bidding-related filters such as a passed hand, a weak two or a 1NT range.
How many deals can Bridge Solver generate?
The web workflow offers targets from 1,000 to 50,000 deals, with account limits controlling the larger runs. Restrictive filters may require many attempts and can finish with fewer accepted deals than requested.
Can I save and share a simulation?
Completed analyses appear in Results. Saved public analyses can be opened through a shareable result URL so a partner, teammate or student can inspect the same assumptions and output.
Does the simulation reproduce human play?
Not exactly. Contract outcomes use double-dummy analysis on each generated deal, so the play calculation assumes complete information and best play by both sides. The distribution of deals is statistical; the play model is double dummy.
Keep exploring
Related bridge analysis tools
Run your own analysis
Test the decision over matching deals
Describe the known hand, encode the auction information and compare the contracts on one matched sample.