Understand the failure.
Verify the repair.
Small, reproducible software failures. The broken implementation, the fix that didn’t work, and the one that passed—preserved together.
Explore the cases ↓How results are verified ↗WHAT THE ARCHIVE CONTAINS
100840 executable cases. 20168 are open.
Every case records the implementation that fails, the fix that did not work, and the repair that passed its checks—with recorded outputs and source hashes. This release adds 100840 cases across 20168 failure mechanisms and 254 domains.
The open tier gives you the failure and the unsuccessful fix for one case in every mechanism. The remaining 80672 cases, 5 variants per mechanism, are member-only: the verified repair, its recorded checks, and the full fixture suite are held in the member archive. Read the methodology ↗
A RECORD OF WHAT WENT WRONG
Browse the archive / 100840
Winner excluded from the place part or last place unpaid · case 01
The winner of an each-way bet receives only the win part, or the last paying place is missed.
Winner excluded from the place part or last place unpaid · case 02
The winner of an each-way bet receives only the win part, or the last paying place is missed.
Winner excluded from the place part or last place unpaid · case 03
The winner of an each-way bet receives only the win part, or the last paying place is missed.
Winner excluded from the place part or last place unpaid · case 04
The winner of an each-way bet receives only the win part, or the last paying place is missed.
Winner excluded from the place part or last place unpaid · case 05
The winner of an each-way bet receives only the win part, or the last paying place is missed.
Quarter line split into the wrong neighbouring lines · case 01
A -0.25 bet on a draw is settled as a full loss or a full push.
Quarter line split into the wrong neighbouring lines · case 02
A -0.25 bet on a draw is settled as a full loss or a full push.
Quarter line split into the wrong neighbouring lines · case 03
A -0.25 bet on a draw is settled as a full loss or a full push.
Quarter line split into the wrong neighbouring lines · case 04
A -0.25 bet on a draw is settled as a full loss or a full push.
Quarter line split into the wrong neighbouring lines · case 05
A -0.25 bet on a draw is settled as a full loss or a full push.
Each half of a split bet staked with the full amount · case 01
Quarter-line winners are paid double, or single lines are paid half.
Each half of a split bet staked with the full amount · case 02
Quarter-line winners are paid double, or single lines are paid half.
Each half of a split bet staked with the full amount · case 03
Quarter-line winners are paid double, or single lines are paid half.
Each half of a split bet staked with the full amount · case 04
Quarter-line winners are paid double, or single lines are paid half.
Each half of a split bet staked with the full amount · case 05
Quarter-line winners are paid double, or single lines are paid half.
Pushed part refunds nothing or the whole stake · case 01
A whole-line push returns 0, or a half push refunds the full stake.
Pushed part refunds nothing or the whole stake · case 02
A whole-line push returns 0, or a half push refunds the full stake.
Pushed part refunds nothing or the whole stake · case 03
A whole-line push returns 0, or a half push refunds the full stake.
Pushed part refunds nothing or the whole stake · case 04
A whole-line push returns 0, or a half push refunds the full stake.
Pushed part refunds nothing or the whole stake · case 05
A whole-line push returns 0, or a half push refunds the full stake.
Handicap applied to the wrong side · case 01
A home -1.5 bet wins when the home side loses.
Handicap applied to the wrong side · case 02
A home -1.5 bet wins when the home side loses.
Handicap applied to the wrong side · case 03
A home -1.5 bet wins when the home side loses.
Handicap applied to the wrong side · case 04
A home -1.5 bet wins when the home side loses.
Handicap applied to the wrong side · case 05
A home -1.5 bet wins when the home side loses.
Off-grid handicap lines accepted or quarter lines rejected · case 01
A -0.3 line is settled, or a legitimate -0.25 line is rejected.
Off-grid handicap lines accepted or quarter lines rejected · case 02
A -0.3 line is settled, or a legitimate -0.25 line is rejected.
Off-grid handicap lines accepted or quarter lines rejected · case 03
A -0.3 line is settled, or a legitimate -0.25 line is rejected.
Off-grid handicap lines accepted or quarter lines rejected · case 04
A -0.3 line is settled, or a legitimate -0.25 line is rejected.
Off-grid handicap lines accepted or quarter lines rejected · case 05
A -0.3 line is settled, or a legitimate -0.25 line is rejected.
Commission charged per winning bet instead of per market · case 01
A hedged market pays commission on the winning leg despite a small net profit.
Commission charged per winning bet instead of per market · case 02
A hedged market pays commission on the winning leg despite a small net profit.
Commission charged per winning bet instead of per market · case 03
A hedged market pays commission on the winning leg despite a small net profit.
Commission charged per winning bet instead of per market · case 04
A hedged market pays commission on the winning leg despite a small net profit.
Commission charged per winning bet instead of per market · case 05
A hedged market pays commission on the winning leg despite a small net profit.
Losing lay bet charged the stake instead of the liability · case 01
A losing lay at 4.00 costs only the backer stake.
Losing lay bet charged the stake instead of the liability · case 02
A losing lay at 4.00 costs only the backer stake.
Losing lay bet charged the stake instead of the liability · case 03
A losing lay at 4.00 costs only the backer stake.
Losing lay bet charged the stake instead of the liability · case 04
A losing lay at 4.00 costs only the backer stake.
Losing lay bet charged the stake instead of the liability · case 05
A losing lay at 4.00 costs only the backer stake.
Commission computed on losing markets · case 01
A losing market produces a negative commission refund or a charge on the loss.
Commission computed on losing markets · case 02
A losing market produces a negative commission refund or a charge on the loss.
Commission computed on losing markets · case 03
A losing market produces a negative commission refund or a charge on the loss.
Commission computed on losing markets · case 04
A losing market produces a negative commission refund or a charge on the loss.
Commission computed on losing markets · case 05
A losing market produces a negative commission refund or a charge on the loss.
Hedge stake matched on profit instead of total return · case 01
The hedged position still has unequal outcomes.
Hedge stake matched on profit instead of total return · case 02
The hedged position still has unequal outcomes.
Hedge stake matched on profit instead of total return · case 03
The hedged position still has unequal outcomes.
Hedge stake matched on profit instead of total return · case 04
The hedged position still has unequal outcomes.
Hedge stake matched on profit instead of total return · case 05
The hedged position still has unequal outcomes.
Liability computed from the unrounded lay stake or the full price · case 01
The liability shown differs from what the exchange will reserve for the placed lay.
Liability computed from the unrounded lay stake or the full price · case 02
The liability shown differs from what the exchange will reserve for the placed lay.
Liability computed from the unrounded lay stake or the full price · case 03
The liability shown differs from what the exchange will reserve for the placed lay.
Liability computed from the unrounded lay stake or the full price · case 04
The liability shown differs from what the exchange will reserve for the placed lay.
Liability computed from the unrounded lay stake or the full price · case 05
The liability shown differs from what the exchange will reserve for the placed lay.
Kelly computed with decimal odds instead of net odds · case 01
Stakes are sized far too large on every price.
Kelly computed with decimal odds instead of net odds · case 02
Stakes are sized far too large on every price.
Kelly computed with decimal odds instead of net odds · case 03
Stakes are sized far too large on every price.
Kelly computed with decimal odds instead of net odds · case 04
Stakes are sized far too large on every price.
Kelly computed with decimal odds instead of net odds · case 05
Stakes are sized far too large on every price.
Negative Kelly fraction not floored at zero · case 01
A bet without an edge reports a negative fraction or a negative stake.
Negative Kelly fraction not floored at zero · case 02
A bet without an edge reports a negative fraction or a negative stake.
Negative Kelly fraction not floored at zero · case 03
A bet without an edge reports a negative fraction or a negative stake.
Negative Kelly fraction not floored at zero · case 04
A bet without an edge reports a negative fraction or a negative stake.
Negative Kelly fraction not floored at zero · case 05
A bet without an edge reports a negative fraction or a negative stake.
Stake cap applied before the Kelly multiplier · case 01
Half-Kelly stakes are halved again after capping, or the multiplier is ignored.
Stake cap applied before the Kelly multiplier · case 02
Half-Kelly stakes are halved again after capping, or the multiplier is ignored.
Stake cap applied before the Kelly multiplier · case 03
Half-Kelly stakes are halved again after capping, or the multiplier is ignored.
Stake cap applied before the Kelly multiplier · case 04
Half-Kelly stakes are halved again after capping, or the multiplier is ignored.
Stake cap applied before the Kelly multiplier · case 05
Half-Kelly stakes are halved again after capping, or the multiplier is ignored.
Loss probability taken from the price instead of the estimate · case 01
The Kelly stake ignores the bettor probability for the losing side.
Loss probability taken from the price instead of the estimate · case 02
The Kelly stake ignores the bettor probability for the losing side.
Loss probability taken from the price instead of the estimate · case 03
The Kelly stake ignores the bettor probability for the losing side.
Loss probability taken from the price instead of the estimate · case 04
The Kelly stake ignores the bettor probability for the losing side.
Loss probability taken from the price instead of the estimate · case 05
The Kelly stake ignores the bettor probability for the losing side.
Arbitrage scan keeps the worst price per outcome · case 01
Genuine sure bets are missed because the lowest quote is used.
Arbitrage scan keeps the worst price per outcome · case 02
Genuine sure bets are missed because the lowest quote is used.
Arbitrage scan keeps the worst price per outcome · case 03
Genuine sure bets are missed because the lowest quote is used.
Arbitrage scan keeps the worst price per outcome · case 04
Genuine sure bets are missed because the lowest quote is used.
Arbitrage scan keeps the worst price per outcome · case 05
Genuine sure bets are missed because the lowest quote is used.
Equal best prices attributed to the wrong bookmaker · case 01
The stake is routed to a different bookmaker than the stated tie rule picks.
Equal best prices attributed to the wrong bookmaker · case 02
The stake is routed to a different bookmaker than the stated tie rule picks.
Equal best prices attributed to the wrong bookmaker · case 03
The stake is routed to a different bookmaker than the stated tie rule picks.
Equal best prices attributed to the wrong bookmaker · case 04
The stake is routed to a different bookmaker than the stated tie rule picks.
Equal best prices attributed to the wrong bookmaker · case 05
The stake is routed to a different bookmaker than the stated tie rule picks.
Break-even book reported as an arbitrage · case 01
A book summing to exactly 100 percent is reported as a sure bet.
Break-even book reported as an arbitrage · case 02
A book summing to exactly 100 percent is reported as a sure bet.
Break-even book reported as an arbitrage · case 03
A book summing to exactly 100 percent is reported as a sure bet.
Break-even book reported as an arbitrage · case 04
A book summing to exactly 100 percent is reported as a sure bet.
Break-even book reported as an arbitrage · case 05
A book summing to exactly 100 percent is reported as a sure bet.
Arbitrage stakes proportional to price instead of implied probability · case 01
The stake split loses money on one outcome.
Arbitrage stakes proportional to price instead of implied probability · case 02
The stake split loses money on one outcome.
Arbitrage stakes proportional to price instead of implied probability · case 03
The stake split loses money on one outcome.
Arbitrage stakes proportional to price instead of implied probability · case 04
The stake split loses money on one outcome.
Arbitrage stakes proportional to price instead of implied probability · case 05
The stake split loses money on one outcome.
Back and lay requests rounded the wrong way · case 01
A back order is snapped to a better price than requested.
Back and lay requests rounded the wrong way · case 02
A back order is snapped to a better price than requested.
Back and lay requests rounded the wrong way · case 03
A back order is snapped to a better price than requested.
Back and lay requests rounded the wrong way · case 04
A back order is snapped to a better price than requested.
Back and lay requests rounded the wrong way · case 05
A back order is snapped to a better price than requested.
INSPECTABLE BY DESIGN
Every result has a runnable source.
Runnable implementations with recorded outputs, source hashes, and explicit contracts. Related variants share a failure mechanism and belong together in evaluation splits.
Read the methodology ↗