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
Passed or failed heights counted as clearances · case 01
A height the athlete passed, or failed three times, is treated as cleared.
Passed or failed heights counted as clearances · case 02
A height the athlete passed, or failed three times, is treated as cleared.
Passed or failed heights counted as clearances · case 03
A height the athlete passed, or failed three times, is treated as cleared.
Passed or failed heights counted as clearances · case 04
A height the athlete passed, or failed three times, is treated as cleared.
Passed or failed heights counted as clearances · case 05
A height the athlete passed, or failed three times, is treated as cleared.
Decathlon points rounded instead of truncated · case 01
Performances just below a points boundary receive the next point.
Decathlon points rounded instead of truncated · case 02
Performances just below a points boundary receive the next point.
Decathlon points rounded instead of truncated · case 03
Performances just below a points boundary receive the next point.
Decathlon points rounded instead of truncated · case 04
Performances just below a points boundary receive the next point.
Decathlon points rounded instead of truncated · case 05
Performances just below a points boundary receive the next point.
Jumps scored in metres instead of centimetres · case 01
Every jump scores zero or is off by a point after conversion.
Jumps scored in metres instead of centimetres · case 02
Every jump scores zero or is off by a point after conversion.
Jumps scored in metres instead of centimetres · case 03
Every jump scores zero or is off by a point after conversion.
Jumps scored in metres instead of centimetres · case 04
Every jump scores zero or is off by a point after conversion.
Jumps scored in metres instead of centimetres · case 05
Every jump scores zero or is off by a point after conversion.
Middle-distance time minutes not converted to seconds · case 01
A 4:30.00 1500m scores as if it were 34 seconds or loses hundredths.
Middle-distance time minutes not converted to seconds · case 02
A 4:30.00 1500m scores as if it were 34 seconds or loses hundredths.
Middle-distance time minutes not converted to seconds · case 03
A 4:30.00 1500m scores as if it were 34 seconds or loses hundredths.
Middle-distance time minutes not converted to seconds · case 04
A 4:30.00 1500m scores as if it were 34 seconds or loses hundredths.
Middle-distance time minutes not converted to seconds · case 05
A 4:30.00 1500m scores as if it were 34 seconds or loses hundredths.
Track base computed in the field orientation · case 01
Running events score zero or slow times still earn points.
Track base computed in the field orientation · case 02
Running events score zero or slow times still earn points.
Track base computed in the field orientation · case 03
Running events score zero or slow times still earn points.
Track base computed in the field orientation · case 04
Running events score zero or slow times still earn points.
Track base computed in the field orientation · case 05
Running events score zero or slow times still earn points.
Point deductions applied to the wrong fighter · case 01
A fighter penalised for fouls gains ground on the cards.
Point deductions applied to the wrong fighter · case 02
A fighter penalised for fouls gains ground on the cards.
Point deductions applied to the wrong fighter · case 03
A fighter penalised for fouls gains ground on the cards.
Point deductions applied to the wrong fighter · case 04
A fighter penalised for fouls gains ground on the cards.
Point deductions applied to the wrong fighter · case 05
A fighter penalised for fouls gains ground on the cards.
Two-to-nil with a drawn card labelled unanimous · case 01
A win with one judge scoring a draw is announced as unanimous or split.
Two-to-nil with a drawn card labelled unanimous · case 02
A win with one judge scoring a draw is announced as unanimous or split.
Two-to-nil with a drawn card labelled unanimous · case 03
A win with one judge scoring a draw is announced as unanimous or split.
Two-to-nil with a drawn card labelled unanimous · case 04
A win with one judge scoring a draw is announced as unanimous or split.
Two-to-nil with a drawn card labelled unanimous · case 05
A win with one judge scoring a draw is announced as unanimous or split.
Majority draw labelled as a split draw · case 01
A 1-0-2 card is announced as a split draw.
Majority draw labelled as a split draw · case 02
A 1-0-2 card is announced as a split draw.
Majority draw labelled as a split draw · case 03
A 1-0-2 card is announced as a split draw.
Majority draw labelled as a split draw · case 04
A 1-0-2 card is announced as a split draw.
Majority draw labelled as a split draw · case 05
A 1-0-2 card is announced as a split draw.
Evens decimal price converted to a negative moneyline · case 01
A 2.00 price is displayed as -100 instead of +100.
Evens decimal price converted to a negative moneyline · case 02
A 2.00 price is displayed as -100 instead of +100.
Evens decimal price converted to a negative moneyline · case 03
A 2.00 price is displayed as -100 instead of +100.
Evens decimal price converted to a negative moneyline · case 04
A 2.00 price is displayed as -100 instead of +100.
Evens decimal price converted to a negative moneyline · case 05
A 2.00 price is displayed as -100 instead of +100.
Moneyline magnitude truncated or banker-rounded · case 01
A 2.005 price shows +100 instead of +101.
Moneyline magnitude truncated or banker-rounded · case 02
A 2.005 price shows +100 instead of +101.
Moneyline magnitude truncated or banker-rounded · case 03
A 2.005 price shows +100 instead of +101.
Moneyline magnitude truncated or banker-rounded · case 04
A 2.005 price shows +100 instead of +101.
Moneyline magnitude truncated or banker-rounded · case 05
A 2.005 price shows +100 instead of +101.
Favourite moneyline computed with the underdog formula · case 01
A 1.50 price is shown as -50 instead of -200.
Favourite moneyline computed with the underdog formula · case 02
A 1.50 price is shown as -50 instead of -200.
Favourite moneyline computed with the underdog formula · case 03
A 1.50 price is shown as -50 instead of -200.
Favourite moneyline computed with the underdog formula · case 04
A 1.50 price is shown as -50 instead of -200.
Favourite moneyline computed with the underdog formula · case 05
A 1.50 price is shown as -50 instead of -200.
Decimal price of 1.00 not rejected · case 01
A 1.00 price raises a division error instead of returning invalid.
Decimal price of 1.00 not rejected · case 02
A 1.00 price raises a division error instead of returning invalid.
Decimal price of 1.00 not rejected · case 03
A 1.00 price raises a division error instead of returning invalid.
Decimal price of 1.00 not rejected · case 04
A 1.00 price raises a division error instead of returning invalid.
Decimal price of 1.00 not rejected · case 05
A 1.00 price raises a division error instead of returning invalid.
Decimal price parsed as a binary float · case 01
Prices exactly on a half-dollar boundary round the wrong way.
Decimal price parsed as a binary float · case 02
Prices exactly on a half-dollar boundary round the wrong way.
Decimal price parsed as a binary float · case 03
Prices exactly on a half-dollar boundary round the wrong way.
Decimal price parsed as a binary float · case 04
Prices exactly on a half-dollar boundary round the wrong way.
Decimal price parsed as a binary float · case 05
Prices exactly on a half-dollar boundary round the wrong way.
Negative moneyline profit computed from the wrong ratio · case 01
A -200 favourite is expanded as a 3.00 decimal price.
Negative moneyline profit computed from the wrong ratio · case 02
A -200 favourite is expanded as a 3.00 decimal price.
Negative moneyline profit computed from the wrong ratio · case 03
A -200 favourite is expanded as a 3.00 decimal price.
Negative moneyline profit computed from the wrong ratio · case 04
A -200 favourite is expanded as a 3.00 decimal price.
Negative moneyline profit computed from the wrong ratio · case 05
A -200 favourite is expanded as a 3.00 decimal price.
Moneylines below 100 accepted · case 01
A "+50" or "-99" line is expanded instead of rejected.
Moneylines below 100 accepted · case 02
A "+50" or "-99" line is expanded instead of rejected.
Moneylines below 100 accepted · case 03
A "+50" or "-99" line is expanded instead of rejected.
Moneylines below 100 accepted · case 04
A "+50" or "-99" line is expanded instead of rejected.
Moneylines below 100 accepted · case 05
A "+50" or "-99" line is expanded instead of rejected.
Whole-number fractional odds rendered without denominator · case 01
Even money shows "1" instead of "1/1" and unreduced forms leak through.
Whole-number fractional odds rendered without denominator · case 02
Even money shows "1" instead of "1/1" and unreduced forms leak through.
Whole-number fractional odds rendered without denominator · case 03
Even money shows "1" instead of "1/1" and unreduced forms leak through.
Whole-number fractional odds rendered without denominator · case 04
Even money shows "1" instead of "1/1" and unreduced forms leak through.
Whole-number fractional odds rendered without denominator · case 05
Even money shows "1" instead of "1/1" and unreduced forms leak through.
Decimal price rounded with binary floats or truncated · case 01
A -800 line shows 1.12 instead of 1.13.
Decimal price rounded with binary floats or truncated · case 02
A -800 line shows 1.12 instead of 1.13.
Decimal price rounded with binary floats or truncated · case 03
A -800 line shows 1.12 instead of 1.13.
Decimal price rounded with binary floats or truncated · case 04
A -800 line shows 1.12 instead of 1.13.
Decimal price rounded with binary floats or truncated · case 05
A -800 line shows 1.12 instead of 1.13.
Implied probability uses one formula for both signs · case 01
Underdog probabilities come out above 50 percent.
Implied probability uses one formula for both signs · case 02
Underdog probabilities come out above 50 percent.
Implied probability uses one formula for both signs · case 03
Underdog probabilities come out above 50 percent.
Implied probability uses one formula for both signs · case 04
Underdog probabilities come out above 50 percent.
Implied probability uses one formula for both signs · case 05
Underdog probabilities come out above 50 percent.
Odds-on suffix stripped but the price not inverted · case 01
"2/1 on" is priced as 3.00 instead of 1.50.
Odds-on suffix stripped but the price not inverted · case 02
"2/1 on" is priced as 3.00 instead of 1.50.
Odds-on suffix stripped but the price not inverted · case 03
"2/1 on" is priced as 3.00 instead of 1.50.
Odds-on suffix stripped but the price not inverted · case 04
"2/1 on" is priced as 3.00 instead of 1.50.
Odds-on suffix stripped but the price not inverted · case 05
"2/1 on" is priced as 3.00 instead of 1.50.
Upper-case odds text rejected · case 01
"EVS" and "2/1 ON" are rejected or mispriced.
Upper-case odds text rejected · case 02
"EVS" and "2/1 ON" are rejected or mispriced.
Upper-case odds text rejected · case 03
"EVS" and "2/1 ON" are rejected or mispriced.
Upper-case odds text rejected · case 04
"EVS" and "2/1 ON" are rejected or mispriced.
Upper-case odds text rejected · case 05
"EVS" and "2/1 ON" are rejected or mispriced.
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 ↗