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
Aki maximum-likelihood b-value: log10 e numerator · case 01
b-values are about 5.3 times too large.
Aki maximum-likelihood b-value: log10 e numerator · case 02
b-values are about 5.3 times too large.
Aki maximum-likelihood b-value: log10 e numerator · case 03
b-values are about 5.3 times too large.
Aki maximum-likelihood b-value: log10 e numerator · case 04
b-values are about 5.3 times too large.
Aki maximum-likelihood b-value: log10 e numerator · case 05
b-values are about 5.3 times too large.
Aki maximum-likelihood b-value: minimum event count · case 01
A single event above Mc produces a b-value.
Aki maximum-likelihood b-value: minimum event count · case 02
A single event above Mc produces a b-value.
Aki maximum-likelihood b-value: minimum event count · case 03
A single event above Mc produces a b-value.
Aki maximum-likelihood b-value: minimum event count · case 04
A single event above Mc produces a b-value.
Aki maximum-likelihood b-value: minimum event count · case 05
A single event above Mc produces a b-value.
Aki maximum-likelihood b-value: standard error · case 01
Reported uncertainties shrink far too quickly with catalogue size.
Aki maximum-likelihood b-value: standard error · case 02
Reported uncertainties shrink far too quickly with catalogue size.
Aki maximum-likelihood b-value: standard error · case 03
Reported uncertainties shrink far too quickly with catalogue size.
Aki maximum-likelihood b-value: standard error · case 04
Reported uncertainties shrink far too quickly with catalogue size.
Aki maximum-likelihood b-value: standard error · case 05
Reported uncertainties shrink far too quickly with catalogue size.
Maximum-curvature completeness magnitude: bin assignment · case 01
Magnitudes such as 1.46 fall in the 1.4 bin, shifting Mc down.
Maximum-curvature completeness magnitude: bin assignment · case 02
Magnitudes such as 1.46 fall in the 1.4 bin, shifting Mc down.
Maximum-curvature completeness magnitude: bin assignment · case 03
Magnitudes such as 1.46 fall in the 1.4 bin, shifting Mc down.
Maximum-curvature completeness magnitude: bin assignment · case 04
Magnitudes such as 1.46 fall in the 1.4 bin, shifting Mc down.
Maximum-curvature completeness magnitude: bin assignment · case 05
Magnitudes such as 1.46 fall in the 1.4 bin, shifting Mc down.
Maximum-curvature completeness magnitude: tie-break direction · case 01
Ties choose the higher bin and Mc is overstated.
Maximum-curvature completeness magnitude: tie-break direction · case 02
Ties choose the higher bin and Mc is overstated.
Maximum-curvature completeness magnitude: tie-break direction · case 03
Ties choose the higher bin and Mc is overstated.
Maximum-curvature completeness magnitude: tie-break direction · case 04
Ties choose the higher bin and Mc is overstated.
Maximum-curvature completeness magnitude: tie-break direction · case 05
Ties choose the higher bin and Mc is overstated.
Maximum-curvature completeness magnitude: curvature correction · case 01
Mc is 0.2 units too low, admitting incomplete bins.
Maximum-curvature completeness magnitude: curvature correction · case 02
Mc is 0.2 units too low, admitting incomplete bins.
Maximum-curvature completeness magnitude: curvature correction · case 03
Mc is 0.2 units too low, admitting incomplete bins.
Maximum-curvature completeness magnitude: curvature correction · case 04
Mc is 0.2 units too low, admitting incomplete bins.
Maximum-curvature completeness magnitude: curvature correction · case 05
Mc is 0.2 units too low, admitting incomplete bins.
Epicentral distance and station selection: latitude cosine units · case 01
East-west distances are wildly wrong while north-south ones are fine.
Epicentral distance and station selection: latitude cosine units · case 02
East-west distances are wildly wrong while north-south ones are fine.
Epicentral distance and station selection: latitude cosine units · case 03
East-west distances are wildly wrong while north-south ones are fine.
Epicentral distance and station selection: latitude cosine units · case 04
East-west distances are wildly wrong while north-south ones are fine.
Epicentral distance and station selection: latitude cosine units · case 05
East-west distances are wildly wrong while north-south ones are fine.
Epicentral distance and station selection: angular distance conversion · case 01
Distances in degrees are reported in radians.
Epicentral distance and station selection: angular distance conversion · case 02
Distances in degrees are reported in radians.
Epicentral distance and station selection: angular distance conversion · case 03
Distances in degrees are reported in radians.
Epicentral distance and station selection: angular distance conversion · case 04
Distances in degrees are reported in radians.
Epicentral distance and station selection: angular distance conversion · case 05
Distances in degrees are reported in radians.
Epicentral distance and station selection: tie ordering · case 01
Station lists are alphabetical instead of nearest first.
Epicentral distance and station selection: tie ordering · case 02
Station lists are alphabetical instead of nearest first.
Epicentral distance and station selection: tie ordering · case 03
Station lists are alphabetical instead of nearest first.
Epicentral distance and station selection: tie ordering · case 04
Station lists are alphabetical instead of nearest first.
Epicentral distance and station selection: tie ordering · case 05
Station lists are alphabetical instead of nearest first.
Epicentral distance and station selection: haversine half angle · case 01
North-south separations are doubled.
Epicentral distance and station selection: haversine half angle · case 02
North-south separations are doubled.
Epicentral distance and station selection: haversine half angle · case 03
North-south separations are doubled.
Epicentral distance and station selection: haversine half angle · case 04
North-south separations are doubled.
Epicentral distance and station selection: haversine half angle · case 05
North-south separations are doubled.
Wood-Anderson amplitude simulation: mean removal · case 01
Traces with a DC offset integrate into huge drifting displacements.
Wood-Anderson amplitude simulation: mean removal · case 02
Traces with a DC offset integrate into huge drifting displacements.
Wood-Anderson amplitude simulation: mean removal · case 03
Traces with a DC offset integrate into huge drifting displacements.
Wood-Anderson amplitude simulation: mean removal · case 04
Traces with a DC offset integrate into huge drifting displacements.
Wood-Anderson amplitude simulation: mean removal · case 05
Traces with a DC offset integrate into huge drifting displacements.
Wood-Anderson amplitude simulation: trapezoid integration · case 01
Displacement amplitudes shift by half a sample of velocity.
Wood-Anderson amplitude simulation: trapezoid integration · case 02
Displacement amplitudes shift by half a sample of velocity.
Wood-Anderson amplitude simulation: trapezoid integration · case 03
Displacement amplitudes shift by half a sample of velocity.
Wood-Anderson amplitude simulation: trapezoid integration · case 04
Displacement amplitudes shift by half a sample of velocity.
Wood-Anderson amplitude simulation: trapezoid integration · case 05
Displacement amplitudes shift by half a sample of velocity.
Wood-Anderson amplitude simulation: Wood-Anderson gain · case 01
Local magnitudes are 0.13 units high.
Wood-Anderson amplitude simulation: Wood-Anderson gain · case 02
Local magnitudes are 0.13 units high.
Wood-Anderson amplitude simulation: Wood-Anderson gain · case 03
Local magnitudes are 0.13 units high.
Wood-Anderson amplitude simulation: Wood-Anderson gain · case 04
Local magnitudes are 0.13 units high.
Wood-Anderson amplitude simulation: Wood-Anderson gain · case 05
Local magnitudes are 0.13 units high.
Wood-Anderson amplitude simulation: peak definition · case 01
Negative-going pulses are ignored.
Wood-Anderson amplitude simulation: peak definition · case 02
Negative-going pulses are ignored.
Wood-Anderson amplitude simulation: peak definition · case 03
Negative-going pulses are ignored.
Wood-Anderson amplitude simulation: peak definition · case 04
Negative-going pulses are ignored.
Wood-Anderson amplitude simulation: peak definition · case 05
Negative-going pulses are ignored.
Clipping-aware amplitude pick: clip threshold inclusivity · case 01
A record touching 95% of full scale is used for magnitude.
Clipping-aware amplitude pick: clip threshold inclusivity · case 02
A record touching 95% of full scale is used for magnitude.
Clipping-aware amplitude pick: clip threshold inclusivity · case 03
A record touching 95% of full scale is used for magnitude.
Clipping-aware amplitude pick: clip threshold inclusivity · case 04
A record touching 95% of full scale is used for magnitude.
Clipping-aware amplitude pick: clip threshold inclusivity · case 05
A record touching 95% of full scale is used for magnitude.
Clipping-aware amplitude pick: half swing · case 01
Amplitudes are double and magnitudes 0.3 high.
Clipping-aware amplitude pick: half swing · case 02
Amplitudes are double and magnitudes 0.3 high.
Clipping-aware amplitude pick: half swing · case 03
Amplitudes are double and magnitudes 0.3 high.
Clipping-aware amplitude pick: half swing · case 04
Amplitudes are double and magnitudes 0.3 high.
Clipping-aware amplitude pick: half swing · case 05
Amplitudes are double and magnitudes 0.3 high.
Clipping-aware amplitude pick: adjacent swing versus range · case 01
Non-adjacent extremes several cycles apart are combined into one swing.
Clipping-aware amplitude pick: adjacent swing versus range · case 02
Non-adjacent extremes several cycles apart are combined into one swing.
Clipping-aware amplitude pick: adjacent swing versus range · case 03
Non-adjacent extremes several cycles apart are combined into one swing.
Clipping-aware amplitude pick: adjacent swing versus range · case 04
Non-adjacent extremes several cycles apart are combined into one swing.
Clipping-aware amplitude pick: adjacent swing versus range · case 05
Non-adjacent extremes several cycles apart are combined into one swing.
Clipping-aware amplitude pick: final sample turning point · case 01
A swing that ends at the window edge is missed.
Clipping-aware amplitude pick: final sample turning point · case 02
A swing that ends at the window edge is missed.
Clipping-aware amplitude pick: final sample turning point · case 03
A swing that ends at the window edge is missed.
Clipping-aware amplitude pick: final sample turning point · case 04
A swing that ends at the window edge is missed.
Clipping-aware amplitude pick: final sample turning point · case 05
A swing that ends at the window edge is missed.
Preferred magnitude selection: deep event threshold · case 01
Events at exactly 70 km prefer mb over Ms.
Preferred magnitude selection: deep event threshold · case 02
Events at exactly 70 km prefer mb over Ms.
Preferred magnitude selection: deep event threshold · case 03
Events at exactly 70 km prefer mb over Ms.
Preferred magnitude selection: deep event threshold · case 04
Events at exactly 70 km prefer mb over Ms.
Preferred magnitude selection: deep event threshold · case 05
Events at exactly 70 km prefer mb over Ms.
Preferred magnitude selection: Ms saturation floor · case 01
An Ms of exactly 5.5 is passed over for ML or mb.
Preferred magnitude selection: Ms saturation floor · case 02
An Ms of exactly 5.5 is passed over for ML or mb.
Preferred magnitude selection: Ms saturation floor · case 03
An Ms of exactly 5.5 is passed over for ML or mb.
Preferred magnitude selection: Ms saturation floor · case 04
An Ms of exactly 5.5 is passed over for ML or mb.
Preferred magnitude selection: Ms saturation floor · case 05
An Ms of exactly 5.5 is passed over for ML or mb.
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 ↗