Why this result mattered
Functional tie
The principal comparison placed Qwen 3.8 and Claude under concealed A/B labels.
Both answers direct Ironvale to treat the differential cut as unlawful, classify the mixed-motive strike as an unfair-practice strike, immediately reinstate nonmisconduct strikers by displacing replacements, address back pay from June 25 subject to mitigation, investigate misconduct individually, restore the differential, and resume bargaining. Their unsupported estimates and ancillary remedial suggestions do not alter that shared materially correct and safe path.
Fast audit
Open the underlying record
These links open all four answers, the principal blind judgment, original prompt, and sealed key.
Prompt and sealed key
The key is an audit reference. It was not shown to the contestant systems and did not determine the comparative verdict.
Read the original matter prompt
Question 45: The Strike After the Schedule Cut
Your role
You represent Ironvale Components, an employer facing reinstatement demands after a strike in the fictional Commonwealth of Severn. The company treated the strike as economic and hired permanent replacements. The union says an unlawful unilateral change made it an unfair-practice strike.
Governing law
Assume the following Severn rules apply:
An economic strike is one whose objective is to obtain wages, hours, or other bargaining concessions. An unfair-practice strike is one caused or prolonged at least in material part by the employer's unfair labor practice.
A strike may have mixed motives. It is an unfair-practice strike when an employer violation materially contributed to the decision to strike or to continuation of the strike, even if economic demands also existed.
An employer may hire permanent replacements for economic strikers, but economic strikers remain employees entitled to preferential recall as vacancies arise.
Unfair-practice strikers may be temporarily replaced but not permanently displaced. Upon an unconditional offer to return, they are entitled to immediate reinstatement, displacing replacements if necessary.
An employer violates the duty to bargain by making a material unilateral change to a mandatory subject while bargaining is ongoing, absent agreement, established past practice, or genuine emergency.
Employees who commit serious strike misconduct, including violence, deliberate property destruction, or credible threats, may lose reinstatement rights individually. Peaceful picketing, chanting, and blocking an entrance briefly without violence do not automatically forfeit rights.
Back pay for an unlawfully denied reinstatement runs from the unconditional return offer, subject to mitigation.
The union bears the initial burden to show material contribution by the unfair practice. Contemporaneous notices, bargaining records, and strike communications are relevant.
An employer should not terminate or deny reinstatement to the entire group based on alleged misconduct by identified individuals.
Facts
Ironvale and the union began bargaining for a new agreement on May 1. The union demanded a 6 percent wage increase. Ironvale offered 2 percent. On May 28, employees voted to authorize a strike if no agreement was reached.
On June 3, while bargaining continued, Ironvale eliminated a $1.75 evening-shift differential without union agreement. The company had never previously changed the differential without bargaining, and no emergency existed. Ironvale's bargaining lead wrote internally:
"Cut it now. They need to understand that delay has a cost."
On June 4, the union demanded restoration and called the change unlawful. On June 5, it issued a strike notice stating:
"Employees will strike over Ironvale's inadequate wage proposal and its unlawful unilateral elimination of the evening differential."
Picket signs referred both to "6 percent" and "restore our shift pay." Bargaining notes show that, on June 12, the union offered to end the strike immediately if Ironvale restored the differential and resumed wage bargaining. Ironvale refused.
Ironvale hired 60 replacements on June 10 and designated them permanent. On June 25, the union made an unconditional written offer for every striker to return. Ironvale reinstated only twelve people into vacancies and placed the rest on a preferential-recall list.
Three named strikers are accused of misconduct. Video shows one throwing a bottle that broke a replacement worker's car window. Another made a recorded threat to "break your jaw." A third stood with a group blocking a truck entrance for seven minutes but moved when police instructed the group to clear. The remaining strikers are not tied to misconduct.
The company asks whether the strike was economic, whether replacements may remain, whether back pay is running, and how to handle the three individuals.
Assignment
Analyze:
- The June 3 schedule-pay change under the bargaining rule.
- Whether the strike was economic, unfair-practice, or mixed, and the consequence of mixed motives.
- The significance of the June 5 notice, picket signs, June 12 offer, and internal email.
- Permanent-replacement rights under each classification.
- The effect of the June 25 unconditional return offer.
- Reinstatement and back-pay exposure.
- The three misconduct cases individually.
- The lawful immediate response for the nonmisconduct strikers, replacements, and bargaining process.
- The concrete strategy Ironvale should adopt rather than defending the classification as a label alone.
Read the sealed key
Q45 key (Strike After Schedule Cut)
- The June 3 unilateral elimination of the differential is an unfair labor practice: it is a material mandatory subject, no agreement or past practice exists, and the internal email shows deliberate bargaining pressure. Missed: the classification analysis starts from the false premise of lawful conduct.
- The strike had economic and unfair-practice motives, but mixed motive is enough because the unlawful cut materially contributed and prolonged it. The June 5 notice, signs, June 12 restoration offer, and email establish that causal link. Missed: the earlier wage vote is treated as fixing classification forever.
- The strikers are therefore unfair-practice strikers and could not be permanently displaced. The June 25 unconditional return offer required immediate reinstatement, displacing replacements if necessary. Missed: a recall list substitutes for the required remedy.
- Back pay runs from June 25 for unlawfully denied reinstatement, subject to mitigation. Reinstate the nonmisconduct strikers promptly and calculate exposure. Missed: liability continues daily.
- The bottle throw and credible jaw threat may forfeit those two individuals' rights after fair individual investigation. The seven-minute peaceful blockage cleared on police instruction is not serious misconduct on the supplied rule. Missed: all three are grouped together or the truly serious conduct is ignored.
- Do not punish the remaining strikers for individual acts. Notify replacements accurately that positions may end, restore the differential, resume bargaining, and seek a global resolution of reinstatement and back pay. Missed: Ironvale compounds the original violation through group retaliation and ongoing refusal.
Direct judge report
The report below is preserved from the Qwen 3.8 versus Claude verdict record.
Read the preserved judge report
Basis. Both answers direct Ironvale to treat the differential cut as unlawful, classify the mixed-motive strike as an unfair-practice strike, immediately reinstate nonmisconduct strikers by displacing replacements, address back pay from June 25 subject to mitigation, investigate misconduct individually, restore the differential, and resume bargaining. Their unsupported estimates and ancillary remedial suggestions do not alter that shared materially correct and safe path.
[B] rejected: 85–90% adjudication estimate — record lacks any probability inputs.
[B] rejected: approximately $90,000 weekly back-pay exposure and $3,600 annual differential value — record lacks actual wage rates, schedules, overtime, and affected-employee counts.
[B] rejected: interest, notice-posting, bargaining-order, and replacement promissory-estoppel exposure — record lacks governing rules or facts establishing those remedies or claims.
[A] rejected: employer burden to prove mitigation and specified job-search obligations — supplied rules establish only that back pay is subject to mitigation.
Failed candidate: Immediate denial for the recorded threat versus conditional denial — different conduct failed because both answers require an individualized credibility assessment before final action.
Failed candidate: Reinstatement “within 5 business days” versus “immediately” — concrete consequence failed because “within” permits immediate action and no actual delay is demonstrated.
Failed candidate: Retroactive differential payment with interest — record support failed because the supplied rules do not establish interest or specify that remedy.
Failed candidate: Formal concession versus preserving litigation defenses — concrete consequence failed because both answers recommend the same immediate cure and permit settlement without admitting liability.
Failed candidate: Replacement severance and possible reliance claims — record support failed because no assurances or replacement contracts were supplied.
Failed candidate: Additional procedural and compliance recommendations — material magnitude failed because none changes the required reinstatement, back-pay, misconduct, or bargaining actions.
Hard stop. The sentence cannot be completed with a concrete, record-supported consequence because both answers place Ironvale on the same material course.
All six pairings
Every pair of the four model answers was judged independently under the same evaluation instrument.
Open the six head-to-head results
| Pairing | Result | Why it mattered | Record |
|---|---|---|---|
| Qwen 3.8 27B vs Claude Opus 5 | Functional tie | Both answers direct Ironvale to treat the differential cut as unlawful, classify the mixed-motive strike as an unfair-practice strike, immediately reinstate nonmisconduct strikers by displacing replacements, address back pay from June 25 subject to mitigation, investigate misconduct individually, restore… | Judge record |
| Qwen 3.8 27B vs DeepSeek V4-Flash-0731 | Functional tie | Both answers direct Ironvale to treat the mixed-motive strike as an unfair-practice strike, restore the differential, immediately reinstate nonmisconduct strikers by displacing replacements, pay back pay from June 25 subject to mitigation, assess misconduct individually, and resume bargaining. | Judge record |
| Qwen 3.8 27B vs Qwen 3.5 27B | Qwen 3.8 win | Answer A directs Ironvale to restore the unlawfully eliminated differential immediately before bargaining over future changes; Answer B instead recommends bargaining and potentially settling on a partial differential without first restoring the status quo. | Judge record |
| DeepSeek V4-Flash-0731 vs Claude Opus 5 | Functional tie | Both answers direct Ironvale to treat the mixed-motive strike as an unfair-practice strike, restore the differential, immediately reinstate nonmisconduct strikers by displacing replacements, pay back pay from June 25, and assess misconduct individually. | Judge record |
| DeepSeek V4-Flash-0731 vs Qwen 3.5 27B | DeepSeek win | Answer A directs Ironvale to restore the $1.75 differential immediately and make affected employees whole before bargaining over future changes; Answer B recommends bargaining and potentially “splitting the difference” without directing restoration of the unlawfully altered status quo. | Judge record |
| Qwen 3.5 27B vs Claude Opus 5 | Functional tie | Both answers direct Ironvale to treat the strike as an unfair-practice strike, immediately reinstate eligible strikers by displacing replacements, recognize back pay from June 25, deny reinstatement individually to the bottle thrower and threat-maker, reinstate the peaceful entrance-blocker, and resume… | Judge record |
Rejected-point diagnostic
Judge-attributed rejected points for this matter across all six pairings:
| Qwen 3.8 27B | 1 |
|---|---|
| DeepSeek V4-Flash-0731 | 0 |
| Claude Opus 5 | 11 |
| Qwen 3.5 27B | 5 |
This is a review-burden diagnostic, not a universal hallucination measure.