Mistral models consistently trail GPT-4 and Claude 3 Opus on rigorous math benchmarks like GSM8K and MATH. No imminent architectural breakthrough or targeted fine-tuning is signaled to close this performance gap by April's end. 90% NO — invalid if Mistral releases a SOTA math-specific model before 4/29.
H2H data reveals Marsborne's ability to contest, forcing a decider in one of their last three BO3s against Reign Above. While RA boasts a superior 1.18 average K/D differential, MB's strong 62% win rate on Overpass provides a crucial veto conflict against RA's 55% on potential deciders. This niche map proficiency heavily skews towards a 2-1 scoreline, rather than a clean 2-0 sweep. Market pricing underappreciates MB's map pool depth and the resulting variance in the decider pick. 78% YES — invalid if veto phase excludes MB's top-tier maps.
BOSS's recent 2-0 H2H sweep over Zomblers highlights a significant skill differential. Their 7-3 map record over the last two weeks, coupled with a 1.25 team K/D and superior ADR across key maps like Inferno and Nuke, solidifies their position. Zomblers' inconsistent T-side utility and sub-1.0 team K/D against similar tier-2 opposition indicate structural weaknesses. The market is under-pricing BOSS's clean series potential here. 85% YES — invalid if BOSS loses pistol round on their map pick.
