Duke (1) vs Siena (16)

Not much daylight here—Duke -28.25 versus Vegas -28.5 is essentially a pass. Metis isn’t trying to rewrite reality; it just acknowledges that Duke’s elite efficiency (128 adj OE, 89.1 adj DE) and Siena’s historically porous numbers (107.1 OE, 109.2 DE) line up almost perfectly with Vegas’ oversized assumption. The slight half-point edge to Vegas is likely just the market giving a nod to hype over data—Duke’s subpar 3PT% (35.1%) and free throw inefficiency (72.4%) are baked into Metis’ simulation, which still spits out a number within half a possession.

If you lean Metis here, it’s because the model isn’t dazzled by narrative. Siena’s meh tempo (64.6) and lackluster offensive rebounding (30.3%) mean they can’t exploit Duke’s weaknesses, and Duke’s defensive lockdown (opponent eFG% 46.2, 2nd nationally) caps any upset hope. Essentially, Metis is just saying Vegas isn’t wrong—they’re just rounding up to make bettors feel something. Stick with the model if you like math over mood.