Scenario Lab

Change how much each component drives the composite, then re-run selection to compare against the base weights. Every result is a projected reordering under different assumptions, not a win probability.

Presets

The calibrated base weights: the blend that best reproduced the committee's picks across 2014-2024.

Normalized weights

Resume
40%
Predictive
30%
SOR
20%
SOS
10%
ScenarioBase

Adjust one or more weights to create a scenario. Base weights match the current run, so there is nothing to compare yet.

Reweight the model

Shift how much the four components drive the composite, then run a scenario to compare against the base run.

After running a scenario, this panel shows

  • Field changes
  • Seed changes
  • Bubble movement
  • Bracket impact