FROZEN PREDICTION

Eastern Michigan vs Michigan State

My CFB Baseline · 2026 Week 2 · revision 2 · Independent

Open this immutable run

Eastern Michigan by 2.3

Why this pick?

Game controlModerately leans away

Avoiding bad lossesSmall effect

Close-game gritHome data unavailable

Defensive toughnessModerately leans away

Scoring punchModerately leans away

Transfer portal impactSmall effect

Recruiting strengthModerately leans home

Returning productionModerately leans away

Team talentModerately leans home

Big-game winsModerately leans away

Recent formSmall effect

Wins away from homeNeutral effect

Tough schedulesStrongly leans away

Statement winsSmall effect

WinningModerately leans home

Show the technical details

Frozen 2026-09-09T06:26:39.256689-05:00 · Snapshot 2 · Home win probability 0.4289 · 80% margin interval -19.0 to 14.3.

Provisional calibration: real-season accuracy is not established.

Model configuration used

Configuration: Revision 2
Engine version: 3.7.0-frozen-classification
Evidence cutoff: Sep 8, 2026 · 6:43 PM CDT

Factors unavailable across the whole snapshot are excluded. Their weight is redistributed among available factors, so the effective model can differ from the configured sliders. Missing evidence for just one team remains neutral. The tables show saved weights; omitted factors are not reconstructed from current model settings.

Factor key	Normalized share
`adj_margin`	0.114286
`avoid_bad_losses`	0.042857
`close_game_wins`	0.028571
`defense_ppg`	0.071429
`offense_ppg`	0.071429
`preseason_portal`	0.042857
`preseason_recruiting`	0.071429
`preseason_returning`	0.057143
`preseason_talent`	0.071429
`quality_wins`	0.085714
`recent_form`	0.057143
`road_neutral_wins`	0.042857
`sos`	0.085714
`statement_wins`	0.042857
`win_pct`	0.114286

Retained calibration/context	Frozen value
`version`	provisional-all-slates-policy-v1
`calibration_kind`	documented_policy
`validation_status`	not_historically_validated
`applied_home_field_points`	2.5
`applied_fcs_baseline_z`	-1.5
`residual_scale`	13.0
`residual_scale_source`	documented_policy_not_validated
`intercept`	0.0
`difference_coefficient`	7.0
`home_field_coefficient`	2.5

These values came from this immutable run, not the model’s current settings.

The complete selected factor list was not retained for this older run. Omitted factors cannot be reconstructed reliably.

Factor	Home evidence	Away evidence	Normalized share	Home effective share	Away effective share	Weighted difference
Game control	Measured	Measured	0.114286	0.11904761904761904	0.12345013477088948	-0.16851190683552225
Avoiding bad losses	Measured	Measured	0.042857	0.044642857142857144	0.04629380053908356	0.029643740921473614
Close-game grit	Data unavailable	Measured	0.028571	0.02857142857142857	0.03086253369272237	0.029183946955640125
Defensive toughness	Measured	Measured	0.071429	0.0744047619047619	0.07715633423180593	-0.14485961289407473
Scoring punch	Measured	Measured	0.071429	0.0744047619047619	0.07715633423180593	-0.08162951452333495
Transfer portal impact	Measured	Measured	0.042857	0.0375	0.03214285714285714	-0.03637719036007794
Recruiting strength	Measured	Measured	0.071429	0.0625	0.05357142857142857	0.055426248481525796
Returning production	Measured	Measured	0.057143	0.049999999999999996	0.04285714285714286	-0.09012812915884354
Team talent	Measured	Measured	0.071429	0.0625	0.05357142857142857	0.06183834791099059
Big-game wins	Measured	Measured	0.085714	0.08928571428571429	0.09258760107816712	-0.18960653998071153
Recent form	Measured	Measured	0.057143	0.05952380952380952	0.06172506738544474	0.019030342603088765
Wins away from home	Measured	Measured	0.042857	0.044642857142857144	0.04629380053908356	0.0
Tough schedules	Measured	Measured	0.085714	0.08928571428571429	0.09258760107816712	-0.3053508628545185
Statement wins	Measured	Measured	0.042857	0.044642857142857144	0.04629380053908356	0.001782557460382403
Winning	Measured	Measured	0.114286	0.11904761904761904	0.12345013477088948	0.12942703940887143

Missing team data was represented neutrally in the calculation; it was not a measured neutral result. Measured equal or offsetting inputs can also have zero effect.

This is the exact saved prediction. Current model settings and current prediction selection are not used here.