Experimental AI-assisted analytics, predictions and ratings are estimates, not official results.

Model scorecard

Accuracy Report

Every prediction is graded against what actually happened, always using only the results that were available before that match, so nothing is scored with hindsight. Chronological walk-forward: each match is predicted using only matches completed before it (warm-up of 6 qual matches per event). No final rankings or later results leak into a prediction.

Experimental, AI-assisted analytics

FRCinsights uses AI-assisted experimental analytics. Predictions, rankings, ratings, and insights are not guaranteed to be accurate and should not be treated as official results or professional advice. Data may be incomplete, delayed, miscategorized, or incorrect. FRCinsights is a new project under active development, and accuracy will improve over time.

Model apex-0.5.1 · 16913 predictions across 206 events · 2026-03-03 → 2026-05-02

Release validation, apex-0.5.1

18,788 matches tested · calculated 2026-08-21

Every completed 2026 match replayed in time order, each one predicted using only results available before it.

Overall accuracy

78%

Championship-strength

76%

Score MAE

55.7pts

Calibration error

0.004

Live season scorecard

The same checks run continuously on this season's results, so these move as new matches finish.

Winner accuracy

77%

Score MAE

55.8pts

Win-probability score

0.155

Calibration

0.009

Confidence calibration

When APEX says 70%, does the favorite win ~70% of the time? Closer bars = better calibrated.

Stated confidence	Actual win rate	Sample
50–60%	57%	3565
60–70%	66%	3145
70–80%	76%	2909
80–90%	87%	3012
90–100%	96%	4282

Baseline comparison

APEX vs naive predictors on the same matches.

50 / 50 coin flip53%

Better win-loss record67%

Higher rank (by record)67%

Recent-performance average70%

Best-predicted events

FIT District Waco Event92%n=59

CA District Ventura County Event90%n=81

FIM District West Michigan Event presented by Grand Valley State University89%n=89

St. Louis Regional88%n=78

Canadian Pacific Regional87%n=83

FIT District McAllen Event86%n=44

Worst-predicted events

Einstein Field60%n=10

Newton Division63%n=134

FIRST in Michigan State Championship - Hemlock Semiconductor Division64%n=91

FIRST Indiana State Championship64%n=86

FIRST in Michigan State Championship - Aptiv Foundation Division65%n=89

FMA District Robbinsville Event66%n=67

Where the model was most wrong

High-confidence predictions that missed.

Playoff Match 7 R2Favored Red at 98% , Blue won2026NCCMP

Qualification 75Favored Red at 98% , Blue won2026NVLV

Playoff Match 1 R1Favored Red at 98% , Blue won2026ONCMP2

Qualification 69Favored Blue at 98% , Red won2026MABOS

Playoff Match 1 R1Favored Red at 98% , Blue won2026GADAL

Finals 1Favored Red at 98% , Blue won2026SCCHA

Qualification 74Favored Red at 98% , Blue won2026MICMP1

Playoff Match 1 R1Favored Red at 98% , Blue won2026TUAK2

Playoff Match 7 R2Favored Red at 98% , Blue won2026MIMAS

Qualification 57Favored Red at 98% , Blue won2026ONBAR

Model limitations

Small samples early

Week 1 ratings rest on very few matches and swing the most. Confidence scores reflect this.

Cold-start teams

Teams with no current-season matches use heavily-regressed provisional estimates with wide uncertainty.

Rare events

Mechanical failures, no-shows, and unusual fouls are hard to predict and inflate score error.

Regression to the mean

APEX deliberately discounts single lucky/unlucky matches, so it can lag a genuine breakout for a match or two.

FRCinsights does not claim its predictions are guaranteed. Accuracy improves as more data arrives and the model is refined.