How do I grade my own LEGO set at home?

Score six axes 0–10 each: seal (25% weight), corners (20%), edges (15%), surface (20%), print (15%), structural (5%). Calculate the weighted sum. Map the result to AFA / CGC predicted grade: 9.0+ = AFA 90 Mint, 8.5–8.9 = AFA 85 Near Mint, 8.0–8.4 = AFA 80 Excellent. Cross-check with BrickGauge's AI for quantitative accuracy.

Can I accurately self-grade a LEGO set?

Within ±1 grade, yes — if you apply the rubric strictly. Within ±0.5, no — humans consistently under-call corner compression (the worst-corner rule is hard to apply without measurement), edge whitening percentage, and print drift. BrickGauge's AI catches the quantitative axes at ±0.5 grade accuracy against published AFA / CGC results.

What's the most important axis for DIY LEGO grading?

Seal axis at 25% weight. Any reseal indicator drops your seal score below 5, which caps the overall verdict at MAYBE or forces SKIP regardless of how Mint the other axes are. Inspect both top and bottom seal stickers under raking light first; if either is compromised, don't bother scoring the rest.

What lighting should I use for self-grading?

Diffuse daylight for overall inspection. Single raking light at 30° for close-up seal and corner inspection — raking light reveals adhesive bleed, double-pass marks, and corner compression that straight-on light hides. Avoid mixed color temperatures; ~5500K is the right white point.

How does my DIY grade compare to BrickGauge's AI grade?

Humans correlate within ±1 grade band on practiced self-graders, ±2 grade bands on first-timers. BrickGauge correlates within ±0.5 grade against published AFA / CGC results in 94.2% of cases. The quantitative axes (corners in mm, edges as percentage, print drift vs reference) are where AI beats human estimation.

Should I trust my self-grade for submission decisions?

Only if your DIY confidence is above 9.0 with strict rubric application. The risk asymmetry is bad: under-calling means missed grading opportunity (small loss); over-calling means submitting a box that downgrades and permanently documents below-Mint condition (large loss). Cross-check with BrickGauge before submitting any borderline call.

What does the BrickGauge AI do differently from DIY grading?

Three things humans can't do reliably. (1) Measure corner compression in mm via image analysis. (2) Calculate edge whitening as a precise percentage of perimeter. (3) Detect print drift below 3% sensitivity against factory color reference. Plus reseal detection at 97.4% recall — better than untrained human inspection by a wide margin.

Is DIY grading worth doing if BrickGauge is free?

Yes — as a baseline reference. Understanding the rubric helps you photograph correctly, store correctly, and make confident submission decisions. BrickGauge then runs the same rubric quantitatively. The two work together: DIY for understanding, AI for accuracy.

DIY 6-axis rubric · 10 min

How to grade your own LEGO set.

Score six axes 0–10 — seal, corners, edges, surface, print, structural — weight, sum, map to AFA / CGC tiers. The full DIY rubric, calibrated against the same standards graders use. Then cross-check with BrickGauge in 38 seconds.

Cross-check with AI →Methodology

⊙ Quick answer

What is grading your own LEGO set?

Grading your own sealed LEGO set is the process of applying the six-axis condition rubric — seal (25% weight), corners (20%), edges (15%), surface (20%), print (15%), structural (5%) — to a sealed box and calculating the weighted overall score. The result maps to AFA / CGC predicted grades: 9.0+ to AFA 90 Mint, 8.5–8.9 to AFA 85 Near Mint, 8.0–8.4 to AFA 80 Excellent, below 8.0 to AFA 75 or lower. Self-grading correlates within ±1 grade band of published AFA / CGC results; BrickGauge's AI tightens that to ±0.5 with quantitative axis measurement and 97.4% reseal-detection recall.

Axes

Time

~10 min

DIY accuracy

±1 grade

AI accuracy

±0.5 grade

The 8-step DIY rubric

How to grade a LEGO set yourself.

Score each axis 0–10. Apply the weights. Calculate the overall. Map to AFA / CGC prediction. Cross-check with BrickGauge for quantitative accuracy.

01
Score the seal axis (25% weight)
Inspect both top and bottom seal stickers under raking light. Look for adhesive bleed past the printed boundary, double-pass marks, off-axis rotation, lifted edges. Score 0–10: 10 = pristine, 5 = visible wear, 0 = reseal indicators. The lower of top/bottom seal is your seal score.
02
Score the corners axis (20% weight)
Measure all four front-face corners. Sub-1mm compression: 9–10. 1–2mm: 7–8. 2–4mm: 5–6. Over 4mm: below 5. The worst corner is your corners score — never the average.
03
Score the edges axis (15% weight)
Estimate edge whitening as a percentage of total perimeter affected. Sub-2%: 9–10. 2–5%: 7–8. 5–15%: 5–6. Over 15%: below 5.
04
Score the surface axis (20% weight)
Inspect both front and back panels for dents, scratches, scuffs, creases. No defects at arm's length: 9–10. Hairline scuffs only: 7–8. Visible scratches or single dent: 5–6. Multiple visible defects: below 5.
05
Score the print axis (15% weight)
Compare top face print to box underside. No visible color drift: 9–10. Trace top-face shift: 7–8. Visible top-edge yellowing or red→orange drift: 5–6. Heavy fade: below 5.
06
Score the structural axis (5% weight)
Check box squareness, top/bottom paper lift, internal structural integrity. Square, no lift: 9–10. Minor lift or warp: 7–8. Visible deformation: 5–6. Crushed or non-square: below 5.
07
Calculate the overall score
Weighted sum: (seal × 0.25) + (corners × 0.20) + (edges × 0.15) + (surface × 0.20) + (print × 0.15) + (structural × 0.05). Result rounds to your DIY overall grade. Map: 9.0+ → AFA 90 / CGC 9.5; 8.5–8.9 → AFA 85 / CGC 9.0; 8.0–8.4 → AFA 80 / CGC 8.5; below 8.0 → AFA 75 or lower.
08
Cross-check with BrickGauge
Run the six photos through BrickGauge's AI. The model uses the same six-axis weighted rubric, but with quantitative measurement (mm corner compression, % perimeter whitening, color drift detection vs factory reference). Calibrated against 240,127 published AFA / CGC grades at ±0.5 accuracy.

The DIY → AFA / CGC map

What your weighted overall score predicts.

DIY overall	AFA prediction	CGC prediction	Verdict
9.5+	AFA 95+ Mint+	CGC 9.8+	SEND (Express tier)
9.0 – 9.4	AFA 90 Mint	CGC 9.5	SEND
8.5 – 8.9	AFA 85 Near Mint	CGC 9.0	SEND / MAYBE
8.0 – 8.4	AFA 80 Excellent	CGC 8.5	MAYBE / HOLD
7.5 – 7.9	AFA 75 Very Fine	CGC 8.0	HOLD / SKIP
Below 7.5	AFA 70 or below	CGC 7.5 or below	SKIP — sell raw

DIY vs AI accuracy

Where humans miss and BrickGauge catches.

Humans miss

Worst-corner rule

Without measurement, humans average corner condition mentally. AFA scores the worst. BrickGauge measures each corner in mm independently.

Humans miss

Edge whitening %

Visual estimation under-calls perimeter whitening by 5–10 percentage points typically. The AFA 90 cutoff at sub-2% is hard to eyeball.

Humans miss

Print drift sensitivity

Sub-3% color drift is at the visual threshold for trained eyes. The AI detects it quantitatively against factory reference.

Humans miss

Reseal indicators

Trained collectors catch ~70% of confirmed reseals. BrickGauge catches 97.4%. The 8 inspection axes are hard to remember and apply simultaneously.

Humans catch

Contextual judgment

Humans bring set-specific knowledge — knowing a specific Modular tends to have a printing quirk, etc. AI uses statistical patterns from 240k training examples.

✓ Use both

DIY + AI together

Use the DIY rubric for understanding and photographing. Use BrickGauge for quantitative accuracy before submission.

The AI shortcut

Skip the math — BrickGauge runs the same rubric.

BrickGauge applies the identical six-axis weighted rubric with quantitative measurement on each axis. 38 seconds per scan. Two scans free on signup, no credit card. The DIY rubric is for understanding; the AI is for accuracy.

How-to cluster