Accuracy
How accurate is the CrypToadz model?
Every score comes from the newest chronological hold-out; the target sale is excluded from training, comparables, and its causal floor proxy.
Median accuracy (100−APE)
89.3%
Mean accuracy (100−APE)
84.0%
Within ±5%
28%
Within ±10%
48%
Within ±20%
72%
Bias
+7.7%
Accuracy by market tier
Floor, middle-market, and grail sales are scored separately so aggregate accuracy cannot hide a weak segment
| Tier | Sales | Median accuracy | Within ±10% | Bias |
|---|---|---|---|---|
| floor | 1379 | 90.6% | 52% | +5.0% |
| middle | 93 | 55.5% | 11% | +38.6% |
| grail | 175 | 87.2% | 40% | +12.3% |
Weekly median accuracy
Chronological hold-out accuracy by sale week
Error distribution
Signed hold-out error; positive means over-predicted
Scored hold-out sales
60 out-of-sample predictions frozen at evaluation time
| Item | Tier / evidence | Date | Sale | Model | Error | Accuracy |
|---|---|---|---|---|---|---|
| #1981 | floor · high | Sep 13, 2026 | 0.500Ξ | 0.324Ξ | -35.2% | 64.8% |
| #1800 | grail · high | Sep 12, 2026 | 0.267Ξ | 0.360Ξ | +34.8% | 65.2% |
| #2300 | grail · high | Sep 12, 2026 | 0.270Ξ | 0.378Ξ | +40.0% | 60.0% |
| #2497 | floor · high | Sep 12, 2026 | 0.277Ξ | 0.360Ξ | +30.0% | 70.0% |
| #755 | floor · high | Sep 12, 2026 | 0.279Ξ | 0.270Ξ | -3.3% | 96.7% |
| #755 | floor · high | Sep 12, 2026 | 0.270Ξ | 0.280Ξ | +3.7% | 96.3% |
| #755 | floor · high | Sep 11, 2026 | 0.280Ξ | 0.344Ξ | +22.9% | 77.1% |
| #3794 | floor · high | Sep 11, 2026 | 0.318Ξ | 0.305Ξ | -4.1% | 95.9% |
| #4805 | floor · high | Sep 11, 2026 | 0.292Ξ | 0.269Ξ | -7.7% | 92.3% |
| #3482 | middle · medium | Sep 11, 2026 | 0.420Ξ | 0.644Ξ | +53.3% | 46.7% |
| #4087 | floor · high | Sep 10, 2026 | 0.279Ξ | 0.277Ξ | -0.8% | 99.2% |
| #5701 | floor · medium | Sep 10, 2026 | 0.256Ξ | 0.334Ξ | +30.5% | 69.5% |
| #1628 | floor · high | Sep 10, 2026 | 0.270Ξ | 0.342Ξ | +26.7% | 73.3% |
| #4087 | floor · high | Sep 9, 2026 | 0.278Ξ | 0.302Ξ | +8.6% | 91.4% |
| #4852 | grail · high | Sep 9, 2026 | 0.275Ξ | 0.354Ξ | +28.7% | 71.3% |
| #6339 | floor · high | Sep 9, 2026 | 0.279Ξ | 0.332Ξ | +19.0% | 81.0% |
| #2849 | floor · high | Sep 9, 2026 | 0.285Ξ | 0.290Ξ | +1.8% | 98.3% |
| #877 | floor · high | Sep 8, 2026 | 0.304Ξ | 0.303Ξ | -0.5% | 99.5% |
| #4135 | grail · high | Sep 7, 2026 | 0.800Ξ | 0.959Ξ | +19.9% | 80.1% |
| #6947 | floor · high | Sep 6, 2026 | 0.278Ξ | 0.371Ξ | +33.5% | 66.6% |
| #1853 | floor · high | Sep 6, 2026 | 0.371Ξ | 0.281Ξ | -24.3% | 75.7% |
| #3132 | floor · high | Sep 6, 2026 | 0.362Ξ | 0.276Ξ | -23.8% | 76.2% |
| #877 | floor · high | Sep 5, 2026 | 0.300Ξ | 0.382Ξ | +27.3% | 72.7% |
| #5660 | floor · high | Sep 5, 2026 | 0.274Ξ | 0.270Ξ | -1.4% | 98.6% |
| #6921 | floor · high | Sep 4, 2026 | 0.267Ξ | 0.364Ξ | +36.3% | 63.7% |
| #5660 | floor · high | Sep 4, 2026 | 0.269Ξ | 0.273Ξ | +1.5% | 98.5% |
| #5365 | middle · medium | Sep 4, 2026 | 0.269Ξ | 0.700Ξ | +160.2% | 0.0% |
| #1551 | floor · high | Sep 3, 2026 | 0.271Ξ | 0.360Ξ | +32.8% | 67.2% |
| #365 | floor · high | Sep 3, 2026 | 0.272Ξ | 0.339Ξ | +24.6% | 75.4% |
| #3717 | floor · high | Sep 2, 2026 | 0.284Ξ | 0.331Ξ | +16.6% | 83.5% |
| #2740 | floor · high | Sep 1, 2026 | 0.289Ξ | 0.280Ξ | -3.0% | 97.0% |
| #4805 | floor · high | Aug 31, 2026 | 0.276Ξ | 0.345Ξ | +25.0% | 75.0% |
| #2321 | floor · high | Aug 31, 2026 | 0.272Ξ | 0.343Ξ | +26.1% | 73.9% |
| #1726 | floor · high | Aug 31, 2026 | 0.310Ξ | 0.376Ξ | +21.3% | 78.7% |
| #5293 | floor · high | Aug 31, 2026 | 0.280Ξ | 0.305Ξ | +8.9% | 91.1% |
| #2016 | floor · high | Aug 30, 2026 | 0.282Ξ | 0.283Ξ | +0.4% | 99.7% |
| #4661 | floor · high | Aug 30, 2026 | 0.300Ξ | 0.340Ξ | +13.3% | 86.7% |
| #5541 | floor · medium | Aug 30, 2026 | 0.283Ξ | 0.365Ξ | +29.0% | 71.0% |
| #830 | floor · high | Aug 29, 2026 | 0.283Ξ | 0.352Ξ | +24.4% | 75.6% |
| #229 | floor · high | Aug 29, 2026 | 0.283Ξ | 0.361Ξ | +27.6% | 72.4% |
Model cryptoadz-1789305347-2026-08-03.2 · evaluated through 9/13/2026. Method details on the methodology page.