Accuracy
How accurate is the Otherdeed for Otherside 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)
94.6%
Mean accuracy (100−APE)
88.0%
Within ±5%
49%
Within ±10%
65%
Within ±20%
84%
Bias
-0.5%
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 | 1641 | 95.7% | 69% | -1.0% |
| middle | 40 | 60.5% | 10% | -5.9% |
| grail | 267 | 89.9% | 49% | +3.0% |
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 |
|---|---|---|---|---|---|---|
| #89071 | floor · high | Aug 3, 2026 | 0.100Ξ | 0.104Ξ | +4.1% | 95.9% |
| #93174 | floor · high | Aug 3, 2026 | 0.100Ξ | 0.108Ξ | +8.1% | 91.9% |
| #70271 | floor · high | Aug 3, 2026 | 0.100Ξ | 0.113Ξ | +13.1% | 86.9% |
| #47375 | floor · high | Aug 2, 2026 | 0.093Ξ | 0.100Ξ | +8.0% | 92.0% |
| #59803 | floor · high | Aug 2, 2026 | 0.093Ξ | 0.102Ξ | +10.0% | 90.0% |
| #56056 | grail · high | Aug 1, 2026 | 0.099Ξ | 0.099Ξ | +0.3% | 99.8% |
| #84896 | floor · high | Aug 1, 2026 | 0.099Ξ | 0.102Ξ | +3.2% | 96.8% |
| #53452 | floor · high | Aug 1, 2026 | 0.099Ξ | 0.098Ξ | -1.2% | 98.8% |
| #56056 | grail · high | Jul 31, 2026 | 0.096Ξ | 0.099Ξ | +2.9% | 97.1% |
| #53452 | floor · high | Jul 31, 2026 | 0.098Ξ | 0.096Ξ | -1.8% | 98.2% |
| #77110 | floor · high | Jul 31, 2026 | 0.125Ξ | 0.102Ξ | -18.4% | 81.6% |
| #88336 | floor · high | Jul 31, 2026 | 0.109Ξ | 0.103Ξ | -5.3% | 94.7% |
| #88336 | floor · high | Jul 31, 2026 | 0.100Ξ | 0.111Ξ | +11.2% | 88.8% |
| #84896 | floor · high | Jul 31, 2026 | 0.100Ξ | 0.103Ξ | +3.1% | 96.9% |
| #3172 | floor · high | Jul 31, 2026 | 0.135Ξ | 0.116Ξ | -14.2% | 85.8% |
| #3172 | floor · high | Jul 30, 2026 | 0.111Ξ | 0.110Ξ | -0.7% | 99.3% |
| #3172 | floor · high | Jul 30, 2026 | 0.102Ξ | 0.121Ξ | +18.6% | 81.4% |
| #51990 | grail · high | Jul 29, 2026 | 0.113Ξ | 0.098Ξ | -13.5% | 86.5% |
| #88086 | floor · high | Jul 29, 2026 | 0.091Ξ | 0.101Ξ | +10.6% | 89.4% |
| #96922 | grail · high | Jul 29, 2026 | 0.099Ξ | 0.096Ξ | -2.8% | 97.3% |
| #81134 | grail · high | Jul 29, 2026 | 0.210Ξ | 0.131Ξ | -37.6% | 62.4% |
| #66972 | floor · high | Jul 28, 2026 | 0.099Ξ | 0.099Ξ | +0.2% | 99.8% |
| #88073 | floor · high | Jul 28, 2026 | 0.199Ξ | 0.119Ξ | -40.2% | 59.8% |
| #51446 | floor · high | Jul 28, 2026 | 0.104Ξ | 0.101Ξ | -2.5% | 97.5% |
| #66972 | floor · high | Jul 28, 2026 | 0.095Ξ | 0.107Ξ | +12.2% | 87.8% |
| #53 | floor · high | Jul 28, 2026 | 3.750Ξ | 0.125Ξ | -96.7% | 3.3% |
| #91060 | floor · high | Jul 28, 2026 | 0.110Ξ | 0.109Ξ | -0.9% | 99.1% |
| #84743 | floor · high | Jul 28, 2026 | 0.212Ξ | 0.123Ξ | -42.0% | 58.0% |
| #96922 | grail · high | Jul 28, 2026 | 0.095Ξ | 0.103Ξ | +8.5% | 91.5% |
| #94111 | grail · insufficient | Jul 28, 2026 | 0.700Ξ | 0.875Ξ | +25.0% | 75.0% |
| #69951 | floor · high | Jul 28, 2026 | 0.114Ξ | 0.105Ξ | -7.9% | 92.1% |
| #15336 | grail · low | Jul 28, 2026 | 1.360Ξ | 0.977Ξ | -28.2% | 71.8% |
| #86851 | grail · high | Jul 28, 2026 | 0.165Ξ | 0.105Ξ | -36.3% | 63.7% |
| #50909 | floor · high | Jul 27, 2026 | 0.105Ξ | 0.099Ξ | -5.7% | 94.3% |
| #25134 | floor · high | Jul 27, 2026 | 0.097Ξ | 0.104Ξ | +7.1% | 92.9% |
| #28537 | floor · high | Jul 27, 2026 | 0.116Ξ | 0.117Ξ | +0.5% | 99.5% |
| #30593 | floor · high | Jul 27, 2026 | 0.097Ξ | 0.085Ξ | -11.9% | 88.1% |
| #88275 | floor · high | Jul 27, 2026 | 0.100Ξ | 0.106Ξ | +6.3% | 93.7% |
| #93174 | floor · high | Jul 27, 2026 | 0.100Ξ | 0.106Ξ | +6.3% | 93.7% |
| #51990 | grail · high | Jul 27, 2026 | 0.104Ξ | 0.106Ξ | +1.9% | 98.1% |
Model otherdeed-1785727811-2026-08-03.2 · evaluated through 8/3/2026. Method details on the methodology page.