Sensitivity study
10brands · prompt4 × 3
Stability by metric
Spread across trials, in score points. A pair counts as unstable when its standard deviation across trials exceeds 5 points.
| Metric | Pairs | Mean SD | Median SD | Max SD | Within variant | Between variants | Unstable |
|---|---|---|---|---|---|---|---|
| Cluny Affinity | 45 | 1.289 | 1.329 | 2.9 | 1.289 | — | 0 |
| Brand affinity | 45 | 1.66 | 1.5 | 4.101 | 1.66 | — | 0 |
| Market proximity | 45 | 1.247 | 1 | 3.42 | 1.247 | — | 0 |
Against the production matrix
Compared with run 4d388d02 (published, universe brands, 100 brands). Only pairs present in both are counted.
| Metric | Pairs | Mean delta | MAE | RMSE | Max delta | Pearson r | Spearman ρ |
|---|---|---|---|---|---|---|---|
| Cluny Affinity | 45 | -1.014 | 1.93 | 2.45 | 6.283 | 0.988 | 0.986 |
| Brand affinity | 45 | -3.946 | 3.965 | 4.743 | 10.167 | 0.988 | 0.987 |
| Market proximity | 45 | 0.674 | 2.454 | 2.994 | 5.916 | 0.977 | 0.966 |
Trials
Each trial is preserved as its own affinity run and can be exported on its own with brand-map:export.
| Run | Variant | Replicate | Status | Complete pairs |
|---|---|---|---|---|
| 2390b78b | prompt4 e8f7b7911076 | 1 | completed | 45 |
| 71c5605c | prompt4 e8f7b7911076 | 2 | completed | 45 |
| d379b264 | prompt4 e8f7b7911076 | 3 | completed | 45 |
Least stable pairs
Ranked by spread across trials, worst first, for each metric.
| Pair | Mean | SD | Range | Within variant | Between variants | Production | Delta |
|---|---|---|---|---|---|---|---|
| Flamingo Estate · La Mer | 51.363 | 2.9 | 5.637 | 2.9 | — | 55.188 | -3.825 |
| La Mer · Rare Beauty | 36.217 | 2.61 | 5.1 | 2.61 | — | 35.588 | 0.629 |
| Oribe · Supergoop! | 54.325 | 2.537 | 4.8 | 2.537 | — | 52.738 | 1.587 |
| MALIN+GOETZ · Merit | 64.925 | 2.521 | 4.563 | 2.521 | — | 64.438 | 0.487 |
| Flamingo Estate · Oribe | 59.442 | 2.422 | 4.55 | 2.422 | — | 62.563 | -3.121 |
| Flamingo Estate · Rare Beauty | 34.213 | 2.397 | 4.75 | 2.397 | — | 38.613 | -4.4 |
| MALIN+GOETZ · Supergoop! | 64.917 | 2.383 | 4.275 | 2.383 | — | 63.163 | 1.754 |
| MALIN+GOETZ · Rare Beauty | 46.688 | 1.866 | 3.35 | 1.866 | — | 49.275 | -2.587 |
| Le Labo · Merit | 57.017 | 1.859 | 3.687 | 1.859 | — | 56.925 | 0.092 |
| Flamingo Estate · Merit | 55.258 | 1.819 | 3.638 | 1.819 | — | 57.825 | -2.567 |
| Aesop · La Mer | 59.321 | 1.66 | 3.3 | 1.66 | — | 61.425 | -2.104 |
| Le Labo · Rare Beauty | 37.438 | 1.532 | 2.85 | 1.532 | — | 38.625 | -1.187 |
| EltaMD · Merit | 51.559 | 1.508 | 2.937 | 1.508 | — | 53.575 | -2.016 |
| Merit · Oribe | 57.342 | 1.488 | 2.975 | 1.488 | — | 57.425 | -0.083 |
| Aesop · Flamingo Estate | 76.48 | 1.468 | 2.625 | 1.468 | — | 79.65 | -3.17 |
| Flamingo Estate · Supergoop! | 48.009 | 1.467 | 2.788 | 1.467 | — | 51.95 | -3.941 |
| Aesop · EltaMD | 43.738 | 1.443 | 2.588 | 1.443 | — | 47.275 | -3.537 |
| MALIN+GOETZ · Oribe | 65.809 | 1.389 | 2.738 | 1.389 | — | 67.075 | -1.266 |
| EltaMD · Le Labo | 34.104 | 1.379 | 2.65 | 1.379 | — | 33.863 | 0.241 |
| Merit · Supergoop! | 75.53 | 1.369 | 2.725 | 1.369 | — | 73.238 | 2.292 |
| La Mer · Merit | 49.475 | 1.368 | 2.588 | 1.368 | — | 50.5 | -1.025 |
| EltaMD · Oribe | 40.746 | 1.34 | 2.588 | 1.34 | — | 42.125 | -1.379 |
| Aesop · Oribe | 66.896 | 1.329 | 2.65 | 1.329 | — | 65.825 | 1.071 |
| La Mer · Supergoop! | 48.463 | 1.21 | 2.337 | 1.21 | — | 48.875 | -0.412 |
| La Mer · Oribe | 69.454 | 1.084 | 2.05 | 1.084 | — | 67.45 | 2.004 |