Event workspace
Results
published
Publication revision 1 · Evidence cutoff 2026-09-30 14:20 UTC · Digest a70d13ad59f6cb08d016a252b639fc44688f5c90be943f038d221756aa0c603c. Correction history.
- Method
- additive-judge-effects+shrunk-scale(kappa=2, floor=0.35, ceiling=2.5, rounds=12)
- Converged
- yes
- Grand mean
- 3.566
- Residual SD
- 0.564
- Rankings agree
- not enough comparisons to say
- Tiers the evidence supports
- 3 - first and second are not separated
| # | Tier | Project | Track | Ballots | Raw mean | Adjusted | 95% range | Was | Moved |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 1 | Dry Relay | trk_02 | 3 | 4.11 | 4.919 | 4.14 – 5.70 | 4 | ▲ 3 |
| 2 | 1 | Salt Ledger | trk_02 | 4 | 4.33 | 4.859 | 4.23 – 5.49 | 1 | ▼ 1 |
| 3 | 1 | Salt Kiln | trk_04 | 3 | 4.00 | 4.600 | 4.22 – 4.98 | 7 | ▲ 4 |
| 4 | 1 | Deep Beacon | trk_02 | 3 | 3.78 | 4.492 | 3.71 – 5.27 | 12 | ▲ 8 |
| 5 | 1 | Copper Orbit | trk_04 | 2 | 3.67 | 4.452 | 4.03 – 4.88 | 13 | ▲ 8 |
| 6 | 1 | Copper Kiln | trk_02 | 3 | 3.89 | 4.395 | 3.69 – 5.10 | 8 | ▲ 2 |
| 7 | 1 | North Drift | trk_04 | 5 | 3.80 | 4.338 | 3.99 – 4.68 | 10 | ▲ 3 |
| 8 | 1 | Iron Switch | trk_08 | 3 | 4.33 | 4.203 | 3.52 – 4.88 | 2 | ▼ 6 |
| 9 | 1 | Dry Harbour | trk_03 | 4 | 3.83 | 4.112 | 3.47 – 4.76 | 9 | - |
| 10 | 1 | Glass Beacon | trk_03 | 2 | 3.50 | 4.085 | 1.80 – 6.37 | 22 | ▲ 12 |
| 11 | 1 | Still Beacon | trk_08 | 2 | 4.17 | 4.022 | 2.95 – 5.10 | 3 | ▼ 8 |
| 12 | 2 | Dry Harbour | trk_03 | 5 | 3.33 | 3.806 | 3.30 – 4.31 | 32 | ▲ 20 |
| 13 | 2 | Flat Meadow | trk_02 | 3 | 3.44 | 3.786 | 3.12 – 4.46 | 27 | ▲ 14 |
| 14 | 2 | Small Loom | trk_06 | 3 | 3.56 | 3.751 | 2.85 – 4.65 | 18 | ▲ 4 |
| 15 | 2 | Salt Loom | trk_07 | 4 | 4.08 | 3.668 | 3.22 – 4.12 | 5 | ▼ 10 |
| 16 | 2 | Hollow Signal | trk_06 | 3 | 3.56 | 3.598 | 2.70 – 4.50 | 17 | ▲ 1 |
| 17 | 2 | Salt Drift | trk_08 | 3 | 3.67 | 3.590 | 2.91 – 4.27 | 15 | ▼ 2 |
| 18 | 2 | Glass Signal | trk_04 | 3 | 3.44 | 3.551 | 2.95 – 4.15 | 24 | ▲ 6 |
| 19 | 2 | Slow Trail | trk_01 | 3 | 4.00 | 3.538 | 2.97 – 4.11 | 6 | ▼ 13 |
| 20 | 2 | Small Meadow | trk_03 | 3 | 3.56 | 3.522 | 2.95 – 4.10 | 16 | ▼ 4 |
| 21 | 2 | Open Beacon | trk_07 | 3 | 3.44 | 3.360 | 2.85 – 3.87 | 25 | ▲ 4 |
| 22 | 2 | Flat Thread | trk_08 | 3 | 3.44 | 3.342 | 2.66 – 4.02 | 26 | ▲ 4 |
| 23 | 2 | Small Relay | trk_06 | 2 | 3.67 | 3.300 | 1.81 – 4.80 | 14 | ▼ 9 |
| 24 | 2 | North Compass | trk_02 | 3 | 2.89 | 3.272 | 2.57 – 3.98 | 40 | ▲ 16 |
| 25 | 2 | Dry Bridge | trk_04 | 3 | 3.11 | 3.248 | 2.64 – 3.86 | 37 | ▲ 12 |
| 26 | 2 | Green Switch | trk_07 | 3 | 3.78 | 3.211 | 2.67 – 3.75 | 11 | ▼ 15 |
| 27 | 2 | Loud Ledger | trk_01 | 3 | 3.44 | 3.156 | 2.59 – 3.72 | 28 | ▲ 1 |
| 28 | 2 | Deep Compass | trk_03 | 3 | 3.33 | 3.095 | 2.44 – 3.75 | 30 | ▲ 2 |
| 29 | 2 | Salt Ferry | trk_01 | 3 | 3.56 | 3.091 | 2.52 – 3.66 | 19 | ▼ 10 |
| 30 | 2 | Paper Anchor | trk_03 | 2 | 3.50 | 3.085 | 1.59 – 4.58 | 21 | ▼ 9 |
| 31 | 2 | Paper Thread | trk_08 | 3 | 3.22 | 2.990 | 2.31 – 3.67 | 35 | ▲ 4 |
| 32 | 3 | Flat Relay | trk_01 | 2 | 3.33 | 2.988 | 2.39 – 3.59 | 33 | ▲ 1 |
| 33 | 3 | Green Lantern | trk_07 | 5 | 3.40 | 2.897 | 2.45 – 3.34 | 29 | ▼ 4 |
| 34 | 3 | Dry Compass | trk_01 | 3 | 3.11 | 2.766 | 2.20 – 3.33 | 36 | ▲ 2 |
| 35 | 3 | Paper Harbour | trk_07 | 3 | 3.11 | 2.709 | 2.20 – 3.22 | 38 | ▲ 3 |
| 36 | 3 | Open Kiln | trk_07 | 2 | 3.50 | 2.700 | 1.91 – 3.49 | 20 | ▼ 16 |
| 37 | 3 | Slow Quarry | trk_08 | 3 | 2.89 | 2.633 | 1.95 – 3.31 | 41 | ▲ 4 |
| 38 | 3 | Slow Loom | trk_01 | 2 | 3.00 | 2.572 | 1.84 – 3.31 | 39 | ▲ 1 |
| 39 | 3 | Warm Beacon | trk_05 | 5 | 3.47 | 2.551 | 2.20 – 2.90 | 23 | ▼ 16 |
| 40 | 3 | Quiet Anchor | trk_05 | 3 | 3.33 | 2.346 | 1.96 – 2.73 | 31 | ▼ 9 |
| 41 | 3 | Amber Hours | trk_05 | 3 | 3.22 | 2.344 | 1.96 – 2.73 | 34 | ▼ 7 |
How much of this order the evidence supports
Tiers
3 levels across 41 projects.
A tier holds the projects that are not separated from the one at the top of it, so awarding on a tier's leader is defensible where awarding on the printed place is not. That is a statement about the ballots and not about the work: it means the panel did not give enough information to tell those entries apart, and a place chosen inside a tier is a judgement somebody has to make rather than a number this portal produced.
- First vs second
- not separated by this evidence
- Separation
- 1.44
- Strata
- 2.25
- Reliability
- 0.67
- Interval
- 95%
- Ballots per project
- 3.1
Reliability is the share of the spread in these scores that is real rather than noise: run the event again with a different panel of the same size, and this is roughly how much of the ranking would survive. Separation is the same quantity as a ratio, and the tier count comes from it directly by the Wright and Masters strata formula.
Where the spread came from
Between projects - 49.7%
Between judges - 40.9%
Unexplained - 9.4%
The first share is the one that should be largest: it is disagreement about the projects, which is what a judging round is for. The second is disagreement about the judges, and it is the share normalization exists to remove - a large one means the adjusted column above is doing real work rather than decorating the raw one. The third is everything neither explains.
- Between projects
- 0.863
- Between judges
- 0.783
- Residual
- 0.564
- One score's error
- 0.389
The same three quantities as spreads on the rubric's own scale, for anybody checking the arithmetic. They are shares of their own sum rather than an exact partition, which is what any fit on an incomplete design can offer.
What qualifies these numbers
- Judge C1 filed only 1 ballot; their scale was not estimated and is held at 1.0.
- Judge W filed only 2 ballots; their scale was not estimated and is held at 1.0.
- Judge T filed 4 ballots with a fitted scale of -1.75 — they are barely separating projects. Their information weight is 0.12, so they move the ranking very little, but consider adding a review rather than relying on this.
- Judge L filed only 2 ballots; their scale was not estimated and is held at 1.0.
- Judge J filed 3 ballots with a fitted scale of 0.00 — they are barely separating projects. Their information weight is 0.16, so they move the ranking very little, but consider adding a review rather than relying on this.
- Judge E filed only 2 ballots; their scale was not estimated and is held at 1.0.
- Judge F filed 3 ballots with a fitted scale of -1.94 — they are barely separating projects. Their information weight is 0.12, so they move the ranking very little, but consider adding a review rather than relying on this.
- Judge C filed only 2 ballots; their scale was not estimated and is held at 1.0.
- Judge M filed 3 ballots with a fitted scale of -0.33 — they are barely separating projects. Their information weight is 0.12, so they move the ranking very little, but consider adding a review rather than relying on this.
- Judge Q filed 4 ballots with a fitted scale of -0.06 — they are barely separating projects. Their information weight is 0.12, so they move the ranking very little, but consider adding a review rather than relying on this.
- Judge B filed only 1 ballot; their scale was not estimated and is held at 1.0.
- Judge N filed only 2 ballots; their scale was not estimated and is held at 1.0.
- Judge V filed only 2 ballots; their scale was not estimated and is held at 1.0.
- Duels are enabled but none have been decided, so there is nothing to check this ranking against.
- First and second place are not separated: Dry Relay at 4.92 and Salt Ledger at 4.86, against a 95% interval of plus or minus 0.78. Another review of both is worth more than a tie-break rule.
- Panel reliability is 0.67 and separation 1.44, so this ranking supports about 2 distinguishable levels. More reviews per project is the only fix; no amount of arithmetic manufactures evidence.
Judges are named positionally - "Judge B" - and not by account, deliberately: these caveats are published, and an appraisal of a volunteer is not something this portal publishes. The organizer's dashboard names them, because that is where the person who can act on it reads.
Appeals
Teams may file a private appeal within 7 days of publication. Only organizers and the appealing team can inspect an open appeal message.
File a private appeal
Submit a private appeal for your team's submitted project. Organizers will review the evidence and may publish a corrected result revision.
The same figures as JSON: this page's API route.