The Fusion Tax: A Diagnostic Analysis of When Hybrid Retrieval Hurts
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
A September 2026 study examines the costs of hybrid retrieval, which combines lexical BM25 with dense neural search and is widely recommended on the strength of average gains.
Score
0.67total under rubric v2.0
- Significance0.57
- Reader interest0.72
- Audience breadth0.48
- Practical relevance0.77
- Substantiated0.79
- Technical depth0.84
- New information0.87
- Counts against: Promotional0.03
- Counts against: Overclaim0.19
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| Significance | 0.64 | 0.35 | 0.72 | 0.57 | 0.37 | +0.25 | +0.143 |
| Reader interest | 0.66 | 0.60 | 0.90 | 0.72 | 0.30 | +0.20 | +0.144 |
| Audience breadth | 0.48 | 0.40 | 0.55 | 0.48 | 0.15 | +0.10 | +0.048 |
| Practical relevance | 0.82 | 0.70 | 0.78 | 0.77 | 0.12 | +0.15 | +0.115 |
| Substantiated | 0.84 | 0.75 | 0.78 | 0.79 | 0.09 | +0.10 | +0.079 |
| Technical depth | 0.83 | 0.85 | 0.85 | 0.84 | 0.02 | +0.10 | +0.084 |
| New information | 0.93 | 0.70 | 0.99 | 0.87 | 0.29 | +0.10 | +0.087 |
| Promotional | 0.06 | 0.03 | 0.01 | 0.03 | 0.05 | −0.15 | −0.005 |
| Overclaim | 0.21 | 0.25 | 0.12 | 0.19 | 0.13 | −0.15 | −0.029 |
| Sum of contributions | +0.666 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.666 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN —
Recorded for context only, checked 24 Oct 2026, 08:54 UTC. It played no part in the ranking.