A blending judgment becomes reproducible when it is tested on labeled small samples with controlled dilution, measured ratios, comparative tasting, and written notes before full-batch scale-up.
Measurement does not remove palate variability, order effects, dilution error, or context bias. A successful small trial may still behave differently after marrying, proofing, filtration, or scale-up, so confirmation remains necessary.
Bench trials protect stock, preserve alternatives, and create a record that can be repeated or scaled. Elliott’s workflow adds a search structure: reduce the candidate pool, sketch several formulas, branch only promising directions, evaluate blind, and adjust intense components in smaller steps. Sensory judgment remains central, but it becomes auditable.
Strengthened with the Four Roses / Brent Elliott source cluster on 2026-08-29.
Durable whiskey knowledge promoted from a fully processed literature note. Maturity remains Developing until independent corroboration is connected.
Atomic statement
A blending judgment becomes reproducible when it is tested on labeled small samples with controlled dilution, measured ratios, comparative tasting, and written notes before full-batch scale-up.
Rationale and implications
Bench trials protect the base stock from irreversible changes, allow competing formulas to be compared side by side, and create a record that can be repeated or scaled. The method does not eliminate sensory judgment; it gives that judgment an experimental structure.
Counterpoints and boundaries
Measurement does not remove palate variability, order effects, dilution error, or context bias. A successful small trial may still behave differently after marrying, proofing, filtration, or scale-up, so confirmation remains necessary.
Evidence trail
- Current source: Hyde 2022, PDF sheets 187–203.
- Literature Note: How to Distill — full-source synthesis
- Claim: linked in the Claims relation
- Evidence status: exact source locator verified; cross-source corroboration still required
Review checklist
Manska 2025 — qualified application
Chapter 24 supports measured component trials, recorded ratios, rest and final-proof checks. Preserve native-proof and normalized-proof sessions separately. The author's unnamed numerical case studies are unverified; a one-person repeat is not an independent panel. No physical trial or recipe was changed.
Analysis: Distilled Spirits Sensory Science — critical reading and application. Located evidence:
Consumer blend trials need traceable component and session records. Existing atomic statement and maturity retained.
September 12 transcript evidence — qualified application
One of the strongest methodological sources: build a baseline, run contrasting starting strategies, retain a control and record what each change actually did.
How to Blend Your Whiskey Like a Master Distiller — analysis ·
YT-08 — located evidence and claim cautions
These are fully read supplied transcripts, not audio-verified videos or controlled experiments. Existing atomic statement and maturity are unchanged; missing recipe quantities and mechanistic claims remain restricted.
Jackson 2017 — September 28 source audit
Jackson's blend comparison (supplied PDF596–597, Fig8.18) supports investigating mixtures but does not show every blend beating its best component or isolate complexity as the cause. Pair measured formulas with component controls, independently prepared repeats and separate descriptive/liking outcomes. Wine evidence does not determine whiskey marrying time, serving proof or scale-up stability.
Provenance: Wine Tasting — sensory mechanics and evaluation methodology. Existing maturity and other source relations are preserved.
Richardson2024 — scope of contribution
PDF116supports tasting in the final beverage matrix,257–258controlled comparisons, and279batch records. These support an experimental habit but do not independently establish whiskey marrying times, blind-panel methods or scale-up equivalence. Existing whiskey-specific evidence and Developing maturity retained.
[SRC-131] Taste the ingredient in its final beverage matrix
Hyde complete-copy audit qualification
The measured-trial principle survives the source audit, but Hyde PDF201–202 contains a scale-up error and inconsistent bottling total.2/15/8ml gives8/60/32%;600ml requires48/360/192, not48/300/192. Equal water additions190 do not normalize unequal ABVs. Require independent arithmetic and final-proof checks; nominal volume math omits contraction. Existing maturity and other sources retained.