Savorant Research · Study protocol

Glass shuffle or pourer?.

Does blinding yourself with the glass shuffle give the same results as having someone pour?

Status
Open
Protocol
Version 1.0, 2026-10-18
Consent
S.shuffle.v1
Investigator
John Schuster, Drunken Unicorn Productions
Preregistration
Being posted to the Open Science Framework; the link will appear here
Ethics review
Not yet reviewed; we intend independent review before any results are published
Target
35 tasters with at least 18 rounds in each mode on the same pair
Runs until
The target is reached, or 2029-10-18

Why this study

A triangle test is only valid if the taster can’t see which glass is which. With a pourer that is guaranteed by design. With the glass shuffle it depends on the shuffle: a taster who half-remembers where the marked glass went is no longer blind. If solo results run higher than results with a pourer, the shuffle is leaking; if they match, the shuffle is a valid way to test yourself.

What taking part means

Run triangle rounds on the same pair both ways over time: some with a pourer, some with the glass shuffle. Your triangle rounds, including which blinding mode each round used. The app has recorded the mode only since this study opened, so only rounds from then on are used.

To join: in the app, turn on research contribution in Settings → Research, then open Glass shuffle or pourer? under Research studies, read it, and agree. You can leave with one tap.

Hypotheses

Design

Observational, within-taster comparison of two blinding modes on the same pair.

Who can take part. Adults 21 or older in the United States who use Savorant, have research contribution turned on, and have joined this study in the app.

Outcomes

Primary. Per-taster difference in proportion correct, glass shuffle minus pourer, on the same pair.

Secondary and exploratory.

Sample size

35 tasters with at least 18 rounds in each mode on the same pair (TOST, margin 0.10, assumed standard deviation of the difference 0.20, α = 0.05, power 0.80, true difference 0: n = 34.3, rounded up). Recruitment target 50.

Analysis

Two one-sided tests (TOST) for equivalence with margins of ±10 percentage points, α = 0.05; mean difference reported with a 90% CI (the interval that matches TOST).

Only data filed after a taster joins is used.

When a study reports more than one secondary test, p-values are adjusted with the Holm method.

Limitations we know about

Tasters choose which mode to use and when, so modes are not counterbalanced, and practice on the pair may favor whichever mode comes later.

Standards every Savorant study follows

The consent you agree to

This is the exact text shown in the app before you join.

What it’s forSavorant offers two ways to run a blind triangle: someone else pours, or you blind yourself with the glass shuffle. This study compares results between the two, so tasters know whether a solo result means the same as one run with a pourer.

What you doRun triangle rounds on the same pair both ways over time: some with a pourer, some with the glass shuffle.

What’s usedYour triangle rounds, including which blinding mode each round used. The app has recorded the mode only since this study opened, so only rounds from then on are used. Nothing else from your account.

How longThe study runs until it reaches its sample size or until 2029-10-18, whichever comes first.

RisksThis study involves tasting whiskey. Take part only if you are 21 or older and it is safe for you to drink. Pours can be small, you can spit instead of swallowing, and never taste before you drive. If you have any reason not to drink alcohol, please don’t take part. The other risk is privacy, and the next sections say how we protect it.

BenefitsThere is no direct benefit to you and no payment. What you get is the app as it already is; what the field gets is data nobody has.

Your choiceTaking part is voluntary and changes nothing else in the app. You can leave this study any time in Settings with one tap; leaving stops all future inclusion. Results already published can’t be withdrawn, but they never identify you.

PrivacyResearch uses your coded taster ID, never your name, email or handle. The analysis dataset leaves out anything that identifies you, and no group smaller than five tasters is ever reported. When results are published, we may post a de-identified dataset so other researchers can check the work.

PublishingWe will publish what we find, as group results, including if it shows nothing.

Ethics reviewThis study has not yet been reviewed by an independent research ethics board (IRB). We intend to have it reviewed before any results are published. If that review changes what we ask of you, we will ask you again. Once a board is appointed, its contact for questions about your rights will be listed here.

Who runs itPrincipal investigator: John Schuster, Drunken Unicorn Productions (maker of Savorant). Savorant’s maker runs this study and sells the app, so we have an interest in the results. That is why the full plan is published before any data is analyzed, and why results are published either way.

The full planEvery hypothesis, the sample size and how the data will be analyzed are published before analysis at savorant.com/studies/shuffle/.

ContactQuestions about this study: palate@savorant.com.

Amendments

All open studies · The evidence review · Privacy policy · palate@savorant.com