How much of a liquid yeast pack is still alive?
Every viability calculator answers this with one number, and the number depends entirely on which published curve it borrowed. There are three in circulation. Brewer's Friend and the old Mr Malty rule say a pack loses 0.7 % of its cells a day — 21 % a month, in a straight line — so it's empty at 143 days. BeerSmith takes the same 21 % but compounds it: 96 % fresh, then 79 % of whatever was left, month after month. White Labs published a table for its PurePitch pouch that says something else entirely. Side by side:
| Pack age | Cautious (linear) | BeerSmith (exponential) | White Labs PurePitch |
|---|---|---|---|
| 1 month | 79 % | 76 % | 99 % |
| 2 months | 57 % | 60 % | 98 % |
| 3 months | 36 % | 47 % | 90 % |
| 4 months | 15 % | 37 % | 84 % |
| 5 months | 10 % | 30 % | 79 % |
| 6 months | 10 % | 23 % | 72 % |
Linear: 1 − 0.007 × days, floored at 10 % here (Brewer's Friend lets it reach zero). Exponential: 0.96 × 0.79^months (BeerSmith forum). PurePitch: White Labs' published points at whole months (99.2 / 98.1 / 90.3 / 84.3 / 79.4 / 71.6 %), interpolated. Computed by this page's engine.
At three months the three models say 36 %, 47 % and 90 %. That's not a rounding disagreement; it's a 2.5× spread in how much yeast you have, from calculators that all look equally confident. The tool above shows whichever model fits the pack, and always prints the other two underneath, because the honest answer to "how viable is my pack?" is a range with a reason.
Which model is right?
The measurements say: none of them, reliably — and the packaging matters more than the curve. Escarpment Labs counted its own strains weekly under a microscope at 3 °C and found most still above 90–95 % at 60 days, where the cautious rule puts a pack at 58 %; they declined to extrapolate past 60 days. Asheville Brewers Supply stained two packs made on the same day and tested at 7.5 months: the White Labs pouch was 92 % viable, the Wyeast smack pack 4 %. No model above predicts either number — the cautious rule has been sitting at its 10 % floor since day 129, and the PurePitch table, extrapolated past the six months White Labs published, says about 60 %. The pouch keeps oxygen out; the smack pack's inner nutrient sac is designed to be broken and then the clock runs fast. So the tool's default is the maker's own table for a PurePitch pouch and the cautious rule for everything else, and it flags any age past what the model's source actually published.
A curve is a guess about the average pack; yours has been through a warehouse, a courier and a shop fridge you didn't see. Wyeast's own advice is the practical test: a smack pack four months or more from its best-by should swell within 3–5 hours of breaking the sac. For any pack, a starter is both the fix and the test — if the first liter shows nothing by 36 hours, the count was lower than any model said. Store cold (2–4 °C; Wyeast says 1 °C is ideal), never frozen, and use a smack pack within two weeks of receipt if you can.
What low viability costs: packs, or a starter
This is where the viability calculators stop and this one keeps going. A 5.5-gallon ale at 1.050 needs about 193 billion cells at the textbook 0.75 million per mL per °Plato. A fresh 100-billion pack is 52 % of that — half the pitch, which is why the by-the-book answer for one pack is a starter even on day one. At three months the same pitch costs 6 packs by the cautious model, 5 by BeerSmith's and 3 by the White Labs table. The same batch as a lager wants 387 billion — 4 fresh packs — and the cautious model says 11 of them at three months, which is not advice anyone follows. The pitch-rate page draws the line at 4 packs: past that, the pack is the problem, and the choice is a stepped starter or fresher yeast. The tool applies the same line.
The result shows the cells you're short, not just a percentage, and "Size the starter" carries the pack's live count, the batch and the OG to the starter calculator in its own units, so the starter is sized on this pack rather than on a fresh one. If you've saved an equipment profile, the batch and yeast type are already filled from it.
Dry yeast: no date math
Dry yeast doesn't follow any of these curves. The manufacturers guarantee a minimum count to the best-before date — Fermentis prints ≥ 6 billion cells per gram — and independent counts commonly land at 10–20. So an 11 g sachet is 66 billion guaranteed, about 110 typical, up to 220, and the right question isn't "how old is it?" but "is it inside the date, and was it kept cool?" Past the date there is no published curve; rehydrate it, watch for activity, and treat the guaranteed count as optimistic. The tool's dry mode counts the sachet at the figure you pick and skips the date entirely, which is the honest thing to do with a number nobody has measured.