Packet Germination vs. Your Own Test: Sizing a Reserve
Two germination numbers are sitting in front of you: the figure printed on the packet and the count from your own small test. They disagree. The useful move is not to split the difference. Run the same seed-demand division twice, once with each rate, keep both results visible, and write down in advance which one governs your sowing decision and how large a labeled reserve you will add on top.
Averaging the two rates hides the disagreement that made you look. Keeping them separate shows how much the planning quantity moves, and it forces you to state whether the packet claim or your own sample is the better evidence for this lot, this season, and this bed. The arithmetic below is a planning worksheet, not a germination standard, not disease certification, and not a promise of field emergence.
Why the two numbers are not the same kind of evidence
A packet germination figure is a claim about a tested sample of a lot. It is not a count you performed, and it is not a field-emergence prediction. The University of Minnesota Extension notes that the percentage of germination and seed purity is governed by law, which places the printed figure inside a labeling framework rather than making it a casual marketing number. The same page observes that the fresher the seed, the greater the chances that all the seeds will still be viable, and that fewer and fewer seeds from a packet will germinate over time.
Your own test is a different object. It is an observed count in a small sample under your conditions, on your date, with your counting rule. It describes those seeds and that method. It does not describe the whole packet, and it does not describe your garden bed.
Neither number is establishment. A seed that germinates on moist paper in a warm room can still fail in soil. Crusting, drying, damping off, cutworms, frost, and transplant shock all sit between "germinated" and "a plant you can use." The Extension page describes damping off as the death of tiny seedlings due to pathogens at the surface of the potting mix, which is exactly that gap. Keep germination and establishment as separate inputs, and treat neither as disease certification.
Because the two germination numbers answer different questions, averaging them produces a figure that describes nothing in particular. A label claim and a small-sample observation are not two measurements of one quantity. Run them separately.
Running the supplied example twice
The supplied example uses a hypothetical small sample of 17 germinations from 25 seeds, a target of 40 plants, and a separately stated 70% establishment assumption. The observed sample rate is 17 ÷ 25 = 0.68. The combined planning fraction is 0.68 × 0.7 = 0.476. Dividing the target by that fraction gives 40 ÷ 0.476 ≈ 84.03, which rounds upward to 85 whole seeds. The table inserted after this section records exactly those steps.
Read that table as a worksheet, not a test result. The 0.68 is your observed sample rate in this hypothetical; the 0.7 is an assumption you supply, not something the test measured. Rounding upward matters because you cannot sow part of a seed.
Now run the same division with the packet's printed figure in place of 0.68, holding the target and the establishment assumption fixed. If the packet claims, say, 0.90, the combined fraction becomes 0.90 × 0.7 = 0.63, and 40 ÷ 0.63 ≈ 63.49, which rounds to 64 seeds. The gap between 85 and 64 is the disagreement made visible. It is not an error to be smoothed away; it is the decision you actually face.
If your own records already give usable plants divided by all seeds sown, that single number is a combined rate. Multiplying it by a germination percentage again would count the same losses twice, so do not stack the two.
Illustrative example using explicitly stated assumptions; not a field test or product specification.
Write the tie-break rule before you sow
Decide in advance which rate governs, and write the reason down. A workable rule for a home garden: use your own observed sample rate when the sample was large enough to be informative and the method was consistent, because it reflects this lot under conditions closer to yours. Use the packet figure when your test was very small, when the method was rushed or inconsistent, or when the packet carries a recent test date and lot context that your sample cannot match.
State the boundary explicitly. For example: "If my sample used at least 20 seeds and the method matched my written protocol, the observed rate governs. Otherwise the packet figure governs, and I note the uncertainty." Choose the threshold before you count, not after you see which number you prefer.
Do not average. Averaging a label claim with a small-sample observation produces a number that is neither, and it destroys the ability to see how much the planning quantity moved. If you want a middle position, get it by choosing one rate and then adjusting the reserve, which is a labeled, visible step.
Record both rates, both resulting quantities, the rule you applied, and the date. When the season ends, compare the outcome with the prediction. That record is what makes next year's choice better than a guess.
Sizing a reserve as a separate, labeled step
A reserve is not a correction for a bad germination figure. It is a deliberate buffer for thinning, culling weak seedlings, a second sowing, or a poor stand. Add it after the division, as its own multiplication, and keep it visible so you can see how much of the total is buffer rather than expectation.
For the supplied example, 85 seeds is the planning quantity before any reserve. A 10% reserve gives 85 × 1.10 = 93.5, which rounds up to 94 seeds. A 20% reserve gives 85 × 1.20 = 102 seeds. Neither is more correct; the size should reflect how costly a short stand would be for you and how much seed you are willing to spend.
Keep the reserve separate from the establishment assumption. Establishment describes the share of germinated seedlings you expect to become usable plants. A reserve describes extra seed you are willing to sow to absorb uncertainty. Combining them into one number hides which part is measurement and which part is caution.
If you are sowing into a bed with known problems — crusting soil, heavy slug pressure, a history of damping off — a larger reserve is a reasonable response. Say so in your notes rather than quietly inflating the establishment figure.
What neither number can tell you
Neither the packet figure nor your sample predicts field emergence. Both describe germination under conditions that are not your bed. Soil temperature, sowing depth, moisture, crusting, pests, and weather sit between a germinated seed and a standing plant, and none of them appear in either number.
A small sample is also uncertain in a way a single number hides. Twenty-five seeds is a small sample, so the true rate for the packet could be meaningfully higher or lower than 0.68. If you want a tighter estimate, test more seeds or run a second sample from the same lot and compare the two counts before deciding.
A USDA Plant Hardiness Zone does not help here. It summarizes average annual extreme minimum winter temperature and is mainly a screen for perennial winter survival, as the USDA hardiness map guidance explains. It says nothing about germination, sowing depth, or establishment in your bed. For the distinction between winter hardiness and frost timing, see the frost dates vs. hardiness zones guide.
Finally, a germination figure is not a planting date. Timing comes from the packet's maturity wording, your local freeze context, and the current forecast. The seed packet guide explains how to read the maturity starting point, which is a separate question from how many seeds to sow.
A short worksheet you can reuse
Write these lines down before you sow, and fill them in as you go.
1. Packet claim: the printed germination figure, plus any test date or lot context the packet provides.
2. Your sample: seeds tested, normal seedlings counted, date, and method.
3. Observed sample rate: germinated ÷ tested.
4. Establishment assumption: your own labeled figure, with a note on whether it comes from records or is a conservative guess.
5. Combined planning fraction: sample rate × establishment assumption.
6. Planning quantity: target plants ÷ combined fraction, rounded up.
7. Reserve: a separate percentage, applied after step 6, with the reason written beside it.
8. Tie-break rule: which rate governs, and why, decided before counting.
Run steps 3 through 7 twice if the two germination numbers disagree, once for each rate. Keep both results in the record. The point of the worksheet is not to produce one authoritative number; it is to make every input visible so you can see which one drove the result and revise it next season.
Sources and method
- University of Minnesota Extension: starting seeds indoors — retrieved 2026-09-28
- USDA: using plant hardiness maps — retrieved 2026-09-28
AI-assisted writing with automated source, calculation and editorial checks. No hands-on testing or professional certification is claimed. Check the stated assumptions before applying this guide.