Sow Forty-One Seeds to Get Thirty Plants

Test the packet you actually have

Seed loses viability with age, at a rate that varies enormously by species — parsnip and onion fade within a year or two, while tomato and cucumber stay usable far longer — and the percentage printed on a packet describes the lot when it was packed, not the packet in your drawer three winters later. Ten seeds folded into damp kitchen paper inside a closed bag, kept somewhere warm, gives a usable answer in one to three weeks: count what sprouts, multiply by ten, and you have this year's rate for this packet. It costs ten seeds and removes the guesswork from every calculation that follows.

Germinating is not surviving

The 10 % margin in this calculation covers seeds that fail to come up, and nothing else. Damping-off in a crowded tray, a cold night after planting out, slugs on the first true leaves and the ordinary attrition of transplanting all happen after germination and can easily take another tenth. If the thirty plants genuinely matter — a hedge, a crop you have promised, a bed that has to be full — raise the buffer to 25 or 30 % rather than adjusting the germination rate downwards, because the two failures have different causes and mixing them into one number hides which one to fix.

Frequently asked questions

Why round up instead of to the nearest seed?

Because the two errors are not symmetrical. One extra seed costs almost nothing and, if it comes up, gets thinned or given away; one seed short means the bed is not full and there is no way to fix it until next season. Any calculation whose output is a whole number of physical things you cannot obtain later should round up as a matter of course — the same reasoning applies to tiles, fence posts and floorboards.

Should I sow two seeds per cell instead?

It is the practical version of the same arithmetic and it works well for anything expensive to raise individually. At 82 %, a cell with two seeds fails entirely only when both fail — about three times in a hundred instead of eighteen — so almost every cell gives at least one plant, at the cost of thinning the doubles. Where seed is scarce or precious, sow singly and use the calculation instead; where seed is cheap and space is the constraint, doubling up is faster and needs no counting.