If you eat the big potatoes and replant the small ones, do you breed for small potatoes?

Something about potatoes makes it not apply, and it is a nice piece of biology: planting a potato is cloning, not breeding.

A seed potato is a tuber — a piece of the parent plant. Growing it produces a genetically identical plant. Every tuber from one plant carries exactly the same genes, so choosing the small ones rather than the large ones selects nothing at all. There is no genetic variation between them to select on.

What makes one tuber big and another small on the same plant is position, resources and chance during that season, none of which is heritable because nothing is being inherited — it is the same individual continuing.

So the character's strategy is genetically neutral. Over any number of seasons the size distribution stays where the variety and the conditions put it.

There is a practical caveat that makes the strategy slightly worse than neutral, though, and it is worth knowing for a real garden.

Small tubers carry less stored energy. A seed potato feeds the shoot until it establishes roots and leaves. A very small one gives the plant a weaker start, which in poor conditions or a short season can mean a smaller crop — not for genetic reasons, purely because of the packed lunch.

The usual advice is to plant medium tubers, or to cut large ones into pieces each with a couple of eyes and let the cut surfaces dry before planting. That gets the energy without wasting the biggest potatoes on the ground.

So: eat the big ones by all means, and use medium rather than tiny ones for seed.

The genuinely serious problem with saving your own seed potatoes is not size at all — it is disease.

Potatoes accumulate viruses through the season, carried mainly by aphids, and those viruses pass into the tubers. The plant may look fine and still be carrying several. Replant those tubers and the next generation starts already infected, and the crop declines noticeably over a few years.

This is why certified seed potatoes exist and why they are grown in specific regions — often cooler and windier places where aphid pressure is low — and inspected. It is not a marketing arrangement; it is the actual reason the crop stays productive.

So for the fictional case, with no aphids and a closed system, saving tubers indefinitely is fine. In a garden, refreshing your seed stock every few years is the thing that matters, and it matters far more than which size you replant.

Worth adding the contrast that makes the point stick: this reasoning would apply to a crop grown from true seed.

Save seed from your smallest tomatoes each year and you really are selecting, because those seeds carry a genetic sample from that specific plant. Do it for a decade and you will have smaller tomatoes.

That is why the same intuition is correct for most of the garden and wrong for potatoes, garlic and anything else propagated vegetatively. The question is always: am I planting a seed or a piece of the parent?