There are far more subnets than anyone can follow closely, and the list keeps changing. A watchlist is the answer most of us settle on: a handful of subnets you understand well enough to notice when something changes. This post is not a list of picks. It is how we decide which subnets deserve a closer look, and what we check first, so that you can build a watchlist of your own.
It is tempting to begin with a ranked table and pick from the top. A ranking, though, answers one question, and usually the wrong one. A subnet can lead on emission and be fragile, or sit near the bottom and be doing interesting work. So we start from a different set of questions: what does this subnet produce, who checks it, and how healthy is the market around it?
When a subnet catches our eye, this is what we look at first. It fits on one screen, and each item points to a page you can open on taostats.
- Is it doing real work? Read what the miners are asked to produce, and find an output you can judge for yourself. If you cannot tell what the subnet makes, that is information too.
- Who validates it? Look at how many validators there are, how stake is spread between them, and whether their VTrust is clustered. A subnet run by one dominant validator is a different bet from one with a broad set.
- Where does the incentive go? Check whether it is spread across many miners or concentrated in a few. See for how to read it.
- How deep is the pool? Depth decides what a trade of your size costs. Estimate your price impact before you decide anything else.
- Is emission steady? Watch the subnet's share of emission over a few weeks, not a few hours, and note whether it moves with the price or against it.
- Is anyone building? Developer activity is tracked per subnet. A quiet repository is not proof of anything, but a busy one is worth a second look.
- How is the neuron churn? Registration cost and deregistrations show how contested the subnet's slots are, and how often participants turn over.
Most of that checklist lives in a few places.
A watchlist only works if you come back to it. Three habits make it stick.
- Write down why. For each subnet, note in a sentence what made you add it. When the reason stops being true, that is your cue to remove it.
- Set a review rhythm. Weekly is plenty for most people. Daily tends to turn a watchlist into a price ticker.
- Keep it short. Five subnets you know well beat thirty you skim.
If you would rather automate the first pass, the API exposes the same inputs: subnet snapshots, pool prices and reserves, and the metagraph. Our shows how to pull them, and a small script that flags a change in pool depth or validator count is a good first project.
New subnets arrive regularly, and the pull of the new is strong. The same checklist applies to them, with one adjustment: there is less history, so there is less to confirm and more to take on trust. Give a new subnet more time on the list before it earns any of your capital, and let the checklist, not the excitement, decide.
