Trending Tokens Web3 Block for Svelte

Each slider holds a raw weight from 0 to 100 in steps of 5. A helper turns the three into shares that add to exactly 100 using the largest remainder method, the score is the sum of each signal times its share, and tokens sort by score with the typed order as the tie-break. If all three sliders are at 0 the three signals count equally, and the labels say so.

The table remembers the order the default weights give, computed once from the same helper, and shows how many places each token has moved from it. The first load always reads Same for every row, which is the check that the helper agrees with itself.

They are typed numbers in LINES. In your app compute them from your own data, such as mention counts against a trailing average, volume against a baseline and price momentum over a window, scale each to 0 to 100, and keep the criteria in the interface. This panel states that its signals are samples and is not a recommendation.

Everything. No request is made, nothing is stored, and the tokens are invented. To wire it, load the signals per token from your backend, keep the weights in your user's settings if you want them remembered, and recompute the rank on the client as here.

Yes. Each slider has a name, arrow keys move it by 5, and a separate polite message after you stop moving it says the weights and who is first. The table has a caption, column heads and row heads, and Reset is a button that answers when there is nothing to reset.

The three sliders stack, the signal bars column is hidden and the rest of the table keeps rank, token, change and score. The numbers behind the bars stay in the table's screen reader text.