tinybacktest

If you want to compare the previous altcoin season to the current one, I don't think you should recreate a historical basket using only the projects that survived until today.

Looking at the historical snapshot on CMC from November 7, 2021, you can see the rankings, names, symbols, market caps, prices, and circulating supplies from that time. I think starting from a list with a fixed date like that would be a good approach.

For example, let's say you started with 100 in A, B, and C, and later they became 200, 50, and 0 respectively—this is just a hypothetical, not actual snapshot return data. If you calculate including all three, you went from an initial 300 to 250, which is about -16.7%. But if you calculate it as if you only picked A and B from the start, 200 became 250, resulting in +25%. It’s the same situation, but the results vary significantly depending on which ones you keep in the calculation.

I set C to 0 just as an example. I'm not saying we should automatically count delisted coins as 0. To compare properly, I think we need to set the list of projects and their weights at the start, and track delistings or mergers separately. We should also mark instances where price data is unavailable, because if you just remove them, the returns might look better than they actually were.