🏅 Top traders

Now that trading has closed, I will run the script to calculate the all time profit/loss of all the accounts on manifold. I verified the script does something sensible. Thanks again to chris.

@zzlk I ran the script at more or less exactly midnight 2023-09-06 00:00 PDT

There are 40672 total users
  Of these, 29375 (72.2%) have nonzero profit.
    Of these, 13036 (44.4%) are in positive profit

So, most users do in fact lose mana from trading.

Here's a Python script that I think will answer this question, assuming that you are happy with treating a user's profit not being exactly equal to zero as a reliable enough indicator that they have placed at least one bet.

I can imagine it's possible that users that have placed a single bet and immediately reversed it might have exactly zero profit, but even then possibly not - since mana balance is a floating point number, floating point rounding may mean even reversing a trade may not result in exactly zero balance. So I think profit exactly equal to zero is probably a pretty good indicator of not having made a bet.

Downloading data on all bets to check for sure whether a user has made a bet is a bit more intensive, but could probably be done if necessary.

https://gist.github.com/chrisjbillington/f86f77499493680002cdf125db9bc0c7

user_profit.py

GitHub Gist: instantly share code, notes, and snippets.

@chrisjbillington Thanks for posting this, I was going to write something in rust but this is much easier. Once the market closes I'll run it just to be sure then resolve.

Trading is zero sum, right? If I am making mana from trading someone else is losing it?

@KevinBurke There's market subsidies for every new trader on a market which come from Manifold and also an initial subsidy from the market creator when a market is made, so it's positive sum.

This is very much yes, but closer than people might guess. A quick run over all current users gives:

There are 39749 total users
  Of these, 27278 (68.6%) have made at least one bet
    Of these, 12193 (44.7%) are in positive all-time profit

Edit: the above was checking user['lastBetTime'] to determine whether a user had made a bet. Confusingly, some users have positive all-time profit despite lastBetTime being absent from their info. If you instead treat nonzero all-time profit as indicating someone has traded (is there any other way one's profit can be nonzero, other than trading?), then you get a slightly different number:

There are 39749 total users
  Of these, 28851 (72.6%) have profit not exactly equal to zero.
    Of these, 12927 (44.8%) are in positive profit

@chrisjbillington I guess that boosts claimed count towards profit.

@AnT I'm pretty sure they don't

Money or mana?

@Arinbjorn Mana, edit: updated the title to make that clearer