cross-posted from: https://lemmy.world/post/49878453

Solo project I’ve been building this month, wanted to share here. The rules:

You get $10,000 and two secret years of real S&P 500 history, drawn from 1928 to 2019 (daily data, dividends included; your cash earns T-bill rates while you hide). One button: BUY or SELL. Your opponent is a couch. It buys at the start and never moves.

Every finished game is audited by 1,000 monkeys. It’s a permutation test 1,000 replays with your same time in the market and same number of trades, on random days. Beat 90% of the monkeys and your win is stamped TIMED THE MARKET. Anything less is stamped LUCKY, in red ink, on your settlement document.

About 84,000 games have been played so far. The couch has won 63% of them. Roughly 1 game in 9 ends in a win that survives the monkey audit.

There’s also a daily hand, same secret window for everyone, streaks, and a public ledger of humanity vs the couch.

If you have time, I’d love feedback, and I will answer any methodology questions, thanks!

  • Caveman@lemmy.world
    link
    fedilink
    English
    arrow-up
    1
    ·
    6 days ago

    I would like a slider mode. It’s very hard to do any sensible strategy with an all or nothing approach. If market goes up by 30% I want to progressively sell and vice versa

    • noir_chat@lemmy.worldOP
      link
      fedilink
      English
      arrow-up
      2
      ·
      6 days ago

      it’s over two years so while not long term, it’s not short either. lots of ups and downs. but i’m thinking of adding a career mode with 2x10 years, start with 10k, see where you end up kind of thing.

      • Caveman@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        5 days ago

        That one would be almost impossible to beat unless you recognise the the boom/bust cycle. It sounds interesting

        • noir_chat@lemmy.worldOP
          link
          fedilink
          English
          arrow-up
          2
          ·
          5 days ago

          yeah, sometimes i think i know the cycle and still get it wrong. too similar and not too similar at the same time. glad you liked it mate