Does winning probability of a day trading system resets every day?

Discussion in 'Trading' started by bpr, Feb 3, 2020.

  1. minmike

    minmike

    100% win rate is a gigantic red flag. Either it means you are doubling down/ martingaling. Something that works in a short trading contest but not over the long term.
    Or, in very rare circumstances, you should reduce parameters because you are missing very good trades that might only have a 75% win rate.
     
    #11     Feb 3, 2020
    comagnum and murray t turtle like this.
  2. Sekiyo

    Sekiyo

    Yes it's a red flag.
    Even if tiny is the risk of ruin, it does add over time and ultimately will materialize.
    If you play the Russian roulette 100 times, you're sure to hit the bullet.

    However I'd take 100% over 75% any day because there's no Drawdown.
     
    #12     Feb 3, 2020
    murray t turtle likes this.
  3. %%
    That, mm;
    + 100% win rate is not really believable for long really. Even for an investor, could have had a 100% win rate if you scaled in QQQ about $30, possible; few investors would want to stop investing @ only $30, on QQQ trends/QQQ drawdowns/ QQQ profits .
    BY the way even in that old 3 month trade contest; market were very strong trends /not random. AND he never traded with his profits- that's one reason he won;in other words , 100k into 900k,[+ win bonus] with fixed size.When he won 100k on 200k, same contest ; he only traded with $ 100k, fixed size, not the 2ook.
     
    #13     Feb 3, 2020
    comagnum likes this.
  4. comagnum

    comagnum

    Any perceived probability you think you have will have variance, every so often you will have a string of consecutive losers much larger than you expected.

    With win rate of 50% over 1,000 trades there is a chance of having one event with 13 consecutive losers. This shows the importance of sizing down during losing periods.

    upload_2020-2-3_13-34-39.png
     
    #14     Feb 3, 2020
    murray t turtle likes this.
  5. %%
    That;
    + a 93% hit rate, can still be wrong 7 months in a row. NOT that it will play out that way much. But for those experienced have seen patterns like that,[if you moniter monthly charts] wrong 7 months in a row, even if its a good money maker, over the years.
    Comangnum;
    Or another hi% pattern about 88% right, hard to lose money with 88%, right??
    Stock Traders Almanac perhaps first noted the positive JAN = positive year/ 52 weeks pattern. .Great 88% pattern, sometimes better; but when it does fail like 1987, 2009, it fails real BIG// LOL...................................................................................................................
     
    #15     Feb 3, 2020
    Sekiyo likes this.
  6. Sekiyo

    Sekiyo

    upload_2020-2-4_5-13-38.png

    It happens for 1 in xyz :
    upload_2020-2-4_5-22-55.png
     
    Last edited: Feb 3, 2020
    #16     Feb 3, 2020
  7. bpr

    bpr

    This makes sense mathematically so basically every trade is independent so probability resets at every trade

    It is strange though… to arrive at win percentage of a system we do back test group of trades in diff periods and take average...but we expect all individual trades to have that probability independently
     
    #17     Feb 5, 2020
  8. bpr

    bpr

    Lets not divert from the topic.
    Nobody is talking about 100% win rate.

    Lets consider a deviation. Lets say one day you have 80% win vs the system average 60%
    Does next day trade proabilty suffer ?
     
    #18     Feb 5, 2020
  9. bpr

    bpr

    Agree the win probability of any system will vary lets say plus minus 5% or 10%
    Let’s say some day u get big deviation …so the next day/days should revert or not ??

    Consecutive loosers is another topic ...
     
    #19     Feb 5, 2020
  10. minmike

    minmike

    Did you read the thread? Murray did talk about 100% win rate.
    You have had some very intelligent responses here. There is no universal truth here. Backtest on you system and see for yourself if it does.
     
    #20     Feb 5, 2020