Positive but low accuracy

[sawtooth500 33]

So both yesterday and today I had positive days, but yesterday only 33% of my trades were green and 30% of my trades were green today. I feel like I have a tendency to take "moonshot" trades, stuff that doesn't work out a lot and I get stopped out, but when it does work, I make bank.... 

What are your opinions on it? I mean yeah I'm positive, but I'm really concerned that I'm positive more because of luck than what I'm doing being a viable strategy. In the past, I've gone down this road and it always seemed to lead to a day of big losses....

Ultimately, what's more important - having a good accuracy, or a positive P&L?

[KyleK29 260]

Average R-return of your trades is way more important. If you risk $100 and return $150, that was 1.5R. If you do this for 10 trades, you are averaging 1.5R per risk. A more important metric is to then calculate a Return Performance, which takes into account different Risks per trade (e.g you may have risked $100 on one trade, and $50 on another trade, the Risk-unit is the not same).

As a note, accuracy by itself is a very misleading metric. For example, how do you define a "win"? How do you define a loss? It's really only a relevant metric if your trading strategy is to get out 100% at -1R or +1R. If you do it as most people will start with -- win is trade PNL >= $0 and loss is trade PNL < $0 you end up with a few problems.

1) B/E trades will skew your numbers depending on how they fall (did B/E end up -$0.20 or did B/E end up +$1.01, etc.).

2) You can end up with losses being mainly full 1R stop outs and wins being on-average less than 1R. If you had 50% distribution you would look at it say "well, I should be even PnL wise" .. but you aren't. You're most likely in the hole, a lot.

 

This page somewhat explains how you can calculate them into each other to get a more accurate metric. I have a Sheets calculator somewhere that will take in your average values and return the required win% needed to be profitable, but I have to find it. https://wiki.timetotrade.com/Risk_Reward_Win_Loss_Ratios#:~:text=Reward to Risk Ratio%3A The,50% of the Average Risk.

Another explainer: https://www.tradeciety.com/how-to-use-reward-risk-ratio-guide/

 

 

 

 

 

Edited July 9, 2020 by KyleK29