How are evaluated the strategy?
We attribute a note between 0 and 10 to strategies shared in the store. The closest it is to 10, the most the strategy is worth to look at.
This article explain how these notes are computed.
When a strategy is shared, we backtest it on a bull, bear and range market.
For each backtest we have
- the number of trades
- the net profit in the quote asset (asset quote = EUR on BTC/EUR)
- the pending profit in the quote asset
- the average gain per trade in the quote asset
- the win rate, i.e. the ratio between profitable trades and the losing one
- the performance against the market
From this data, we get the note by doing:
+1 if the win rate is above 50%
+1 if the average gain is positive
+1 if the net profit is above 0
+1 if the pending profit is above 0
+1 if the net profit + the pending profit is above 0
+1 if the net profit is superior to the pending profit
+1 if there is more than 3 trades over
+1 if the max drawdown is less than 10% of the initial volume
+1 if the performance against the market is above 0
+1 if the performance against the market is above 100 %
If the strategy isn't evaluate, we display a "?".
Backtest are done on 10k candles. You should know the backtest duration significantly change according to the strategy unit time (UT). If the UT is 1m, then the backtest is done over a week. If the UT is 1h, then the backtest is done over one and half year.
To choose the markets bull/bear/range, we have classified all available markets. You can read more about our classification here: https://blog.botcrypto.io/fr/2020/10/28/classement-des-strategies/.
The note is definitive.
This note is not pertinent for semi-autonomous strategies (see label manual) that require a manual order from the user during the strategy execution.
This note help you to sort and find quickly valuable strategies.
This note doesn't give any warranty! It reflect the strategy performance on a market that wasn't choose randomly.
Check the article about labels to learn more about the different type of strategies in the store: https://botcrypto.crisp.help/en/article/labels-16sch6h/?1605792165941.
This article explain how these notes are computed.
Strategy evaluation
When a strategy is shared, we backtest it on a bull, bear and range market.
For each backtest we have
- the number of trades
- the net profit in the quote asset (asset quote = EUR on BTC/EUR)
- the pending profit in the quote asset
- the average gain per trade in the quote asset
- the win rate, i.e. the ratio between profitable trades and the losing one
- the performance against the market
From this data, we get the note by doing:
+1 if the win rate is above 50%
+1 if the average gain is positive
+1 if the net profit is above 0
+1 if the pending profit is above 0
+1 if the net profit + the pending profit is above 0
+1 if the net profit is superior to the pending profit
+1 if there is more than 3 trades over
+1 if the max drawdown is less than 10% of the initial volume
+1 if the performance against the market is above 0
+1 if the performance against the market is above 100 %
If the strategy isn't evaluate, we display a "?".
Testing conditions and disclaimer
Backtest are done on 10k candles. You should know the backtest duration significantly change according to the strategy unit time (UT). If the UT is 1m, then the backtest is done over a week. If the UT is 1h, then the backtest is done over one and half year.
To choose the markets bull/bear/range, we have classified all available markets. You can read more about our classification here: https://blog.botcrypto.io/fr/2020/10/28/classement-des-strategies/.
The note is definitive.
This note is not pertinent for semi-autonomous strategies (see label manual) that require a manual order from the user during the strategy execution.
This note help you to sort and find quickly valuable strategies.
This note doesn't give any warranty! It reflect the strategy performance on a market that wasn't choose randomly.
Check the article about labels to learn more about the different type of strategies in the store: https://botcrypto.crisp.help/en/article/labels-16sch6h/?1605792165941.
Updated on: 23/11/2020