[SHARE] Simple RSI BULL/BEAR strategy

[whoodat]

  Screenshot from 2018-05-12 11-30-27.png (Size: 28.17 KB / Downloads: 213)
  Screenshot from 2018-05-12 11-30-27.png (Size: 28.17 KB / Downloads: 213)
  Screenshot from 2018-05-12 11-30-27.png (Size: 28.17 KB / Downloads: 213)

@Ryancoin,

You have to match exactly what was posted, the from/to date down to the second, the candle size, the warm up time, and all the paper trader settings.

@Henry151,

I briefly looked into this. Looks like some issue with Gekko handling the response from the Binance API. I have to dive deeper to know but it is likely because the API is returning duplicate results.

@whoodat,

Use TradingView to manually scan through whether the strategy you are using should have made a trade or not. I'm assuming you set a candle size that is available on Trading View or you have the pro account and you can adjust to whatever candle size you want.

I set the candle size to 15 minutes. In 11 hours it made no trades something must be wrong.

[crypto49er]

@whoodat

For me, even with 5 minute candles, the warm up time is 20 hours. At 15 minutes, you probably have to wait days. You might want to tun this in cli do it doesn't have to warm up.

[whoodat]

@whoodat

For me, even with 5 minute candles, the warm up time is 20 hours. At 15 minutes, you probably have to wait days. You might want to tun this in cli do it doesn't have to warm up.

So it won't make the first trade until it has reached it's warm up period....so if I set the warmup to 1000 to match what the SMA, it wouldn't make the first trade for over 1month?

[whoodat]

@whoodat

For me, even with 5 minute candles, the warm up time is 20 hours. At 15 minutes, you probably have to wait days. You might want to tun this in cli do it doesn't have to warm up.

So it won't make the first trade until it has reached it's warm up period....so if I set the warmup to 1000 to match what the SMA, it wouldn't make the first trade for over 1month?

How does running it in the CLI take away the warm up period? Also, the longer the "warm up" period does this mean the "smarter" it will trade do to the amount of time it has had to review old data?

[crypto49er]

@whoodat

For me, even with 5 minute candles, the warm up time is 20 hours. At 15 minutes, you probably have to wait days. You might want to tun this in cli do it doesn't have to warm up.

So it won't make the first trade until it has reached it's warm up period....so if I set the warmup to 1000 to match what the SMA, it wouldn't make the first trade for over 1month?

How does running it in the CLI take away the warm up period? Also, the longer the "warm up" period does this mean the "smarter" it will trade do to the amount of time it has had to review old data?

Running in CLI will let it use the data that you already downloaded. At least that's what I read from here. I personally haven't tried it myself. As for the bot trading smarter, well, per tommiehansen, if you don't let it wait, it will just assume bear trend until it has enough data. So imo, it won't run efficiently but could be ok for the next day or two since we are in a mini bear trend right now.

[whoodat]

@whoodat

For me, even with 5 minute candles, the warm up time is 20 hours. At 15 minutes, you probably have to wait days. You might want to tun this in cli do it doesn't have to warm up.

So it won't make the first trade until it has reached it's warm up period....so if I set the warmup to 1000 to match what the SMA, it wouldn't make the first trade for over 1month?

How does running it in the CLI take away the warm up period? Also, the longer the "warm up" period does this mean the "smarter" it will trade do to the amount of time it has had to review old data?

Running in CLI will let it use the data that you already downloaded. At least that's what I read from here. I personally haven't tried it myself. As for the bot trading smarter, well, per tommiehansen, if you don't let it wait, it will just assume bear trend until it has enough data. So imo, it won't run efficiently but could be ok for the next day or two since we are in a mini bear trend right now.

I see, also is there a way to specific how much of the balance I want to trade. Say I have 10g in the account but want to only try 1g of it. Do I need to remove 9g of it before testing or is there a way to specify how much to test it with? Also, am I reading that correctly, where if I put in the warm up period to match the SMA long and it is 1000, then the warm up period will really be an entire month?

[crypto49er]

You have to remove the 9g, no way to specify trade amounts on gekko. I suggest using a shorter sma long than 1000. Most backtests show you don't need that much. At 15 minute candles and 1000 sma long, you should only need to wait 11 days. I say only as if that's short. I know it's a long time. Definitely look to using smaller candles and a shorter sma long.

[whoodat]

You have to remove the 9g, no way to specify trade amounts on gekko. I suggest using a shorter sma long than 1000. Most backtests show you don't need that much. At 15 minute candles and 1000 sma long, you should only need to wait 11 days. I say only as if that's short. I know it's a long time. Definitely look to using smaller candles and a shorter sma long.

What do you mean only as if that's short?

[crypto49er]

being sarcastic. That's a very long time. I'm barely okay with having to wait 26 hours. I'm probably going to switch to CLI.

[whoodat]

being sarcastic. That's a very long time. I'm barely okay with having to wait 26 hours. I'm probably going to switch to CLI.

What I'm trying to understand is, if you give it a shorter warm up period without much data how reliable really is the strategy? What size candles have worked bet for you?