You don't have to share all your secrets, but can any anons share their experiences with bot trading?

>8-10% gains daily

(1.1)^365 = approximately 1.3 quadrillion dollars

I think Zenbot is open source, not endorsing anything, I have never used it so it could be a scam for all I know.

if a bot is public it won't make you money

Not python, but I guess I could try using it.

see i don't even know where to look for information since i'm pretty sure any bot advice is just the competition trying to undercut me.

>I think theres a clue somewhere in there

Just get the bot to make the same kind of decisions you'd make. But faster.

Depends on what you're using it for.
Arbitrage is kind of hard to fuck up.

Enough larping. I am writing a framework, phd student in statistical machine learning. It's currently doing about ~1-2% in a 8-12 session. The main thing you will learn is that there is just extremely many events that can ruin a session, API errors on the exchanges, random market turns.

The problem is not so much the signal often but the immaturity of the infrastructure, so most of what you do is dealing with special cases on ultra volatile events, api failures, backend lags, etc.

followup: The thing you have to do is to treat this just like any other experiment in statistics/machine learning. I dont think about it as money, it's just an optimisation experiment with a lot of constraints.

I repeat. if a bot is public it won't make you money.

>machine learning

how is machine learning meme relevant here?

what about chatbots? anyone have experience using one in tandem with your trade bot?

Markets, on a short enough time frame, can be modelled as random processes with drift. It really comes down to thinking about it in the sense of information theory/entropy. What kind of information can there be about a certain time frame? E.g. longer time frame macro events, news, general trend, ultra short term time frame (few seconds) purely random with drift created longer term correlations. ML is just applying these modeling techniques.

i know targets like 3% profit or a stop-loss at 5% are a good starting point, i just have no idea when to enter trades.

t. year-long holder who wants to trade

wonder if there is a good bot on the darkweb that would take a percentage of the profit it made you.. I'd be down for that

>hurrrrrrrrrrrr joshpeeeeee

aren't some coins more stable trading than others?

identifying these is part of the optimisation problem but that is relatively basic statistics (analysis of variance regarding volume, price change, news). My last answer, have to get to experiments and as they say, dont do anything you are good at for free

yeah i didn't expect to learn the real valuable stuff on an ethiopian sticker exchange. i appreciate the insight though. i guess i should do some manual trading for a while before i take the bot plunge.

I have a trading strategy in my mind and would like to write a bot to do it. I don't have much programming experience, though. Is there an existing bot framework that I can use to start from?

Final answer: I don't really know. You can obviously google frameworks, but I wrote mine from scratch because I dont believe you will get good results unless you understand every part of the design, why things are done how they are done - especially regarding order execution, how certain events are handled, etc.

exactly. anything that comes close to resembling a trading algo is suspect to me. i think i'll just look into trade entry points and do it the old fashioned way for a while.

I use this bot for my indicators but I trade manually.

satoshibox.com/z8hggvejve7eppf8ymjzy4ik

i've seen this shilled on here so much, and the image names are always the exact same format. desu i will never visit that link, but i'm sure some retards will.

There are some youtube vids that can explain you how to do it, I mostly followed the steps from rweBollo and then tinkered with things. But my strategies are fucking shit apparently.

its satoshibox you newfag, it might be a scam but it won't give you aids when you click the link

see:

Just learn python. It is not that hard.

If you cannot program you probably have no experience in making trading strategies either - so prepare for it being bullshit.

Everyone thinks writing a bot is as easy as
>if price_up:
> buy
>else:
sell

Have fun.

I wasn't referring to the bot itself, I meant the website is safe to visit. But I see now that you weren't talking about it like that.

>if price_up:
> buy
>else:
> sell

For 1m-5m intervals it pretty much boils down to this.

i'm sure the website is safe, but paying to download a mystery-box .zip is probably the last thing i will ever do with my BTC

I learned programming in C and matlab, also learned some phyton a few years ago. I think I could implement my strategy pretty well. My problem isn't that I don't know how programming works, but I'm struggling with how to start.

I'm currently reading docs.bitfinex.com/v2/docs, I think I will take a look into one of these open source libraries, for example github.com/Crypto-toolbox/btfxwss. There's already my first question: What's the difference between websocket and REST protocols? Which one would you rather use?

Rest is http request.
Websocket is bidirectional communication.

>arbitrage is kind of hard to fuck up

lol... says the guy who has only arb'd with lunch money.