Hey guys, its strange but did you notice people on biz rarely talk about trading bots?
Some say that its because if you have a good strategy and you code it on a working bot you use it for your personal profit and never share it, well that's understandable, in fact I've made some bots with simple strategies myself, and they work well.
But what about the public bots I mean those that are for sale. Are they any good? I mean this market is full of people that cannot afford them so if they are any good they may produce lots of profits. Do people avoid talking about them because they fear they will be used by the masses and no be longer profitable, or just because they are expensive bullshit?
If you go on biz or on YouTube you see tons of people shilling shitcoins or making up predictions, but never an honest review about that amazing trading bot that earns a lot of profit. It's because they are shit or because they are to good?
I am inclined to think they may be overpriced shit, because if not their creators would't be selling them in the first place if they were very good. What do yo think?
>He's asking too many questions >I need this thread terminated immediately
David Johnson
I've written my own, setting it up is not too hard, making them trade in a way that makes consistent profit is what I still need to figure out.
Jacob Ward
bot's cant predict the market. no one can
Austin Young
It's still too hard to make them work consistently, when btc was going up consistently late last year it was best because they bought alts and sold back to btc to increase your stack in satoshis. But now bots lose more money than manual trading because the market is so wacky. Another thing you have to look out for is so many bots are scams, "Gunbot" being the biggest one that is heavily shilled on bitcointalk but so many people say isn't even a working program.
Adrian Williams
LMAO thanks I almost fell for the gunbot meme for the constant shilling on bitcointalk. What about Haasbot is a scam too or it's legit?
I'm just not that sure about this. You don't need to predict it completely, just make a good number of decent enough bets.
Bentley Green
I use profit trailer / profit feeder. Going on a week now, about 1-3% profit per day. If you make $100 per day with a bot thats $36,500per year.
I like knowing its running and watching it work while I am at my job.
Wyatt Roberts
Because the more people using bots the less of an edge we have against you. now fuck off nobotter.
Ryder Wright
That's what is so hard about them
Ever wonder why biz is just people posting pink wojaks and betting on monopoly money and never about entrepreneurship?
Landon Howard
You either have to write it yourself or have a lot of capital and connections. They work, but they don't just mint millions every day. They might give you 50-150% advantage over the market. In a bear market you would still be lucky to break even.
Joseph Barnes
Is this legit?
Benjamin Harris
Bots only work if you have considerable sway over the market or if you're doing illegal things. They're not useful to win money.
Easton Watson
Are they using function approximation on some intervals to maybe "guess" what will happen?
I spent the last year writing my own bot that can execute any strategy with IB's API.
That was the easy part. The hard part is finding a good and consistent strategy. As someone with little financial knowledge it was quite hard for me. I eventually threw all my TA and a few other parameters into a neural net and that has been working well so far.
Liam Wood
>buying bots You total dumb fuck No one sells a bot
anyone with classical experience is making a killing in these exchanges
95% of the vol on GDAX is botted. badly.
Botting the chop on there is easy rn since the botters have no classic algotrading background
still full of kiddies
Oliver Gray
Any difference in performance in the bear market?
Oliver Ross
Can you explain more about the neural net?
I've got a number of bots running but always wanting to learn about more signals I can include.
Juan Kelly
My data set was the S&P 500 for 2016-now. It is a day trading bot and does well on up and down days.
My neural net takes in 38 parameters. It has been my experience the more parameters the better. I have tried most Neural Net libraries and even wrote one myself. The best I have come across is Infer.net.
Brandon Parker
How much do you use for training data? All you can find, a year, a month?
Jonathan Jenkins
have you tried pytorch before? I'm writing one that uses this pytorch.org/docs/master/nn.html#lstm and takes the last N trades and prices plus the average price of the next N trades and prices as input
Daniel Carter
I'm curious, on a bull run obviously it's easier to make gains. What about this bear market? I assume your get stoplossed out. For the parabolic dips do you just shut down after a certain amount of losses? Can the bots be programmed to buy a reversal?
What parameters are used, are they right equivalent to typical TA?
Thomas Brown
Not used pytorch. Looks like a decent Lib though.
Jordan Roberts
I bought that bot some time ago. It made like 1% sometimes, and other times it just entered in a money burning mode when it bough high and sold low and if I didnt stop it it on time would made me broke. Never trust java software made by an anonymous dev in a forum. At that time it was sold for 0.5 BTC no refunds, only resell allowed. What a scam.
Jaxson Wood
I've put together some bots which execute on hand coded strategies and have the same question as .
What's the idea behind incorporating a neural net into a trading bot? Do you just a combination of indicators with dynamic parameters and train the bot with historical chart data and let the bot figure out what parameters are best?
Samuel Gomez
>the bot bought high and sold low >the bot bought high and sold low >the bot bought high and sold low >the bot bought high and sold low
