Botfags General

what kind of gains are you making with it user? how long did it take to write and do you have a programming background?

Uh well, I have built a few bots so far in a mishmash of C++ and PHP. They all hang out in a slack clone I run at home. Here's the current roster:

Marco
>downloads market data from exchanges (tickers, chart data, etc)
>shovels everything into a local SQL database for fast querying

Luna
>analyzes market data continuously looking for specific patterns that indicate good-time-to-buy
>90% success rate at picking coins that make 2% profit within 30 minutes

Alice
>takes tips from Luna and turns them into exchange orders by figuring out a good bidding price, how much to buy etc.
>also implements stop loss if things don't turn out the way we hoped
>takes 10% tax on all profits and spends it on BitBeans. these are immediately sent to my bean farm for sprouting

Bob
>talks to exchanges, posts trades and maintains a local copy of all my current account info / balances etc
>hopelessly in love with Alice

Minty
>monitors my PoS coin wallets for stake rewards and announces them in chat

Wally
>monitors my "long term savings" BTC wallet and announces in chat when a deposit is made

Cute.

I would like to write a bot that tries to maximize a coin by riding its waves using technical analysis to determine buy and sell points.

I have a lot of ETH sitting on an exchange, and I'm getting tired of hodling it. If the price won't go up anymore, I might as well trade to make the stack grow. There's more than one way to increase value.

Gunbot looks interesting

take all my monies, can you clone this setup for me? how much?

Currently a steady 3% per day, my bot is parametrized to sell 1/2 of the buy amount after 3% and 1/2 after 5% profit, if the price falls 2% below the buy price it cuts losses and sells.

It took a while before I had the coinigy API because I had to implement all exchanges manually but it has gotten better since then.

Definitely some complexity involved, I am a professional software engineer and I do have to sit and think about it.

Currently it's 100% Python, I am thinking of switching to something more adapted to async like JS but not sure

Relatively close setup to this one except all in Python and I didn't give them cute names

Interesting.

Ive been trying to figure out a good % to sell at. I might copy this sell point.

My buy point was just on downturns but I probably need an actual formula for that instead of just the past % changes.

All these parameters are hard to figure out, it would be worth simulating to optimize them but the complexity becomes very hard to handle with a single developer.

Some values I found empirically:
- Up to 5 open orders at a time, 0.1 BTC each
- Focus on coins with a volume >240 BTC/24h
- Buy when there has been 5+ consecutive trades with increasing price and the ratio between open buy orders and open sell orders is >0.25 (I need to add rules relative to the order book, i.e avoid coins with buy walls and sell walls too close to the current price)
- Orders timeout after 2h

I anybody here has figured out something that works well I would be happy to hear about it

Sorry user, my bots are very personal. :)

If it appeals to you, I urge you to build something similar for yourself. And if you don't have the skills yet, start working on acquiring them right now. I promise it'll be worth it.

After it detects the pump isn't it too late already to put money in because it's probably on the way down when you get the signal?

No, it notices pretty early.

Some pumps have too much velocity, and there's no way to catch those. However, there are plenty of more gradual increases that can be caught pretty early. These slow rollers are nice and safe and easy to catch.

From today, SYS is a good example of a wave I can easily ride. OK from the past few days is another one.

That's really wholesome. How's sticking 10% of your gains into bitbeans going? Worth it?

I like that you have your family in mind. I hope they are grateful.

Anyone else an arbitrage fag here?

Been writing and testing some scripts for over a month in Python and tests show positive results, but when I actually try to execute them I always miss the opportunity because it comes and goes so fast. Tried doing arbitrage on one exchange for different pairs so maybe that's why. Going to switch to an arbitrage strategy between different exchanges but the fees and waiting times are really shitty.

A different script I wrote monitored pump n dump tier coins on shitty exchanges and plotted them all on Excel (like 300 coins) and the cells would be colour coded if certain conditions were met. I never pursued that any more though because I wanted to put my energy into arby. Might try it again though

Arbitrage seems to be a good way to make systematic 2% gains, so yes it is to pursue.
Timing is key though and probably hard to get right

lots of brainlets in this thread

Sure, it's working out pretty well! Here's a screen shot of my farm monitor right now. ("Plant" == "Deposit" obviously.)

My bots plant beans a couple of times per day (~900 beans each time, depending on market price.) I also get a couple of sprouts per day (~1000 beans each time.)

I seeded it with about 1 million beans that I got around ~50 sats a while back, so things are looking pretty good. :)

if it's genetic then how can it be learned?

Read up on jury nullification.

i just write bots in the +2100 member bitpam cryptocurrency trading/mining/coding discord server desu, would be cool if all the anons worked together here and made something awesome, we're doing something right now

discord.gg/ddFaYDN

Why are you shilling this so hard? What do you gain?

Probably trying to get people who know what they are talking about in a children's discord to benefit from it, I am wondering as well

I get to talk with more people and escape from the life where my father hits and yells and me and my brother all day so I have a place where I feel like home. I want to build a cool cryptocurrency community where we can all make it and have fun.

Make others happy and you will be happy is the way of life i follow.

idk man its just feels right to build a cool community

I like your naming scheme. I don't trade coins instead focusing on stocks. I use a ml algo I wrote to do the pattern finding, the results are then output for me to manually decide which stocks to buy and sell for the day. Unfortunately I wrote the data gathering part around the Yahoo icharts which have been discontinued with no warning, so I haven't tried it yet. Instead of bots mine is a series of programs.

Scout
>Gathers the historical data, currently not functioning because of fucking Verizon buying Yahoo and dropping icharts

Farmer
>Builds the ml network, these can reach tens of gigs in size depending on how many inputs I give it, normalized history, roi, moving averages, and whatever else I feel like. This is a training program, it is not used in day to day trades, only when doing training on bulk historical data.

Researcher
>Finds the best set of parameters for the current NT3 network to get the best accuracy.

Worker
>Runs through current data to generate predictions.

Colony
>Converts the raw predictions into buy/sell signals and sorts based on predicted movement.

Stellar Converter
>Accepts orders and watches the market to manage risk so you don't have to monitor it constantly. Orders are input by hand because news articles are not taken into account during the prediction phase yet.

Which slack clone? Mattermost? rocket.chat?

Aren't some of those bots distributed with source code?

Mattermost!

Let's turn that into a product and sell a monthly sub fee! Email me at [email protected], let's do this

Forgot to mention - I wrote some software that pumps out realtime candlestick data. May be more prrformant than querying the API since it leverages the exchanges websocket streams

Do you hold them or sell periodically?

nothing can be learned it can only be copied and enhanced nobody truly understands what they are doing

Hello
If anyone is interested I would love to be able to purchase a working and profitable over a few Weeks bot. I won't be able to learn to program I'm not smart enough but at least I admit it. I have a lot of capital to work with so this could be extremely profitable if I could just find someone to work with. My email is [email protected] I'm in Boston. Would also do a joint venture so we both make good monthly profit and willing to split it with you for your hard work on making us both money.