Topic: New bitcoin trading bot: looking for feedback (Read 3511 times)

If you are interested in a collab for a trading language, let me know. I'm working on java stuff, though. But the trading language should not be affected by this.

Could someone explain how this would this work? I've seen talks of it a while ago, but people seem to be turned down by the speed of withdraws from exchanges. I know that you'd have to have money in both (or multiple) exchanges, but the orders are not necessarily instantly processed so wouldn't you'd have to project the price at an exchange from a few tenths of a second to a few seconds ahead of the time of the buy to compensate for the trading lag?

The market will provide you loud and clear feedback.

I am hoping to move to an open beta within the month. The alpha-testing group has not had too many issues so far. I have not even set up a static page anywhere for the site.

1.  arbitrage between different currency types
2.  maybe a hosted option
3.  store a history of each trade for review
4.  semi-intelligent trading (don't make a trade if its going to lose money)
5.  would be GREAT if it had really easy interface with detailed explanation of what the affect of adjusting each configuration will do...  not just a link to bitcointalk.org thread with 100 pages of discussion on the bot...

@Stephan: OK, I'll PM you.

@minzie: That looks interesting Any idea when it will be open for public?

I have been working on exactly that; a simple bot, with a simple interface, that can run on any exchange, with any currency. It is currently in closed alpha testing, and I am working with a graphic artist to put a nice face on the web content.

https://bitcointalksearch.org/topic/lazy-i-bitcoin-trading-bot-alpha-testing-271825

Features:
Testing Mode - the bot uses a test ledger to conduct trades. The balance can be reset at any time.
Spread Calculator - Automatically shows projected fees based on configuration options.
Speed Sim - Back test simulator to test trading strategy against an entire body of price points.
24/7 operation - Bots run unless instructed to not run.
Transaction Log - A record of all bot transactions.
Multiple Bots - When complete, operators will be able to use multiple bots, on multiple accounts, on multiple exchanges, using multiple currencies.

Active, single bot proof of concept running online. (Loads slow....just a proof of concept):
http://www.minzie.com/Lazy/


@AscotV, the second link contains most of the code for the proof of concept bot, although it has not been updated since June. Also, the language is irrelevant, in spite of what anyone says, and would be dictated by your intended purpose. I also agree that most bots that I have seen are far too complex to be approachable by new operators.

I have to disagree. It is almost exactly as easy as it looks, although it can take a tremendous amount of time. Buy low and sell high is a pretty simple algorithm that grade school children understand. Everything else is built around determining what 'low' is, and what 'high' is in order to trigger a transaction. The proof is in the fact that I have one running, and it works. If I can do it, than anyone can.

About which trade bot are you talking here? Is it by accident my trade bot? If it is my trade bot you could contact me by PM to talk. I am open for any suggestion/improvement.

If you decide to make (and sell) your own trade bot i really recommend to talk to me too. I am around sinds May 2013 and i was the first here to break the scams of the trade bots here. I can help you get started, show you the problems which i have had and so on. Things are not that easy as it looks, trust me.

It seems to me like people trust web services that are closed source and opaque far more than client applications, even if the web applications get more "dangerous" data inserted, so my tip would be to work on a 24/7 hosted bot platform that allows people to sign up, input their API key(s) and select a strategy. It would also eliminate the need to keep a PC running at night which is something that few people of your target audience do probably.

Thanks for your response.


Well, because I was looking for a trading bot like the one I'm planning to make now, and didn't found one. If somebody can point me to such a trade bot (remind my target audiance: no experts) and I think it's sufficient, I won't reinvent the wheel. But for now, I agree there are many "wheels" already, but no one that really meets the needs of my intended users.


I have no complaints about the trading strategies of the bot I bought. However, I think I can make it more user friendly, stable, and add some extra features.

Ofcourse another possibility is to ask the developer of the bot I bought, if he needs help... (It's closed source, so I cannot just improve the bot myself). But if the answer is no, the only option left is to make my own bot... No?

There are multiple open source bots available, why should anyone use your closed source one?

Also bots are only as good as their strategy... and of course their connection to the API/exchange. I personally would recommend to take a look at existing bots and trading libraries already out there before you start (again) to implement the Bitstamp API from scratch. If you think you ca do better than the bot you bought, do you mean that you can have a nicer interface and/or more stable program running or do you mean that you have ideas for better trading strategies?

Hi all,

I am working on a simple bitcoin trading bot (verly early stage). For now it will only support Bitstamp, but I'll add other exchanges later.
The goal is to make a trading bot for people that or no trading expert, nor a cryptocurrency expert. I bought already a trading bot, but I think I can do better. I would like some feedback from the community on some questions I have. Or anything that can help me, or make the bot better is welcome.

First of all: Am I wasting my time? If so, please tell me now

Platform: I'm developing the bot using C#. I know many of you would prefer Java, so it would be cross-platform. I know Java, and made already some projects in Java a few years ago, but at work I'm developing in C# now and I've got fond of it, and would prefer not having to switch between languages every few hours (at work C# and at home Java). Also, because of the target audiance, I think the fact it will only run on Windows won't be that big of a deal.

Open source/closed source: All of you would strongly prefer open source. First of all so everybody can check the code for any password stealers or whatever. Secondly because it fits with the idea of Bitcoin. I understand these arguments. But if I'm spending my free time on this project I would really like something in return. I don't have any experience with donations, but I think, certainly for my target audiance, this would be very little. I don't expect to make a living out of it, or even earn a normal hour rate for my work. I don't do it only for the money. But I would be dissappointed if I spend most of my free time on it for several months and then get like 0.5 BTC in return (in total). I have also some other Bitcoin related software ideas, and if I get something in return, this would stronly motivate me to realise these other projects too. So I would prefer to make it closed source and sell it for something like 0.5 to 1 BTC. Unless you can convince me otherwise


Verifiability: If it's open source, this is trivial. But if i keep the source secret, is there any other way I can prove, or convice the community (you) the software is not a scam and safe?

Feature requests: What did you miss in other bots, or what feature should be in every bot? What do you want in the software?

And finally: would you buy it? Would you donate? If not, why not? Please be honest

Any other remarks I didn't ask for are also welcome

Regards,
AscotV