MINOR FIX AVAILABLE: for those running on windows 10 64 bit that are experiencing an issue that would stuck the bot making it not updating prices.
Please if you dont have this issue, you dont need this update. This update is for people like ewibin and/or others that reported their bot stucking at price updates.
PM me to get the patch.
Thanks!
~Gun
Topic: 🔥🔥🔥🔥🔥[GUNBOT] Automatic Poloniex Profit Generator🔥🔥🔥🔥🔥 - page 182. (Read 508402 times)
Gun, calm down buddy you are going to give you are self an aneurysm, lol.
You are prolly right and being 98yo is not a good thingYour, bot is still the best bot available for sale here, just trying to offer constructive criticism.
Yes it is: say it loudBack to work now:
First you never said 15 minute was ever the only profitable with the 2EMA
Yes it is specified in the Instruction manual on top of this thread but most pairs do perform better with 15 yes.
Yes they do: the 2EMA strategy works ONLY with a 15 minutes strategy because it needs the volume to work. A 5 minutes candlesticks has not enough volume to generate a responsive enough long and short EMA.My point with having you use the 5Min chart on tradingview strategy i put together, is how far back it goes on the 5 min chart for backtesting,
It is a false backtesting because you must do the backtesting with the same EMA period you use live but by all means use the 15Min chart you are going to end up with same situation of sitting for weeks.
In fact using the 15 min chart looking at the trade history on backtesting shows it would of been stuck waiting on XMR to recover from October 2 till December 27th. That is a ton of missed trade opportunities.
In fact using the 15 min chart looking at the trade history on backtesting shows it would of been stuck waiting on XMR to recover from October 2 till December 27th. That is a ton of missed trade opportunities.
Correct! That's why i deprecated the aggressive and moderate strategy to only allow the 1000trades: the difference is easy and straight:
You dont sit anymore missing trade opportunitise by the moment you place a sell order immediately allowing the bot to continue trading using the availabel balance that you still have. EVENTUALLY: only a small portion of your balance will get "stuck", not all yrou balance.
Today a user (i will find his name later and post here his questions) PMed me and told this situation.
Wanna knwo what is his limit indeed? Think abotu it for one moment: the REAL limit is the % of balance you allow the bot to trade per pair.
I could have suggested him to enable Bloody Mary! Sorry. it is against my principles lol, even if i coded it. So wanna knwo what i told that guy? I told him: stop yrou bot and raise your max % balance to 40%.
Look what he wrote me back after a while (found him while writing):
How long should you wait out this swing? Like I said in an earlier thread I have been testing this against a bot I wrote with autoview that is actually capable of taking small losses and moving on. Comparing the two accounts is a stark difference. My bot in autoview had long dug out of this latest swing while gunbot still has pairs hodl'ing.
Sorry i dont support the trade at loss: i understand the importance of it and i coded it in the bot upon users requests but here it is not a matter of creating liquidity by tradign at loss: raise your max% balance as only a small portion of yrou coins are being hold.it seems like the smarter move is to sell so as to continue trading.
As i said: no. it still will get stuck with same scenario eventually.
It is not the trade getting stuck: it is you getting stuck on the "need to sell asap" strategy (again) If you look at the strategy I just put together on tradingview you can see what I am talking about when you look at pairs like XMR or ETC on say the 5 min chart.
Wrong chart: wont work on 5 minutes candlesticks. Stated many times in the thread and stated to you too when i gave you my porting of gunbot to tradingview: the 2EMA strategy is profitable ONLY on the 15 minutes period They would be stuck without any trades sitting on huge losses since early January late december.
What you call "huge losses" is the normal swing that happens every week since january/december with teh price btc/usd. Again the "loss" terms is coming from a worng point of view of the "need to sell asap" The 1000Trade from what I can tell doesn't seem to be much smarter without me intervening
And here is the mistake: today on the telegram group i posted the bloody mary (which in my opinion isnt the solution to the btc/usd swing, it is just a piece of code i've thrown there for those like you taht asked a "sell asap" exit). Lost 0.001 and gained back in 5 hours. Right now my today earning is at 1.76 vs 1.74 of the post bloody mary of today.
HOLD!!!!!!
Edit: Didn't buy it? I paid you for it.
No wait that wasnt for you lol. Of course you paid.
EDIT: HOLD!
So further review of this bots strategy, and after observing it for a while I think I see a flaw in the current implementation that needs to be dealt with. See: https://www.tradingview.com/script/ofJIry2F-Gunbot-Noloss/
Basically, gunbot has no means to ever exit a bad trade and depending on the pair can get stuck until manual intevention. I know there is bloody mary strategy, but with how I understand that works I don't think that completely solves this problem per-say. I am thinking I personally am going to back away from this bot for now and trade with my own strategies until I see a better way to deal with bad trades.
Perhaps instead of just creating a sell order at a certain percentage above the buy order, it would be better to continue to watch the order and if the indicators are that of a sale position, but falling and not close yet to the original sell order placed it replaces the sell order with a sell at the current price instead. It would mean the bot would take losses occasionally, but better than staying forever in a bad trade.
THat's what the old strategies Aggressive and Moderate did: while you wait for the sell order to be placed, you woudl get stucked with buy opportuinities.Basically, gunbot has no means to ever exit a bad trade and depending on the pair can get stuck until manual intevention. I know there is bloody mary strategy, but with how I understand that works I don't think that completely solves this problem per-say. I am thinking I personally am going to back away from this bot for now and trade with my own strategies until I see a better way to deal with bad trades.
Perhaps instead of just creating a sell order at a certain percentage above the buy order, it would be better to continue to watch the order and if the indicators are that of a sale position, but falling and not close yet to the original sell order placed it replaces the sell order with a sell at the current price instead. It would mean the bot would take losses occasionally, but better than staying forever in a bad trade.
The point of view is being able to get no emotions and have a "daily" trading strategy plus a middle/long term strategy with those coins that you call "bad trades"
Those are not bad trades man, those are coins that you bought cheap and dont have the patience to wait for the sell order.
How long should this been repeated?
EDIT: or yes, trade them manually. By manually it is not defined here what you call "manual intervention". THere shoudnt be a manual intervention at all in a strategy so easy and simple like this: price goes down both ema1 and ema2 > the bot buys. Price goes up both ema1 and ema2 > bot sells.
Getting rid of the emotion to "not be able to wait for the price to sell"...is the hardest part. As i always repeat: that's why you bought a bot uh.
Thanks for your feedback, glad to hear that it's fire-and-forget, as that's really what I'll need for the small-scale hobbyist purpose I intend (and also to mitigate emotion on my side).
e.g.
today rebootet the router and all tasks have stopped and never revived
Due to an issue that I don't think ever got resolved with Windows 10 and gunbot I actually have my bots running from a batch file with the batch file closing and relaunching the bots every 30 minutes. I then set this to run as a service using nssm. This actually works out because if I decide I need to make a change to the bots I can modify the config files and the changes will take effect within 30 minutes. Doing something similar would fix your issue if your internet goes out.
some Linux users are using it with nodemon as well: it restarts the bot at every *.js change (so even params.js and config.js)
Actually the error i'm seeing in case of a lost connection are ECONNECT and Socket Hang Up. I could try to handle the socket problem with node.js https parser and handle the ECONNECT error with a file change (for example) so nodemon would immediately restart it (even a dummy.js file would be good).
If you want to test it, you can install it with
Code:
npm install nodemon
Then launch the bot with
Code:
nodemon 1000trades.js
Then try to change any settings in config.js or params.js and you will see your bot auto-restarting
I find that with the not, it performs better with coins that go up and down regularly, but essentially stay the same.
With coins that are on a continuous down/up trend, the not encounters problems.
On down trends: The bot will continuously buy until it reaches its max, and then wait to sell ~2% above what it bought for (using default trading logic). If the price keeps going down, the bot will just hold and hold until the value goes under 40% of what it was originally (If I'm not mistaken)
On up trends: The bot will buy at the lower EMA value, which will happen because up trends are not 100% green, and then sell ~2% above. Then it will try again to buy more coin, but since the price is going up it will have to buy for more than it sold for. This will keep happening, and you will make profit, but for trading a pair like USDT/BTC, which is mostly bullish, you will have a hard time (especially recently). I had to stop my bot and trade manually to make some of my investment back.
I would recommend the two pairs XMR/DASH, USDT/ETH
With coins that are on a continuous down/up trend, the not encounters problems.
On down trends: The bot will continuously buy until it reaches its max, and then wait to sell ~2% above what it bought for (using default trading logic). If the price keeps going down, the bot will just hold and hold until the value goes under 40% of what it was originally (If I'm not mistaken)
On up trends: The bot will buy at the lower EMA value, which will happen because up trends are not 100% green, and then sell ~2% above. Then it will try again to buy more coin, but since the price is going up it will have to buy for more than it sold for. This will keep happening, and you will make profit, but for trading a pair like USDT/BTC, which is mostly bullish, you will have a hard time (especially recently). I had to stop my bot and trade manually to make some of my investment back.
I would recommend the two pairs XMR/DASH, USDT/ETH
Thanks for your feedback, glad to hear that it's fire-and-forget, as that's really what I'll need for the small-scale hobbyist purpose I intend (and also to mitigate emotion on my side).
e.g.
today rebootet the router and all tasks have stopped and never revived
If you need to reboot your router, it is safe to reboot your bot as well: he is able to fetch orders from poloniex and start it over from where it left.
Thanks for your feedback, glad to hear that it's fire-and-forget, as that's really what I'll need for the small-scale hobbyist purpose I intend (and also to mitigate emotion on my side).
e.g.
today rebootet the router and all tasks have stopped and never revived
Any reviews of GunBot from real users? 
I have used this as well as made my own bots using autoview and tradingview. I have been running a test for over a week now between my bots and gunbots on two different accounts with same starting value and trading same pairs. Gunbot does well and will make decent profit, but it does easily get stuck when BTC rally's like it did this week. Gunbot is very aggressive and constantly making trades. My own strategies are not so much, but end up with roughly the same amount of profit each day.
The difference this creates is like this week gunbot has much higher unrealized losses vs my bots in tradingview as mine just wasn't in as many positions when BTC rallied as gunbot was.
If you are looking for a ready made solution that isn't a ping pong bot and actually trades the indicators gunbot is the way to go. If you are tinker, good with scripting languages, and have a strategy of your own that you want to automate try something like autoview chrome extension.
Actually arrived here after seeing Gunbot on signatures. Thanks for sharing your side of using this bot. I've never, ever considered using a bot but I have to admit after reading this thread and others on minor successes, I wonder if there's anything to be said for bots to aggressively use and stick to your strategies.
I also wonder, in the hands of a total newbie who will have the discipline to follow instructions from OP, what performance this bot would give? Any bot newbies ready to share?
Im newbie and using Gunbot. Just bought it few days ago before BTC mooned again, but those 3 days the bot did trade in profit. Now i stopped to wait BTC shakeout, then will start it again. And yes there is nothing! nothing difficult - is easy as a pie! Plug&play. It dont require huge BTC depo to ROI himself. So i like it. Also s you see here you can find preferable pair to enter in bots config for more profits. Will post my review later when will have more stats. Gunthar youre awesome brain!
Thanks for your feedback, glad to hear that it's fire-and-forget, as that's really what I'll need for the small-scale hobbyist purpose I intend (and also to mitigate emotion on my side).
Don't forget to share your review and stats - I won't be the only one looking forward.
These show up occasionally but regularly. Some more about timeouts but that's just polo being flooded (or slow) or my internet connection.
I had it already several times that all tasks (Linux Terminator) had stopped at the same time and never recovered...
only a Strg+C and restart helping
dude just put it on some virtual machine (amazon aws for example). Its 99.8% up and works 24/7
These show up occasionally but regularly. Some more about timeouts but that's just polo being flooded (or slow) or my internet connection.
I had it already several times that all tasks (Linux Terminator) had stopped at the same time and never recovered...
only a Strg+C and restart helping
See if in the err.txt you see something like ECONNECT or Socket Hang Up or similar messages. I had some users seeing that. It should be fixable with some tricks on node.js but i'd need to reproduce that.
These show up occasionally but regularly. Some more about timeouts but that's just polo being flooded (or slow) or my internet connection.
I had it already several times that all tasks (Linux Terminator) had stopped at the same time and never recovered...
only a Strg+C and restart helping
I would use 2 different accounts to trade the same coin against different pairs. Anyway: bought at? And sold at ? 
bought at???because was also gain of other pairs
unfortunately sold all at 925 USDT
crystal clear now, thanks! So me trading 200% altogether is is totally fine 
what I meant is, for how much % of my BTC balance will it buy? my currently available (not on Orders) or everything (including on Orders)? (I might not by thinking straight here and the second one is not even possible!-/)
The formula is on available BTC balance and altcoin "on orders" balance: if the altcoin "on orders" balance is more that MAX_BALANCE % you set of BTC avaliable balance, it stops to buy for that pair.
Example:
You have 1BTC of available balance and you bought already 0.1BTC worth of Monero, and you set MAX_BALANCE to 20%: it buys!
Notice: available BTC balance is shared with all pairs! So, i.e.:
It your bot bought in the previous Monero example, you have now 0.9BTC of available balance. If a trade occurs in ETH it will be 20% max of 0.9BTC
And so on...
Hope it helped.
~Gun
First, as mentioned before, on what is the "MAX_BALANCE" p/P based on? Currently available or total?
Currently it is based "onOrders" altcoin balance: if you invested more than what you set in % on MAX_BALANCE it would stop to buy for that pairwhat I meant is, for how much % of my BTC balance will it buy? my currently available (not on Orders) or everything (including on Orders)? (I might not by thinking straight here and the second one is not even possible!-/)
Second, I'm unsure if "bloody Mary" is working correctly in my installation. You mentioned something like a "sales.txt" will be created?! I don't have that.
And i haven't seen Mary intervene! (which is a good thing I suppose.)
order Number and rate/amount are stored in local variables. It would create some text files if it need to "move" orders to a different rate level (and it would be a temp storing tho). I'm glad your BM didnt have to act: mine is definitely disabled And i haven't seen Mary intervene! (which is a good thing I suppose.)
good!-) all through I dont believe in your theory "never sell without a profit"!-) Not, if you dont live forever or got the patience of a snail tamer
Code:
BTC_LBC *** POLONIEX CALLBACK | returnCompleteBalances ***Please do not make more than 6 API calls per second.
BTC_DCR *** POLONIEX CALLBACK | returnCompleteBalances ***Nonce must be greater than 1740117903113001. You provided 1740117902976001.
trying the first one. Honestly I was a bit afraid that the second one happened because API calls got mixed up! I still use leo.... occasionally (different Key) for its bloody amazing visualisation! But I limit myself to starting it once a day to check since using your bot!-)
Also, would it be possible to add Timestamps to the logs?
Timestamps are already in the logs. You meant in err.txt? I can do that yes, that file is mostly used by me when users ask me some debug but is ok, i can add that.yes err.log. to see if those errors overlap with "other things" that I'm doing. And it would make them more useful anyway!-)
Thanks! And i'm a happy costumer btw!-)
Thanks! I'm listening to all suggestions with the right attntion to see what we can do to make this bot even smartier ~Gun
I definitely want the most clever bot out there
(way to many "quote" calls!-) First, as mentioned before, on what is the "MAX_BALANCE" p/P based on? Currently available or total?
Currently it is based "onOrders" altcoin balance: if you invested more than what you set in % on MAX_BALANCE it would stop to buy for that pairSecond, I'm unsure if "bloody Mary" is working correctly in my installation. You mentioned something like a "sales.txt" will be created?! I don't have that.
And i haven't seen Mary intervene! (which is a good thing I suppose.)
order Number and rate/amount are stored in local variables. It would create some text files if it need to "move" orders to a different rate level (and it would be a temp storing tho). I'm glad your BM didnt have to act: mine is definitely disabled And i haven't seen Mary intervene! (which is a good thing I suppose.)
Code:
BTC_LBC *** POLONIEX CALLBACK | returnCompleteBalances ***Please do not make more than 6 API calls per second.
BTC_DCR *** POLONIEX CALLBACK | returnCompleteBalances ***Nonce must be greater than 1740117903113001. You provided 1740117902976001.
Also, would it be possible to add Timestamps to the logs?
Timestamps are already in the logs. You meant in err.txt? I can do that yes, that file is mostly used by me when users ask me some debug but is ok, i can add that.
Thanks! And i'm a happy costumer btw!-)
Thanks! I'm listening to all suggestions with the right attntion to see what we can do to make this bot even smartier ~Gun
I've taken quite a "hit" with the BTC rise!-/ But there's hope..!
But my weekly stats aren't as "rosy" as your daily ones.
I have a couple of questions however.
First, as mentioned before, on what is the "MAX_BALANCE" p/P based on? Currently available or total?
Second, I'm unsure if "bloody Mary" is working correctly in my installation. You mentioned something like a "sales.txt" will be created?! I don't have that.
And i haven't seen Mary intervene! (which is a good thing I suppose.)
Third, I do have some errors in err.txt. Possibly no real effect on trading but I'd like to eliminate them anyway if possible.
These show up occasionally but regularly. Some more about timeouts but that's just polo being flooded (or slow) or my internet connection.
Also, would it be possible to add Timestamps to the logs?
Thanks! And i'm a happy costumer btw!-)
But my weekly stats aren't as "rosy" as your daily ones.
I have a couple of questions however.
First, as mentioned before, on what is the "MAX_BALANCE" p/P based on? Currently available or total?
Second, I'm unsure if "bloody Mary" is working correctly in my installation. You mentioned something like a "sales.txt" will be created?! I don't have that.
And i haven't seen Mary intervene! (which is a good thing I suppose.)
Third, I do have some errors in err.txt. Possibly no real effect on trading but I'd like to eliminate them anyway if possible.
Code:
BTC_LBC *** POLONIEX CALLBACK | returnCompleteBalances ***Please do not make more than 6 API calls per second.
BTC_DCR *** POLONIEX CALLBACK | returnCompleteBalances ***Nonce must be greater than 1740117903113001. You provided 1740117902976001.
These show up occasionally but regularly. Some more about timeouts but that's just polo being flooded (or slow) or my internet connection.
Also, would it be possible to add Timestamps to the logs?
Thanks! And i'm a happy costumer btw!-)
How does Bot reacts in case of Huge Pump and Huge Dump...
unfortunately has now sold all my BTC with pair USDT/BTC I would use 2 different accounts to trade the same coin against different pairs. Anyway: bought at? And sold at ?
How does Bot reacts in case of Huge Pump and Huge Dump...
unfortunately has now sold all my BTC with pair USDT/BTC Jump to: