Reddit botting, fingerprints etc

Pavilion

Regular Member
Jr. VIP
Joined
Jul 25, 2012
Messages
463
Reaction score
118
Hello to all Reddit bots :)
I ran into a problem that Reddit started banning my bots en masse.

Maybe some of you know what hardware parameters reddit checks and how it detects bots in general?

I know that Reddit checks navigator properties but I'm not sure what else. Maybe WebGL or Canvas?

If anyone has information on this I would be grateful!
 
You are talking about creating accounts? if yes, doesn't matter.
If you create an account with good human actions, you will not have any problems, In this way, I create reddit accounts without any problems.

Not only creating an account, but also such behavior in general is required.
 
What do you mean human actions? Could you plesse explain. Mouse movements? Timeouts?
 
They check IP, OS version, browser version/type, OS language, keyboard language, date and time.
My theory is that they check canvas and other options. Try to make your user agent blend with others as much as possible.

This website is useful for checking fingerprints: amiunique.org/fingerprint
 
They check IP, OS version, browser version/type, OS language, keyboard language, date and time.
My theory is that they check canvas and other options. Try to make your user agent blend with others as much as possible.

This website is useful for checking fingerprints: amiunique.org/fingerprint
Blending with others, using latest Chrome and Windows. But still it somehow detecting, it used to be ok 2 years but the latest update has ruined everything.
 
Blending with others, using latest Chrome and Windows. But still it somehow detecting, it used to be ok 2 years but the latest update has ruined everything.
Are you using selenium with chromedriver for automation?
 
Are you using selenium with chromedriver for automation?
Yes, it's not detectable (at least was), using advanced settings and checked with some services. I'm only changing WebGL and hardware concurrency.
I also checked Reddit source files (javascript) and ajax requests they sending - nothing special to worry about. I really don't know what's the reason of banning. Emulating human behaviour, clicking everywhere, even searching the syte, typing etc
bot.sannysoft.com - everything is green
using 4g proxies
 
Last edited:
Yes, it's not detectable (at least was), using advanced settings and checked with some services. I'm only changing WebGL and hardware concurrency.
I also checked Reddit source files (javascript) and ajax requests they sending - nothing special to worry about. I really don't know what's the reason of banning. Emulating human behaviour, clicking everywhere, even searching the syte, typing etc
bot.sannysoft.com - everything is green
using 4g proxies
There has to be some pattern then or simply maybe your proxies are not good anymore. (reddit started flagging their IPs for some reason).
 
There has to be some pattern then or simply maybe your proxies are not good anymore. (reddit started flagging their IPs for some reason).
I don't think it's a pattern since everthing is randomized. Ips.... 4g proxies? Many other users are using them.
I'm testing now BAS software, let's see how it goes.
 
I don't think it's a pattern since everthing is randomized. Ips.... 4g proxies? Many other users are using them.
I'm testing now BAS software, let's see how it goes.
I meant your provider not whole 4g proxies. He may be selling them IPs to 10 users from which 4-5 use reddit the same way you do.
But yeah if it is not that then it is something with software. Best of luck!
 
There is definitely a common link between your accounts, that probably is what is being detected.
 
Yes, it's not detectable (at least was), using advanced settings and checked with some services. I'm only changing WebGL and hardware concurrency.
I also checked Reddit source files (javascript) and ajax requests they sending - nothing special to worry about.
Interesting. How deeply did you check those source files though? Aren't they all minified and obfuscated?

It wouldn't be hard to implement canvas checks. They have libraries for doing this. What's hard is to actually detect if Reddit is doing those checks (at least correctly).
 
Resold 4G proxies are surprisingly dirty. Have you put them through IP/ fraud detection services?
 
How important to have a confirmed, not a fake email in the profile? I believe Reddit just tightened there.
 
They check IP, OS version, browser version/type, OS language, keyboard language, date and time.
My theory is that they check canvas and other options. Try to make your user agent blend with others as much as possible.

This website is useful for checking fingerprints: amiunique.org/fingerprint
Alsontry to give your account real usernames , and put avatars or pictures ,, this some how worked for me , also remeber to act as human .. as bots you can also make them do less actions at first and increase interaction as karma and age increase
 
We have a massive ban wave since January. Even during farming them. We use ADS power and static residential proxies and seems to me that Reddit doesn't like artificial fingerprint.
 
its actually easier to make accounts manually now then artificially. cause when you realize no
They check IP, OS version, browser version/type, OS language, keyboard language, date and time.
My theory is that they check canvas and other options. Try to make your user agent blend with others as much as possible.

This website is useful for checking fingerprints: amiunique.org/fingerprint
in reddits terms of service they say what they chek. canvas isn[t mentioned
 
Back
Top