Evil users

[suchmoon]

I'm assuming users who have "Never" as their last active date, like this here:

https://bitcointalk.org/index.php?action=profile;u=2237312

... are mostly those who got hit with the "evil IP" fee and didn't want to pay it.

Edit 2020-10-17: Probably not quite as simple. "Evil" (proxybanned) users can have a valid active date so some of them might be in the "No Login" and some in the "No Posts" column below.

• No Login - Users who never logged in, i.e. literally have "Never" as their last active date.
• No Posts - Users who never made any posts (includes the "No Login" users).
• Banned - how many users of those who registered during that month have been banned so far (not the number of bans during the month). May or may not have posted - probably did though, why else would they be banned.
• Active - users who made at least one post in the last 30 days, i.e. mid September to mid October 2020.

Here are month-by-month numbers of total new registrations vs evil the stuff mentioned above:

Month          Total  No Login    %  No Posts    %  Banned    %  Active    %    
-------------- ------ -------- ----- -------- ----- ------ ----- ------ ----- 
2017 January    13354       39  0.29     8913 66.74    160  1.20     72  0.54 
2017 February   12677       46  0.36     8163 64.39    172  1.36     67  0.53 
2017 March      14760       58  0.39     9403 63.71    183  1.24     91  0.62 
2017 April      15807       65  0.41     9745 61.65    268  1.70    111  0.70 
2017 May        23978     4745 19.79    14986 62.50    436  1.82    150  0.63 
2017 June       32734     7175 21.92    19847 60.63    495  1.51    164  0.50 
2017 July       41011     8716 21.25    27360 66.71    746  1.82    232  0.57 
2017 August     42254     8169 19.33    25251 59.76    807  1.91    264  0.62 
2017 September  46186     6626 14.35    26585 57.56   1265  2.74    269  0.58 
2017 October    89443    10872 12.16    58890 65.84   2317  2.59    361  0.40 
2017 November   97286    14398 14.80    67891 69.78   2350  2.42    282  0.29 
2017 December  191455    19925 10.41   154978 80.95   2224  1.16    289  0.15 
2018 January   224546    18907  8.42   177141 78.89   3653  1.63    518  0.23 
2018 February  112211    16595 14.79    82134 73.20   2431  2.17    312  0.28 
2018 March      97644    10922 11.19    69256 70.93   3319  3.40    306  0.31 
2018 April      86898     8418  9.69    63182 72.71   3135  3.61    248  0.29 
2018 May        91994     9827 10.68    65575 71.28   2797  3.04    245  0.27 
2018 June       77681     7340  9.45    55926 71.99   3022  3.89    260  0.33 
2018 July       69840     7510 10.75    52658 75.40   2157  3.09    221  0.32 
2018 August     54384     8293 15.25    41898 77.04   1471  2.70    206  0.38 
2018 September  45362    11212 24.72    34907 76.95   1852  4.08    165  0.36 
2018 October    39354     6487 16.48    30526 77.57   1787  4.54    165  0.42 
2018 November   35127     8598 24.48    26569 75.64   2094  5.96    129  0.37 
2018 December   23504     3445 14.66    17687 75.25   1229  5.23    114  0.49 
2019 January    22232     3039 13.67    15977 71.86   1017  4.57    142  0.64 
2019 February   16202     2155 13.30    11850 73.14    880  5.43    140  0.86 
2019 March      18931     4680 24.72    14402 76.08    992  5.24     87  0.46 
2019 April      17910     2709 15.13    11909 66.49   1249  6.97    185  1.03 
2019 May        24473     3576 14.61    16363 66.86   1988  8.12    211  0.86 
2019 June       22356     2550 11.41    15723 70.33   1529  6.84    158  0.71 
2019 July       18913     2953 15.61    13307 70.36   1181  6.24    177  0.94 
2019 August     16451     2407 14.63    12256 74.50    754  4.58    163  0.99 
2019 September  14561     2798 19.22    11106 76.27    669  4.59    124  0.85 
2019 October    17891     2778 15.53    12767 71.36    688  3.85    164  0.92 
2019 November   16356     2441 14.92    12256 74.93    846  5.17    166  1.01 
2019 December   17600     2518 14.31    12626 71.74   2052 11.66    175  0.99 
2020 January    13620     2807 20.61    10570 77.61    488  3.58    205  1.51 
2020 February   18974     3069 16.17    15752 83.02    474  2.50    225  1.19 
2020 March      15152     2386 15.75    12536 82.73    491  3.24    194  1.28 
2020 April      12871     2348 18.24    10031 77.93    539  4.19    177  1.38 
2020 May        13004     2216 17.04    10271 78.98    422  3.25    277  2.13 
2020 June       10326     1753 16.98     7975 77.23    400  3.87    309  2.99 
2020 July       11795     2322 19.69     9302 78.86    361  3.06    330  2.80 
2020 August     12309     2160 17.55     9819 79.77    308  2.50    471  3.83 
2020 September  11737     2066 17.60     9259 78.89    277  2.36   1528 13.02

Charts because who doesn't love charts... Y axis is percentage in all charts:


Edited 2020-11-30 to fix a broken image


Edited 2020-11-30 to fix a broken image

TBH I'm not really certain what this means. Trending up? ATH soon? Should we invest in evil futures?


Edited 2020-11-30 to fix a broken image


Edited 2020-11-30 to fix a broken image

[1715402809]

1715402809

[1715402809]

1715402809

[jackg]

In % we've been kinda low for a while.

It probably discourages users to have to pay the evil fees to register via tor but newbie jail was rejected before t
Evil fee implementation so it'll probably be the same again now...

[The Sceptical Chymist]

That time period in the months in and around October 2017 were effin' awful for the forum.  I recall that when I was doing a lot of reporting of shitposts (to soothe my nerves) and trying to link alt accounts, I kept seeing that many of them had registered right around that time.  I think the later months of 2017 were when a lot of ICO bounties were being created, and it brought out all the shitposters who no doubt created at least several alt accounts each. 

It was also when bitcoin was exploding in price, so it was extremely hard to find a spot in a regular bitcoin-paying signature campaign.  Anyway, that's why I think there were so many accounts registered between the summer and tail end of 2017.

I don't know a lot about what that "evil fee" is, though I've certainly heard of it.  That's for when you get IP banned, is it not?  Well, in any case it also doesn't shock me that there's a lot of "evil" accounts generated in 2017.  And boy, the numbers have stayed quite low ever since the merit system kicked off, eh?

[suchmoon]

I don't know a lot about what that "evil fee" is, though I've certainly heard of it.  That's for when you get IP banned, is it not?

Here's some info:

When you register, the IP that you used when you submitted the registration form is used to calculate your evilness. The more frequently this IP or its neighbors were banned, the more evil is associated with your account. The amount of evil associated with an IP decays slowly over time, but the amount of evil associated with an account does not. You must pay or be manually whitelisted to enable posting on one of these "banned" accounts.

Here are some stats:

Evil	% new users
0	53
0-1	35
1-10	4.4
10-20	0.80
20-50	2.2
50-100	1.3
100+	2.9

Currently each unit of evil requires a payment of 4023 satoshi. You only need to pay something if you have 1 or more, though.

It looks like the system was in place well before 2017 so perhaps the last active date wasn't always set to "Never" for those users back in the day. Or more people paid the fee when Bitcoin was cheaper.

[notblox1]

What about undercover hidden evil members who got banned, but they purchase new accounts, use new IP, and continue with their evil campaigns?
If we include them than your chart would look even more bullish

[1miau]

TBH I'm not really certain what this means. Trending up? ATH soon? Should we invest in evil futures?

My technical analysis skills are very limited, but let's try:



Looks like make or break SOON! 

I would not invest because it's infinite supply, isn't it?   

No financial advice of course, always DYOR!!!

[icopress]

It looks like the system was in place well before 2017 so perhaps the last active date wasn't always set to "Never" for those users back in the day. Or more people paid the fee when Bitcoin was cheaper.

Here is the information for 2015 and 2018 but you must have seen it.

What about undercover hidden evil members who got banned, but they purchase new accounts, use new IP, and continue with their evil campaigns?
If we include them than your chart would look even more bullish

These are isolated cases that do not contribute to the overall picture. In addition, the option of paying for a VPN service for such users will be much more acceptable than paying a fee for each blocked account. (I'm talking about the fact that this data does not include information related to VPN IP addresses).

Here's some info:

suchmoon, Do you know what is the time range for saving evil IPs? I know that IP dims over time, but I don’t know how long it will take for the IP to turn white.

[suchmoon]

suchmoon, Do you know what is the time range for saving evil IPs? I know that IP dims over time, but I don’t know how long it will take for the IP to turn white.

I have no idea and I don't think theymos would divulge that info - it could probably be used to game the system.

[Little Mouse]

Where did you get the data? Are they public? I never knew the existence of the data publicly    Is there any valid reason why evil IP has suddenly increased in May 2017 while it was too low at April? Any changes made on the evil on IPs criteria at that moment?

I would not invest because it's infinite supply, isn't it?   

No, limited as yearn finance I think and that's why there are more evil on a lot of IPs. The % is moving slowly and that's good for future, same thing happening in bitcoin too.

[tranthidung]

It will be a good support if theymos shows the 2020 ban-map to compare with ban-maps in 2015 and 2018 (with same code for ban-map).

• https://bitcointalk.org/banmap201510.png
• https://bitcointalk.org/banmap201805.png

Last time, in May 2018, theymos wrote a topic when he created the second ban-map update. All Internet evil, mapped: evil score visualization 2

That time period in the months in and around October 2017 were effin' awful for the forum.

It is the period you are mentioning and charts look like flash crashes happened but statistics on those have never recovered well. Fortunately, it is good for the forum. 

[suchmoon]

Where did you get the data? Are they public?

I literally checked which profiles have "Never" as their last active date, like I mentioned in the OP. I don't know if that's entirely accurate but it seems unlikely that so many users would just register and never login for some other reason.

[Yogee]

This topic by LoyceV appeared when I searched about evil fees https://bitcointalk.org/index.php?topic=5175905.0
A $58 fee to be able to post is just insane and that was 2019. That could have been higher for accounts that were created in 2017.

......

That is an excellent TA 1miau hehe.

[Rikafip]

I literally checked which profiles have "Never" as their last active date, like I mentioned in the OP. I don't know if that's entirely accurate but it seems unlikely that so many users would just register and never login for some other reason.

While majority of those accounts might be inactive due "evil fee", there is one very active group that is constantly using old accounts with no prior posting history, Vitor Services and I guess some of those might be theirs, waiting to be activated. They sell thread bumping, and for some reason they prefer using old accounts and every time those accounts get tagged or banned, they just replace them with fresh batch. It is like they were thinking in advance and years ago created thousands of them, and then just activating when and if needed. They are probably  still creating new ones.

For example, you can find those accounts I am talking about currently active in these two threads. Accounts created in 2018, no activity whatsoever and then suddenly waking up few years later and going straight to "work". If those get tagged they wake up new ones, rinse and repeat.
[ANN] Baex dApp - Trade traditional stocks directly through the blockchain
[ANN] ZYX NETWORK - A GROUND-BREAKING POS-BASED PRODUCT FOR A WIDE AUDIENCE

[Pmalek]

Here's some info:

When you register, the IP that you used when you submitted the registration form is used to calculate your evilness. The more frequently this IP or its neighbors were banned, the more evil is associated with your account. The amount of evil associated with an IP decays slowly over time, but the amount of evil associated with an account does not. You must pay or be manually whitelisted to enable posting on one of these "banned" accounts.

I wonder how big that neighborhood is? I remember reading stories of users registering on the Bitcointalk forum for the first time and they were presented with an evil fee. That is a good way to get rid of unwanted members, but unfortunately it keeps legit people from being part of the forum as well.

Question is, how well does the evility perform under VPNs, proxies, and socks? I guess free VPNs can't be used since most of them would probably have been misused by someone by now. 

[CryptocurencyKing]

This is a statistic that I don't think is being considered much as it should be in terms of influx of new users in to the forum. Fee payment serves like a deterrent to most users with a need to join the forum but for some reasons, can't afford the fee and even if they do, they are yet to discover the leverage the forum offers it's users and the potentials found there in.
This fee payment could be the main reason behind most brand new dormant accounts remaining dormant. I don't know but, is this fee payment really a good thing, considering the fact that it to some extent hinders new prospective forum users from joining the forum. Not like the fee is levied on every new account as I believe there are criterias for this and influx of prospective users doesn't seem like a major concern for but, is  is it really or so important that these fees be levied?

[suchmoon]

For example, you can find those accounts I am talking about currently active in these two threads. Accounts created in 2018, no activity whatsoever and then suddenly waking up few years later and going straight to "work". If those get tagged they wake up new ones, rinse and repeat.
[ANN] Baex dApp - Trade traditional stocks directly through the blockchain
[ANN] ZYX NETWORK - A GROUND-BREAKING POS-BASED PRODUCT FOR A WIDE AUDIENCE

Interesting. So I checked a few users:

https://bitcointalk.org/index.php?action=profile;u=1746040
https://bitcointalk.org/index.php?action=profile;u=1747918
https://bitcointalk.org/index.php?action=profile;u=1748332

All registered in 2018, all started posting in 2020. These accounts did not have "Never" as their last active date when they were dormant so I'm still fairly confident that "Never" means something different than just a reserve force for future spamming... likely the evil fee.

[PrimeNumber7]

Where did you get the data? Are they public?

I literally checked which profiles have "Never" as their last active date, like I mentioned in the OP. I don't know if that's entirely accurate but it seems unlikely that so many users would just register and never login for some other reason.

This will not predict if someone has to pay a fee. When you sign up for an account, you will get a message saying there was a "cookie error" and that if you just signed up, you need to login. At this point, there is no indication you need to pay a fee, and the profile last active date is "never". Once you login, there will be a message at the top of the page that you are unable to post with a link to the proxyban page that tells you how much you need to pay to get unproxybanned; once you login, your profile will no longer indicate your last active date is "never".

A profile indicating they have "never" been active means the person has never logged into their account, and does not know if they need to pay an evil fee or not.

I just confirmed the above procedure a minute ago and the above is correct.

[icopress]

Interesting. So I checked a few users:

https://bitcointalk.org/index.php?action=profile;u=1746040
https://bitcointalk.org/index.php?action=profile;u=1747918
https://bitcointalk.org/index.php?action=profile;u=1748332

All registered in 2018, all started posting in 2020. These accounts did not have "Never" as their last active date when they were dormant so I'm still fairly confident that "Never" means something different than just a reserve force for future spamming... likely the evil fee.

This is what I was asking about ... when I wrote about time slots.

Judging by this data, all these accounts were blocked immediately after registration due to "evil IP" for this reason, "Never" was as the last active date.

The beginning of publications in 2020 can be explained by the fact that after a time (a year or three) the blacklists were dropped, the owner of these Alts came in and saw that the accounts were no longer blocked.

[suchmoon]

The beginning of publications in 2020 can be explained by the fact that after a time (a year or three) the blacklists were dropped, the owner of these Alts came in and saw that the accounts were no longer blocked.

AFAIK once an account is hit with an evil fee it never gets reduced for that account. So even if you log in after two years you'd still have to pay it. Not sure if you register a new account from the same IP if it could get a lower fee, assuming the IP address was not involved in any more bans in that time. But I don't think registering a bunch of accounts and reactivating them later is a viable strategy to avoid the evil fee.

I'll add a couple more columns to the table in the OP that might be more useful - number of users who made at least one post and number of still-active users.

[icopress]

You may be interested in: Applications and mapping algorithms

AFAIK once an account is hit with an evil fee it never gets reduced for that account. So even if you log in after two years you'd still have to pay it. Not sure if you register a new account from the same IP if it could get a lower fee, assuming the IP address was not involved in any more bans in that time. But I don't think registering a bunch of accounts and reactivating them later is a viable strategy to avoid the evil fee.

I'll add a couple more columns to the table in the OP that might be more useful - number of users who made at least one post and number of still-active users.

Probably this happens only in those cases when the account itself is banned, then you will have to pay a little for registration or activation through the same IP address.

Below is a quote that I understand as confirmation that it takes time for the IP to turn white, This probably only applies to IP addresses that have not been associated with evil in the future. Unless the accounts are blacklisted or permanently blocked, this can be a viable strategy to avoid being charged.

Do evil points ever fade away

Yes. This map is pretty short-term. And I'm probably not going to update it. But it's still interesting IMO.