Bitcoin Forum
May 05, 2024, 01:29:45 AM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Poll
Question: Should analysis that could potentially identify alt accounts be posted, or should the 'privacy' of individuals take priority?
Post the analysis
No, don't post the analysis

Pages: [1] 2 »  All
  Print  
Author Topic: Trust List Analysis - Ethics  (Read 1551 times)
dogie (OP)
Legendary
*
Offline Offline

Activity: 1666
Merit: 1183


dogiecoin.com


View Profile WWW
October 03, 2015, 12:47:19 AM
 #1

Hello all,

I was spending some time with the new Excel and decided to code something that analysed the trust list dumps in a different way to the visualisations. Essentially this tool is able to identify multiple accounts which may belong to the same owner. The results are statistically significant however should still only constitute one piece of evidence of a link between accounts.

Rather than going into detail, I wanted to ask what people's thoughts are on the privacy issue surrounding trust lists. One person I contacted felt it was a personal intrusion that I had looked at their list (I've looked at all lists), even though the information was public and actively shared. Additionally, this tool may 'out' alt accounts which aren't nessecarily doing anything wrong or may indicate a link when there isn't one.

Vote above, I'll create a new thread if it comes to posting it.


tldr: made a tool that may identify alt accounts, should it be posted?

There are several different types of Bitcoin clients. The most secure are full nodes like Bitcoin Core, which will follow the rules of the network no matter what miners do. Even if every miner decided to create 1000 bitcoins per block, full nodes would stick to the rules and reject those blocks.
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
1714872585
Hero Member
*
Offline Offline

Posts: 1714872585

View Profile Personal Message (Offline)

Ignore
1714872585
Reply with quote  #2

1714872585
Report to moderator
BitcoinEXpress
Legendary
*
Offline Offline

Activity: 1210
Merit: 1024



View Profile
October 03, 2015, 01:08:24 AM
 #2

Hello all,

I was spending some time with the new Excel and decided to code something that analysed the trust list dumps in a different way to the visualisations. Essentially this tool is able to identify multiple accounts which may belong to the same owner. The results are statistically significant however should still only constitute one piece of evidence of a link between accounts.

Rather than going into detail, I wanted to ask what people's thoughts are on the privacy issue surrounding trust lists. One person I contacted felt it was a personal intrusion that I had looked at their list (I've looked at all lists), even though the information was public and actively shared. Additionally, this tool may 'out' alt accounts which aren't nessecarily doing anything wrong or may indicate a link when there isn't one.

Vote above, I'll create a new thread if it comes to posting it.


tldr: made a tool that may identify alt accounts, should it be posted?


Trust list are public info.

There is no "privacy invasion issue".




~BCX~
poeEDgar
Sr. Member
****
Offline Offline

Activity: 299
Merit: 250



View Profile
October 03, 2015, 01:10:30 AM
 #3

Why would people put alts on their trust list?

Unless you were on DT... and wanted to farm trust.

Seems to me that it's just compiling information that is already available.

Quote from: Gavin Andresen
I woulda thunk you were old enough to be confident that technology DOES improve. In fits and starts, but over the long term it definitely gets better.
dooglus
Legendary
*
Offline Offline

Activity: 2940
Merit: 1330



View Profile
October 03, 2015, 01:25:39 AM
 #4

One person I contacted felt it was a personal intrusion that I had looked at their list (I've looked at all lists), even though the information was public and actively shared.

Trust lists are public. I don't see how anyone could have a problem with you sharing them, or the results of analysing them.

I say post your results.

Just-Dice                 ██             
          ██████████         
      ██████████████████     
  ██████████████████████████ 
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
    ██████████████████████   
        ██████████████       
            ██████           
   Play or Invest                 ██             
          ██████████         
      ██████████████████     
  ██████████████████████████ 
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
██████████████████████████████
    ██████████████████████   
        ██████████████       
            ██████           
   1% House Edge
botany
Legendary
*
Offline Offline

Activity: 1582
Merit: 1064


View Profile
October 03, 2015, 02:04:08 AM
 #5

Posting is the way to go.
If you don't, somebody else might try and do the analysis.
Athertle
Sr. Member
****
Offline Offline

Activity: 252
Merit: 250


Go figure! | I'm nearing 1337 posts...


View Profile WWW
October 03, 2015, 02:05:10 AM
 #6

One person I contacted felt it was a personal intrusion that I had looked at their list (I've looked at all lists), even though the information was public and actively shared.

Trust lists are public. I don't see how anyone could have a problem with you sharing them, or the results of analysing them.

I say post your results.

Exactly. If they thought it was a "personal intrusion" then why haven't they complained about the publicity of trust lists? IMO, analysis of public information of any kind should be public as well.

Blazed
Casascius Addict
Legendary
*
Offline Offline

Activity: 2128
Merit: 1119



View Profile WWW
October 03, 2015, 03:38:46 AM
 #7

One person I contacted felt it was a personal intrusion that I had looked at their list (I've looked at all lists), even though the information was public and actively shared.

Trust lists are public. I don't see how anyone could have a problem with you sharing them, or the results of analysing them.

I say post your results.

This exactly...post it!
ndnh
Legendary
*
Offline Offline

Activity: 1302
Merit: 1005


New Decentralized Nuclear Hobbit


View Profile
October 03, 2015, 06:18:41 AM
 #8

Why not? Post it.

But I didn't get that part. Alts are added into their trust list Huh
shorena
Copper Member
Legendary
*
Offline Offline

Activity: 1498
Merit: 1499


No I dont escrow anymore.


View Profile WWW
October 03, 2015, 06:25:25 AM
 #9

I votes yes, I just hope this will not lead to more drama.

Im not really here, its just your imagination.
HolgerDansk
Member
**
Offline Offline

Activity: 68
Merit: 10


View Profile
October 03, 2015, 10:35:17 AM
 #10

I votes yes, I just hope this will not lead to more drama.

Of course it will lead to more drama.  For those that are outed for using alts, they'll scream and cry.  Similarly, those that are identified as alts and really aren't, will scream and cry.

There definitely will be some drama - some from those that are busted, and some from those that are innocent - but they'll both be singing the same tune.  So how will anyone know what the truth is.  The OP clearly states, it may identify alts.

I have a few questions about this:

How can anyone absolutely prove the accuracy of this kind of analysis?  Without knowing that, I can see that innocent accounts could fall victim to being wrongly identified.

Even though the raw data is publicly available, is it reasonable to examine personal trust habits of all BCT users?  And then to share publicly that analysis?


To make a comparison, just because a person goes out in public, to work, to the store, to the post office, to dinner, (all in the public view mind you), it is not deemed legal (nor appropriate in my opinion) to follow them about and inspect their personal life.  So much so even, that there are laws against that, at least in my country, the USA.

I wholeheartedly believe, that I would be charged with stalking, if just wanted to do my own in-depth analysis of my insurance agent, and followed him about to see who else he interacts with and who he trusts, by following him to different appointments with clients, when he goes to lunch and follow him to where he lives.  I don't think the US Courts would find anything impressive in a defense of, "I was just performing my own trust analysis of my insurance agent.  I didn't invade his privacy.  I only followed him around in public."

Here’s another quick comparison.  Just because I can walk up and down my residential street and see lots of windows open and without any curtains drawn, does not make it okay for me to walk up to each window and take a picture of the inside of the houses on my street (publicly viewed information).  What if I then take those photos, make a compiled list (analysis) of the details of say, everyone's televisions (or whatever), then buy billboard space and list all of that information for public view.  Is that okay?

Here's my last question to everyone: are you okay with this analysis being used for personal purposes other than identifying alts?

There's another current thread about, is doxxing okay.  There seems to be a debate about that.  But after all, it's just public information right?  So it's no big deal to share it, right? :-)  (let's not have a doxxing discussion here).


I truly and genuinely compliment Dogie for publicly asking this question, since I'm at least one person that brought this question to mind.  So I am curious to know people's response as well.   

From my personal experience, earlier this week, because of this analysis, Dogie politely PM'd me and asked why I had removed him from my personal trust.  As we exchanged some PM's I was amazed to find out that my personal trust list was publicly available.  I hadn't complained about it, because I didn't know.  Next I was amazed to know that someone had been looking at the trust of little ol' me.  A rather quiet person on this forum.  And finally I was amazed that I was being essentially asked, why don't you trust me.  It was a private decision that I made.  Now I was being asked about my private decision? 

So yes, I did feel that it was an intrusion of my privacy. 

Would you consider that an intrusion of privacy?



If they thought it was a "personal intrusion" then why haven't they complained about the publicity of trust lists?

As I mentioned above, I didn't know that my personal trust list was publicly available.  To whom & how would you suggest I complain to?  I would like to.

I don't think that my personal trust list should be publicly available.  Who I trust or don't trust, should be my choice to share.  I think the way it should be, is if you want to know trust opinions of a user, send the a PM and ask.  Isn't that how it is in real life?  For me to know who you trust or don't trust, I'd have actually ask you.  I really would need to ask specifically, do you trust DavidT or JohnnyL? (etc)  That's exactly what I have to do in real life, is ask.  I don't have some way to quietly peer into someone's data and find out.


IMO, on the positive side of this, I would love to see this analysis (if it can be proven out) be an additional tool of proving alts, available to admins or staff only.  Not publicly.  If they can use the information, to further a review into specific, possible, alt accounts, let them do that, with discretion, using other input as well.  Dogie mentioned in the OP, that this may help identify alts and that it should be but one piece of evidence (great wording btw Dogie, and thank you for qualifying the possible limitations of the analysis).  It's not an absolute.

As I read that, I read that there is possibly some margin of error or something similar.  With that in mind, I don't think the results should be shared publicly, because it is way too easy (and likely here on BCT) for it to turn into a witch hunt, where innocent people will be trampled along with alt abusers.

Dogie, thanks for posting the question.  It was your analysis, so it was your question to ask publicly.


My request would be, if this was to be posted publicly, that it only be done, with the blessing of Theymos (in a public post).  (it's really his forum, let him decide if something like this should be publicly shared and let him publicly let us know he has approved that).
BitCoinDream
Legendary
*
Offline Offline

Activity: 2324
Merit: 1204

The revolution will be digital


View Profile
October 03, 2015, 11:00:28 AM
 #11

One person I contacted felt it was a personal intrusion that I had looked at their list (I've looked at all lists), even though the information was public and actively shared.
Please name & shame this person publicly.

Additionally, this tool may 'out' alt accounts which aren't nessecarily doing anything wrong or may indicate a link when there isn't one.
I'd urge to provide the tool to the community as well, so that anyone can independently verify your results.

tldr: made a tool that may identify alt accounts, should it be posted?
Of course you should post it.

Vod
Legendary
*
Offline Offline

Activity: 3696
Merit: 3070


Licking my boob since 1970


View Profile WWW
October 03, 2015, 11:39:34 AM
 #12

Here’s another quick comparison.  Just because I can walk up and down my residential street and see lots of windows open and without any curtains drawn, does not make it okay for me to walk up to each window and take a picture of the inside of the houses on my street (publicly viewed information).  What if I then take those photos, make a compiled list (analysis) of the details of say, everyone's televisions (or whatever), then buy billboard space and list all of that information for public view.  Is that okay?

If you walked right up to their window, that would probably not be OK, as you would be on private property.  A person has a reasonable expectation of privacy on their own private property.  

However, if their curtains are wide open, and one can see their television from the street, they can't have a reasonable expectation of privacy on what they are watching.

I'm looking forward to this analysis.  I voted to post it.  

Note:  This post broke my 10,000 post marker.  Sad

https://nastyscam.com - landing page up     https://vod.fan - advanced image hosting - coming soon!
OGNasty has early onset dementia; keep this in mind when discussing his past actions.
dogie (OP)
Legendary
*
Offline Offline

Activity: 1666
Merit: 1183


dogiecoin.com


View Profile WWW
October 03, 2015, 11:41:49 AM
 #13

Please name & shame this person publicly.
He posted above you.


I'd urge to provide the tool to the community as well, so that anyone can independently verify your results.
I don't propose to share the tool as it could be mis-used or edited into a form to fit a particular narrative, and its a .xlsm so you couldn't be sure it didn't contain malicious code until it was executed.


But I didn't get that part. Alts are added into their trust list Huh
Its both more and less complicated than that, and I probably won't go into too much detail on the methodology in order to keep up its usefulness for now.

This version of the analysis has suggested 6 pairs then (very) weakly 100 or so other pairs. The next version will suggest many more pairs strongly and so actually be useful, but its got to be done from scratch so it may take a few nights to complete. Unless I here any compelling arguments against, this will get posted.

redsn0w
Legendary
*
Offline Offline

Activity: 1778
Merit: 1042


#Free market


View Profile
October 03, 2015, 01:24:24 PM
 #14

I voted for :

Post the analysis

It will be really interesting (in my opinion) to read it.
sho_road_warrior
Member
**
Offline Offline

Activity: 114
Merit: 10

PMs blocked, send answers to main.


View Profile
October 03, 2015, 03:30:08 PM
 #15

@HolgerDansk I find your comparissons ill fitted. I do aggree that nothing revealed here should be taken as concrete proof, but I still think the results should be made public. A possible compromise I see would be to only reveal issues regarding default trust members up to level 2 as they have significant influence within the system. As such their ratings and choices should be discussed openly.

@dogie could you provide any details on the method(s) you used? I think it would help to understand how significant the results are.

┏(-_-)┛┗(-_- )┓┗(-_-)┛┏(-_-)┓
dogie (OP)
Legendary
*
Offline Offline

Activity: 1666
Merit: 1183


dogiecoin.com


View Profile WWW
October 03, 2015, 03:41:59 PM
 #16

@dogie could you provide any details on the method(s) you used? I think it would help to understand how significant the results are.
If I do then this will probably be a one off - it'll inform those who could potentially be revealed in the future how to reverse engineer it to protect themselves from this analysis. The methodology is also not too important with this kind of thing and that'll make more sense once its posted. Essentially the clever stuff just deals with the 10,000s of accounts and the output is relatively simple.

tmfp
Legendary
*
Offline Offline

Activity: 1932
Merit: 1737


"Common rogue from Russia with a bare ass."


View Profile
October 03, 2015, 03:57:57 PM
Last edit: October 03, 2015, 04:16:05 PM by tmfp
 #17

How many times can I vote "no" in the poll, just the once with my main account or...?
j/k

Of course you should be free to post it if you want to.

Edit:
I suppose the best way to see if it works would be to run it on yourself.
Would you be proposing to open the new thread with that as proof?

 Wink

Extraordinary Claims require Extraordinary Evidence
Quickseller
Copper Member
Legendary
*
Offline Offline

Activity: 2870
Merit: 2298


View Profile
October 03, 2015, 04:47:55 PM
 #18

I don't think this is a good idea for a number of reasons:
(in no particular order)
1) Even if you do not publish the methodology of how you reach the conclusion that two people are potentially an alt of someone else, by posting 500 sets of potential alts (for example) then someone others can find similarities in their trust lists to determine what your methodology is. So by posting your conclusions, you are in effect also posting how you reached your conclusion (it will probably will still be difficult for most to run their own simulations though).

This presents a number of issues, primarily that it makes it possible to impersonate someone else. Someone who wishes to scam could potentially make it appear as if they are a potential alt of someone trusted with their trust list, and leave other subtle clues that they are an alt of someone trusted, then someone trading with that person would potentially be more comfortable in doing a direct trade with that person while being under the assumption that they are trading with someone trusted when they are not.

The opposite is true as well as someone already outed as a scammer could try to impersonate someone who is not a scammer in order to remove credibility from someone they know is looking at this information. As above, they can leave subtle cues they are the person they are trying to impersonate.

2) Many people here do not know (or they ignore) the difference between the words "potential" and "probable" so they will probably take it as fact when you say that person "A" is an alt of person "B". However since there is nothing wrong with having alts, you are not saying with certainty that both are the same person and if both "A" and "B" are respected/reputable currently, then there will be no reason for either of them to respond to this claim, then when "B" scams someone, others will think that "A" is the same person as "B" and might question as to why they did not deny this claim when it was initially made.

3) Piggybacking on one of HolgerDansk's points is that a trust list is very similar to something that is going on inside one's house while they have their blinds open. An individual looking at that person's trust list is somewhat similar to them looking inside their window, however if I see someone looking in my window then I can choose to close my blinds if I so wish. I would say that publishing these conclusions would be similar to broadcasting what you see in this open window for everyone to see, even if they choose to close their blinds later.

4) This will lead to more drama and will be one more thing for scammers to use to sling mud at people they do not like.


I didn't vote in the poll, as you especially should know that forum polls are useless.
monbux
Legendary
*
Offline Offline

Activity: 1736
Merit: 1024



View Profile WWW
October 03, 2015, 05:21:37 PM
 #19

Post the analysis, I see nothing wrong with it.
I'm just curious to see if it's accurate or not.
btc-facebook
Legendary
*
Offline Offline

Activity: 1862
Merit: 1015


View Profile
October 03, 2015, 05:33:06 PM
 #20

So it means, alt can be reveal if only someone put his/her own trust into his/her alt by this method ?
Let see this..bring it on
Pages: [1] 2 »  All
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!