DefaultTrust changes

[suchmoon]

I'm more and more thinking the entire Trust system is too complicated.

How about creating an opt-in for DT1? If users who don't want to be on DT1 reach it anyway without their knowledge, shouldn't the Trust system prevent this from happening?

I think it's easy enough to opt out... send a PM to theymos and ask to be blacklisted. I would assume that most users don't want to opt out so making it opt-in would be more cumbersome.

[theymos]

Thinking of the two "voter" criteria...

- You must have at least 10 people directly trusting you each with an earned merit of at least 10, not including merit you yourself sent. These "votes" are limited.
- You must have at least 2 people directly trusting you with an earned merit of at least 250, not including merit you yourself sent. These "votes" are limited.

Somebody PMed me pointing out an example of someone included in DT1 only because they were trusted by a permabanned member. I was aware that permabanned members were allowed to be voters, but I thought it'd be too rare to matter. But now I'm leaning toward excluding them.

Also, I've increasingly been feeling that something should be done about inactive voters. Currently, a user can be inactive for years, but still contribute toward those two criteria. The specific way I'm leaning toward addressing this is to require that voters must have received merit from at least 2 distinct users in the last 3 years.

I did a pretend reshuffle, and compared to the 136 users who were eligible for DT1 last time, with the above 2 changes, these users were no longer eligible: bavicrypto, be.open, Best_Change, comit, digicoinuser, ekiller, ezeminer, finaleshot2016, Gunthar, Harkorede, Heisenberg_Hunter, hybridsole, joniboini, Koal-84, Lachrymose, mandown, and witcher_sense. (Some of these users may have become ineligible for reasons other than the above two changes; I just compared the output today after the changes to the output on Feb 1.)

If instead it was "voters must have received merit from at least 3 distinct users in the last 3 years", then these users would also become ineligible: Baofeng, MinoRaiola, and Russlenat. I'm leaning toward 2, mainly just on the principle of preferring a larger list generally.

What do people think of these two changes?

How about creating an opt-in for DT1? If users who don't want to be on DT1 reach it anyway without their knowledge, shouldn't the Trust system prevent this from happening?

To be eligible, you have to have 10 people in your trust list, which is pretty unusual. So I think that PMing me to opt out is good enough.


While I'm here, here's a thought that's been floating around in my head for a while: A while ago I realized that if you are among the ~136 people eligible for DT1, then your probability of being in DT (DT1 or DT2) after the cross-DT1 exclusions are applied is the CDF of the hypergeometric distribution, with the shape of the function largely dependent on [the number of people eligible for DT1 trusting you minus those excluding you] (ie. your "net DT1 inclusions"). If your net DT1 inclusions is very positive, then you will almost always be in DT; if your net DT1 inclusions is very negative, then you will almost never be in DT; if it's within a few of zero, then it'll switch randomly back and forth. This got me thinking that maybe it'd be more elegant to determine either DT1 or perhaps even all of DT by just directly computing a probability distribution based on net-inclusions (perhaps with other inputs) for every user, and then shuffling according to these distributions every month. But it's just a vague idea in my mind at this point; I don't plan to do anything with this soon, and I might never pursue it.

[EcuaMobi]

...But now I'm leaning toward excluding them.

I agree on this, mainly because there was obviously a reason for them to be banned so I'd trust them (and their trust lists) less.

However I don't see why inactive users should be excluded. Their previous opinions/trusts/lists aren't less valid just because they're inactive. Maybe I'd trust their opinions/trusts/lists less if they were active but didn't get any merit during those 3 years.
If the change is to be done and the choose is between 2 or 3 distinct users then I'd vote for 2.

[JayJuanGee]

Thinking of the two "voter" criteria...

- You must have at least 10 people directly trusting you each with an earned merit of at least 10, not including merit you yourself sent. These "votes" are limited.
- You must have at least 2 people directly trusting you with an earned merit of at least 250, not including merit you yourself sent. These "votes" are limited.

Somebody PMed me pointing out an example of someone included in DT1 only because they were trusted by a permabanned member. I was aware that permabanned members were allowed to be voters, but I thought it'd be too rare to matter. But now I'm leaning toward excluding them.

Also, I've increasingly been feeling that something should be done about inactive voters. Currently, a user can be inactive for years, but still contribute toward those two criteria. The specific way I'm leaning toward addressing this is to require that voters must have received merit from at least 2 distinct users in the last 3 years.

I did a pretend reshuffle, and compared to the 136 users who were eligible for DT1 last time, with the above 2 changes, these users were no longer eligible: bavicrypto, be.open, Best_Change, comit, digicoinuser, ekiller, ezeminer, finaleshot2016, Gunthar, Harkorede, Heisenberg_Hunter, hybridsole, joniboini, Koal-84, Lachrymose, mandown, and witcher_sense. (Some of these users may have become ineligible for reasons other than the above two changes; I just compared the output today after the changes to the output on Feb 1.)

If instead it was "voters must have received merit from at least 3 distinct users in the last 3 years", then these users would also become ineligible: Baofeng, MinoRaiola, and Russlenat. I'm leaning toward 2, mainly just on the principle of preferring a larger list generally.

What do people think of these two changes?

I personally agree it is a good idea that members should be active to be included as part of the DT1 algorithmic trust calculation, at least active in the sense of receiving merits like you mentioned as a way of determining activity, and I am not opposed to either the 2 merits or the 3 merits idea... but I would prefer if you would consider the possibility of describing your timeline in terms of four years in order to try to stay with something in the ballpark of the bitcoin cycle (what we perceive to be the bitcoin cycle), which might cause you to have to increase the merit requirement to 4 or some other number that might seem largely fair without being overly inclusive or overly exclusive.

While I'm here, here's a thought that's been floating around in my head for a while: A while ago I realized that if you are among the ~136 people eligible for DT1, then your probability of being in DT (DT1 or DT2) after the cross-DT1 exclusions are applied is the CDF of the hypergeometric distribution, with the shape of the function largely dependent on [the number of people eligible for DT1 trusting you minus those excluding you] (ie. your "net DT1 inclusions"). If your net DT1 inclusions is very positive, then you will almost always be in DT; if your net DT1 inclusions is very negative, then you will almost never be in DT; if it's within a few of zero, then it'll switch randomly back and forth. This got me thinking that maybe it'd be more elegant to determine either DT1 or perhaps even all of DT by just directly computing a probability distribution based on net-inclusions (perhaps with other inputs) for every user, and then shuffling according to these distributions every month. But it's just a vague idea in my mind at this point; I don't plan to do anything with this soon, and I might never pursue it.

Probably I am not refreshed enough about this whole DT establishing dynamic, yet I have noticed that through the years, the quantity of members eligible on the eligibility list has not really changed very much, and surely if you make the above changes to exclude non-active (non merit receiving) members, then the eligibility list will shrink towards almost all eligible being included as you mentioned, so surely it seems that it would be better to have an eligibility list that was closer to 200 members rather than 100 members to then to have a poole upon which to randomly pick so that not all 100% would be included every single month.. or alternatively you could just set the system that no matter the number eligible, that ONLY half (or maybe 2/3..or some percentage that makes sense) would be on DT1.. and maybe if the eligibility were to go up to 200 members, then it would then be at 50% so that you never go over 100 with the actual monthly selection, so then perhaps the number of selected DT1 for each month would end up ranging between 50 and 100 each month, but figuring out some way to get more members to be eligible if there might be some reason that 100 is a preferred target number for DT1.

For sure, I don't claim to be deeply enmeshed with these ideas, and so I am just throwing out some ideas that might not exactly make sense from your perspective of the current situation or what you are wanting to fix, and if there might be needs to expand some of the members that are eligible or not (so that random selection ends up being a portion on the overall eligible list rather than all of them).. since maybe over the past 7-ish years (since the latest major changes) there has not been a lot of variety in members that are regularly on DT1, even though maybe you might not consider the numbers or who is on to be really broken either since the forum has largely been operating with this somewhat rotating (but mostly the same or very similar 100 monthly DT1 members) DT1 group for the past nearly 7 years..

[ibminer]

Thinking of the two "voter" criteria...

- You must have at least 10 people directly trusting you each with an earned merit of at least 10, not including merit you yourself sent. These "votes" are limited.
- You must have at least 2 people directly trusting you with an earned merit of at least 250, not including merit you yourself sent. These "votes" are limited.

Somebody PMed me pointing out an example of someone included in DT1 only because they were trusted by a permabanned member. I was aware that permabanned members were allowed to be voters, but I thought it'd be too rare to matter. But now I'm leaning toward excluding them.

Also, I've increasingly been feeling that something should be done about inactive voters. Currently, a user can be inactive for years, but still contribute toward those two criteria. The specific way I'm leaning toward addressing this is to require that voters must have received merit from at least 2 distinct users in the last 3 years.

I did a pretend reshuffle, and compared to the 136 users who were eligible for DT1 last time, with the above 2 changes, these users were no longer eligible: bavicrypto, be.open, Best_Change, comit, digicoinuser, ekiller, ezeminer, finaleshot2016, Gunthar, Harkorede, Heisenberg_Hunter, hybridsole, joniboini, Koal-84, Lachrymose, mandown, and witcher_sense. (Some of these users may have become ineligible for reasons other than the above two changes; I just compared the output today after the changes to the output on Feb 1.)

If instead it was "voters must have received merit from at least 3 distinct users in the last 3 years", then these users would also become ineligible: Baofeng, MinoRaiola, and Russlenat. I'm leaning toward 2, mainly just on the principle of preferring a larger list generally.

What do people think of these two changes?

On the surface these seem like good changes, I'd vote for the power of 3 myself.

I don't really see a good reason for inactive users to continue to be on vote for DT1, they would not be keeping up-to-date on their lists and maintaining it, so I'd be fine with that.

While I'm here, here's a thought that's been floating around in my head for a while: A while ago I realized is that if you are among the ~136 people eligible for DT1, then your probability of being in DT (DT1 or DT2) after the cross-DT1 exclusions are applied is the CDF of the hypergeometric distribution, with the shape of the function largely dependent on [the number of people eligible for DT1 trusting you minus those excluding you] (ie. your "net DT1 inclusions"). If your net DT1 inclusions is very positive, then you will almost always be in DT; if your net DT1 inclusions is very negative, then you will almost never be in DT; if it's within a few of zero, then it'll switch randomly back and forth. This got me thinking that maybe it'd be more elegant to determine either DT1 or perhaps even all of DT by just directly computing a probability distribution based on net-inclusions (perhaps with other inputs) for every user, and then shuffling according to these distributions every month. But it's just a vague idea in my mind at this point; I don't plan to do anything with this soon, and I might never pursue it.

Are you basing this only on net-inclusions and not exclusions at all?  What are the "other inputs" that would go into this?

~hypergeometric distribution~

Careful, you might be giving Elon the next selling point on a future Tesla, lol.

[Free Market Capitalist]

I too see no reason why permabanned members should be able to vote, however statistically rare. To me law (or rules, if you prefer) should be based on principles, not statistics.

Same thing about inactive members. Even taking the rule of having received merit from 3 users in 3 years the total number of DT1 stays above 100.

...mainly just on the principle of preferring a larger list generally.

This is not shared by the community IMO. I have only heard complaints saying that there are too many people in DT.

[LoyceV]

I did a pretend reshuffle, and compared to the 136 users who were eligible for DT1 last time, with the above 2 changes, these users were no longer eligible: bavicrypto, be.open, Best_Change, comit, digicoinuser, ekiller, ezeminer, finaleshot2016, Gunthar, Harkorede, Heisenberg_Hunter, hybridsole, joniboini, Koal-84, Lachrymose, mandown, and witcher_sense. (Some of these users may have become ineligible for reasons other than the above two changes; I just compared the output today after the changes to the output on Feb 1.)

I'll try to find the reason for each of them:
bavicrypto: hasn't posted in the past 30 days
be.open: is included by 3 users with >250 Merit plus 3 users with >10 Merit (based on the pretend new rules)
Best_Change: drops from 10 to 8 qualifying inclusion-votes
comit: drops from 12 to 9
digicoinuser: drops from 10 to 7
ekiller: drops from 29 to 9
ezeminer: ................. I'll stop here. Manually checking them takes too much time.
finaleshot2016
Gunthar
Harkorede
Heisenberg_Hunter
hybridsole
joniboini
Koal-84
Lachrymose
mandown
witcher_sense

Based on my (incomplete) check above, it's not a bad idea

What do people think of these two changes?

I think it's an improvement, but it's also going to be very easy to abuse: sending Merit to a few users every 3 years can be enough for some users to be eligible again. That might be a reason to make it Merit from 3 instead of 2 users, although it doesn't rule out abuse.

If your net DT1 inclusions is very positive, then you will almost always be in DT; if your net DT1 inclusions is very negative, then you will almost never be in DT

That's a good thing, right?

if it's within a few of zero, then it'll switch randomly back and forth.

Lauda comes to mind, ranging from DT2-strength (-8) to (+7) and everything in between for years.

This got me thinking that maybe it'd be more elegant to determine either DT1 or perhaps even all of DT by just directly computing a probability distribution based on net-inclusions (perhaps with other inputs) for every user, and then shuffling according to these distributions every month. But it's just a vague idea in my mind at this point; I don't plan to do anything with this soon, and I might never pursue it.

I assume that would mean there will be less users with "negative DT1-strength" on DT1. Now, most of the time only ~92 out of 100 DT1-members can vote for DT2, while the others can only vote for DT1. Having 100 "positive" DT1-members could be an improvement indeed.

However I don't see why inactive users should be excluded. Their previous opinions/trusts/lists aren't less valid just because they're inactive. Maybe I'd trust their opinions/trusts/lists less if they were active but didn't get any merit during those 3 years.

Their Trust and Trust list will still be equally valid, but their vote on other users to reach DT1 will be ignored. They can still be on DT2 themselves.

I don't really see a good reason for inactive users to continue to be on DT1, they would not be keeping up-to-date on their lists and maintaining it, so I'd be fine with that.

Inactive users already can't be on DT1:

- You must have been online sometime within the last 3 days.
 - You must have posted sometime within the last 30 days.

Are you basing this only on net-inclusions and not exclusions at all?

My assumption: net-inclusions = inclusions minus exclusison.

[ibminer]

Inactive users already can't be on DT1:

Good point, I've edited to clarify!

Are you basing this only on net-inclusions and not exclusions at all?

My assumption: net-inclusions = inclusions minus exclusison.

Based on the "net-" piece, this is hopefully a safe assumption. I guess I'm not understanding the multiple outcomes used for a probability distribution here, but I assume it is still factoring the random selection in some way.

[PrimeNumber7]

Thinking of the two "voter" criteria...

- You must have at least 10 people directly trusting you each with an earned merit of at least 10, not including merit you yourself sent. These "votes" are limited.
- You must have at least 2 people directly trusting you with an earned merit of at least 250, not including merit you yourself sent. These "votes" are limited.

Somebody PMed me pointing out an example of someone included in DT1 only because they were trusted by a permabanned member. I was aware that permabanned members were allowed to be voters, but I thought it'd be too rare to matter. But now I'm leaning toward excluding them.

Also, I've increasingly been feeling that something should be done about inactive voters. Currently, a user can be inactive for years, but still contribute toward those two criteria. The specific way I'm leaning toward addressing this is to require that voters must have received merit from at least 2 distinct users in the last 3 years.

I did a pretend reshuffle, and compared to the 136 users who were eligible for DT1 last time, with the above 2 changes, these users were no longer eligible: bavicrypto, be.open, Best_Change, comit, digicoinuser, ekiller, ezeminer, finaleshot2016, Gunthar, Harkorede, Heisenberg_Hunter, hybridsole, joniboini, Koal-84, Lachrymose, mandown, and witcher_sense. (Some of these users may have become ineligible for reasons other than the above two changes; I just compared the output today after the changes to the output on Feb 1.)

If instead it was "voters must have received merit from at least 3 distinct users in the last 3 years", then these users would also become ineligible: Baofeng, MinoRaiola, and Russlenat. I'm leaning toward 2, mainly just on the principle of preferring a larger list generally.

What do people think of these two changes?

Receiving merit means that you are making high quality posts (according to the intentions of the merit system), while being included in someone's trust list means that you give accurate trust ratings (according to the intentions of the trust system). The two will generally contain the same people, however, they might not always. An old user might come online occasionally to trade, while not make any posts.

I do agree that with age, the DT system needs to be modified to prevent, for example, someone who is included by 50 inactive users from staying on DT after deciding to scam their trading partners after a history of honest trading.

Over time, people devote less and less time to the forum, and some are difficult to get in touch with.

One possible option would be to modify the trust system such that ratings decay over time. Users could have the option to specify a date of a transaction (with a blank date defaulting to the date of the rating), and older ratings could be semi-hidden and/or have a lessor effect on a user's trust score.

[theymos]

I decided to exclude banned users, and to exclude users if they haven't had at least 4 different people send them merit in the last 4 years. This'll be what I use in the reshuffle happening in the next couple of days.

I changed the meritor criteria from my original idea because the new criteria has roughly the same result, but it addresses JayJuanGee's preference for 4 years, and LoyceV's concern that getting 2 distinct meritors is too easy to game.

Some stats illustrating the difference:

Code:

Users with 10+ earned merit: 13549
After deleting banned users: 11313
--
After applying a 2 meritors in 3 years test: 3973
After applying a 3 meritors in 3 years test: 3499
After applying a 2 meritors in 4 years test: 5098
After applying a 3 meritors in 4 years test: 4556
After applying a 4 meritors in 4 years test: 4119 ***
After applying a 5 meritors in 4 years test: 3684
After applying a 6 meritors in 4 years test: 3315

So the pool of eligible voters will go down from 13549 to 4119, though this won't actually have as much of an effect on the final output as you might think, since most of those eligible voters don't actually have anything in their trust lists. After the change, as of now, these users lose their eligibility for DT1: be.open, Best_Change, digicoinuser, finaleshot2016, Harkorede, Heisenberg_Hunter, joniboini, Koal-84, KWH, mandown, Russlenat, and stoos.

If your net DT1 inclusions is very positive, then you will almost always be in DT; if your net DT1 inclusions is very negative, then you will almost never be in DT

That's a good thing, right?

Yes, I think it's a good behavior. Though since the current shape of the probability function was basically an accident, the exact shape is maybe not ideal. Perhaps it should be more or less steep, or asymmetrical (eg. maybe people with -2 net inclusions should have a lower probability of inclusion than people with 2 net inclusions have of exclusion).

[Timelord2067]

Looks reasonable.

Just checking it's still one hundred selected each month from the pool of eligible participants?

[theymos]

Just checking it's still one hundred selected each month from the pool of eligible participants?

Right, that hasn't changed. I think that the shuffling each month is very good, so if the pool of eligible members gets too low, I'll probably either reduce the number selected each month or loosen the criteria somehow to increase the number eligible. But I think that the current ~120 eligible is OK.

[Timelord2067]

Just checking it's still one hundred selected each month from the pool of eligible participants?

Right, that hasn't changed. I think that the shuffling each month is very good, so if the pool of eligible members gets too low, I'll probably either reduce the number selected each month or loosen the criteria somehow to increase the number eligible. But I think that the current ~120 eligible is OK.

Not sure reducing from 100 to a lower number will benefit the entire forum (can't recall how many were DT1 before this lottery system was adopted) a small pool of users could game the system if it were too low a number.

If there is evidence of alts gaming the system by meriting users to make their 4 from 4 quota will you take action to exclude them from selection?

[theymos]

If there is evidence of alts gaming the system by meriting users to make their 4 from 4 quota will you take action to exclude them from selection?

Yes: if anyone sees that, please make a topic in Meta about it.

[jamyr]

If there is evidence of alts gaming the system by meriting users to make their 4 from 4 quota will you take action to exclude them from selection?

Yes: if anyone sees that, please make a topic in Meta about it.

This would be one of the time I would like to improve your answer Boss T.

You may be wanting to say, if the evidence is heavy enough then, go. However, please use your best discretion as always.

no offense to your WhOliness TimeLord

💡 WhOliness - a personal term I just made up today, for addressing(with compliment) TimeLord2067 doing TimeLord2067

[Timelord2067]

If there is evidence of alts gaming the system by meriting users to make their 4 from 4 quota will you take action to exclude them from selection?

Yes: if anyone sees that, please make a topic in Meta about it.

This would be one of the time I would like to improve your answer Boss T.

You may be wanting to say, if the evidence is heavy enough then, go. However, please use your best discretion as always.


no offense to your WhOliness TimeLord

None taken - I asked a question and it was answered.  ^{For example, I could ask why the so called "self scratchers" aren't excluded (there's more than you might imagine).}

If people choose not to believe evidence, then you have to ask yourself why they are in denial which is a discussion for another thread.

[jamyr]

If there is evidence of alts gaming the system by meriting users to make their 4 from 4 quota will you take action to exclude them from selection?

Yes: if anyone sees that, please make a topic in Meta about it.

This would be one of the time I would like to improve your answer Boss T.

You may be wanting to say, if the evidence is heavy enough then, go. However, please use your best discretion as always.


no offense to your WhOliness TimeLord

None taken - I asked a question and it was answered.  ^{For example, I could ask why the so called "self scratchers" aren't excluded (there's more than you might imagine).}

If people choose not to believe evidence, then you have to ask yourself why they are in denial which is a discussion for another thread.

Does it "meta" now?

Today I dont think that they do, but when Im gonna look for a campaign to join, maybe they'll "meta" again.

[bill gator]

If people choose not to believe evidence, then you have to ask yourself why they are in denial which is a discussion for another thread.

Perhaps it is the rigidity that defines you, albeit admirably - old friend, that provokes a better question for us to explore (in another thread); If people choose to believe evidence in contradiction to proof, then you have to ask yourself the same, eh?

[Timelord2067]

If people choose not to believe evidence, then you have to ask yourself why they are in denial which is a discussion for another thread.

Perhaps it is the rigidity that defines you, albeit admirably - old friend, that provokes a better question for us to explore (in another thread); If people choose to believe evidence in contradiction to proof, then you have to ask yourself the same, eh?

On rare occasions, I've had contrary evidence presented to me and I have modified my reports to match.  The howlers and nay-naysayers choose not to remember that - rather to develop a closed mentality bordering on psychosis.

[bill gator]

On rare occasions, I've had contrary evidence presented to me and I have modified my reports to match.  The howlers and nay-naysayers choose not to remember that - rather to develop a closed mentality bordering on psychosis.

It's easier to dismiss the caricature they've created

If there is evidence of alts gaming the system by meriting users to make their 4 from 4 quota will you take action to exclude them from selection?

Yes: if anyone sees that, please make a topic in Meta about it.

But who holds our feet to the fire? (That's you, timelord <3)

The new change is a +1 from me, gator approved

"On rare occasions", made me genuinely smile, thanks for that.