I checked it out. Congo and Cook Island each have two members. Finland has 17,422 members.
In order to bring the number of Finnish members into line with these two, only two to say ten members should be admitted or write cards. In this case one should orientate oneself on the lowest number of members in order to create equality. Or you can first advertise in Africa to attract members. Then you can also involve other countries again.
1 Like
@Elikoa, I would give your post 5 hearts if I could!
And for those people who want the ability to control what they receive, the forums provide a way to participate in many round robins.
5 Likes
Too bad that postcards are not sent by countries and regions, but by individual users.
Being a part of this community is subjective from the beginning - we all chose it and stayed for some reason. I said that it works well within the idea of the project, which is explained in FAQs. If you don’t like the idea of the main site (which is very clear about not being a collector’s site), there are many games in the forum. Sustainability, once again, depends on the goal of site’s creators. If the goal was to discriminate users from certain countries, I’d subjectively have a very big problem with that and choose not to be a part of it. There’s a difference between pointing out minor technicalities to improve how the site works and completely disregarding and ignoring everything you read and agreed with when joining postcrossing.
10 Likes
And here are 5 hearts back from me! 
You @doryfera used a very interesting word/concept here: control. Ultimately, this is what we are talking about/what the original post was about - the algorithm and lack of control we have on it + wondering if there is, essentially, a way to understand it (and understand it can lead to…controlling it?).
Some people want, need to control the countries they interact with. Some don’t, and the lack of control is for me the most attractive part of Postcrossing. I don’t do many tags or swaps because I don’t love to know what is coming, though of course sometimes I do for the sake of interacting in a different way.
When we go out for food, I love it when someone orders for everybody and we share (which is common in Chinese culture, for example). I often don’t even get interested in the ordering process, I sit back and wait for what is coming. Some people hate that, they want to choose their own dish and eat it alone - fair enough. So the sharing and randomness is official Postcrossing, and the single personal dish is the swaps? And indicating that “I eat anything except broccoli” is when people choose to tick “don’t send to my own country”. I guess “repeated countries” is a restaurant that specialises in only a few dishes 
18 Likes
Very good word control.
But I have a different idea on it. It is not the users that want to control the countries of address that they get, but the users want an algorithm with without any preference, which is a control from the Postcrossing team.
These users just want to know whether the algorithm has any preference to different users. I do not know why you think these users want to control the algorithm, instead.
For those who really wants to know how algorithm works, Ana already kind of reply here.
And I must say I agree with her. When you start a Postcrossing account you sign up to receive a card for each one you send. Unless this doesn’t happen to you, I don’t understand what there is to complain.
For me it’s fair, those who send more cards receive more cards!
If you don’t like this random model, as several persons already said, there’re swaps, games, tags, whatever…
We can discuss how random a computer-modeled random algorithm is - but it seems there isn’t where the conversation is going… 
edit: something to read about randomness in computers 
8 Likes
Because that’s what it seems like, to me. I can see your point when you say it but had not thought about it. I think I see an algorithm as something you can’t control/can’t have preferences because it has been programmed to do one thing randomly and that is it.
Also I would venture to say that the Postcrossing team does control it because they created it. Obviously if it didn’t work then they’d have to re-create it, but that would also mean there wouldn’t be any members, not even from Germany.
I understand the algorithm so: If you activate the button “Check if you would like to also send and receive postcards to and from your own country.” then you will send and receive less cards from Germany becuase the will rplace by cards from japanese cards.
1 Like
Well, I have an suggestion.
We can have a look at the title of this thread. It is talking about the algorithm itself.
Therefore, I tried hard to make the discussion to the algorithm itself, but we can see, many people are talking about their feelings of receiving cards from the Three Great Countries.
Too add more, I need to say that, you mentioned that the Postcrossing team created this algorithm. I think you are perfectly right. However, since they created the algorithm, they can control this algorithm by many methods, such as adjusting the parameters, or add a list of countries that user cannot sent postcards to. This is a black box to users. We can only know some results, but the details are controlled and adjusted by the team.
As for whether this algorithm works, apparently it is not a binary result. It can be made better. That’s why this thread is opened.
Just take an example, the US President is elected by US people. However, it does not mean every one is satisfied with him, or the President can do nothing to make his work better.
2 Likes
For me sending to the same countries feels more comfortable in in the meaning of control. I have experience and know what travel times to expect. Of course, there are always exceptions.
So control can also be found in the no-variety in the countries as well as in variety.
4 Likes
I know. (I’m not supporting this idea.) The members should of course also be active, because otherwise maybe only Finnish would be active, and people complain about not getting variety 
Another option would be to leave out the countries with only few members. And only allow countries who can gather for example 2000 active members (and then we should define activity, and people should agree to send certain amount cards or their account would be ended and next one in queue were allowed to join). No a good idea.
Also in situations like mail strike, natural disaster, epidemic making mail delivery difficult, should this just stop, because not all countries can join equally.
It also doesn’t mean there is a problem.
Here we send to people (most often), not politically arranged areas.
There are other variables in this; when certain amount of cards are registered, the member gets more slots to send. Therefore, when German sends to German, they get free slots very fast, and more addresses goes to German members. So, actually forbidding German member to send to another German, would help in a way. If all their cards were travelling long times, no addresses for Germans would become free.
(I don’t support this, or any restrictions or compulsory actions.)
7 Likes
The results of the algorithm can indeed be tweaked by the creators, either by adjusting parameters or adding/removing rules on how to generate an address for the sender.
However, to me there’s no evidence in this thread that the algorithm is unfair and should be made better. For the specific example given by the thread opener: I have not seen data showing that newbies get more diverse countries in the same amount of requested addresses then long-term members. Nothing to improve there.
I believe that the postcrossing team do their best to create a fair game for any member. And that there might have been times where in some specific cases the results were less fair, but they had 16 years to improve that. It’s fixed by now!
The only thing now influencing the results is the activity of members in all the countries, their settings when requesting an address and the postal system itself (i.e. blocked routes and delivery times).
13 Likes
Sorry, I also don’t like idea for minimum level at the member of any country. Why put off members just because postcard writing isn’t that popular in the country? Every member counts.
I found postcrossing through an advetisement. That would be a way of attracting members in rare countries.
1 Like
Of course no data is seen, because a normal user cannot get the full data.
To know whether there is a preference to newbies, it needs a huge amount of data, or check the algorithm design.
“not seen data showing that newbies get more diverse countries in the same amount of requested addresses then long-term members” does not mean newbies not get more indeed, because normal users never know the data.
That’s why I hope the Postcrossing team to have a clarify. We normal users can only guess and feel by a small range of data.
I am sorry but do you have evidence on this point?
1 Like
yes this is what i meant earlier. and mail within germany is super fast so slots would open faster as well. unless you let germany only send to germany.
but that’s all theoretical obviously. i’ve said it before but i think the option ‘send to repeated countries’ should just disappear and everyone will get the address first in line, how it’s suppose to be.
i think it only looks like that because if a newbie gets basically any country besides germany or russia it’s new and exciting. if you’re here for a few years and have sent hundreds of cards it takes longer to get a new country.
but i don’t see the ‘problem’ that everyone is talking about as it brings me great joy to write cards to people and receive cards from people. thought that was the whole point of postcrossing but i guess i missed something.
5 Likes
You are correct. There are many factors in this situation.
Maybe an A/B test can tell us the result.
Indeed, it seems to originate from a feeling and just might be explained by this theory instead of a fault in the design of the system.
To find the fault in the system, we’d have to find examples. @Ryuuen you indicate that we do not have the total data set, but I think it can be constructed if you wish. As users when we think something is wrong in the system, we should build our case with examples and share that with the team that can go and find the fault in the system.
No, I don’t. I only have about the same amount of years of experience in my work to explain algorithms and their results to people. And finding the errors in them.
I won’t bet my life on stating that there are no more small errors to be found, but the design flaws should have been fixed by now or the team is aware and there is a good reason to not change them. Only when adding big changes like travel mode, the risk of design flaws increases again.
Building database by user will lead to a heavy load to server. I think you understand this too. Since sever of this site is not so strong (sometimes out of service), I do not think it is a possible way to build user’s own data.
Instead, if the team explains the algorithm, I think it would be the most economical way to answer the topic’s question. See the tile, I think “yes” or “no” from the team can easily answer the question.
Well, I need to say, I do not have the motivation to find out the result. I just find many discussions here are out of the topic, the algorithm itself.
Meanwhile, I do not regard it as a fault. I regard it as motivation with a specific purpose.
For opinions from others, you want a evidence or data from them. And for your own opinion, you have no evidence but experience. Emmm 
1 Like
If we want to have a problem fixed in the algorithm, we need to get the facts together to show what is the problem, was what I was saying with my statement on data/evidence.
If all we do here is venting opinions, I indeed need no evidence.
When somebody truly believes the system is unfair, I want to help find out the reason behind that as it can be so many type of reasons.
As for your wish to have the postcrossing team explain the algorithm, someone else already quoted their reaction on that from another thread.
I noticed that quotation. Thank you for your informing again.
I am sorry for my poor English in advance, but does the contents in that quotation explains whether the algorithm has preference to different users when choosing address? I did not see the details. I think my poor English is to blame.