The Health of Postcrossing

Postcrossing General Topics
42

Hello! Statisical analyst here.
The idea of evaluating the overall activity level by rank #100 is strange to me…that observation is WAY in the upper 99.9+th percentile.

Really, we should be watching the median number of postcards (the exact 50th percentile).
That would be much more useful in my humble opinion.

Now, that is still a very simplistic way to describe the activity level, because it ignores the fact that Postcrossers differ:

  1. Newer postcrossers have lower Travelling limits, which means their maximum values are severely restricted compared to longtime postcrossers.

  2. Volume will also be a function of average travel time. Chinese postcrossers will therefore send much fewer cards than German members who enjoy speedy transit time.

Also, I think we’re ignoring some important aspects of the “health” of a hobby:

NET GROWTH, a function of:
New member accounts.
Attrition (accounts that close)
“silent attrition” (accounts that go inactive)

AVERAGE Utilization:
actual Traveling cards / maximum Travelling allotment

7 Likes
43

Anthropologist/historian with zero understanding of statistics (but some familiarity of social behavior over time)…I assume factors affecting member’s recent participation would include the war in Ukraine, temp closure of drawing officials to China, and rising prices for food/gas/rent as well as inc easing stamp prices etc. These factors could contribute to lower participation across all member levels.

4 Likes
44

Absolutely! But they would affect different measures in different ways. The Ukraine war, and the sanctions against Russia that followed, would affect total message volume but maybe not net member growth.

1 Like
45

Alas behavior is a good (but not perfect) predictor of future behavior. Have you considered a behavioral model with perhaps multiple regression analyses?

1 Like
46

Hello Mike and everyone else who responded,

Thanks for your thoughts. This is interesting :slight_smile:
I will give my opinion on this matter.

Obviously, the LAST 60 DAYS rating includes mainly those players who have the maximum number of slots. According to Postcrossing rules, today it is 100 slots.
It is easy to see that the LAST 60 DAYS rating consists mostly of ALL-TIME TOP-100 participants.
If we take a closer look at the statistics of the number of sent/received postcards for most ALL-TIME TOP-100 leaders, we will see that approximately 5-6 years after entering Postcrossing, their values reach a maximum and then they begin to decrease linearly. And this is observed in all the leaders who play at the limit of activity.

It is this reduction that underlies the change in the geometry of the curve.

What caused this decline?
Let’s say that you are a Postcrosser who has reached the limit of the maximum number of slots in 5-6 years equal to 100.
You will not be able to send another postcard until you register one received. And your number of slots is no longer growing.
The fact is that every day more and more participants from countries remote from its epicenter (Europe) are included in Postcrossing. That is, Postcrossing is spreading all over the world. And the average postcard travel time is on the rise. Therefore, the intensity of sending and receiving Postcrossers with the maximum number of slots is reduced. That is why the value of the number of sent/received postcards per unit of time (60 days) falls among the leaders.

In other words, this decrease is directly related to the increase in the average postcard travel time. And it is growing, as the proportion of participants from external to the large centers of poscrossers concentration in Europe, the USA, etc. is growing.

You’re right. And in the old forum I started with this. But it is considered more difficult and cannot be estimated in the mind. Therefore, I changed it to a simpler criterion after I investigated the dependence of tail wagging on Postcrossing activity.

And as we see now, the nose of the rating is unstable, so it is not suitable in its pure form for calculating some metric associated with Postcrossing activity or its Health.

Maybe. If you offer your methodology for assessing Postcrossing Health, we will have something to discuss :slight_smile:

If you’re into quantitative analysis as much as I am, I invite you to take part in the Next Million Countdowns, where we try to predict when Postcrossing will cross the next million level.

In fact, I invite everyone to the Countdowns without exception.
Very soon we will close the 71 million level. And then we will start with a clean slate the 72 million level.

47

Yes, covid, war and mail blocking at some point became factors accelerating this fall. But the fall itself occurs independently of this. And above I stated my version of what is happening.

Yes exactly. The average postcard travel time drops throughout the game as the game area expands.

When 80% of Postcrossing was Europe, the postcards flew from sender to addressee much faster.
But throughout the time, the proportion of participants from countries scattered around the world is growing and the average time of a postcard in transit is growing.

Therefore, the number of cards sent or received by top participants who have long exhausted their slot limit is creeping down. Namely, such participants form the basis of the rating LAST 60 DAYS, at least nose of the curve.

1 Like
48

Sunday, March 26, 2023 - 23:40 (UTC -5)

Hello, Sergius:

I appreciate your creativity and concern for Postcrossing. I share it and continue my own efforts in various ways.

A lot of information about Postcrossing is “indirect” at best. We are often reduced to what we might call speculative anthropology as we try to figure out what’s going on. Unobtrusive measures, anyone?

But one obvious place to look is the monthly Stats page (Postcards statistics). The phrase “the trend is your friend” readily comes to mind. And the trend is not good.

After nearly parabolic growth during the first eight years, Postcrossing has entered what can only be described as a long-term secular decline with unmistakable lower highs and lower lows. Sure, there are seasonal spikes and there has been some recovery from several “anomalous events” over the last three years. The March, 2023, data has not yet been posted, but I fear that the downtrend will continue. The most active “super” Members (the top 100) can’t carry the whole Postcrossing population.

My review of county-level data in Bucks County, Pennsylvania, and surrounding counties in the Delaware Valley (my local area where I’m concentrating my efforts to promote Postcrossing, e.g., The Bucks County Herald (February 4, 2021 Edition) Features Postcrossing!) shows two very concerning things: there are practically no Postcrossers in this densely-populated, highly-developed territory and the majority of them have not even logged on in y-e-a-r-s.

A look at the state-level data for Pennsylvania and New Jersey shows the same pattern. We’re not able to see how many total Postcrossers are registered in these states, but of the first 1000 people shown, the last login < 1 month stops at about 500 (Pennsy) and 300 (Jersey). Clearly, the vast majority of people in these states are not “active” in any reasonable sense.

We can’t really determine it precisely, but it looks as though a relatively small number of people are sending most of the postcards each month. Personally, I can’t say that’s “healthy” enough. Which is why I work hard to promote Postcrossing here. But the little we can piece together gives a picture of small but intense participation (both in the “operational” side of Postcrossing (Send and Receive) and the Forum) that is slowly but surely losing the battle to grow Postcrossing, at least as measured by total cards sent each month.

Keep up the good work.

Best regards,

Michael

5 Likes
49

Hello, Michael (@AMDGIHS2019),

I can definitely say that:
The TOP-100 accounts for about 3.7% of all postcrossing postcard registrations.
The TOP-1000 ~ 21% of all registrations.
The TOP-2250 ~ 31% of all registrations.
The TOP-5000 ~ 43% of all registrations.
The TOP-10000 ~ 54% of all registrations.
…

The total number of active accounts has not changed in two years and is approximately 800,000 members.
The organizers delete inactive accounts, so there are about the same number of arrivals as they are deleted.

That is, the proportion of really active players is extremely small.

I think @paulo and @meiadeleite can confirm this :slight_smile:

1 Like
50

Active - meaning users who have logged in within the last 30 days - is around 15%. (13,7% in Feb in AT :austria:). Better wording is maybe still existing? number of created?

2 Likes
51

I mean these.

Members
Members874×768 46.1 KB

I guess that these are the participants whose addresses participate in the choice of the recipient and send something themselves.

I do not remember where I read that some accounts that are not used for a long time are deleted. Or perhaps it only applies to new accounts that have never been used. I do not remember.
I can only say that the number of participants grew at one time, but it has been at the level of about 800,000 for two years now.

52

Yes, accounts are deleted sometimes, most of the time by the users themselves (I think). I also don’t know about the variation around 800k.

But given our internal wording here, active users have logged in within the last 30 days - the rest are temporary or eternal inactive profiles. That’s why I’m suggesting ‘profiles created’

PS: I’ve found this number from 2017, so over the last year there was some increase at least

pc_numbers_21102017_1812
pc_numbers_21102017_1812345×312 48.7 KB

53

Is this not a product of the very system designed ?
I have exceeded 250 cards sent/received but have just 14 slots. This is a severe restriction on participation in increasing numbers by newer members.

2 Likes
54

It’s a long story.
At one time I suggested changing the curve of free slots versus the number of postcards sent.
Blue – the rule we have now.
Orange – the rule I suggested.

New rule
New rule840×520 77.4 KB

This would increase the dynamics of Postcrossing. Get out of the stagnation that has been observed over the past 10 years. And it would increase the motivation and enthusiasm of the average postcrossers.
But this idea did not find support from the creators of the game.

3 Likes
55

No, they don’t.

Only unused accounts are deleted and only if the users wish to or if the accounts are left for some reasons and there’s no reaction from the user.

They will never delete inactive accounts! Otherwise around 1/3 to 1/2 from all accounts would’ve been gone.

2 Likes
56

Of course not.
I expressed myself incorrectly.
This refers to accounts that were created, but did not show any activity for a long time. Or like that.
I can’t find where I read about it.

57

Aaah, okay. Yes, now I understand what you meant :slight_smile:

58

A few months ago I counted all the users that have logged in at least once in the past month, and that had sent at least one card since joining postcrossing. My figure was around 65000.

A few of them didn’t actually send any cards in the past month, but are still active in the sense that they check the website occasionally.

I have a different theory regarding the slowdown: I think that a new-ish user will draw most of the time (depending a lot on their drawing patterns too) seasoned, reliable users - top users that are owed large numbers of cards at the time - so their cards will be registered promptly. As one goes into their thousands, one catches more and more often newbies that might have lost interest, people that are registering with delays, people from countries with slower mail systems and so on. So top users eventually get longer travel times and more expired cards per trimester.

And sort of a corollary - the ratio of very active, longterm postcrossers that joined in recent years is much smaller than those of olden years. I think it’s because if you’re a certain age group and postcrossing is something for you, you’re way more likely to have heard about it already by now. There are people still discovering it, and some will stay, of course - but it’s been almost 20 years since its launch, plenty of time for most people to learn it exists, and try it out if they are so inclined.

2 Likes
59

More slots would not stimulate me…I have 25 open now, sending these all out would cost approx $2 each. I can’t justify spending $50 every 2 or 3 weeks. I try to do 5 cards a week, that’s all the time and money I have for now

I am curious if you have another account? This one shows you have only sent 30 cards in 4 years, wondering why you are not more active?

2 Likes
60

I’m not very interested in sending and receiving postcards.
I’m interested in exploring Postcrossing as a phenomenon.

2 Likes
61

Observer bias? :joy:

1 Like