I have been researching Postcrossing for a long time to assess its current activity and trying to find a parameter by which it would be easy and simple to assess the “Health” of Postcrossing.
Perhaps, some of us remember my post on the Beaufort scale in the old forum: https://forum.postcrossing.com/viewthread.php?tid=126985
This was my first attempt at getting such an assessment. It was funny, but not quite objective
After many observations over a long time, I came to the conclusion that the Health of Postcrossing can be very easily estimated by looking at the number of postcards sent over the Last 60 days by the participant №100 at the end of the list. https://www.postcrossing.com/stats/topusers
Unlike the leaders of this list, the number of the postcards sent for which are very volatile and sometimes go against what is really happening, for the participants at the end of this list, the values are quite accurately correlate with the upward or downward trend of Postcrossing.
Thus, I have compiled a new scale (Health % depending on the number of postcrards sent by №100), which allows us to quite reliably understand how active and cheerful the Postcrossing at the moment is.
I did not complicate the calculation method and just took the probability distribution graph of the value №100 as a basis, smoothed mathematically.
In other words:
245 or less = 0%
253 = 10%
259 = 20%
263 = 30%
265 = 40%
267 = 50%
269 = 60%
271 = 70%
273 = 80%
276 = 90%
278 or more = 100%
At this moment the №100 = 274 postcards sent.
So, it means the the Health of Postcrossing is about 83%.
That is, Postcrossing is currently very active and this is felt by some other indicators.
The Postcrossing Health will start to decline dramatically starting December 20.
And by mid-January it may drop to 5% on this scale.
Postrossing will recover to 60% only by the end of February.
And by the middle of March, the activity of the game can reach 100%.
Postcrossing, like a living organism, exhibits regular seasonality
This year also, Postcrossing’s health was around 0% in May due to COVID-19.
Over all.
The curve that describes the ranking of these random variables is not random and obeys the rank distribution.
And the value that stands, say, in the 95th place cannot be very different from the value in the 94th or 96th places. They are seamlessly related to each other and to the rest of the values in the aggregate.
When Postcrossing is active enough, then the tail of this curve with its entire series of values goes above, if inactive, then in the same way below.
In any case, by the single value at the 100th place, we can roughly tell what value will be at the 50th place. Therefore, knowing the only value, say at the 100th place, it is possible to estimate at what level above or below the entire curve passes. And thus assess how active Postcrossing is.
I have been collecting statistics on ranking data for more than a year and I know in what interval the values fluctuate at all places.
The value at the 100th place fluctuates in the range 230…284 in accordance with the law of probability distribution that I gave in the first post. And on the basis of which I built a scale for assessing the activity or health of the game.
The number of postcards sent over the Last 60 days by the top-postcrosser №100 at the moment is 252. https://www.postcrossing.com/stats/topusers
That means the Health (or Activity) of Postcrossing is about 9%.
This is the usual Christmas-New Year drawdown that occurs every year.
According to my calculations, the activity of the game will recover up to 60…70% only by the end of January.
№100 = 253
According to the method outlined above, Postcrossing is now in stagnation and activity or health can be assessed as only 10% on the proposed scale.
Postcrossing will come to life only by March, according to my observations. Let’s see.
Glad to know this ‘stagnation’ is a thing and not just my account…whose cards are going nowhere fast .
Have you only got data for the years you were a member? I’m interested to know if the ‘stagnations’ over the covid years have been more or less intense than over the non-covid years?
I’m going to crawl back into my stagnant pond now…and wait for a card to be registered .
I know interpretation is dangerous, but to what do you attribute the seasonal stagnations?
I know my own cards are slowed down thanks to (1) the colossal volume of Christmas mail, which Canada Post is struggling to keep up with during the Omicron surge (2) the aftermath of the floods here in British Columbia, which have resulted in a serious postal bottleneck, and (3) from my side alone, a feeling of fatigue and (figurative!) hangover from the frenzies of the Christmas season
The main pattern is that the highest activity is in March and the lowest in December. The rest is more or less evenly adjusted for the number of days in a month
I believe that the more worries people have, the less time they have for entertainment. There is usually a lot of worries at the end of the year, also in September - when the summer ends, children go to school.
The high activity in March is probably emotional in nature - it becomes warm, more light, nature come alive and people also come to life.
Hm, maybe I don’t get everything of you statistics, but when I see how many postcards only German postcrossers are writing in December, also because of “Cards for Literacy”, I don’t believe that December has really such a low activity.
I do kind of get it. Normally, my postcards to Germany take 2-4 days, but this month, they seem to have taken 7 days+, with nothing less than 4 days. This month, I have posted 11 postcards to Russia, but not a single one has been registered, when in other months, I could be looking for them to be registered after 11 days.
Hopefully, in the new year, things will speed up. I’m looking forward to March!
It is regrettable when people talk about what they do not understand.
Statistics are just mathematics. It’s neither bad nor good. It is important to understand what interests it serves. And if it is in the hands of thieves and liars, then it’s sad. But in this case, it is not the statistics that should be blamed, but precisely those thieves and liars who deceive other people.
It is sad that Mark Twain does not understand this and thus misleads the rest.
I think Mark Twain understood quite well; statistics, even when used with the best of intentions, can deceive. Hence the title of Darrell Huff’s best-selling book, “How to Lie with Statistics.”
.
.
