I don’t think the acceleration in registration is due to World Postcard Day.
I monitor the activity of a number of Internet resources on a completely different topic and see an increase in activity there as well.
2 Likes
17th Oct 2022, 16:10 UTC
2 Likes
Now I am finalizing a new system for predictions evaluating.
It will not use the Diviner’s Talent Rate as before, but will use the Angular Deviation of the forecast from the fact.
And as a result, the Average Angular Deviation will be calculated for all Participant’s Predictions of the Countdown if they made several predictions.
The smaller the angle, the more accurate the Participant’s shoots.
Such an approach in predictions evaluating is more clear and logical.
I don’t know exactly when we will pass the 69 million level.
But, if we assume that this will happen on October 24, 2022 at 23:00 (this is my 5th prediction in this countdown), then the accuracy of the participants’ predictions can be ranked like this:
| Participant of Countdown | Average Angular Deviation Theta, (°) |
|---|---|
| @shanx | 1,38 |
| @Stevyy | 1,81 |
| @greenskull | 2,93 |
| @anon95639744 | 3,15 |
| @Florallle | 3,24 |
| @Herbstsommerwind | 5,34 |
| @reiselustig | 6,73 |
| @tesc | 7,14 |
| @caknuck | 7,89 |
| @shootingstar7 | 16,52 |
| @ChocCrossie | 24,39 |
| @B00 | 33,37 |
Shanx’ and Stefan’s predictions look the most accurate assuming that we will pass the 69 million level on the evening of October 24th.
And as it often happens, @Florallle and me (@greenskull) are somewhere in the middle. Although it should be noted that @Florallle took first place in the last countdown.
Feel free to refine your predictions. That’s more interesting 
5 Likes
I know correlation is not causation, but in this case I am convinced WPD has a big influence. Users writing an/some extra card/s, users waiting from start of September to write, users feeling new motivation.
Yes, northern autumn may have an effect too, or the urgent need to ignore world news, but here, in postcrossing, WPD was/is definitely a factor.
The final stretch begins to start …
99,576 to go
22.10.2022
20:22 UTC
a card from FR to EE (to a user moving to DOM)
5 Likes
It would be interesting to know have you already calculated the date of reaching 70 million level?
Personally, I begin to roughly estimate offhand the next level in the middle of the previous one. That is, for about a couple of weeks now I have been analyzing and estimating the timeframe for passing the 70 million level.
1 Like
Your calculations are very inspiring and put me to shame. I mean, not shame but I enjoy the detail albeit over my head 
I’ve thought about 70 million already too - glad I’m not the only one.
4 Likes
I’m not sure why I ranked lowest, I look like I’m closest to the actual date. I missed it by one day by my calculations
2 Likes
We’re not there yet (89,738 to go) and won’t be today or tomorrow (miracle aside). So you’re guess
is not the closest one.
Why you’re lower than shootingstar7 that’s @greenskull’s task to explain 
Happy postcrossing, write, send, register and have fun
2 Likes
We’ll see
2 Likes
Methodology for assessing the accuracy of forecasts:
Obviously, the shorter the Distance to the Moment of Truth, with which the prediction is made, the smaller the possible Error.
Therefore, to evaluate each prediction, NOT the Error in its pure form is taken into account, but the ratio of the Error to the Distance from which it was made.
The arc tangent of this ratio will give the value of the angular deviation of the forecast from the fact. The smaller the angle, the more accurate this prediction.
For all predictions of the participant in the countdown, the average value of the angular deviations is calculated. The lower the average value, the better the overall predictions of the participant in this countdown.
This prediction was made on 13/10/22 at 21:31 (UTC)
If we accept that the Moment of Truth will come on October 24 at 23:00,
then the Distance is (24/10/22 23:00 - 13/10/22 21:31) = 11,0618 days,
and the Error is (24/10/22 23:00 - 17/10/22 16:10) = 7,2847.
Error/Distance = 7,2847/11,0618 = 0,658545625485906
Arctan (0,658545625485906) = 33,37° – Angular Deviation
For the case of @shootingstar7:
Error/Distance = 9,4063/31,7229 = 0,296514505294283
Arctan (0,296514505294283) = 16,52° – Angular Deviation
If the Moment of Truth comes not on October 24, but earlier, then @B00’s Angular Deviation will be less, of course.
3 Likes
@greenskull Since my prediction expired, can I make another prediction or will the small angle deviation make my prediction worthless?
7 Likes
I understood that! Good explanation.
3 Likes
Of course, you can make predictions whenever you want.
Even taking into account the fact that there is not much time left until the Moment of Truth, your prediction will improve the result.
And please don’t take all this too seriously. It’s just a game and entertainment
2 Likes
21 oct 15:00 utc
2 Likes
20 October 2022 at 0200 UTC 
2 Likes
62740 to go 
So maybe tommorrow is the day
2 Likes
This above all! It’s just a game.
1 Like
23 October 2022 at 08:00 (UTC)
And I will take the liberty of suggesting that the first three places will take:
- @Stevyy (prediction: October 22, 20:22 UTC, made on Aug 19, 08:12 UTC)
- @Florallle (prediction: October 21, 10:25 UTC made on Aug 23, 14:35 UTC)
- @shanx (prediction: October 26, 10:26 UTC, made on Aug 24, 15:46 UTC)
These are the most accurate shots given the long distances from which the shots were taken.
The Angular Deviations of the shots do not exceed 3° (degrees).
But the order may change Let’s see…
3 Likes
Getting closer! Wed., 19 October 17:56 US Central Daylight Saving Time (UTC -6)
- 68,953,590 postcards have been received because of this project
This also means my guess was wrong 
But, I don’t mind
1 Like
44330 to go
1 Like