Articuno Day Shiny Rate in JAPAN

The report of the shiny probability survey in Articuno Day.

Survey Results

postsWins on Raidsshinyprobability

※As of August 5 8:30pm:

Therefore, it is suggested that the probability of shiny in Articuno day was 5%, which is half probability of Zapdos Day.

The data of posts

Please look at this spread sheet

※the window above may be difficult to see on mobile phone. Please open from the link.

How to investigate


Considered from the good quality result of the Zapdos Day shiny rate in Japan last time, I did not set a schedule of investigation in advance this time, but just in case I announced on Twitter at the last minute and asked people to take a screenshot of the picture before the event.

However, there was an obvious difference between the Articuno day’s shiny probability and Zapdos’s. Then, after the event, I asked on Twitter for some Twitter replies and DM about ”how many times you tried” and “how many shiny you saw”.

Moreover, I got some answers from the people around Tivoli, the founder of this site.

Condition of answer

①Screenshot of Articuno picture book before the event
②Screenshot of Articuno picture book after the event
③Screenshot of all shiny Articuno on PokemonGO (this is only required when you see the shiny)

Those screenshots of two or three above were considered as evidence.
Also, I asked people to send some information even if they do not have screenshots.

Collection of Range

・mainly from the link above, all the replies on Tivoli’s twitter account.
・all the DM for Tivoli’s twitter account.
・all the replies for this Website's twitter account.
・the members of Kyukyoku Kouryaku (ultimate strategy book) DB (team DB).
・Friends outside Twitter of Tivoli (Team Investigation).

How to gather

All the information above are gathered.
And there is no intentional data.
Speaking of the evidence, the ones that met the conditions above, the ones that we considered as trustable even though it only has a few evidences, and the ones given by people who usually connect Tivoli’s Twitter and if it was trustable, we considered it “true” and if it was not, “false”.
※False doesn’t mean it is not true but it is just lack of evidence.

The details of investigation result

The data based on enough evidence

postsWins on raidsshinyprobability

The data without evidence

postsWins on raidsshinyprobability


As well as the Zapdos Day Survey Result last time, the data with evidence tends to show many shiny data, and the data without evidence tends to show less shiny data.
The deviation like this is considered due to psychological reason. I wrote about it the article of Zapdos Day Survey Result.
In the voluntary survey investigation like this, the participants are generally not selected in advance. So, the deviation is not avoidable because the result affects people’s mentality.
At this point, this method of collecting the data, I ask people for the data visible, and basically I demand the evidence but also accept the data without evidence, is excellent way because it can keep the data reliance, the deviation of data counteract each other, and finally the data become homogeneous in total.

Moreover, since I did not make a big announce in advance this time, not so many people take a screenshot. Therefore, there are more data without evidence and the total result become less slightly.

Comparison to Zapdos Day

The probability is half

In comparison to the Zapdos Day Shiny Probability was 10 %, the Articuno day Shiny Probability was half, 5 %.

In other countries?

In Japan, the Articuno day was postponed because of west Japan rains, and after the Zapdos Day, the Articuno day was held.
However, in other countries, Articuno day was held before the Zapdos Day.

About the Articuno shiny probability at that time, the web site TheSilphRoad shows the survey similar to this time. It says it was 4 %, 76/1873

I’m not sure how this survey was taken, but it is possible range even if the probability was 5 %. And if the data was voluntary taken, there can be deviation because of the psychological factor.

In any case, if it was 10 %, the survey above is not possible.
As this reliable investigation this time shows, I suggest the Articuno day shiny rate in abroad would be 5 % as well.


Why was there differences between Articuno day’s shiny rate and Zapdos Day’s?

Apparently, there is no rational reason to set this differences of each probability before the first Articuno day
Therefore, the Zapdos day rate was increased after the Articuno day in abroad which rate was considered too low.

It must be discussed whether the rate of the Articuno day in Japan should be changed or not among the management team.
They must think that the rate in abroad was too low, however, the postponement was because of the safety. It is not good to change the rate in Japan which make t\inequality between Japan and others.

Finally, it is suggested that they decided to set the shiny rate 5 % as well instead of the period was set to be 6 hours considering the terrible heat in Japan.

The probability that can see the shiny more than once

If the probability of shiny were 5%, the probability that can see shiny one per a raid would be like this;

The probability that can see it on the free pass at fifth time would be about 22.6%.

In order to see it at the half probability, 14 times raid is needed.

In order to see it more than 90%, 45 times raids is needed.
But still one person in ten cannot see it.

Fractional Chart per 1 time is here







トーク マイページ ユーザー設定 ログアウト