Trainers,
over the past few weeks, we’ve been collecting data (time and date) on Evolution Item drops, in search for an answer to a simple question: are evolution item drops influenced by time of day?
The theory we tested was that “a correlation exists between the current time of day (local) and Evolution item drops”. Our research showed no confirmation of that, quite the contrary – evolution item drops are perfectly random and well distributed.
You can view the collected data (1300-ish entries) here, but do note that the data was not collected in a controlled fashion. Again, even when we manually tried to clean up the data, it still showed the same random results as previously.
More information
There was a number of other interesting questions that were asked (will same PokéStop drop always on the same time, is the Trainer ID included in any calculation, etc…), but this research does not answer them.
We are inclined to believe that none of those questions will ever have affirmative answers, as seemingly, the evolution items drops are completely random.
Sure, one can say that our research methods were messy, not scientific enough or even argue that we’re completely wrong.
There is a number of trainers who did report drops happening at the same time, but the quantity of contradicting reports made it quite obvious that these are just coincidences, and not proof of anything concrete.
Moreover, a redditor gave an excellent summary (Could time be related to X?) of how and why it is unlikely that Niantic bases drops (or anything) on time of day:
1) If goal is to have uniform distribution and equal chance, then simple random function will be superior in results and less bug prone.
2) If goal is to have uneven, but fair distribution that for example allows players with low spawns/pokestop density, then they have demonstrated to do it otherwise (with daily/weekly bonuses for example).
3) If players figure it out, then it will be used and abused feature. That is very risky because it would reward people in-the-know a lot, while being COMPLETELY invisible to others (Unlike nests for example which are un-named, but highly visible mechanic). So why do it this way to begin with?
4) It would suck for people who are off-schedulle for these timers or mechanisms.
5) Timezones exist. Does each timezone get its own timer? Or do people in off-timezone suffer? Again, source of developper headache and player frustration.
6) There are arguments which support it with alternating source – for example, random seed being same, so results being same. That one argument has some merit, but will not help in anyway to predict results. It would also be notably difficult to pull off because any lag would wreck ability to send requests to server to achieve random generator initialization in exact same MILLISECOND.
7) There are deterministic features that look random in main series games (like shinyness). TRANSFERRING them to pogo makes some sense, but that does not mean that
In other words, the null hypothesis did not stand – there is no correlation between time of day and evolution item drops.
