Single post in DLD #3 [Look out for that tree!]

Forum Index > Core > Announcements > Dev Log Discussion > DLD #3 [Look out for that tree!] >

Garthic [Jamie]

👍 5

Posted 24/Feb/2020 17:26:39 (4 years ago)
Edited 24/Feb/2020 17:27:15 (4 years ago)

long quote

QUOTE originally posted by BlankSmile

[Question]

QUOTE originally posted by Niet

On a typical day before the update, the number of Eggs hatched site-wide would vary between 20k and 35k - with the higher numbers occurring on Niet days usually. Total interactions made would range from 10M to 15M. -snip- Total interactions didn't change at all. It's still 10M to 15M per day. But Eggs hatched... 100k to 120k. That's three to four times as many Eggs hatched.

I'm curious about these stats that were posted in the Dev Log. Where did those numbers come from? Peachi and I keep track of the bonus counters and the data we have based on Ravyne bonus is not even remotely similar to the numbers that were given in the Dev Log. Stats according to Ravyne day: Feb 5 (Niet) - 80k eggs hatched Feb 6 (Shazi) - 56k eggs hatched Feb 7 (None) - 40k+ eggs hatched (bonus reached)* Feb 9 (None) - 63k eggs hatched Feb 12 (Shazi) - 74k eggs hatched Feb 13 (Niet) - 89k eggs hatched Feb 14 (Sei) - 67k eggs hatched Feb 15 (Garthic,Uzumi,Dusky,Sei) - 70k eggs hatched Feb 16 (None) - 49k eggs hatched Feb 17 (Shazi) - 51k+ eggs hatched (bonus reached)* Feb 19 (None) - 54k eggs hatched Feb 20 (None) - 55k eggs hatched Feb 21 (Niet,Dusky) - 90k eggs hatched Feb 22 (Niet) - 89k eggs hatched Feb 23 (Garthic,Shazi,Sei) - 77k eggs hatched *Note that on days where Ravyne bonus was reached, the number of eggs hatched are more than what's shown From this data we can gather that there's not actually all that much of a change in the number of eggs hatched. I'm sure there may be a bit of an increase, but it's definitely not as much as being said in the Dev Log. This is of course, assuming that Ravyne day is 100% accurate in the number of eggs hatched per day. I believe it should also be considered that when new updates to any game is implemented, users are more likely to participate in an effort to test out said update. This is something that can also skew data as this is often only a temporary increase in activity.

Unfortunately, Ravyne can't count all of the eggs due to race conditions (see: Two people doing the same thing at the same time which results in one getting dropped). This is why you have a lower number than us for recent days. THAT SAID. We've realised an issue with the stats. The database doesn't track eggs hatched, we had to reconstruct that from timelines. This has some problems. Shinies count double, but we considered that minor enough to not matter because it's being done across the board - we realised however, that there is a much bigger problem. If something is released to the shelter. Then it stays there for the week. It gets deleted, right? That means no timeline. Which means it's not counted in the stats we generated because, as I said, it uses timelines. Timelines get deleted when a Pokémon gets deleted. That's why you got higher numbers than us for the older dates. The stats generated were incorrect. This means that the stats that we generated for the past week were accurate, because everything is still there. (sans shinies counting double, ofc!) Everything older than a week - Ravyne is definitely much more accurate, but it's also still wrong. The 30k discrepancy for the 90k to 120k shows as much. That's much too big to be accurate. This means we need to amend what we stated about the stats. The general upward trend of them still stands to the positive effects that can be seen, but it's not something we can use categorically as proof due to the inaccuracies presented in the data. We are amending the statement made in the Dev Log regarding the scale of the increase and data utilised because we want to be clear that we made a mistake there. Thank you bringing this to our attention. This was something that we accidentally overlooked. While I'm not happy that we wound up providing incorrect information, I am glad that it was found out relatively quickly so that we can correct the information as fast as possible. I'm really sorry about this, we messed up here. Regardless of whether or not the trend is still somewhat showing, what we provided was just wrong.

The titles just rhyme for the most part. References where applicable. In this case, we're referencing "George of the Jungle".

Loading...

Single post in DLD #3 [Look out for that tree!]

long quote

QUOTE originally posted by BlankSmile

QUOTE originally posted by Niet