Page 3 of 28

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 20th, '22, 22:13
by Amura
Thank you everybody for your support! :qstar:

OK, just gonna share some results already, because seeing them is quite cool.
On average we are getting one donut every 5,5 posts.

But there seems to be a significative deviation.
AutobotDen and me are slightly under 5 posts, blackbriar over 6 and Akili Li over 7!



PS: The event is probably over soon, I guess the caravan will arrive shortly after the event is over?

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 21st, '22, 21:11
by Akili Li
Don't we have another week still on the Winter '22 event? I thought it ran through January 28th, maybe I mistook it.

Anyhow, interesting to see that....
speculating on possibilities:
1) People remember to post when reminded by a frozen donut pop-up reward which will skew it shorter
2) There's an additional percentage chance bonus/reward after X amount of time, which means lurkers will see that impact them at a significantly higher percentage of their overall posting, and so have a shorter post-to-donut ratio
3) We don't have stats over a large enough, long enough time period and it's truly just random amounts that haven't smoothed out yet to show the true average
4) I'm simply unlucky
5) There's a reduction in rate-of-return for frozen donuts after a certain number of posts, so babblers like me will see their ratios become lower over time
6) Probably other possibilities I haven't thought of yet

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 21st, '22, 21:42
by Amura
*checks announcement*
It says January 23th.
So we have until Sunday.
And I'm assuming the Caravan will probably arrive right after it...



I'm quite inclined to think it's mostly reason (3) although reason (1) may have an impact indeed.
Or maybe it's just all (4) LOL

But I don't think there is any complicated formula for this, because it would make the code way more complex - and what's the point?
I would bet starkad kept it as simple as he could.

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 21st, '22, 21:51
by Akili Li
I don't know enough about coding to tell what's actually hard and what's relatively simple. You already have to have it check one thing (whether the caravan is present or not, since that changes the percentages and the reward), so how hard would it be to have it check another thing (when the poster's last post was)?
I know that's already tracked because if you look at profiles you can see stats like that. So the information is already there, right?
But...
How hard is it?
That, that's a mystery...

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 21st, '22, 22:45
by Amura
I'm no expert but I've done a little bit of coding, and it's a good practice keeping the code as clean as possible.
Considering we only have one person coding at least he does not have someone else messing with his code, and the things you mention would not make the code particulary messy (it's just a check which changes an output), so yeah, I think they would not be very hard for starkad.

But from my point of view those checks do not add any value to the code, so unless Firn asked specifically for them starkad would not consider them.
Would Firn ask for such a thing? That would be the mistery.


PS: There are more checks implemented. For example, you don't get any donuts in the Spam forum.
And technically creating a thread is not the same as posting, so there could be a different code there too.

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 21st, '22, 22:52
by Akili Li
Ooooh, good points, I hadn't considered those things!

Also, if we only have until Sunday I want to make sure to pick up more extras.
I was expecting to see lots and lots of the Winter '22 items in the market already but there really hasn't been much, which is surprising to me and tells me I should be getting more for the people who are having trouble.
I should pull out that egg timer again.

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 22nd, '22, 02:54
by blackbriar
It's usually very difficult to do scaled increases like that and I've never really seen it in posting based sites. We only just started looking at stats so the data isn't complete.

I think at one point we also speculated that longer posts could earn higher if not more frequent rewards. I don't know how we would test for that though...

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 22nd, '22, 09:54
by Amura
I don't think we can properly test that.
I mean, we would need some people making only large posts and other ones making only short posts. But who would only do long or short posts?

Also: What is long or short? How many characters or words or whatever?


I keep thinking they went the un-complicated way.
Because they just want to promote posting, but encouraging only people who make long posts would not add much activity to the site.
This again would make the code more complex (although it would also be doable) for no gain.

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 22nd, '22, 19:01
by blackbriar
Well short would be spam sized posts most likely so that actual conversation would be rewarded. I'm pretty sure that the code is nothing complicated though.

Re: Frosted donuts statstics - Let's do it! :D

Posted: Jan 22nd, '22, 23:24
by Amura
No, not hard.
As I said before, I just think that if you start adding so many "ifs" then it gets messy. And you only want messy code for a good reason.
I doubt whether they would consider all those cases a reason good enough - but I don't think we'll ever find out.


Well, I currently have 11124 posts and 378 donuts.
I'm betting this is my last update, cause I'm expecting the Caravan to arrive soon.
Maybe even tomorrow, who knows!