[A] Removing the Page then Post Roll

[slimycrow]

It would only be weeding it out if they Rolled a page, and the fifteen people on that page got it. Every page has an equal chance of getting chosen on each page roll.

The only thing that would affect the math is people posting that are not flaffle related, but the flaffle threads usually ask people not to do that.

But the math doesn't lie.  

[Revolutionary Roniel]

Slimycrow... check the previous flaffle threads where this has occurred. There are always the first and last pages where there are not 15 posts of entrants per page. Therefore, the odds for those people would be better.

Say there are 3 pages of people posting. If there are 7 posts of rules and descriptions and stuff on the first page and then 8 posts of people entering, 15 posts of people entering on the second page, and 1 post of someone entering on the third page... The people on the second page would have a much lower chance of getting picked, the people on the first page would have a better chance, and the person on the third page would have an awesome chance.  

[USS_Galactica]

Revolutionary Roniel

Slimycrow... check the previous flaffle threads where this has occurred. There are always the first and last pages where there are not 15 posts of entrants per page. Therefore, the odds for those people would be better.

Say there are 3 pages of people posting. If there are 7 posts of rules and descriptions and stuff on the first page and then 8 posts of people entering, 15 posts of people entering on the second page, and 1 post of someone entering on the third page... The people on the second page would have a much lower chance of getting picked, the people on the first page would have a better chance, and the person on the third page would have an awesome chance.

Yeah. This happened with the last flaffle. There was only one entry on page one, and it did get rolled.  

[mindsend]

Statistically, it actually doesn't Lower your chances of winning, nor does it increase it. If 200 people enter for one soquili, you have a 1 in 200 chance no matter if its rolled as a huge list, of if its rolled page then post.

It just means no one has to actually spend hours making a list, when the colorist running it could be working on Soquili instead, and people have a higher chance of being left out by accident.

Now, if its, say, 3 soquili, a new page should be rolled for each one, and it should NOT be "One page gets three chances".


I'm not very good at math, understandably, so if someone can prove that it reduces everyone's chance at winning compared to one roll=one pony, however its basic enough math that I think its all psychological; Instead of "BAM YOU LOSE" Its "BWHAHAHA YOU LOSE BUT 15 PEOPLE HAVE A CHANCE TO WIN", When in reality, its the same exact chance.  

[LOLTERNATIVE]

I have to agree with what Roni said. Rolling by page and then by post isn't fair. everyone has the same 1/x (pages) chance of being rolled but not everyone has the same 1/15 chance of being chosen.

Like Roni said, not EVERY page has 15 posts exactly. So while the page roll possibility is the same, the decision of the winner isn't. Just try and get on the first and last pages and your chances higher.

Example

15 pages.
page 1 = 6 posts = 1/6 chances for being chosen as a winner.
pages 2 - 14 = 15 posts = 1 / 15 chance.
page 15 = 8 posts = 1/8 chance.

Everyone has the same chance for their PAGE being picked but not their post.

The Plush event had a large list and it seemed to be handled relatively well and that had multiple entries per person.

tl;dr; 6 posts on page one regarding tickets, 15 posts on page 2, people on page one have a 1/6 chance of being WINNERS people on page 15 have 1/15.
THE PAGE ROLL IS THE SAME CHANCE BUT NOT WINNING.

It's basically like, aim for first and last page and you have a better chance if your page is rolled. the chances aren't lowered for pages but chances are RAISED for those who get on pages that don't have 15 posts.  

[LOLTERNATIVE]

I don't think it was implied that the staff weren't doing enough I just think it's a general concern that the odds are different when it comes to the sort of roll. It doesn't matter that the odds mean nothing if the page isn't rolled, the odds are still different and that's that.
1/6 =/= 1/15

and so what if they post early? does that mean people who have jobs or other life commitments / live in different timezones should be punished?

i'd love to be able to post first but sadly most of the stuff starts at 5am for me and since i wake up at 4:30 am weekdays i don't want to wake up early on Saturday. I shouldn't have to, I should just be able to post wherever I want and not be upset for not getting a post on the first or last page.

If there was a way to do these rolls where it was fair, i'd be all for it. but i can't see that happening, sadly. unless you counted all the applicants, created a new thread and made pages with even numbers which again is too much effort and there's a chance of getting a number that won't divide right. o_O;

i know that updating lists is a hassle, believe me. but in such a big shop i think the customer's concerns should matter. :3

in the end, though, it's up to the staff and that's just my opinion.  

[Kamiki]

The way to "fix" the whole "people on the first and last page have a slightly higher chance to win" is simply roll 1-15 for ANY page that's rolled. If it's an invalid post (ie, the introductory post), then roll again.

Ta Dah - list is back to being equally fair for everyone heart  

[saedusk]

I disagree with the rolling 1-15 regardless solution, since the invalid posts are still ineligible to win, which effectively returns the odds to whatever they were in the first place. You may have a 1/15 chance of being rolled, but there's also, say, a 1/6 chance that an invalid post is rolled. So, to me, it still seems like there would be a 1/9 chance of winning, not necessarily of being rolled.
And depending on how many blank spots there are on the page, it may lend itself to multiple re-rolls, which takes more time.

My own solution idea is this:

By deleting all the extra posts on the front page, you can eliminate it as an option, and then it's possible to have a simple chart to give to the colorists to help them determine lists and winners by rolling only numbers.

Ex: There are 10 pages. Page 1 has no one, pages 2-9 are full, page 10 has 6 posts.

The Chart
Page 2 = 1-15
Page 3 = 16-30
Page 4 = 31-45
Page 5 = 46-60
Page 6 = 61-75
Page 7 = 76-90
Page 8 = 91-105
Page 9 = 106-120
Page 10 = 121-135

So, the colorist could look at the chart and know that they have to automatically roll 1-120, since pages 2-9 are full. And then they add the 6 posts from page 10, so the final roll is 1-126.

Then, the colorist rolls a 25, for example. They look back at the chart, see that 25 is on page 3, go to page 3 and count down from the top: 16... 17... 18... etc. Number 25 is the 10th post on page 3.

It is more work than simply rolling a page and then rolling a post number. However, the provided list can be as long as anyone wants to make it and can be reused over and over. This method eliminates the need for the colorists to make their own lists. They just have to do a bit of math and counting. :3

Since I was thinking about this, I figured I'd go ahead and post it. xD Sorry if it seems confusing.  

[ATh e a r t]

Hey you guys. There's a reason why I didn't use math in this. I've taken upper division at my university. YES. The math is understood, but as I said, this isn't about the MATH, it's about the weeding out.

I've pointed out SEVERAL times that there is a difference between the first and second roll DESPITE the odds being the 'same'. I'm not talking about ODDS. I'm talking about the ROLLS themselves and WHAT THEY DO.

I will provide another example that I have previously used.Note again, that this argument is beyond the math reasoning. Think a little deeper than just the math, kay?

Say that you are at a fair. There's a tub of tickets that you can pick from to win a prize. The man running it says that you can ONLY choose through a tube, meaning that you could only pick from A CERTAIN AREA. You can't shake the tub for it to be completely random. And that winning ticket is at the very bottom, in a corner, FAR from the tube. Can't ever get to that ticket, right? Even if you dig in with the tube, you can't shake it to get around all of the other tickets. Ho will you get to that elusive ticket? You can't, meaning that it's made sure that one ticket has no chance of being picked.

Tub = Flaffle
Tickets = People
Tube = Page roll
Final chosen ticket = Winner

Yes, the odds can be calculated the same, but the fact is, just because the odds are the same, doesn't mean that every single person got their name in the raffle for a Soq. I've highlighted different times that what the first roll does is different from what the second roll does. They have different meanings.

YES THEY ARE FOR THE SAME RAFFLE, BUT THEY DO NOT DO THE SAME THING.

I feel horribly ignored in people reading the first post... I'm not grawr though, promise D: !

Edit; I do agree with what sae said, because that eliminates the weeding out. It's not about the odds, but about the actual fact that everyone should get their named rolled for a Soq. The way Saie pointed it out puts everyone one into one single poll, not constant individual pools.  

[moineau bavarde]

I disagree, I think the odds are much more important... and they aren't even. Say 20 people enter:
1 on first page (1/3 odds of page getting picked x 1/1 odds of getting rolled = 1/3 odds of getting picked)
15 on second page (1/3 odds of page getting picked x 1/15 odds of getting rolled = 1/45 odds of getting picked)
4 on third page (1/3 odds of page getting picked, 1/4 odds of getting rolled = 1/12 odds of getting picked)

1/3, 1/45 and 1/12 are VERY different odds. For any given raffle, the odds for winning can be 15 times higher for certain individuals. That isn't fair at all.

Rolling for page, then rolling 1-15 regardless of how many posts were on the page would work provided that if the "winning" post was illegible, then a new page was rolled, and not a simple post on the old page.

Buuut, really, Sae's solution solves two problems at once because:
-Everyone has equal odds
-The colorists only have to roll once per winner, not counting rerolls, instead of twice per winner (page, post)  

[Graceangel]

I don't think the page/post method should be removed from EVERYTHING. I think things like... referral contests where you have multiple people posting multiple times for a week to a month.. that's a good method. for a single raffle where each person has only posted once; then they shouldn't just roll for a page. that's really not a fair way to choose winners. You should have had a one in 500 chance rather than your chance to win based on the randomness of which page you happened to have been lucky enough to post on.

edit: i think saedusk's suggestion is awesome! could really take a lot of the trouble out of things. so then the only thing that needs to be done is lists.. and as already pointed out on the first page: there are by far more than just a few willing volunteers who'd be more than happy to help construct accurate lists.  

[Sirenz]

Update: The page post rolling is retired from flaffles and raffles and we are doing a list instead by asking for a volunteer at the beginning of the event to help or someone on staff could keep the list.

The only thing that will be used for page post rolling is free thread raffles where an entire thread is rolled when we have a holiday or other type of event because its faster.

Thanks all!!  

[Uta]

All right!

While there really isn't statistically much of a difference between use of either method, the staff have all decided to do away with the page-list method and continue to randomize from a full list. That way everyone has their proper chance and shot for a win, without feeling cut out by choosing a particular page first.