Broken Lottery, Please Rollback

[UkcsAlias]

Depends on the bug. If the bug is plot related, then only roll the one that has Zero as winner. But, If the bugs really because of the flaw in the RNG (thus affecting all the participants), then it would only fair to be rerolled for everyone. Otherwise that would be unfair for participants that didn't win because of the flawed code, and only fair to the dude that happens to win.

Depends on the RNG flaw.

If the flaw was simply that 0 was always considered a participant. Then even if a 0 could be rolled, it was a fair distribution for whoever won. With 4 participants, everyone would have gotten a 20% chance, and 20% chance on the 0 (which would only happen in 1/5 cases). Which is most likely the cause and easily allows a reroll for only those affected.

The other RNG flaw would usualy mean a poor RNG implementation, and favoring certain numbers. Sure, that flaw could get numbers like 0, 8, 16 etc (or offsets like 5,13,19) to show up more regularly, and in that case would require a complete reroll server wide to counter unfairness here. But any decent security oriented RNG would already be fine (sure, they can have minor advantages, but practicaly this would require millions of participants to show). And those RNGs are most of the time already available by default within software. Its unlikely this was the cause.

[dutiona]

a lot of people still won fairly.

Actually no. Assuming there is 2 bidders on a plot, the bidders do not have 50% chance of winning but only 33%, which is not what is displayed nor is fair. And it goes down the more bidder you have.

My FC lost a large plot because the last 30min 2 bidders came and one won. I suffered from a 8.33% lost of chance of winning the lottery, which could have made the difference. I'll never know.

My point is: the freaking code is not even able to count to one correctly, why would I assume that their RNG is fair ? Also, what RNG algorithms did they use server-side ? What pseudo-random generator did they use in their code, was it coded properly ? Any audit on its fairness ? Did they seed it properly ? Did they publish the distribution of the RNG engine for 1B numbers to prove it's a fair one ? I don't think so. Nothing tells me the system was, is and will be fair. For all I know, maybe the number 4 has more occurrence of happening, we'll never know.

[Krimelio]

is unfair that the same Fc can bid multiple times on the same plot

[severan]

I disagree. Why should you have your house rolled back and pray for RNG to bless you again when you have already won? It's 100 % understandable for houses that got absolutely no winners, that sucks a lot and I'm all for rolling back the lottery and re-drawing it ( without allowing new entries btw to not mess up original chances! ) But the houses that have been claimed should not be taken away from the winners again imo. That's just going to make people furious too.

Because the "winner is 0" means that even those who won, probably wouldn't have won at all, if the 0 wasn't being counted in the first place. It changes the entire math.

[moxphlox]

Oh wow this is interesting. Is there anyone here winning their plots with the last lottery number? If there isn't single winner coming out from the last number, then this major F up

I won with lucky entry 4 out of 4.

[Dustytome]

As per the JP Lodestone (google translated):

Currently, in the lottery sale of housings, when checking the signboard after the result announcement period, the following message is displayed and there are multiple reports that there are no winners.

・ Even though there were applicants, "There were no lottery participants." Is displayed.
・ "The winning number is 0. You have lost." Is displayed and it is displayed as if there are no winners.

We are proceeding with the investigation and response while confirming the application on the database as soon as possible. We apologize for the inconvenience and thank you for your patience.

We apologize for any inconvenience caused to our customers.

So they're aware of the issues and they're looking into the cause. The devs are likely to have a fun weekend ahead of them.

[Cerusia]

Because the "winner is 0" means that even those who won, probably wouldn't have won at all, if the 0 wasn't being counted in the first place. It changes the entire math.

Only if 0 counts as a participant. And some people here have won with the last number of listed participants, so we don't know if that is truly the case.

[Themarvin]

As per the JP Lodestone (google translated):


So they're aware of the issues and they're looking into the cause. The devs are likely to have a fun weekend ahead of them.

Still does not justify an error of this margain, I could understand if could not built on the plot due to some wonky error in the code, but not an error as simple as this technically is, it should never had passed QA to begin with and if testet at all, it should never had passed testing either.

[Greyhawk]

As per the JP Lodestone (google translated):


So they're aware of the issues and they're looking into the cause. The devs are likely to have a fun weekend ahead of them.

And of course, they ignore NA Lodestone. We all know they don't work 8-5 in a NOC. Good job guys.

[Jojoya]

And of course, they ignore NA Lodestone. We all know they don't work 8-5 in a NOC. Good job guys.

It's on NA Lodestone now. Usually takes about an extra half hour for localization to translate and webmasters for the NA/EU Lodestones to get messages up.