I realized you were trying to set quorum with the terrible proposal, but
what was the quorum bug itself? I thought it was a side-effect of it
working as intended (a feature not a bug).
On Mon, 25 Sep 2017, Aris Merchant wrote:
> Thank you for your honesty.
>
> -Aris
>
> On Mon, Sep 25, 2017 at 11:45 PM, Alex Smith <ais...@alumni.bham.ac.uk> wrote:
> > On Mon, 2017-09-25 at 23:24 -0700, Aris Merchant wrote:
> >> Speaking of mysterious secrets: ais523, can you tell us why you
> >> submitted the repeal all rules proposal around the time of your
> >> junta?
> >> Sorry if I already asked and forgot about it.
> >
> > First of all, the actual text of the proposal was irrelevant, so I
> > decided to go for something amusing. (Because Agora is Agora, I first
> > triplechecked that it wouldn't do anything if it passed; there were a
> > number of intentional mistakes in it.)
> >
> > The original intention when I filed the proposal was to make it
> > possible to manipulate quorum with it as an emergency counterscam (you
> > couldn't force through a proposal if quorum was high enough). A
> > proposal like that tends to attract a lot of votes, and it would be the
> > next proposal to go through if things went badly wrong during the scam
> > itself. Of course, this wouldn't exactly be a bulletproof counterscam –
> > the votes on it could be retracted by the scamster – but having bribed
> > the Assessor, it seemed that having control over two important parts of
> > the gamestate (the timing of Assessing and the nature of the first
> > proposal to go through with the scam public) would help put me in the
> > best possible position minimize any potential damage. (I wasn't
> > planning to make it publicly known that quorum was relevant until
> > absolutely necessary, and was hoping any potential scamster would miss
> > this.)
> >
> > As it happens, the proposal was distributed much earlier than I'd
> > expected (the Terrible nature of the proposal, in addition to the fact
> > that I'd stated that it was important to the scam and the fact that the
> > Promotor had been stalling other obvious scam proposals at the time,
> > made me think it would be stalled, but it wasn't). So I adapted, and
> > used it as a vessel on which to do quorum manipulation; because it
> > would necessarily attract several votes, and I could retract them, it
> > let me accurately set quorum to any specific value I wanted, which was
> > particularly handy in making sure that the scam worked. Even better, it
> > let me disguise the reason why I was retracting the votes (I could make
> > it look like I was panicking about the proposal potentially actually
> > passing, rather than doing something with no apparent purpose,
> > increasing the chance that people realised that I was trying to set
> > quorum). This was likely a much better plan than my original one (which
> > was kind-of half-thought-out).
> >
> > Incidentally, the specific quorum bug exploited was an intentional bug
> > that I slipped into the quorum rule at the time when I wrote it. Agora
> > was going through a lull, and finding it very hard to make quorum (back
> > then it was based on the number of players, not on the number of voting
> > players), so it was easy to slip in a buggy quorum rule as we badly
> > needed one. This obeyed my standard Agoran practice for proposals
> > ("every proposal that isn't an obvious scam should improve the game;
> > it's just that it's allowed to contain a minor scam at the same time"),
> > incidentally, it's fairly rare that I slip a scam into a proposal over
> > here at Agora, because if I did it too often everybody would vote down
> > my proposals on principle (i.e. "what happened to me over at BlogNomic"
> > ). I decided that passing a proposal with very few votes is the sort of
> > thing that I'd be much more likely to do than anyone else, and so this
> > disproportionately benefitted me (although at the time, I was
> > originally planning to let quorum reduce "naturally" as hardly anyone
> > was playing; this method with retracting votes wouldn't have worked in
> > the ruleset at the time).
> >
> > --
> > ais523
>
