Changing the BitBet resolution process.

Remember that anatomy of a disaster thing coupla years ago ? Well... it happened again.

nanashi_ Could I make an inquiry about BitBet's Freeseas Bet? The first three of the last four bets were "No" bets, so they should not be refunded (though it seems it's too late...).
mircea_popescu link ?

nanashi_ https://bitbet.us/bet/1155/freeseas-inc-nasdaq-free-will-trade-at-least-0-10/
BitBet - FreeSeas INC, NASDAQ:FREE, will trade at least 0.10 USD per share before 1st July 2015. :: 3.26 B (44%) on Yes, 4.19 B (56%) on No | closed 10 hours 21 minutes ago ... ( http://bit.ly/1J62PZ7 )
nanashi_ Should I speak at "mircea_popescu tab" (which is (I think) more private?)
mircea_popescu i do not know you. we can't talk privately for this reason.

nanashi_ ok.
nanashi_ Am I talking to a right person (about Freeseas bet inquiry), mircea_popescu?
mircea_popescu nanashi_ havin' it looked into, a sec.

nanashi_ I am "143Cz" better. I know I lost the bet, but I'm surprised the later three bets (which were all "No") were refunded (which they should not).
mircea_popescu myeah. mod dropped teh ball, an' as a result im out a coupla bitcoins now.
mircea_popescu nanashi_ thanks for pointing it out.

nanashi_ you're welcome...

It actually was 6.5 BTC, or roughly speaking seven times BitBet's income last month. Which sucks.

There are two major problems contributing here. One of them is that there can not in fact be such a thing as errorless process - and I regard a two year interval without this sort of problem a major achievement, rather than any sort of failure. The other is that in general you can not expect any activity to cost less than the maximum liability to which it exposes the practitioner. For instance, if a doctor runs a 1% risk of having to pay a million dollar malpractice settlement, there will be no doctor visits costing less than 10k no matter what.

Leaving aside that moderator error is in principle an unbounded liability, using the aforementioned rule of thumb would promise that the previous 24 months would see a moderation cost of no less than a quarter bitcoin, which would in a majority of them represent a significant portion of BitBet's net profit. This then implies that the current approach is not really sustainable, or perhaps is marginally sustainable - but in any case should be reviewed and revised.

In which spirit, the following changes will be introduced in the resolution process :

1. Bet closes, consequently coming up for resolution.
2. Moderator resolves the bet, indicating through this what sides will win and what sums.
3. The bet is held for two days, during which any interested party may make whatever protestations, in the usual venue.
4. The bet is paid out, after which no further complaints of any kind and for any reason will be entertained.

These changes have the disadvantage that they extend the interval a winning bettor has to wait for his winnings by two days, and also the disadvantage that bettors are held to review results of the bets they participate in during a fixed timeframe. While I can appreciate that disadvantages suck, I see at the present moment no alternative approach that'll offer the cost control benefits of this one.

Feel free to contribute your cents below. Otherwise this change is set to become actual BitBet policy by the end of the month (FAQ will be edited to reflect this once it is the case).