@deagol your random number is: 410

Salt: 4WXhmbHqrBSv8VR6oPnA, round: 2645862 (signature b7d91a198a171255c528ff5ffe82ed33a69608f5c9383ce3fef34c124c8b2ecf1f949cffcf34aa21bd713446e2e9fcbd16779da4fd0516204a919698d77543d12820f2d9622364ca4ab00b82e1f55d6590cd09afc26445aea34e43220b15f9f7)

@deagol you asked for a random integer between 1 and 1000, inclusive. Coming up shortly!

Source: GitHub, previous round: 2645860 (latest), offset: 2, selected round: 2645862, salt: 4WXhmbHqrBSv8VR6oPnA.

@FairlyRandom woops sorry didn’t mean to do that

always high!

@deagol Not high enough for me 🤣

@TenShino someone should post a distribution of like 1000 @FairlyRandom requests

@citrinitas your random number is: 126014

Salt: NEtJWQW1N8qOnzoyJXPi, round: 2645854 (signature a7460a937301422dd2d38aa611d3e08632a7ba3cf6b58a0b3fc8a71637a86c2f0b6153bca10a0af053a9b9e0bc00ab54061af6c842c6ada0f1b15574f4d887fe0519df3e0dd0376fcfe829d3760756c4294d78c5ee2c34a7cdf32189319542c4)

@citrinitas you asked for a random integer between 1 and 130978, inclusive. Coming up shortly!

Source: GitHub, previous round: 2645852 (latest), offset: 2, selected round: 2645854, salt: NEtJWQW1N8qOnzoyJXPi.

@FairlyRandom 130978

@citrinitas voted O-O (weight 40)

@deagol voted h6 (weight 432)

@jack voted dxc3 (weight 128955)

@TenShino voted c5 (weight 1551)

Totals:

O-O: 40

h6: 432

dxc3: 128955

c5: 1551

To be chosen from a random draw 1-130978

1-40 = O-O

41-472 = h6

473-129427 = dxc3

129428-130978 = c5

!VOTE c5

Seems our goal is not to win but some other, so this logically follows through on our bold d4 attack. Surely our benevolent dictator appreciates the infallibility of this plan.

And I think I’m done buying lotto tickets.

@deagol I'm sold! :)

!VOTE dxc3

!VOTE Nc6

Is there a way for someone to retire their previous vote? Say something like, VOTE NULL

@deagol

def extract_vote(txt):
    key = '!VOTE'
    pos = txt.find(key)
    if pos == -1:
        raise ValueError()
    pos += 1 + len(key)
    end = pos
    while end < len(txt) and txt[end] not in {' ', '\n'}:
        end += 1
    return txt[pos:end]

I'm not actually checking that the votes are valid chess moves. I've been intending to manually edit if that comes up, but it hasn't yet.

Sure, we can declare VOTE NULL to mean cancelling any votes you've cast

!VOTE Nc6

@jack is this stockfish L3 v. stockfish L3?

@deagol No, I'm picking moves for reasons I will probably wait until the end of the game to reveal.

Also, not to worry, I don't currently intend to fully take over the game, that would be boring for me and everyone else.

@deagol I believe this is a valid move in Lichess, and stockfish might accept with this position (I would).

Would provide a quick out to @citrinitas (PROB=50% resolution instead of YES/NO) and avoid a costly (in terms of subsidies for each move) prolonged game without the need to resign right away, given we already know the result of this particular experiment posed in the stake market description as:

“we lost -- someone bought a bunch of NO and then bid up bad moves. This is an experiment to see if that can be avoided with good market structure.”

Of course, it’s up to @jack to decide.

Now that we know that the experiment doesn't work at the intended goal, I'm wondering if people want to continue this game, or if they'd prefer to start over with a new market design.

I won't drag on the game if others aren't interested in it, but I'm guessing there will probably be interesting shenanigans to watch as the game progresses, and I intend to do my best to sow chaos and keep things interesting. And I expect there's a good chance I'll defer to the rest of the market's move choices when they align with my own evaluation. Of course, other whales could come along too and mess with things too.

@jack Would you be up for helping out

@citrinitas With the subsidies? I think most of us lab rats… sorry, players, would be happy to continue betting on roulette moves, or try guess your vote, or jump through more elaborate hoops you might devise, as long as the cheese and dopa hits keep on coming. :)

Or maybe we can slightly tweak the current system since I think it wasn't a total failure. Is it realistically worth for a manipulator tying well over M$100k of their capital just to win a few hundred?

For example, how about using diminishing returns for the voting shares, like your x YES votes would count as ln(x) or sqrt(x) or some other concave function for the random die roll? Or a sort of capoed "number of seats at the voting table" to be filled from the least shares held first? I'd love to try some of those even with this current game (though likely unfair for others with big investments in the ongoing stakes market).

In any case, I'm having fun even just watching, so keep at it!

@deagol *capped not capoed

@deagol The market subsidies shouldn't be an issue, because the markets have been paying for themselves with the unique trader bonus.

Using sqrt of vote count is a good idea - see quadratic voting. But dunno if we want to make changes like that to the existing game.

Oh, yeah, I'm doing fine on subsidies. The only reason I won't continue indefinitely is keeping up principles re being hard to exploit. I am slightly annoyed they are showing up as losses on my profile but I can complain to staff elsewhere

Definitely not changing the voting mechanism for the existing game, and I don't think quadratic voting would be able to pass Mikhail's exploitability bar. Someone will start voting with 20 accounts, just for the prestige of having done so

@citrinitas much prestige