The strange case of the six hundred dollars and other stories.

Wednesday, 23 September, Year 7 d.Tr. | Author: Mircea Popescu

Consider the following two propositions :

1. Bitcoin over $600 before November.

2. BTC to top $700 before November.

Knowing the contents are substantially the samei, and knowing that there's apparently a total of 115.01 BTC in liquidity available to finance both of these propositions (distributed on Yes/No however it may distribute for each), what proportion of this liquidity would you expect each proposition will attract ?

I mean, network effects and all, but are you prepared for a 112.8 to 2.21 split ? Fifty-one to one ?

I wasn't, honestly. But let's forget this for a moment, and consider the following matter : given that the "Bitcoin over 6-700 in a month" attracted over 100 BTC in interest, what sum in BTC would you expect to have been wagered on "Bitcoin $250 or above on September 30th" ? One month bet, also using Bitstamp as the reference.

It's a lot closer, so really, more BTC, right ? Because BTC doesn't like being locked up.


It's a lot more balanced, so really, more BTC, right ? Because long/short disparity kills liquidity.


Even given all those counterintuitive nopes, unless you've actually looked at it you won't on your own guess that it failed to gather as little as 1 BTC's worth of interest, which is to say 1% of the longshots. Or for that matter, half a BTC. Or, for that matter, a quarter of a BTC. That's right : we're talking a 575th over here, so small a fraction it's not even a measuring error. The proposition might as well never have happened, as far as the betting public is concerned.

How come ?

Seriously, how the fuck come ? Bitbet is the sort of venue where one can throw a Bitcoin away on the craziest of obscure propositions and you'll be covered. Every timeii, at fair market value, you'll be covered. So what happened here ?

Well... nobody threw the Bitcoin away, that's what happened here!

The bet that did well ? Submitter put half a BTC down. The bet that did poorly ? Submitter put three Bitcents down. The bet that did shockingly horrible ? Submitter put one Bitcent down, and no doubt would have put nothing had I not changed the rules to disallow that last year.

What I suspect is going on here is simply this : the original investment by the author is perceived by the market as a honest signal to the legitimacy of interest in the question being proposed. Consequently, well financed bets receive interest, and disputes as to the correct pricing within the market come to a resolution through the intended process : resource exhaustion in the shape of submitted liquidity.

Poorly financed bets on the other hand do not receive interest, and disputes as to the correct pricing come to a resolution through the usual process : the Jewish handwave. Aaaaah! The original BitBet split is chiseled away into shape, and that's how you end up with 10-20 Bitcent propositions.

The morals of this story as presented would be that :

  • The meaning of words does not matter at all, the only point of any interest in this post-post world we live in being "how much money says so". This situation has been intuited and hinted at on the only remaining venue for intellectual discourse, but it is rare to see such drastic proof of otherwise nebulous concepts (which "nebulous" strictly means "things that are true and we don't want to know about").
  • The actual utility of the guaranteed BitBet seeding is very low : it will take a bet from 0.01 to 0.2 maybe, but that's about all it'll do. There's no leverage in guaranteed behaviours. A point which again has been hinted at - guess why "aid" sent to Africa does little of anything ? Guess why "helping" US schoolchildren learn easier makes them dumber and lazier ?
  • Allowing the poor a voice is a bad idea. This may seem counterintuitive, given the intellectual fashion du jour is to pretend the contrary. Nevertheless, it's not only factual - but not really all that counterintuitive, either. At any rate not more counterintuitive than the proposition that species evolve, or that there's no Russell teacup watching over human evolution from outer space.

I am not going to take any measures immediately, here. But I can't say I'm not mulling some over.

  1. A. They both use Bitstamp as a reference point ;
    B. They both close on the same date ;
    C. 1 gives one week lockout, 2 gives 3 weeks' lockout ;
    D. 1 gives 22k weight as endpoint ; 2. gives 1k weight as an end point.

    One could perhaps argue that a longer lockout encourages more early bets - but then again the contrary case can also be made just as well. One could similarly argue that a higher endpoint weight discourages early bets (the contrary argument can perhaps be made but it seems unpersuasive).

    As it happens these arguments are balancing each other out, as C favours 1 and D favours 2. Whether on the balance there's anything significant left seems to me altogether improbable. Thus I'm going with "substantially the same". []

  2. Two examples off the top of my head : recently I threw a BTC at an obscure scamcoin (some sort of rehashed Ripple or whatever). It was covered 10:1 within minutes. (Some substantial differences among the public as to the proper odds seem to also subsist, so the volume ballooned way over 100 BTC while the actual line was chiseled much closer to 2:1 - a fulminant phenomenon that didn't escape making the news. Because yeah, this sort of thing is exciting.)

    A coupla years ago I threw a BTC on an (at the time) entirely obscure Romanian tennis player for the purely unrelated reason that she was a petite girl with an impressive rack that loved playing tennis and decided to cut it off. I opined (unasked) that this is a stupid idea at the time (she was 17), bringing as argument among other notions the idea that a tennis carreer in the tree is not worth a grindstone to the tit. Coupla years later when she came out of it in superb form, I felt I kinda owe the world reparations, and so I threw a Bitcoin away, by betting it on a really long if flattering shot : that she may win the Roland Garros. She almost did, but not really.

    All that aside : I was covered, ~6 to 1, which seems rather fair to me now and seemed rather fair to me then. []

Category: S.BBET
Comments feed : RSS 2.0. Leave your own comment below, or send a trackback.

2 Responses

  1. You were covered on the tennis bet because the odds were long all the way up until the bet was closed (pre-match). There are much different dynamics with a bet like this where you can have a much higher confidence closer to the deadline. An early bet is not going to pay out 6-1.

  2. Mircea Popescu`s avatar
    Mircea Popescu 
    Thursday, 24 September 2015

    I kinda deliberately picked two very unlike examples. Take a third if you wish, I put 1 on the 700 bet, taking payout from 11:1 to something like 10:1, it was readjusted to 11:1 within... uh look at that, 12 minutes.

    Obviously any individual example I might pick is open to the "hey, this is an individual example which has its own individual characteristics" discussion. Of course it is. I'm not trying to say it isn't, I'm just saying : I've yet to encounter a situation where I wasn't covered to my satisfaction. Obscure bet, long odds, short odds, whatever color you want to put on it.

    Heck, I've put 1k BTC down blind, and it was covered satisfactorily.

Add your cents! »
    If this is your first comment, it will wait to be approved. This usually takes a few hours. Subsequent comments are not delayed.