For lack of a nail

Tuesday, 16 July, Year 5 d.Tr. | Author: Mircea Popescu

Motto :
For want of a nail the shoe was lost,
For want of a shoe the horse was lost,
For want of a horse the message was lost,
For want of a message the battle was lost,
With loss of that battle the kingdom was lost,
And all for the want of a nail in the horseshoe.

I'll do what's not usually done anywhere elsei, which is to say make useful, important and meaningful comments.

So, there's this guy who's done the right things over the course of time, and consequently enjoys my respect. He came up with an advance on the by now classical Bitcoin betting site, which craftily employed the mechanics of the MPOE bond for supplying the needed capital. This design is certainly better than what else we've seen, but it suffers from one fundamental flaw. To quote,

Jun 21 05:44:07 mircea_popescu the genius idea is actually a very bad idea, but anyway.
Jun 21 05:44:25 Bugpowder why is it bad?
Jun 21 05:44:38 Bugpowder it certainly doesn't maximize his profit
Jun 21 05:44:50 mircea_popescu it will wreck the site.
Jun 21 05:44:50 Bugpowder but I think it is optimal for other investors
Jun 21 05:44:53 Bugpowder how so?
Jun 21 05:45:08 mircea_popescu the casino needs money one time and one time only : when it just lost.
Jun 21 05:45:17 mircea_popescu this model ensures he has money one time and one time only : when it just won.
Jun 21 05:45:22 mircea_popescu total mismatch.
Jun 21 05:45:29 Bugpowder I dont think so

There's precisely one fix for this problem : somehow attract large investors who will back you over the long term. For bitter or for worse. So, because I like the guy but much more importantly because I promote the soundii I made a large investment early on, and supported his project. I still think it's a great project, for that matter.

Soon enough the following problem became apparent : his very inflexible capital allocation scheme whereby 1% of the deposited funds were being used on each bet fundamentally meant that if I were to deposit 100`000 BTC to back him up I'd be having a 1`000 wallet at risk on the mechanics of the site and a 99`000 wallet at risk on his head. Obviously the price of risk on the 1k was significantly - perhaps a degree of magnitude - less than the price of our guy's CP, no matter how much I like or dislike him. Moreover, the cost of book for keeping that 100k probably exceeded any profit he'd be likely to receive off the entire discussion (at the time 1% of 1% of gross, or 0.01%).

I pointed this out to the guy, he did comprehend what the problem was, I do believe he intended to implement a solution, but the unfortunate part is that the codebase as he had it didn't lend itself too well to making the sum-at-risk aribitrarily settable. This is a nail, a trifle, a nothing. On top of it the responsibility of coding this correctly, and the endless stream of trifling demands put on him by the "I'm your biggest fan" troop of nothings and nobodies resulted in him not coming up with a solution for about a week or so. And so I withdrew my investment, not for any political reasons, but simply economical : it just didn't make sense. And that was a horseshoe.

Soon enough the site found its first loss. To quote,

The whale came back, and won another whole bunch.

I was watching the betting patterns as it happened, and it was somewhat suspicious.

The pattern was to bet 0.01 until a loss, and then bet big, knowing that the loss made the next roll more likely to be a win.

Then a series of 5 losses in a row came up, and he bet 0.01 on all of them.

Series like that are where the previous gains are meant to be recovered, but if you bet tiny amounts through the losing sequences, you end up winning.

This could be nothing, and I may be overreacting, but it's suspicious enough that I wanted to bring it to investors' attention and let them decide what to do with their investment.

To that end I have set the "max profit" to 0, so not bets are currently being processed. I didn't want to have to do that, but I would rather be safe than sorry. No accounts are frozen, you can still divest, invest, deposit, withdraw, as before. Just no bets are currently being processed until I can look at the data.

I'm sorry to have to do that. It may look like I'm just upset that the house lost for a change, but it's not that. I'm worried the server seeds have somehow been compromised, and that they are being used to win.

I don't know how I can tell if that is the case, or what to do next.

Lots of people in the chat are telling me I've done the wrong thing.

Let the shitstorm commence...

This was statistically unavoidable, obviously.iii But bereft of the support of the almighty investor, that one guy who has the experience and the capital allowing him to say "forget about it" like he means it, our guy went on tilt. And the tilt was the horse.

Tilt means you do stupid shit that's incredible in retrospective. Again, to quote :

07:41:03 (1) last night 'celeste' asked me to cash him out 1300 BTC
07:41:19 (1) I panicked, shut the site down, reduced max bet, checked stats
07:41:31 (1) it all looked ok, so I put it back up and sent him 1300 BTC
07:41:37 (1) but didn't debit his account the 1300
07:41:42 (1) so the 1300 he lost last night was mine
07:41:51 (9006) you kidding me? :|
07:41:55 (1) he really lost it, but it wasn't his
07:42:00 (1) no - that's what happened
07:42:12 (9006) so the 1300 he lost was actually the sites?
07:42:17 (9006) or your own private 1300?
07:42:18 (1) he denies playing it at all, says someone accessed his account
07:42:24 (9006) Oh the lies.
07:42:34 (2619) and, there is no way...
07:42:35 (1) well, it's 1300 that was in his account that I should have subtracted when cashing him out
07:42:42 (1) it's my fault, so I guess it comes out of my pocket

Now you see, that's indisputably the battle. It did start with a nail, a trifle, a nothing at all, but it did not end there. Oh no, did not end there at all. 1`300 BTC, other than being more BTC than the average "member of the community" ever saw together, is also a large enough sum that it'd take the poor guy ~13 mn BTC bet to make back, presuming the servers are freeiv. The total BTC bet in the history of BTC betting ain't quite there yet.v

Does the man have enough BTC to actually cover for the loss and remain liquid ? Perhaps. Does the man actually have the mettle to pick himself up after this and continue rather than hit the bottle or dissolve in whatever other side goop ? Perhaps. He's not the first one, however.

He's not the first one, from my peak I have surveyed at least half a dozen similar situations : people who meant very well, people who were more than a shade or two above their peers, let alone the average, let even more alone the "common man". Exceptional people with remarkable abilities being sunk by completely unexpected, improbable, tiny things coming at them entirely out of (their) left field.

The reason is that Bitcoin is very, but I do mean very very hard. It's harder than business, the fiat kind, and while people routinely make companies on their own there, the failure rate is something upwards of nine tenths in the first year. Bitcoin by yourself however... now that's nonsense. More importantly, it doesn't really need to be that way (nor will it be that way, within a decade or maybe even less).

Be that as it may, I do hope this guy's kingdom does somehow survive the loss of this particular battle.

  1. The reason it's not usually done is twofold. On one hand, the people writing up the comments rarely have anything substantial to do with the things they're commenting on, much along the lines of that observation that "those who can do, the rest talk about doing it". On the other hand it's rarely to the benefit of those doing things to be talking about the situation. []
  2. It may be difficult to swallow, this bitter pill, but it's still the truth : if I don't like you it's always because you're stupid. It's absolutely never the other way around. Yes, I know, it'd be so much easier to sleep at night if it were the case that I say you're stupid because I just don't like you. It's not the case. []
  3. Another problem stemming from his particular capital allocation structure was that the max bet was relatively low, something to the tune of a few hundred Bitcoins. Many people can afford to bludgeon into the ground a casino in that position. I delicately peeled about 20 BTC off of it in the hopes that he'll notice why exactly it's very dangerous to sport low max bets, plenty of other people used their significant wallets to do the same. Somehow I suspect the point wasn't very well understood, even though it's trivial - for instance, why do poker players in the final stages of a tournament play a lot higher much weaker hands ?

    In a heads-up match, the size of your stack is almost as important as the quality of your cards. I chopped one of his legs out in the first hand. Now all I have to do is lean on him until he falls over.


  4. Not in the sense of the gear, which never is free, nor in the sense of electricity, which also isn't free, but in the sense of the cost of security for this thing, which easily surpasses those trifles. []
  5. Satoshi Dice is still above half the total, and Satoshi Dice saw less total value bet in its year of life so far. []
Category: 3 ani experienta
Comments feed : RSS 2.0. Leave your own comment below, or send a trackback.

5 Responses

  1. Very well written and enjoyable read. Thank you.

    I (still) don't understand footnote 3.

    Suppose the house offered bets up to a thousand bitcoins instead of only 200 bitcoins. How does offering you a wider range of bets reduce your chance of walking away with a profit?

    At best it offers you the same chance, since you can chose to self-limit at 200 BTC profit per bet. It seems obvious to me that offering you a superset of the currently available bets cannot possibly reduce your chance of a positive outcome. Yet you appear to be claiming exactly the opposite.

  2. Mircea Popescu`s avatar
    Mircea Popescu 
    Tuesday, 16 July 2013

    It doesn't reduce my chances of walking away with a profit, it improves the likeliness of my walking away with a loss.

  3. ... but only if you deliberately use an inferior strategy. When I double the max bet, you're not obliged to bet bigger. If doing so increases your expected losses then you shouldn't do it.

    But then, you shouldn't be playing -EV dice games at all, by that standard.

  4. Mircea Popescu`s avatar
    Mircea Popescu 
    Tuesday, 16 July 2013

    Possibly the simplest way to put this : you do believe that reducing the max bet has the virtue of improving your risk-of-ruin outlook on the same capital basis, but you do not believe that incresing the max bet has the virtue of hurting the player's risk-of-ruin outlook on the same capital basis. Well why the hell ?

  1. [...] instead count the results of the want of a nail. So, because an airplane that should have been maintained wasn't properly maintained, Romania has [...]

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.