Assuming by  "btc local" you mean "localbitcoin.com", if you're willing to share the account details with me (after removing all the money and what not) -- I'd be willing to login to your localbitcoin account and confirm or deny whether this was the case.

I have no idea who's right or wrong here, but DirectBet is seriously asking someone to upload their passport to imgur wtf?!! Is that a joke?


That's unnecessarily condescending.


That is completely irrelevant. ~47 hours earlier, the user (or the wallet service he uses) unsuccessfully tried to relay a (very large 3323 byte) transaction with 7450 satoshis of fees. Not surprisingly,  it didn't confirm, nor propagate nor had any chance to do so.


That's not true. I have dozens of support tickets per day of people who use misconfigured wallets or wallets that send too little fees. It's a huge stretch to go from "small fee" to "fraudulent intentions".


Not really. I wouldn't say this constitutes a double-spend attempt. As you note transaction 89be96f6cec.. was sent with very low fees, which would have facilitated a possible future double-spend (if the bet lost)


That's 100% demonstrably false. All we saw here was a wallet that tried to tried to send with too little fee and the transaction didn't confirm. blockchain.info's wallet for instance must do that hundreds of times a day. As you may have noticed, bitcoin blocks are full -- what do you think happens to the transactions that never confirm? They're resent..

Yeah, this definitely happens and I think the two motivations for this happening are:

a) The person is hoping that by not releasing the escrow they can get it back
b) The person is a troll


Bithra solves problem a)   (because it explicitly refuses to return escrowed funds to the sender to avoid this attack) but is open to b).  However with b) the person has no financial motivation, so hopefully it doesn't happen too much.

I had some ideas to combat this, like having configurable deposits for both the sender and receiver. And then *after* the deposit obligations are done, they do the escrow. Then if the deal is amicably resolved, everyone gets their deposit back. (if not, people would also lose their deposits). This would mean trolls would lose money themselves. But at the moment, I'm not sure it's worth the complexity.

Thanks for trying it out! It should have appeared instantly! Sorry about that. It seems to be showing now. The nerdy answer is that Bithra itself doesn't really update or is aware of how much is in escrow, if you look at the source of the page you'll see this:



i.e. your browser is querying blocktrail and then displaying the results, so blocktrail was probably lagging or behind. If this continues to happen, I'll switch API providers =)

(And on a side note, if you want to be extra-secure you should check using a different block explorer to make sure blocktrail isn't lying or anything)

Well the beautiful thing about the vast majority of bitcoin casinos is that you don't need to trust them, just learn how to verify the outcomes   

RTP and provably fair are concepts largely orthogonal. RTP is just 1 - houseEdge, while provably fair just means people are given the means to verify the results were ..fair. In order to verify the results were fair, it ends to encode the house edge in the algorithm. So all provably fair systems I'm aware of do allow people to calculate the houseEdge (and thus RTP), but if you just give someone the RTP or house edge it doesn't mean they can verify it's the case.



Absolutely. The only thing that complicates provably fair tends to be when it's multiplayer stuff.

It would seem that way, but it's actually just convincing illusion. The median will converge back to 1.97, but it won't really help you be +EV. As an example, imagine you flip a coin 100 times, and 90% of them land on heads.  If you keep flipping the coin, eventually the amount of heads will be 50%. But that doesn't mean that tails is any more likely to come up =)

I really like the coin example, because people can look at a coin and see it has no memory of the past. What tends to really confuse people is more abstract things like "low game median". E.g. over the last 1000 games, let's say the median is bust was 1.9.  We know the median is going to trend back towards 1.97x -- so surely it's a good time to bet?

Yup, that's right! thanks for posting this.

It's not ideal, but I think this thread is probably the best spot for such scripts. The subreddit never really got much traction, and was kind of abandoned

o.0 weird, I can't reproduce this at all and don't know how it would happen. Looking at the logs doesn't seem like the site was down last night either

I've never played csgo or used steam, so I might be wrong; but I believe this refers only to the official steam market (which gives you money on your steam amount). As I understand it, you can trade and send items for free, which is why 3rd party sites are so popular. So if you avoid the official market, you don't need to pay a commission on transfers (and steam still makes their money by selling items in the first place).


(this is all second hand knowledge from bustabit's chat, so don't fault me if I'm totally wrong)

Yeah, I suspect it is. I think csgocrash is only a ~month old and has already double the amount of users as bustabit (which is ~1.5 years old), they have *way* more users and players online at once. And that's with a much higher min bet, and running with a significantly increased house edge  (2-3%) from bustabit's (0-1%). Their stats also show they've made ~250k USD (?!) so there seem to be a lot of money in it.

That said, they seem quite competent at advertising; while I've done virtually zero.

On a side note, because of csgocrash I've got a lot of new csgo players on bustabit -- so it's worked out very well for me as well =)

They actually bought a copy of the source:
bustabit.com/license.txt

and seem to be doing a lot better than original! Kind of amazing really, I'm glad for their success.

(It's been a while since a version of bustabit was released under an opensource license, but any copied used with that has the condition it must stay under the same AGPL license)

Due to rampant faucet abuse, I've temporarily disabled it. I'll try use it as a bit of an experiment to see if not having a faucet attracts a different crowd =)

In other news, we've passed 350k BTC wagered, which is just incredible! 

The release secret is 100% generated client-side, and NEVER sent to the server. You can verify this by either:
a) Check the source code (a bit tricky, because bitcoinjs-lib is minified and hard to check its not tampered with)
b) Check the network tab in chrome (you'll see only the address is sent)
c) Use the advanced mode, and generate an addresss totally off the site, and only provide the address  (most recommended, if you're concerned)

You can still put money in a released escrow, and re-release it. i.e. your money won't be lost.  But because the escrow is released, it can't be used multiple times (the provably fair system only says "release the entire escrow" it doesn't say "release payments sent before X". So you should use a new escrow for new things.

You can make installment payments to the escrow address, yeah. And when it's released, all the outputs will be sent at once to the receiver. Super dusty payments (e.g. under ~20) won't be used, because they'll add more to the txfee than they would to the output amount

There's no automated way to do so, but with a signed instruction from the sender I would have no objections to provably burning the coins (e.g. sending them to OP_RETURN ). If that becomes a popular request, I can add a "burn escrow" feature =)

Check the other tabs of your history (e.g. withdrawals, transfers) or more relevant: "all". Normally deposits are handled in the same (database) transaction that credits your account, so it's not possible for the (bitcoin) transaction to appear but your account not being credited.

You're definitely right, but let's not let perfect be the enemy of good. The vast majority of transactions that should have been done through an escrow-like service are not. For instance, I recently sold MoneyPot (for several times your 20 BTC example) with no escrow or non-repudiation. The buyer was 100% at my mercy to act honestly. Of course, like the vast majority of deals it went through smoothly, but I'm sure the buyer would have been better off protecting themselves using escrow; but it was never done because of the additional friction. That's the problem that bithra tries to solve, it's a light-weight, low-friction escrow. It's not perfect, it's never going to replace having a human arbitrate, but I strongly believe it's a net positive to the community.

Yeah, it seems incredibly scummy. Searching the forums, it seems they've used this trick multiple times before, once to avoid paying out 20 BTC as well:
https://bitcointalk.org/index.php?topic=651730.msg14194712#msg14194712

And funnily enough, even when looking for the conditions I was unable to find it. It's hidden in the fine print of a completely different page. No one in their right mind would assume making a bet that is allowable by the UI is a punishable offense to the tune of your entire balance.

It hard to see it this as anything but a predatory and scammy way to avoid paying winners (if you lose with large bets, they seem to have no problems taking your money).