Simple Bitcoin Game

[Enochian]

I just had an idea for a Bitcoin game where the patron can prove the Website didn't cheat.  I call it "BitReflector."

The user sends Bitcoin to an address, and the site instantly echos back 5% of the amount.

If the least significant hexadecimal digit of the transaction hash agrees with the least significant hexadecimal digit of the hash of the block containing the transaction, the site keeps your wager, otherwise it refunds it.

At 5%, it takes 20 plays to double your money, but it keeps your wager an average of once every 16 plays, giving the website an edge of 1.25%, which is very generous compared to the house edge of 2% typical of casino gambling games.

Since there is no way for the website to manipulate the block hash, which is generated by miners and is strongly pseudorandom, and this information is publicly available, undetectable cheating is impossible..

This game should be pretty addictive, given that it offers instant gratification and a generous payout, although like all popular games of chance, you will slowly lose money in the long run.

So a few questions...

Has anyone ever done this before?  I see there have been websites like CryptoDouble and others, and casinos with verifiably honest casino games using message digests, but I don't see anything that is this exact scheme.

How legal would BitReflectors be in the US?

[ralle14]

I just had an idea for a Bitcoin game where the patron can prove the Website didn't cheat.  I call it "BitReflector."

The user sends Bitcoin to an address, and the site instantly echos back 5% of the amount.

If the least significant hexadecimal digit of the transaction hash agrees with the least significant hexadecimal digit of the hash of the block containing the transaction, the site keeps your wager, otherwise it refunds it.

At 5%, it takes 20 plays to double your money, but it keeps your wager an average of once every 16 plays, giving the website an edge of 1.25%, which is very generous compared to the house edge of 2% typical of casino gambling games.

Since there is no way for the website to manipulate the block hash, which is generated by miners and is strongly pseudorandom, and this information is publicly available, undetectable cheating is impossible..

This game should be pretty addictive, given that it offers instant gratification and a generous payout, although like all popular games of chance, you will slowly lose money in the long run.

So a few questions...

Has anyone ever done this before?  I see there have been websites like CryptoDouble and others, and casinos with verifiably honest casino games using message digests, but I don't see anything that is this exact scheme.

How legal would BitReflectors be in the US?

This reminds me of luckyb.it their plinko concept is nice and your lucky moves depends on the tx id and such. Also this thread is in the wrong place and should be moved in the gambling section.

[ardi wijaya]

how to get bitcoin quickly. I just played BTC

[shorena]

#1 too much fees
#2 miners can cheat. Probably not worth it right now, but possibly in the future
#3 most casinos have an edge of 1%

[Enochian]

#1 too much fees
#2 miners can cheat. Probably not worth it right now, but possibly in the future
#3 most casinos have an edge of 1%

Fees are currently under 10 cents a transaction.  There are a number of bitcoin games where you don't maintain a balance and each individual play involves sending to an address.  The game operator can absorb the fees on paying out, and you can include multiple TXOUTs in a single transactions to do multiple plays with a single fee.  I agree this will discourage people from making large numbers of small bets.

Miners can certainly cheat by colluding with players to make sure no one ever loses, with only a small increase in the processing power expended.  But really, what is their incentive to do so, if only small amounts of money are involved. 

Ignoring poker, where you play against other players and not the house, the edge for some common casino games are as follows. 

European Roulette - 2.7%
Slots - up to 15%
Blackjack with optimal play and multiple decks - no more than 0.5%

On the average, casinos manage to rake in about 2% of the money wagered, over all games weighted by their popularity.

I looked at luckyb.it.  I can't play it, since they block the US, but it looks pretty engaging, with the graphical display of bets being processed.

[shorena]

#1 too much fees
#2 miners can cheat. Probably not worth it right now, but possibly in the future
#3 most casinos have an edge of 1%

Fees are currently under 10 cents a transaction.

Yes, but not during spam waves or high usage. IMHO on chain gambling is dead, because you pay a fee for every bet und thus remove the chance to play with very small amounts. Your idea would probably have a fixed stake anyway, so its just removing potential customers.

  There are a number of bitcoin games where you don't maintain a balance and each individual play involves sending to an address.  The game operator can absorb the fees on paying out, and you can include multiple TXOUTs in a single transactions to do multiple plays with a single fee.

More outputs means you create a bigger TX, which increases your fee. As operator the fee you pay reduces your income.

  I agree this will discourage people from making large numbers of small bets.

Miners can certainly cheat by colluding with players to make sure no one ever loses, with only a small increase in the processing power expended.  But really, what is their incentive to do so, if only small amounts of money are involved. 

Miners could play themselves. Right now there is next to no incentive due to the blockreward, maybe its not a point as falling blockreward should be covered by higher fees.

Ignoring poker, where you play against other players and not the house, the edge for some common casino games are as follows. 

European Roulette - 2.7%
Slots - up to 15%
Blackjack with optimal play and multiple decks - no more than 0.5%

On the average, casinos manage to rake in about 2% of the money wagered, over all games weighted by their popularity.

I looked at luckyb.it.  I can't play it, since they block the US, but it looks pretty engaging, with the graphical display of bets being processed.

I had the dice casinos in mind that are more common for bitcoin, almost all of them have a 1% house edge.

[Enochian]

I had the dice casinos in mind that are more common for bitcoin, almost all of them have a 1% house edge.

I tried luckyb.it.  Turns out they really can't tell from a transaction if you're in the US.

There doesn't seem to be any large volume of players.   A ball drops like every 15 minutes.

The major problem with the idea is latency.  They don't drop the ball until your transaction has at least 1 confirmation.  Any gambling game where you have to wait 10-20 minutes from the time you place your money on the table, to having the play actually take place, is like pulling molars.  By the time the bet is processed, the person no longer cares and has moved on.

I think you're right that gambling directly on the blockchain is dead.

[Enochian]

I had a little chat with the luckyb.it people this morning, and I need to clarify something in my prior post.

It turns out that if you put a transaction fee of 0.0002 BTC on your bet, they will roll the ball instantly as soon as they see the transaction, and not wait for any confirmations, so the game runs pretty much instantly.

I asked how they avoid double spending under this scenario, and they said that as long as the transaction fee is at least 0.0002 BTC they are immune to double spending, although for security reasons they couldn't tell me exactly why.

They also invited me to try to double spend on their system.

So I was of course wondering whether the miners, upon seeing a transaction fee that high, just take it, and don't look for later transactions with higher transaction fees which may reference the same TXINs.

I don't have the time right now to hack away at their system, and verify their claims, but it seems like an interesting experiment to try.

There's a reason they call them "confirmations."

[pitham1]

I had a little chat with the luckyb.it people this morning, and I need to clarify something in my prior post.

It turns out that if you put a transaction fee of 0.0002 BTC on your bet, they will roll the ball instantly as soon as they see the transaction, and not wait for any confirmations, so the game runs pretty much instantly.

I asked how they avoid double spending under this scenario, and they said that as long as the transaction fee is at least 0.0002 BTC they are immune to double spending, although for security reasons they couldn't tell me exactly why.

They also invited me to try to double spend on their system.

So I was of course wondering whether the miners, upon seeing a transaction fee that high, just take it, and don't look for later transactions with higher transaction fees which may reference the same TXINs.

I don't have the time right now to hack away at their system, and verify their claims, but it seems like an interesting experiment to try.

There's a reason they call them "confirmations."

Luckyb.it has been running for quite some time now. They have a strong community over here as well.
In the past, the minimum transaction fee for instantaneous bets was 0.1mBTC, but they have increased it to 0.2mBTC now.

[Enochian]

Luckyb.it has been running for quite some time now. They have a strong community over here as well.
In the past, the minimum transaction fee for instantaneous bets was 0.1mBTC, but they have increased it to 0.2mBTC now.

They've say that they have been up for 2.5 years.   According to their stats, just under 100k BTC has been bet, and with a house edge slightly over 1%, that's over 1000 BTC in profit, or about $400k.

If that's over the 2.5 years, it's about $160k a year gross.  There probably isn't much left after server and ISP fees, and paying employees, but it seems like a nice little home business for someone.

I still don't see how they avoid people scamming them with double spending if they allow instantaneous bets, to the degree where they practically dared me to try it.  Clearly there is something undocumented about the strategy the miners use to select transactions, if their claims of immunity from double spending are true.

I'm hoping someone can further clarify this.

[adaseb]

Some casinos are like this already where you need to send BTC to an address to make a bet and it causes problems when there are long block times or when people send transactions with the min. fee.

[binaryFate]

Luckyb.it has been running for quite some time now. They have a strong community over here as well.
In the past, the minimum transaction fee for instantaneous bets was 0.1mBTC, but they have increased it to 0.2mBTC now.

They've say that they have been up for 2.5 years.   According to their stats, just under 100k BTC has been bet, and with a house edge slightly over 1%, that's over 1000 BTC in profit, or about $400k.

If that's over the 2.5 years, it's about $160k a year gross.  There probably isn't much left after server and ISP fees, and paying employees, but it seems like a nice little home business for someone.

I still don't see how they avoid people scamming them with double spending if they allow instantaneous bets, to the degree where they practically dared me to try it.  Clearly there is something undocumented about the strategy the miners use to select transactions, if their claims of immunity from double spending are true.

I'm hoping someone can further clarify this.

The 0.0002/kb fee is a requirement for immediate bet processing, but not necessarily sufficient. In practice though, it will be most often sufficient to have the bet played right away, as the staff suggested.
Luckyb.it uses a proprietary algorithm to judge nearly in real-time the risk of double spend associated with an incoming transaction. That algorithm is not disclosed for security reason, and is in fact an evolving beast that is regularly adapted. It must adapts to the conditions that prevail on the entire Bitcoin network (depending on protocol evolution, full nodes behavior which is heterogeneous & miners behaviors which is heterogeneous as well). You need an intimate knowledge of the Bitcoin protocol, and of the evolution of the ecosystem over time, to assess that risk properly. It is pretty hard. So much in fact that Luckyb.it has been thinking since quite some time about offering that risk assessment as a service to other entities.

The high difficulty of being double-spend resistant while still offering a snappy experience to users is probably the hardest technical barrier to entry if you want to design a new on-chain game with a reasonable chance of success.