I was hoping someone first provide simpler sequences (e.g. bet 0.5 at 3x, if lose bet the other 0.5 at 4x: which reduces the  house edge to 0.835%), but looks like you might have nailed the ideal (?) betting sequence , with an effective house edge of  0.70376%.

If no one can beat this, you have a well-deserved 0.1 BTC coming your way

You're right, my language was pretty poorly picked. I should have said: "Maximizing the probability to win 1 BTC" as opposed to "Maximizing the EV".  Your bet does however does use fractions of a satosishis, so I'll give you 0.05 if you tell me your MP Account or bitcoin address.

The core question is still open, and for the people who suggested martingale betting sequences, I'll writes some code to figure out the probabilities a bit later.

Challenge: The person who provides the best way to win 1 BTC using a bankroll of 1 BTC. I will give 0.1 BTC to (payable to their moneypot account).

Assuming the constraints of a dice site such as bitdice.me or pocketrocketscasino.eu: 1% house edge, 9900x max payout, 1 satoshi minimum bet.  Provide a well-defined betting sequence to maximize the chance of winning 1 BTC. If I don't have the skills to analytically analyse your solution, I will be forced to simulate it over a few billion runs. Your betting sequence must be FULLY specified without any ambiguities.

Let me start with the most obvious solution:


Bet 1 BTC @ 2x  ... which gives an EV of -0.01, and a chance of winning at 49.5%

Yes, a bit is a 100 satoshis, a µBTC, 1000⁻¹ mBTC, etc. However, it's important that we as the bitcoin community move to a common unit and name, and it's pretty clear by now that that will be bits, with companies like bitpay moving to it as their default unit (and being already supported by coinbase, blockchain.info, etc.)

It's also a homonym, predating your definition as a unit of money and now rapidly replacing common usage of 'bitcoin' as a millionth of bitcoin.

It comes from the kelly criterion which in the case of a casino, is super simple. A "full kelly" really just means, that the house accepts a bet where the profit is X% of it's bankroll when the house edge is X%. Most dice sites with a 1% house edge, refuse any bets that would put them above a half-kelly (0.5% of their bankroll).

blockage and csmcanarney are right, but let me expand a little on what we're building with vault (which is coming together, but still 1-2 months from production quality).


Vault Value proposition for users:
* A highly usable web-wallet, suitable for day-to-day use with all the features you'd expect for that
* We don't ask who you are, or why you're using vault
* Good (external) privacy, your coins are mixed with everyones coins
* Offchain transfers: Transfer money to people/accounts instantly off-chain.
* Unified login system for gambling sites. Vault is an OAuth2 provider, and you can authorize apps to login and have access up to a specified  amount of money. Not only do you get the unified identity across sites, but you can "Add funds" instantly by just authorizing the app to gamble/spend more (no more waiting for confirmations)

Vault Value proposition for investors:
* Invest your money in a single bankroll, that is used by multiple websites. You only need to trust Vault, and not other gambling operators.


Vault Value proposition for devs/casinos:
* Focus only on building an engaging gambling game. You can even build a gambling game without running your own server!
* Leverage the dirt simple "Login With Vault", no screwing around with creating your own user system, security system, 2FA, forgot password, support requests to reset passwords/emails/2FA, etc., handling withdrawals, deposits and reduce friction for people who already have a vault account
* You can use the flexible provably fair betting API to make bets on behalf of your user. You can verify them in real time, to make sure Vault isn't ripping you off. The provably fair betting API is flexible enough to support games like: dice, roulette, plinko and slots from day one.
* Have users gamble directly against Vault's bankroll. This means on *day 1* you can offer high limits, and have absolutely no worries about variance or solvency, vault takes care of that for you.
* Offload security to Vault. All users balances are kept in vault as is the site bankroll.
* Have a steady source of income. From investors perspective, we always give them a 1/2 kelly for every bet (with a minimum of a 0.1% house edge). When ever you send Vault a bet with a house-edge greater than a 1/2 kelly, vault and your app will a 50-50 profit share. [1].  The core principle here, you get a steady stream of income based on your volume and the odds you offer your users.





[1] Details of the investor/vault/app profit sharing subject to change

I don't have any plans on adding any altcoins, but Money Pot is likely more popular than you imagine. The chat is often dead or Indonesian, but right now there's 88 connections to the game, from about 40 players and weekly wagered is actually higher than both PRC and DiceBit.me. Of course the house margin is considerably lower than both of theirs, so people can play a lot more with the same amount of money. (So maybe a better measure of popularity is the weekly EV of the casino, in which case they'd both be quite a bit higher than MP)

Yeah, definitely need to fix this. Currently the stats page is really database heavy. It's not an issue if only a single person is requesting it, but when bots are crawling it, or lots of users are requesting it the stats pages can start timing out. I've been ip banning a lot of bots that crawl those pages (if anyone wants the data, just ask me! I'm happy giving out full db dumps of plays and games!) but I need to work on just optimizing/precomputing the whole thing

Yeah, when players win I lose (and vice versa of course). It's been a rough swing, but really it doesn't infer with my ability to run the site. The site has a bankroll (which is what I'm willing to lose), and the max profit per game is configured to be 3% of that; so the amount players can win per game has swung between 3.2 to under 1 BTC, and currently sitting at 1.5.

But the nice property here is, as I lose more money the max-win goes down and keeps the house solvent. As long it doesn't drop to the point of being detrimental to game play, there's not too much impact.

Additionally, if I'm being honest even though the site is down 57 from the high a few days ago, the recent changes have me feeling better than ever about running the site. I now have a bounded downside, and thus don't have to worry about waking up to a 200 BTC cashout from a whale raid; and the provably fair scheme is fantastic from the point of view I can ignore any claims of a "rape button" as nonsense and not take them seriously.

Plus, what better way to show how ludicrously low MPs house edge is, than hand out money hand over fist

Wow, thanks for the great turn out! Since increasing the limit, we've doubled our daily wagered volume to between 400 and 500 BTC, and players have netted a total of 44 BTC! Well done everyone!

It was not a silent update, it had been announced multiple times on the chat; however, your premise is right. I should have done a better job making sure no one got burnt by this.

So here's what I'll do. Firstly, I'll refund the losing bets that were than or 0.1 BTC which are:

108063, 144804, 194038 and 260011 -- for a total of 0.706916 BTC

Secondly, I'll add a max-bet field to the strategy editor so people stop implicitly relying on the site limits:

https://github.com/moneypot/webserver/issues/130


And thirdly, in the future any update of the sort, I'll make sure to version it along with a notice so people are aware of what changed, rather than needing to check the chat.

Max bet has been increased to allowing 1 BTC bets, and max win increased to 3% of the bankroll (which is currently 3.12 BTC).

Good luck everyone!

This has totally made my morning, someone has created a third-party verification website to verify Money Pot games in real time.

https://leprchaun.io/

Thanks to who ever did it!

I'm happy to announce that the most requested feature of Money Pot ever is now live. Provably fair distribution of game crashes! I believe this is a big first for any multiplayer gambling game. The technical details are well covered in the seeding event, and in the coming days we'll make it easier to verify the games as well as try get some 3rd party tools verifying games.

Enjoy your game! And never worry again if a "rape button" exists or not!

We now have a client seed: '000000000000000007a9a31ff7f07463d91af6b5454241d5faf282e5e0fe1b3a', thanks everyone who participated in this event.

Just 15 more blocks until block 339300, in which we'll know the client seed. I am just heading out the door right now to go out and have a few beers with old friends: So me pushing the prov branch to production will largely depend on my sobriety when I get back home.

I'm also going to delay the increase wager size until Monday or Tuesday, to give me more time to develop and test limits based on the site current bankroll. The plan now is to start with a bankroll of 150 BTC, max bet of 1 BTC and have a max-profit of 2%. While this is massively above a 1x kelly, extensive simulation proves it's still +EV

The implementation (containing information about the seeding event) is here: https://github.com/moneypot/gameserver/commit/bc1bffe32833a3f73046698677804c216607f952

For anyone wanting auditing the provably fair seeding, this is a great thing to backup.

This is the code that will be used once the miners have mined block 339300

Agree with most of what's been written here. The blurb of this section it no favors:

I think most people here would prefer if there was a childboard for for HYIPS/investments/pyramids/ponzis and sports-picks-monologues

Your story appears to check out, so I'll give you the benefit of the doubt and credit your account with the full amount ( 52936.76 bits) =)

I have added a disclaimer now, and will not credit further cases of this to prevent any abuse =)