Just to make it clear, you can't invest in a particular moneypot casino but rather invest in the global moneypot bankroll, which all players at moneypot are betting against. Keep in mind it's quite a high risk investment, much more so than doing a <= 1x kelly at any other place.

That said, I think a huge mistake people make is trying to find, and invest in only the "best" place. There's so many things that can go wrong (lucky whale, malicious operator, operator error, operator getting hit by a bus, bitcoin crashing etc.) that to put more than say 1% of your net worth into such a scheme would be reckless. What I think is a good idea is to find multiple independent schemes, which you believe are +EV and allocate your funds accordingly. It's a lot more work, but you can transition from gambling-investor to investor =)

Yeah, looks like you guys got it =)

On one end of the extreme, if the bet poses no risk to investors the app will get 50% of the house edge. On the other extreme if the bet poses a maximum, fully-kelly risk to investors (e.g. a binary bet risking N% of the bankroll with an N% house edge) the app will get 0% of the house edge (the bet unmodified goes to investors).

The vast majority of the bets are for tiny amounts, so they tend to look like the app makes "50% of the house edge" but it's not exactly the case. Most devs start to notice this when larger, riskier bets come in


E.g. If you look at this bet: https://www.moneypot.com/bets/61529776  it was placed when the bankroll was a little under 300 BTC. The bet was trying to win ~1.5 BTC  with a 1% house edge (a bit over a half-kelly). And thus the app made a commission of ~63000 bits (Which it would've made regardless of win or lose).  As the bet was of ~30 BTC this translates to about 0.21% commission. MoneyPot itself made another 0.21%  and then from there, the investors took it.

Lot of action happening on MoneyPot. Almost need to create a newsletter with updates.


But some highlights:

* BetterBets is absolutely killing it, bringing in great volume topping 1000 BTC in a single day. You can see some of their stats charted here: http://dicesites.com/betterbets
* Site bankroll has grown to over 300 BTC
* Our largest bet ever of just over 30 BTC was made, which unfortunately was a loss ending helpmywallet's incredible streak on his app.

Sure, you can always restrict what bets you accept. I'm not sure why you'd want to do that though as the absolute worst case is they max-bet every single bet and you make 0 commissions, but it would be still a huge boon to the casino assuming you're not offering a rakeback or something.



Yeah, we do support this. If you want to talk to me on slack I'm more than happy to explain how it would be achieved

It's nothing too unusual about freak bad luck, after all someone has to be the unlucky 1-in-a-million case.

But fortunately PrimeDice provides you with the tools to check if you're just unlucky or were cheated. Give it a read here: https://primedice.com/verify

The simple way to check is before you start gambling save in a notepad the "server seed hash", then set your own custom client seed (and write that down too). Once you're finished playing, hit the reveal server secret button. Copy and paste that server seed into DiceSite's excellent tool:

http://dicesites.com/primedice/verifier

And make sure the results match.  Before you do that for real, claim a few free bits from the PrimeDice faucet and attempt this procedure and make sure you can see how it works.

Honestly, verifying bets is satisfying. You can remove all the doubts from your mind about the site being fair or not.  =)

If you don't care about computational complexity it's trivial. Just enumerate all possible server seeds

Negligible


All depends on how the nonce is stored. Assuming conservatively that it's a signed 32 bit integer, that should be about 2.1 billion

It's a little involved, unfortunately, and there's no simple answer. Probably the best read is: https://www.moneypot.com/faq#what-is-the-investment-option-

Basically any bet that is kelly-compliant is accepted by investors. Sometimes a house edge of 0.001% is all that's needed for that, or other times it could be 10%. Any bet that is not kelly-compliant from the investors perspective is rejected. Any bet that is exactly kelly-compliant, the investors take 100% of the risk and 100% of the rewards. If the bet is *better* than a kelly-complaint bet (from the investors perspective), MP arbitrages the difference and does a 50-50 profit share with the app.

So from an investors point of view, the same bet with a 1% or 10% house edge would have the same amount of risk and reward (but the apps and MP are going to make a lot more from the 10% house edge bet).


And if you had a whale who was max-betting each round on a 1% house edge site, the investors would have an EV of exactly 1%  (and MP and apps would make 0).

I do like the idea of tooltips, it'd be handy if you hovered over 100,000 bits to see "0.1 BTC / $28.08". I'll add that to my todo, along with a setting of which fiat unit you'd like to see the conversion to.


I do like pushing bits though, as I believe it important for the bitcoin community, especially the less technical segment to have a common base unit. I don't believe BTC is suitable for that, and would be happy with either mBTC or bits, but I'd like to throw my weight behind bits as it seems to have the most momentum at the moment.

And from what I've seen on bustabit, people really do end up liking it and not trouble expressing large numbers by using "k bits" and "m bits". We also there allow people to use the suffix "k" to mean thousand. So instead of 100,000 bits, people just write 100k.  I think for most people, it's a lot easier to grok and work with  "1 million x" than it is to grok  "1 millionth of an x" so it does make sense using a very small base unit.

We do support bitcoins in the "send" dialog as often prices are still quoted in bitcoin, so this does makes things a lot easier. But in general, I'd like to stick with bits as much as possible. It's probably selection bias, but most of the users who get used to 'bits' seem to prefer it, and express frustration with the bitcoin unit.

I would however consider making a "bitcoin helper" display. So 12,000,000 bits could show up as:  "12,000,00.00 bits"  and 10 bits could render as "00001.00 bits" to help people eyeball the bitcoin conversion easily.

Unfortunately, you can't really stop bets like that. If they're a legitimate whale who happened to get very lucky, you'll just burn a great relation and a lot of chance to make your money back. On the other hand, if it's someone exploiting the game, they'll just be back on another account, on another ip address, hidden amongst the thousands of other players. This will only make it harder to monitor the person and try figure out how they are exploiting you.

  Well, at least that's better than the person who said he was continually reminded of "bustabitch"

That's the easy part, and intrinsic to how provably fair works. Let's say we know the server seed is "moAQKPIMiPI23Eg", you get to pick your own client seed (lets use "my_seed") and from that we can deterministically generate all future rolls. Here are the first 100:

http://dicesites.com/primedice/verifier?ss=moAQKPIMiPI23Eg&cs=my_seed&ln=100&n=0&ssh=A7B8301BBF0E2ADBFE00636443D5570146478D62E1F57437892BD2FB94ED45C8

Now, the only tricky thing is to not look too obvious ;d

Yeah, there's a few around. Some of the players trade (and sell) the little scripts.  The builtin bot behaves weirdly if you activate it in the middle of a game that you're currently playing, if that's the issue you have?

Is this on your end, perhaps? I'm pretty close to the server, but have latency under 100ms. Regardless, I'd highly advise you play exclusively with the auto-cashout to prevent any latency related issues, like not being able to cashout

It's currently using bitcoin-core 0.11's default dynamic fees, which has  a goal to be confirmed within two blocks. Most of the time it works very well, even when block-explorers say otherwise. I've experimented with using the results of estimate-priority 1, however it's quite unstable and often leads to paying x3 as much fees as required.

However, what I'd like to give users the option of picking their own priority (high, medium, low) but bitcoin-core currently doesn't offer support for it over rpc very well at the moment. There's no way to adjust the estimatefee target over rpc (issue: https://github.com/bitcoin/bitcoin/issues/5796) and settxfee doesn't work very well either (issue: https://github.com/bitcoin/bitcoin/issues/6479). If no one touches these issues, I'll create a patch for it myself when I get the chance

Well, it's an API so you can use your own application logic to control it. E.g. You can easily put a policy, "max bets of 0.1 BTC" and stop it after a certain budget is used. Even the way the api works is "max_subsidy" where you set an exact (max) budget for the bet, so you don't get any nasty surprises.

It's worth noting that MP investors will not accept 0% house edge bets (it doesn't make sense for them, has negative expected bankroll growth and doesn't satisfy the kelly). If you do want to give your users a 0% house-edge bet, what you'll have to do is pay the investors (using the max_subsidy api), it'll end up costing the app approximately N*wager when the person is trying to win N of the site bankroll.

E.g. If someone wants to bet 1 BTC to try win 1 BTC at 0% house edge, it would currently cost the app ~0.0034 BTC payable directly to the investors to accept it.

In the case of multi-outcome bets (e.g. plinko) it's a bit cheaper but still something I'd be very hesitant to do myself as it could work out being pretty expensive for an imperceivable difference for most gamblers.

Just doing a quick re-organization of the cold wallet to make it more robust in the case of getting hit by a bus. I'll publish a new MP cold address in a couple hours.

Edit:  Funds are safely at rest. You can view our cold wallet holdings here:

https://www.moneypot.com/proof-of-assets.txt

which should always exceed our liabilities:

https://www.moneypot.com/proof-of-liabilities.txt

All Quantum Dice's bets are placed against the MoneyPot bankroll. So in effect they do have a 280 BTC bankroll, it's just not exclusive.

First of all, I'll just make it clear this is purely hypothetical. I haven't yet decided anything


Well, that is something the market would need to decide. But I would be totally transparent about networth, bankroll, etc

For instance, someone recently won almost 30 BTC which reduces our assets by that amount. But it's unclear how that would affect a share valuation (perhaps it's so fuzzy that it makes no impact, perhaps people panic sell, perhaps people buy as it's a sign of healthy activity).



Yes, that's a big difference. I would only be interested if there could be a robust secondary market, but a large amount of the value will be derived from expected future returns.


I'm open to suggestions, but I was thinking that I would start off with 100% equity (what I currently have) and have the company start issuing shares, which it sells on the market. The money the company raises will get put into the bankroll, which will in-turn increase the casino limits. Then I could personally buy or sell, just like any other investor. If the company is holding shares at the time of dividends, the money would just go towards the NAV.

Ohh! I didn't know. I'll check that out now, that's great to hear

Edit: bittrex is out: