Topic: Satoshi Dice -- Statistical Analysis (Read 192914 times)

Actually, SD is doing 1.9%. (I forget where I got the 1.98%, maybe it was the actual math or something.) (And I found it, back then losing bets got 0.005x or something.)

Do losing bets still get something back? There would be no reason for the off-chain version to give back on lost bets, because the players can already see that they lost. I believe the whole point of returning 0.005x was so that players knew they lost, because it was on-chain.

The only other dice sites that exist right now, which are worthy of mentioning are Prime, and PRC. The last two big ones were Dicebitcoin and Dice Ninja, both of which are effectively gone. While others exist, there does not seem to be enough volume on them. Coinroll has changed owners, and I don't see much action there, so it's pretty much dead.

So, if Prime increases to 1.9%, PRC might follow, just because Dean seems to be that kind of person, but I don't really know. There is that dicesites.com which lists a whole bunch of them and we can probably ignore the non-bitcoin versions.

The other sites will either increase to just below 1.9% or stay at 1% hoping to get more players.

Whales, will still play on the top 3 sites (SD, Prime, PRC).

PD has been around for a long time, and for the most part have proven to be honest (very few "hiccups" - when hiccups do occur members are fairly compensated, no scammy policies, however "you never know") so I think that some people would gamble on PD with a higher edge verses another site with a lower edge that has not "proven" itself

Yeah, .. I think most players don't really care as long as it's not 10% or 50%. Thus, I am starting with 1.5625%, and only because that number is derived from 1/64 = 0.015625

The most popular land based casino games (in Macau anyway) averages 2% ~ 3%. And the others are all higher. Baccarat is low. Sic Bo is middle. ... I think Black Jack is the lowest on the best player possible.

1% will only work while online bitcoin gaming is not regulated. When you factor in costs and expenses and license fees and taxes and all that, you'll end up just like the land based casinos or you simply shut down.

I'm curious to know that too. I suspect that most gamblers don't really care too much about the edge. All the bars in Canada have state-run gambling terminals where you can play Keno, lotto, some kind of fake poker, etc. Some games have a house edge as high as 50% and yet people still play them constantly.

From http://lotto.bclc.com/keno-and-keno-bonus/prizes-and-odds.html :



That top row though: you have a 1 in 4 chance of doubling your money.

House edge percentages for those 4 games:


Of course, that's different because the state has a monopoly on gambling. If you don't like the 40%-50% house edge they offer, there's no legal alternative other than not playing at all.


I thought the same. Giving 0.1% back in the form of a jackpot for rolling a specific number is easily forgotten, by virtue of the fact that it happens so rarely. It would have been much more visible if you just dropped the house edge to 0.9% instead. That also would have resulted in less variance for the house I think, than paying out the occasional big jackpot.


Did PD really launch that late? JD launched in June 2013 and I thought PD had already been around for a while then.


I saw the biggest two offchain sites at the time (coinroll and primedice) both having a 1% house edge. I didn't see any possibility other than to match it. I didn't want to get into a price war with either site by undercutting them, but also didn't think that charging a higher edge would work. If you guys had set the bar at 1.5% then I'm sure I would have followed suit.

My original concept for JD was that the investors would set their own house edge, like traders set their own price for BTC/USD on the exchanges. The ones offering the lowest house edge would be the ones who got the action, in the same way that the traders offering their BTC for sale at the lowest price on BitStamp are the ones who get matched first.

I didn't figure out a good way to present such a market to the players. It's probably too confusing to have the house edge bounce up and down as investors compete with each other for a bigger share of the betting action, so I didn't end up doing it. That would have been an interesting experiment, to see where the free-market would decide the house edge should be.

We've considered raising the primedice edge for a long time, 1% is extremely risky and often unsustainable without a solid bankroll or low max bet. It would be interesting to figure out the elasticity of the house edge on dice in terms of how much volume/profit a site would lose or gain by shifting the edge. When primedice had a 0.9% edge I don't think it made any difference on volume, however the way we integrated it via jackpot made it still seem like a 1% game with an easily forgettable added incentive.

In my opinion it makes more sense to set a 1.9-2% edge rather than a 1.5% one. People who care about the edge would probably stick to 1% so you'd entirely lose that audience while profiting more off of the audience you did obtain which clearly don't care about the odds as much. That's my opinion at least, if more dice sites shut down/run off with funds we'd probably consider raising edge but no plans to right now. When PD was about to launch we had decided on a 1.5-1.6% edge if I recall correctly, satoshidice had plans to release a new off-chain site around then (may '13) and coinroll had kept a 1% edge due to that reason and we ended up matching it to stay competitive. I wonder what sort of edge bitcoin casinos would have if things didn't progress that way though, if primedice was sitting at a 1.5-2% edge would just-dice and all investment sites have followed suit?

SD, stay at 1.98%, so my 1.56% doesn't look bad. hehe.

I don't know if it's new or not, but I recently discovered:
  https://satoshidice.com/stats/

It includes the following, showing the slow decline in the number of bets as off-chain games started eating SD's lunch:



It will be interesting to see whether the launch of a new off-chain game will bring an increase in betting volume. SD seems to be keeping its 1.9% house edge even though all the competition has a 1% edge.

Means bot for betting on new brand , new design .

That's right.


What do you mean?

Any bot for the new script ?

Hi douglas,

so basically it means no more of this kind of analysis?

The last I heard, SatoshiDice had followed Just-Dice.com's lead (just kidding, PrimeDice.com and CoinRoll.it!) and taken their betting off-chain.

So it's no longer possible to see how well they're doing.

Any updates coming?

Thanks Dooglus

Very interesting. SD has been unlucky but turnover was good this month of August with 143k btc (up until 28th), up from 110k btc last month. At a current value of $118 per btc (bitstamp) expected profit for this month is $344k, up considerably compared to last month where it was only $209k (using $98 as btc value). But still not beating the top months that were $460k (January 2013) and $430k (May 2013). Interesting to see no collapse is happening and a chance for actual growth is again opening up.

Comparing this to JD. For the month of August JD turnover was 205k (up until 27th), expected profit $240k. Last month of July turnover was a staggering 685k btc and expected profit $671k. This means a drop of -72% month to month. Expected profit at Satoshi Dice also collapsed after it's all time high month in January but only by 50% (from $460k to $220k).  

It remains to be seen whether JD will succeed where SD failed: after setting a great month, continuing to grow in expected USD profit over time.


(Am I doing this analyses correct? Does expected profit even matter? If gamblers gamble until they lose, than expected profit will go up or down depending on luck, but actual profit will just be the same. Maybe it is better to forget about expected profit and just look at actual profit?)

It's been a long time, but here's an update:

This is Bitcoin. I don't believe anything until I see 6 confirmations.

Why on Earth would Eric want to buy back the shares he issued? Especially at the insanely increased prices he bought them back at.

And why does he need them off his back? He didn't owe them anything, he was producing them respectable dividends, he just chose not to pay as much attention as one would expect.

Thanks, exactly what I am thinking since the sell announcement.

Anonymous buyer my ass!  Eric paid off the shareholders, threw them a bone to keep them off his back and kept the rest of the profits.  Moving to Panama would not protect him from prosecution and he's smart enough to know that.

If/when this anonymous buyer emerges or SD return in another form I'd gladly apologize.

Unfortunately since the beginning of bitcoin time Bitcoiner have proven to be an awfully gullible bunch.

 

*other tinfoil hat theory that shareowner betting to prop up prices has ceased*