gordonhill
Legendary
Offline
Activity: 1274
Merit: 1006
|
|
December 24, 2016, 01:15:48 AM |
|
|
|
|
|
ranlo
Legendary
Offline
Activity: 1974
Merit: 1007
|
|
December 24, 2016, 01:15:58 AM |
|
I understand max multiplier will be x9900
what you are describing is martingale
My bad. I missed a zero. I was describing a reverse martinglae where you double on win, not on loss. Using a reverse martingale it's possible to turn 1 satoshi into a million satoshis, by doubling 20 times. Since you would be 1000000x'ing your starting balance would you be attempting to break the rules? That was what I was asking. But it sounds like the rules aren't really decided upon yet, so there's no point trying to get clarity until they are. Hey, Dooglus! You're the most qualified person I feel I can ask this to, as you also showed in the past with your analyzation of martingale at different % chances that you have a very solid grasp. My question here is... is there any tangible difference between: 10 bets at y for 10% (10x) 100 bets at 1/10y for 10% (10x)? By its look, the total amount would be equal (meaning the same HE rule would apply in terms of EV), and the total expected reward would be equal (in case 1, we expect to hit 1/10 for 10x = 1.0, whereas the bottom we expect to hit 10% of 100 = 10 at 10x = 100x / 10 = 10x/10 = 1.0). I did remove the HE part of this calculation for ease of reading. Not sure if I'm quite explaining the question clearly enough but hopefully I am. Basically, making the same bet, but with a smaller amount (and more bets) should result, theoretically, in the same outcome, correct? Such that let's say it were capped at 100x. Doing this: If (win) bet wonAmount at 100x if (lose) returntoBase Would be the same as just betting straight at 10,000x? Or is there something I'm missing here? In #1, you'd have a 1%*1% chance, or 0.01% = 1/10,000 chance. In the second, you'd still have a 1/10,000 chance.
|
|
|
|
dooglus
Legendary
Offline
Activity: 2940
Merit: 1330
|
|
December 24, 2016, 01:33:41 AM Last edit: December 24, 2016, 01:44:02 AM by dooglus |
|
My question here is... is there any tangible difference between:
10 bets at y for 10% (10x) 100 bets at 1/10y for 10% (10x)?
Well, let's look at a simpler case: Is there any tangible difference between: a) betting 1 BTC at 50% (1.98x) b) betting 0.5 BTC at 50% (1.98x) twice Both risk 1 BTC, so both have an expected profit of -0.01%. Let's look at the possible outcomes: a) has 50% chance of profit 0.98 BTC and 50% chance of profit -1 BTC b) has 25% chance of profit 0.98 BTC, 25% chance of profit -1 BTC, and 50% chance of profit -0.01 So the maximum and minimum profits are the same, but b) (betting a smaller amount, more times) fills in the middle. The extreme outcomes become less likely the more you subdivide your bet, and the expected outcome becomes more likely. In the extreme, if you bet a single satoshi 100 million times (ignoring rounding), you will be very likely to have a profit very close to the expected profit of -0.01 BTC For the players, their only hope of winning is to beat the expectation, and so their best strategy (assuming they have a fixed amount to bet, and have to bet at a certain chance of winning) is to make a single large bet. We could plot the outcomes against the chance of them happening. Both charts would have the same "center of mass" (ie. the expected profit, -1%), but the chart where the bets were divided up more would be more pointy, with the outcomes concentrated around the expected -1% point. Edit: I missed the end of your question: Such that let's say it were capped at 100x. Doing this: If (win) bet wonAmount at 100x if (lose) returntoBase
Would be the same as just betting straight at 10,000x? Or is there something I'm missing here? In #1, you'd have a 1%*1% chance, or 0.01% = 1/10,000 chance. In the second, you'd still have a 1/10,000 chance.
I think the player is at a disadvantage if you limit them to 100x. They effectively pay 1% of everything they stake. If they can bet 1 BTC at 10,000x, their expected profit is -0.01 BTC. But if they can only bet at 100x, they have to bet 1 BTC at 100x, then 100 BTC at 100x. That second bet has an expected profit of -1% of 100 BTC, ie. -1 BTC. The strategy "bet 1 BTC at 100x, and if I win, bet 100 BTC at 100x" has an expected profit of -0.01 BTC for the first bet, and -0.0099 for the second bet (0.99% chance of having to make the 2nd bet, and an expect profit of -1 BTC if you have to make it, for an overall expectation of -0.0099 BTC), for a total of -0.0199 BTC. That's very nearly twice as bad for the player as if you let them make a single bet.
|
Just-Dice | ██ ██████████ ██████████████████ ██████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████ ██████████████ ██████ | Play or Invest | ██ ██████████ ██████████████████ ██████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████ ██████████████ ██████ | 1% House Edge |
|
|
|
fiscorcle
|
|
December 24, 2016, 06:37:47 AM |
|
My question here is... is there any tangible difference between:
10 bets at y for 10% (10x) 100 bets at 1/10y for 10% (10x)?
Well, let's look at a simpler case: Is there any tangible difference between: a) betting 1 BTC at 50% (1.98x) b) betting 0.5 BTC at 50% (1.98x) twice Both risk 1 BTC, so both have an expected profit of -0.01%. Let's look at the possible outcomes: a) has 50% chance of profit 0.98 BTC and 50% chance of profit -1 BTC b) has 25% chance of profit 0.98 BTC, 25% chance of profit -1 BTC, and 50% chance of profit -0.01 So the maximum and minimum profits are the same, but b) (betting a smaller amount, more times) fills in the middle. The extreme outcomes become less likely the more you subdivide your bet, and the expected outcome becomes more likely. In the extreme, if you bet a single satoshi 100 million times (ignoring rounding), you will be very likely to have a profit very close to the expected profit of -0.01 BTC For the players, their only hope of winning is to beat the expectation, and so their best strategy (assuming they have a fixed amount to bet, and have to bet at a certain chance of winning) is to make a single large bet. We could plot the outcomes against the chance of them happening. Both charts would have the same "center of mass" (ie. the expected profit, -1%), but the chart where the bets were divided up more would be more pointy, with the outcomes concentrated around the expected -1% point. Edit: I missed the end of your question: Such that let's say it were capped at 100x. Doing this: If (win) bet wonAmount at 100x if (lose) returntoBase
Would be the same as just betting straight at 10,000x? Or is there something I'm missing here? In #1, you'd have a 1%*1% chance, or 0.01% = 1/10,000 chance. In the second, you'd still have a 1/10,000 chance.
I think the player is at a disadvantage if you limit them to 100x. They effectively pay 1% of everything they stake. If they can bet 1 BTC at 10,000x, their expected profit is -0.01 BTC. But if they can only bet at 100x, they have to bet 1 BTC at 100x, then 100 BTC at 100x. That second bet has an expected profit of -1% of 100 BTC, ie. -1 BTC. The strategy "bet 1 BTC at 100x, and if I win, bet 100 BTC at 100x" has an expected profit of -0.01 BTC for the first bet, and -0.0099 for the second bet (0.99% chance of having to make the 2nd bet, and an expect profit of -1 BTC if you have to make it, for an overall expectation of -0.0099 BTC), for a total of -0.0199 BTC. That's very nearly twice as bad for the player as if you let them make a single bet. I didn't actually think about the house edge problem before, that's a very good point - of course the same point could be made about a max-profit. Then again there have to be limits at some point, whales wiping out the bankroll is never a good thing
|
|
|
|
dooglus
Legendary
Offline
Activity: 2940
Merit: 1330
|
|
December 24, 2016, 07:10:36 AM |
|
Then again there have to be limits at some point, whales wiping out the bankroll is never a good thing
Yes. You limit the maximum profit per bet using the Kelly criterion.
|
Just-Dice | ██ ██████████ ██████████████████ ██████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████ ██████████████ ██████ | Play or Invest | ██ ██████████ ██████████████████ ██████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████ ██████████████ ██████ | 1% House Edge |
|
|
|
Quickseller
Copper Member
Legendary
Offline
Activity: 2884
Merit: 2327
|
|
December 24, 2016, 08:15:08 AM |
|
We won't bring the multiplier live until we have fully figured it out and give a thorough explanation of what it means.
I think it should be fairly easy to implement a 9999x multiplier betting limit for apps. It is trivial for you to visit each app yourself and check to see if the app is actively encouraging users to circumvent the 9999x multiplier bet limit, or offering players the ability to make a single bet on the app with a >9999x bet multiplier but sending multiple bets to the MP bankroll. If a player himself decides to engage in a betting sequence, that after a series of bets, would result in a total of greater then 9999x win, then that is not the app's fault. I do not see the point in limiting maximum multiplier to 9900x ,
Whenever a player has a big win, bankroll investors (and the public) will be suspicious of cheating, and when a player has a big win while risking very little, bankroll investors (and the public) will be even more suspicious of cheating; this is regardless of the casino, be it happen at MP or one of it's competitors. When players turn a 0.001 bet into 10 BTC enough times (which will happen over a long enough period of bets), investors might decide to withdraw their money, which would result in MP being able to safely accept a lower maximum bet size, hurting the remaining bankroll investors. Edit: I missed the end of your question: Such that let's say it were capped at 100x. Doing this: If (win) bet wonAmount at 100x if (lose) returntoBase
Would be the same as just betting straight at 10,000x? Or is there something I'm missing here? In #1, you'd have a 1%*1% chance, or 0.01% = 1/10,000 chance. In the second, you'd still have a 1/10,000 chance.
I think the player is at a disadvantage if you limit them to 100x. They effectively pay 1% of everything they stake. If they can bet 1 BTC at 10,000x, their expected profit is -0.01 BTC. But if they can only bet at 100x, they have to bet 1 BTC at 100x, then 100 BTC at 100x. That second bet has an expected profit of -1% of 100 BTC, ie. -1 BTC. The strategy "bet 1 BTC at 100x, and if I win, bet 100 BTC at 100x" has an expected profit of -0.01 BTC for the first bet, and -0.0099 for the second bet (0.99% chance of having to make the 2nd bet, and an expect profit of -1 BTC if you have to make it, for an overall expectation of -0.0099 BTC), for a total of -0.0199 BTC. That's very nearly twice as bad for the player as if you let them make a single bet. A player making two sequential 100x payout bets verses a single 10,000x payout bets effectively changes a 1% HE to 1.99% HE. The difference between three sequential 100x payout bets verses a single 1,000,000x payout bet turns a 1% HE into 2.9701% HE. The formula to calculate the effective house edge when multiple small payout bets are multiple together to make effectively one higher payout bet is as follows: Y=1-H NY= Effective house edge H= House edge of a single bet N= number of bets multiple together
|
|
|
|
fiscorcle
|
|
December 25, 2016, 12:17:46 AM |
|
What about some kind of variable kelly option investors? Maybe the site maintains a constant kelly criterion, but investors could invest up to a certain kelly level that theyc an specify?
|
|
|
|
|
BoXXoB
Legendary
Offline
Activity: 2018
Merit: 1108
|
|
December 25, 2016, 02:35:11 PM |
|
What about some kind of variable kelly option investors? Maybe the site maintains a constant kelly criterion, but investors could invest up to a certain kelly level that theyc an specify?
This way each and every investors could determine the risk they want to take, right? And not be controlled by the site. Perhaps this would remove some of the limitations. Please correct me if I got what you meant wrong. The formula to calculate the effective house edge when multiple small payout bets are multiple together to make effectively one higher payout bet is as follows: Y=1-HN Y= Effective house edge H= House edge of a single bet N= number of bets multiple together
Thanks for clarifying this, I always thought that making multiple bets indeed basically increase the house-edge but it's great to have some factual information. Never thought about taking a closer look into it...
|
|
|
|
fiscorcle
|
|
December 26, 2016, 02:39:56 AM |
|
Merry Christmas everyone. Support will be a little slow today and tomorrow (as all of the Moneypot team is celebrating the holidays). We apologize for that but will try to check in from time to time if there are any issues. A lot of good points were brought up and I will address them after the holidays (or if I can sneak in some time). @Fiscorcle, there will be optional kelly multipliers in the future It's a set kelly though, right? Not like a range? Cuz a rang would be dope, but I don't even know if that would mathematically work
|
|
|
|
kolloh
Legendary
Offline
Activity: 1736
Merit: 1023
|
|
December 26, 2016, 03:56:49 AM |
|
Merry Christmas everyone. Support will be a little slow today and tomorrow (as all of the Moneypot team is celebrating the holidays). We apologize for that but will try to check in from time to time if there are any issues. A lot of good points were brought up and I will address them after the holidays (or if I can sneak in some time). @Fiscorcle, there will be optional kelly multipliers in the future It's a set kelly though, right? Not like a range? Cuz a rang would be dope, but I don't even know if that would mathematically work Not sure what you mean by a range and I can't think of how that would actually work, but usually you can select from a preset kelly multiplier ( say 1x, 2x, 3x, 4x, 5x, etc ). It is essentially just weighting your invested amount by that multiplier so if you had a range, it wouldn't know how to weight your invested amount appropriately.
|
|
|
|
JackpotRacer
Legendary
Offline
Activity: 1946
Merit: 1014
All Games incl Racer and Lottery game are Closed
|
|
December 26, 2016, 02:31:28 PM |
|
New Rules are going live in the immediate future:
Maximum win 0.5% of bankroll (once commission is changed, will be closer to 1x kelly -- kelly options will be added at a later date) Minimum house edge of 0.5% (bets below will be rejected by system).
New Measures:
Apps will be cleaned up in preparation for new app levels. Seriously inactive and broken sites will be deleted.
In the new year, the investor credit plan and personal statistics will be available on each player's profile page.
Any usernames deemed to imitate an other account will be banned (for example: using an upper case ' i ' to act as a lower case ' l ' )
what does immediate future mean? please lets have a time and date thx Maximum win 0.5% of bankroll (once commission is changed, will be closer to 1x kelly -- kelly options will be added at a later date)when? Minimum house edge of 0.5% (bets below will be rejected by system)
when? New Measures:................from when on?
|
|
|
|
Bit-Exo.com
|
|
December 26, 2016, 04:11:37 PM Last edit: December 26, 2016, 05:29:26 PM by Bit-Exo.com |
|
It looks like the 0.5% max win is already updated.
|
|
|
|
dooglus
Legendary
Offline
Activity: 2940
Merit: 1330
|
|
December 26, 2016, 04:51:45 PM |
|
New Rules are going live in the immediate future:
what does immediate future mean? please lets have a time and date thx "This pattern is used to refer to a time immediately after the moment of speaking, and emphasises that the event or action will happen very soon" -- http://www.ef.com/english-resources/english-grammar/immediate-future/
|
Just-Dice | ██ ██████████ ██████████████████ ██████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████ ██████████████ ██████ | Play or Invest | ██ ██████████ ██████████████████ ██████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████████████ ██████████████████████ ██████████████ ██████ | 1% House Edge |
|
|
|
|
BettorChain.com
Member
Offline
Activity: 79
Merit: 10
|
|
December 26, 2016, 05:09:54 PM |
|
The HE of these bets is: 0.27%. They should be not accepted when MoneyPot 2.0 will be released. (0.50% will be the minimum house edge)
|
|
|
|
AcoinL.L.C
Legendary
Offline
Activity: 1036
Merit: 1000
|
|
December 26, 2016, 07:46:42 PM |
|
JPR, as dooglus said, it means very soon: it does not mean that moment. I appreciate you are trying to make it sound like we are giving ourselves different rules than others, but be more direct instead of trying to slyly insinuate it.
|
|
|
|
JackpotRacer
Legendary
Offline
Activity: 1946
Merit: 1014
All Games incl Racer and Lottery game are Closed
|
|
December 27, 2016, 04:55:15 AM |
|
JPR, as dooglus said, it means very soon: it does not mean that moment. I appreciate you are trying to make it sound like we are giving ourselves different rules than others, but be more direct instead of trying to slyly insinuate it. no need to be sarcastic but you as MP owner should be the one not to implement or have a lower than 0.5% HE in your own apps and set an example when asking/warning all app owners not to have lower than 0.5% HE
|
|
|
|
BetKing.io
Legendary
Offline
Activity: 1400
Merit: 1021
|
|
December 27, 2016, 11:35:12 AM |
|
Merry Christmas everyone. Support will be a little slow today and tomorrow (as all of the Moneypot team is celebrating the holidays). We apologize for that but will try to check in from time to time if there are any issues. A lot of good points were brought up and I will address them after the holidays (or if I can sneak in some time). @Fiscorcle, there will be optional kelly multipliers in the future You are joking right? Your dev CrazyCraig jumped in my thread complaining that I'm not allowed a holiday! Even though what was actually happening was totally different as I'm sure you are aware
|
|
|
|
CrazyCraig
Sr. Member
Offline
Activity: 501
Merit: 340
Bye Felisha!
|
|
December 27, 2016, 03:30:54 PM |
|
Merry Christmas everyone. Support will be a little slow today and tomorrow (as all of the Moneypot team is celebrating the holidays). We apologize for that but will try to check in from time to time if there are any issues. A lot of good points were brought up and I will address them after the holidays (or if I can sneak in some time). @Fiscorcle, there will be optional kelly multipliers in the future You are joking right? Your dev CrazyCraig jumped in my thread complaining that I'm not allowed a holiday! Even though what was actually happening was totally different as I'm sure you are aware I didn't say you weren't allowed a holiday. I said that you had my trust and that I was waiting paitently and I understand why others have the right to frustrated. Look at the context in which I responded. But truthfully, best believe if our site crashed I would be there responding to it. I wasn't attacking you, just trying to justify why others were upset to someone who was trolling them. I have no beef towards your site or you.
|
|
|
|
|