Topic: Moneypot just took a huge loss? - page 5. (Read 7588 times)

Moneypot (not to be confused with the game Bustabit which was originally called Moneypot) was launched last April, so it has been running for less than a year AFAIK. Yup the implementation of bet limit (just follow what Kelly criterion suggests) has not been changed since the beginning, but in the past no gambler was dare and lucky enough to make such a big bet on plinko and to win such a big amount. It seems no one even knows that the max bet limit on plinko is actually that high in the past.

Perhaps this is a silly question, but I'm curious why are we just now having this discussion? MoneyPot has been up and running for years, and I'm assuming its usage/implementation of the kelly criterion has remained the same over that time. It's just strange to me that previously there had not been any bets that drew this kind of attention to the % of bankroll kelly would allow to be risked.

Is it because of the additional nuances introduced by multi-results with the Plinko play-style? Or just simply due to the fact that in certain scenarios it's not as intuitive as previously thought?

I am not sure that the risk-reward ratio is appropriate in this case.

My point more is that if the person who is offering an investment personally thinks that such investment carries excessive risks then he should either not offer such investment or should disclose his opinion of such.

No. How it is/was setup is that if you make the maximum bet size, then once out of roughly every 30k bets the bankroll will lose ~99% of it's value. So while gamblers are making the other ~29,999 bets, the bankroll will be steadily increasing, as will the maximum bet size. On the 30k^th roll, the bankroll would lose ~99% of it's value and would then be significantly less attractive to gamble against which would make it less likely that it would ever recover.

At gambling you must have just really lucky. You can not just go win viz. You really need luck and at the right time.
You also need to invest much else win you anything because.

No it's not.

As an investor - I would prefer the Bankroll not to be based on any humans personal risk tolerance level.  (if my assumptions are accurate, then especially ranlos)

I would prefer my investment to be subjected to the extreme risk they said it would be.  I'm fully aware the bankroll could get wiped out and the upside that comes with it.  That's one of the biggest reasons I find it appealing.





You've got it backwards.  Volume reduces variance.

Your grammar is weird, but to be clear I do not and never have owned any part of moneypot.


Don't feed the troll.

It's clear nobody understood the risks, not even the guy who coded moneypot...


That is interesting. It took me a while to understand your point, so I guess others may not understand it too.

You noticed than none of the payouts in the array are zero, so if you bet 1 BTC you can't lose the whole 1 BTC no matter which payout you get. And so in a sense you aren't really risking 1 BTC. The losest payout is 0.3x, so the worst that can happen is that you lose 0.7 BTC. So you recalculated the multipliers such that you get the same result from betting 0.7 BTC with your new payouts as you would have got betting 1 BTC with the original payouts. Then there's a zero in the array, and you're truly risking the whole amount you bet.

Calculate the new multipliers as (oldMult - minOldMult) / (1 - minOldMult)
So the 121 jackpot becomes (121 - 0.3) / (1 - 0.3) = 172.4285


Right. The house edge of the original row of payouts is 0.9195%, and after 'normalizing' it, it goes up to 1.3136%. You're losing the same, but betting smaller, and so are losing a bigger percentage of your stake.


Yes. Before normalization the house edge of this line:
    [1.0656, 1.0585, 1.0368, 1.0155, 0.9616, 0.9841, 0.9895, 0.9918, 0.9962,
     0.9918, 0.9895, 0.9841, 0.9616, 1.0155, 1.0368, 1.0585, 1.0656];
is 0.9997%.

After normalization:
    [2.7083, 2.5234, 1.9583, 1.4036, 0.0, 0.5859, 0.7265, 0.7864, 0.9010,
     0.7864, 0.7265, 0.5859, 0.0, 1.4036, 1.9583, 2.5234, 2.7083]
it is 26.0342%.

'Kelly' tells us the same before and after normalization: risk 99.97% of the bankroll. This doesn't seem overly unreasonable. Who is going to play a 26% house edge game at large enough stakes to stand a chance of winning the whole bankroll?

To gain an intuitive understanding of why Kelly tells us to risk more than 26% of our bankroll on a 26% house edge bet, it helps to see the plinko bet as 17 separate bets. For the low payout bets where the ball lands towards the middle we would want to set quite a high max bet, since the payout is a small multiplier of the stake, but for the high payout bets on the outside edges we would want to set a low max bet. We can only set a single max bet, and so we need to find some way of balancing the competing forces. It turns out we maximise the log of the bankroll by giving more weight to the max bet implied by the more common (inner, low payout) outcomes.

As the wikipedia page says:


And so you should never play a plinko game that doesn't have a 0x somewhere on it?

The chance is 1 in 32768 or 0.00305%. As Rmcdermott927 said in https://bitcointalksearch.org/topic/m.13715691 yesterday, the moneypot staff were looking into it to make sure the bets placed were legitimate, which I think is a common practice.

That's an amazing win such a lucky guy. I haven't played Plinko in my life but what is the probability to hit that 121x? And did Moneypot pay him the whole amount because there is another thread where he claims he is still waiting for his withdrawal.

@Dooglus an interesting note:

Taking the payout line that was used:

[ 121, 47, 13, 5, 3, 1.4, 1, 0.5, 0.3, 0.5, 1, 1.4, 3, 5, 13, 47, 121 ]

We naively assumed the house had a 1% edge, but actually it's an illusion.

We can actually normalize this payout line to:


So when betting 0.7 BTC, it's an identical bet! The EV stays the same (-0.01 BTC or what ever), but now we calculate the house edge, we'll figure it out as 0.01 / 0.7 ... or 40% higher.

The "super payout" line you had, just exacerbates the problem. The most you could lose was ~4% of your bet, so with a 1% house edge, the *real* house edge, after you normalize the payouts would be 25%!

So while crazy, it's not as crazy as you'd first think!

--

One take away for plinko players would be that the house edge is rather misleading. You actually pay the house edge on the entire bet, rather than just what you're risking.

Ahh, I see now. He had said that Plinkopot didn't allow higher multipliers but MP does, which led me to believe that the bet needed would be much, much smaller than that. This makes sense. So essentially, a median needs to be reached between BR risk and allowing that thrilling chance (likely through a capped "if max. win > x% BR, decline, even if it meets Kelly" setup), so as to protect investors and still allow risk-takers to get their fix. And we can essentially classify the bet in question as being a rare event, but still something that's concerning solely because of its impact.

So for argument sake, lets use a gambler with infinite money.

The site starts with a 100 BTC bankroll, this means using Dooglus' payouts the gambler would have to bet 1524 BTC to aim at winning the 99.97% of the bankroll. As the gambler loses money, to the house, the more and more he needs to bet. By the time the gambler wins the 99.97% of the bankroll, he will have probably grown the bankroll so much that the remaining 0.03% is far bigger than the initial bankroll.

So even with infinite money, the gambler is still -EV and the casino still have +expected bankroll growth

Can you please tl;dr this for me then? Something I'm not catching on to I guess. If the risked amount is a % of the bankroll (as per Dooglus), that % remains the same. So let's say there's 100 BTC in the BR, and 99.84% can be lost (99.84). Player loses 100 BTC chasing it. Now there's 200 BTC there, but he can chase after 199.68 of it (meaning his 100 + 99.68/99.84 of what he was chasing), and this could go on forever.

A cap would prevent this, is my understanding (based on a few others also breaking it down). But am I misunderstanding something with the above?

And I absolutely thank you for your help better understanding this, and Dooglus as well, . Despite the hiccups, I think it was fairly alerting.

(Also, I wasn't trying to be abrasive or devalue your assistance and/or previous/current knowledge on the subject, but rather explain that if people that deal with this a lot weren't completely aware, then to expect someone -- since he calls out me, specifically -- with no math background to know it all is a bit asinine)

I was going to avoid this thread, but I don't think that's fair. While I indeed underestimated how much the house can "sanely" risk for many plinko bets, I didn't not fully understand the repercussions. In fact, on numerous occasions I have told investors that they could lose (almost) the entire bankroll with a single bet. (The most trivial example: A dice bet with 99.99% house edge can win 99.99% of the bankroll)


FWIW, even if you don't understand the maths you should at least understand the high level purpose of the kelly criterion. This is almost the exact opposite of reality, and what we just spent hours proving on your behalf.

All the bets have a positive EV from the viewpoint of investors and the site, but that doesn't mean investors will certainly get a profit particularly in the short run. I am not sure what you meant by "rules set by us", since there is no rule set by individual investor and all investment are treated the same way following the Kelly criterion, which unlike on some dice sites that you can set your own Kelly multiplier.

I just want to say that... I'm impressed. Even when two of the more mathematically-inclined people on the forums didn't fully understand the repercussions of the Kelly algorithm, you do. From now on, I will be sure to bring any math-based questions I have directly to you, since you're obviously now the most knowledgeable person on the forums.

Regardless of if you are speaking on behalf of MP or not, I think it says a lot when it is one of the owner's opinion that a controversially large winning bet that was allowed to be made was risking more then what he thinks should have been risked.

Since you were offering bankroll investments, it would be your job to understand how Kelly works. If I had to guess I would say that Stunna did not know how probably fair betting works prior to deciding to open PD, however likely deals with it on a daily basis now. He previously has stated that none of the owners of PD are devs. The same is likely true for other technical aspects of PD.

When you are managing the risks of other's money then you should understand these risks before you start offering bankroll investments. This is similar to how am escrow service should know how to keeps funds secure prior to offering his services.

as being an investor for myself, i believe that the policy is made up to profit the investor and the mp. if the casino function does not follow the rules set by us, should we still acknowledge the whole bet and pay the full price???

I am not speaking on behalf of MP, I'm speaking on behalf of myself. Any official correspondence from MP itself will come from the official account. That said, I personally don't even understand Kelly, how it works, and how to break it down mathematically. It's not my area of expertise and I would never insinuate I know anything about it. I'm also not a developer, nor would I ever claim to be. As such, this entire discussion is completely foreign to me, . Things like Kelly are way over my head. Looks like it's above others', too, with Bitvest's hefty loss today (9.2 BTC loss on a 0.1 BTC bet, with a 160 BTC bankroll). Wondering what their mass loss per roll is.

But then the same could be applied to the investors. On the investment page, there is a reminder "Never invest more than you are willing to lose, and do not invest unless you clearly understand the risks and rewards." and on the FAQ page, it tells investors "We use the generalized kelly criterion, to determine if a bet is acceptable or not". I do agree with you that it may not be the best idea for sites to completely follow Kelly formula in this case though.
Yup dooglus has nothing to do with MP while RHaver no longer owns it. What I was trying to say is that even the math gurus didn't realize the potential problem before it happened and I could understand why the current owners of MP didn't know the problem.

It is his responsibility to know. If this was something that he did not fully understand, then he should not have allowed these kinds of bets in the first place.

(AFAIK) dooglus, nor RHaver are the ones who own MoneyPot, therefore it is not their responsibility to make sure that bets do not expose MoneyPot bankroll investors to excessive risk