Hello Guys,

Can the Danish and Italian Players report on the traffic change since the beginning of the month.

Good day.

Danish players play in the dot-com player pool. As Denmark accounts for <5% of total poker revenue for Amaya, it will not have an appreciable impact on traffic.

I can tell you that the 7-day moving avg of traffic in PokerStars Italy has declined every day since June 7, and its major competitor People's has increased each day since that date. That might show some shift in play, but the change is not significant yet. PSIT still sits much higher than the annual low point, which was reached on May 29.

Thank you Hood.

I find it useful to look at the frequency of chest rewards like this:



Just over half the time, everybody gets the bottom two rewards.
99.99% of the time, nobody is winning one of those top rewards. The chances of you hitting the top reward is roughly the same as you being struck by lightning in your lifetime. (see google :-) )

No matter what rewards you win, if you continue to play, over time they will average out to the EV for each chest (see picture).

Using some information from the blog post of player bnoise from Italy, I calculated the average rakeback as follows.
Each red box = €1.54 rake paid (approx). 4x red box = €6.16 rake paid. rakeback = €0.13x4 = €0.52. rakeback = 8.4%

red chest rakeback = 8.4%
blue chest rakeback = 9.8%

(typically around ~4%, but over time averaging to the above percentages)

Because half the time we all get the bottom two rewards, our rakeback will be half these percentages shown above during these times (i.e. often it will be ~4%), and will average out to the percentages above over time (if we can open enough chests).

red chest requires 280 points.
blue chest requires 675 points.
gold chest requires 18,000 points.

Open 4 chests and you are moved up to next chest colour.

It is uncertain whether we will ever be able to open enough chests (to have any chance of hitting the less frequent awards), because as soon as you open 4 chests you are moved up to the next chest colour. You are moved up until you reach one you cannot beat and presumably you are then stuck earning no rakeback until Stars lets you move back down?
The old system had about 20 steps per level, this one seems to have 4 per level. This seems like another way they may limit how much we get back.


Thanks MaxShark, bnoise, Hood (I used your EV calcs ) and everyone else who posted.

Hope people find this helpful. I think the info is right, anyone feel free to correct this or add more info if you have any. Just trying to get it clear and in one place.

So your basically getting nothing. Struggling regs should leave for another site or just give up altogether to pursue other opportunities. More to life than a card game.

DN really needs to make a blog about how these changes are beneficial for the players, smh.

I think these are the exact words decision makers at Stars did employ when planning the changes

so chromestars lose 50% rb ?

I think the calculations are wrong. Don't have a he time now to do it myself, but basically you need to adjust the ev of each chest to the ev where you take out the top price tier. My guess is that the real rakeback won't then be much lower than the 8,4% and 9,8% for red and blue chests.

Ok, so got home and ran the numbers:

EV without top prize:

red: 0,12
blue: 0,39
bronze: 1
silver: 2,65
gold: 6,91
platinum: 21,42

So the calculation above would give for red chests :

Each red = €1.54 rake paid (approx). 4x red box = €6.16 rake paid. rakeback = €0.12x4 = €0.48. rakeback = 7.8%

It does however get a bit worse when you take out the top 2 tiers of prizes (leaving 96% in):


red: 0,09
blue: 0,31
bronze: 0,8
silver: 2,12
gold: 5,52
platinum: 17,14

Giving for red chests: = €1.54 rake paid (approx). 4x red box = €6.16 rake paid. rakeback = €0.09x4 = €0.36. rakeback = 5.8%

why do you take out the top tier?

you are getting the ev of chests without a top prize then, you just weight is accordingly.

because you can only get 4 chests of each level before going to the next level, so chances are you won't be getting many/enough chests in a lifetime to hit the top tier prize and a more realistic rakeback % will be the one without the top tier?

Which Is why it's gets a weight of .01%.

Because something is mathematically insignificant doesn't make it just not exist. Not that any of this matters.

wtf are you all talking about? the whole point of the new system is that amount you need to rake to get a chest differs from person to person, so all calculations are meaningless in the vacuum

lol, i thought it would be at least 10k for the top tier chest top reward so that there would be a gamble attractor for recs. who the heck would gamble for 1k? ppl roll the dice for life-changing numbers

Will this joke be implemented from the 1st July ?

RIP stars

Hopefully alot of the players will move to Party Poker so the 200 fast forward will be up and running...

+1000000

Not to mention that people underestimate the impact of it being all or nothing. I mean assuming they will want you to rake like 180$ or something ridiculous to open a gold/platinium check most of the time you will ignore your 5-7% of the potential rakeback and just quit the session unless the games are good. So quit a few times given it is a daily system you will not play optimally in terms of rake/chests opened.

I mean won't make much of a difference whether you get 4 or 5 % rakeback so they will pocket with it too because no sane person will grind the system out to get 1% more or so.

I think the larger chests offer less rakeback. 7/180 for a gold anyway.

Ya but that's just the cash reward, right? Don't you also get x amount of star coin (redeemable for cash) and some points towards your next chest as well?

can't think of anything more worthless
3% of your 3% rakeback plus rakeback
so 3.09% rakeback
whoopie doo

One of the more surprising aspects of the Kelly Criterion, the system for choosing optimal bet sizes, was that if the variance on a bet is sufficiently high for a given bankroll, the expected bankroll growth is zero. This is because any series of bets that result in the entire bankroll being lost at least once (busting out) results in zero growth. But such a series of bets makes the individual bets themselves operationally equivalent to having a negative EV, even if the single-bet EV is >0.

A poster above was comparing the top tier chest reward to the lifetime risk of being struck by lightening.

If that's in the probabilistic ballpark of getting the top tier prize, almost no one would have a sufficient bankroll for that component of the EV to cause an expected bankroll growth of greater than zero. Thus, that component of the EV can be eliminated from the equation. So, to rebut what you said above, it does make it just not exist.

One way to reality check that idea would be to imagine if someone tried to pay you for a car you were selling for $10k by handing you 18,000 $1 Mega Millions lotto tickets. Would you even consider taking them for payment? How many Mega Millions tickets would you take for a $10k car, if you knew the average return to player was 55% (about what most state lottos return)? For most people the answer would be that they wouldn't take any amount of Mega Millions tickets, especially if the $10k was a significant part of their net worth.

The chance of scoring the jackpot is mathematically insignificant enough that for the cash-strapped car seller, it might as well just not exist.

This, Stars said as much in their own press releases.

edit: the long-term EV of a bet where the bankroll is being significantly overbet is probably more like equivalent to the negative amount of the bet size, or a -100% edge, from the standpoint of bankroll growth. That's for non-replenishable bankrolls. But because the top tier chest prize can only be won, not lost, it seems appropriate to treat it as 0 rather than negative.

This is completely different to the weighting concept that was discussed, and easily contradicted. For instance, let's say 99.9% of the time you get $1 and 0.1% of the time you get $10000. Then the EV is:

0.999*1+0.001*10000=$0.999+$10=$10.999

Then if you take out the top prize EV without weighting, you are left with $0.999. But since the minimum prize is $1, this makes no sense. Hence you must weight the prizes by calculating 1/0.999*$0.999=$1

So yes, weighting does matter. Not by much, but it does.

The fact that it's worth $.999 and not $1 is a miniscule rounding error. The example you used is actually a fairly probable event. When the probability goes to less than 1/1,000,000, the rounding error becomes nanoscopic. I can't claim applying the Kelly Criterion to a component of what amounts to a bonus is perfectly rigorous. What I'm saying is that it makes sense once the variance gets past a certain point, particularly when that component of the bonus is a significant portion of the total EV that most players expect to get out of the games they're in.

I'm just quoting some poster above. Are the probabilities of hitting the top tier .1%, or are they lightning-strike chances, <.0001%? In the latter case, it makes sense to disregard any EV coming from that component of the chests/bonus/rakeback because the likelihood of hitting it in one year of play or one lifetime of full-time play, for many, is low. Are the top tier bonuses with extremely low hit rates 10% of total EV, as in your example, or are they closer to 50%?

If 50% of rakeback is coming from 6+ sigma events, I wouldn't count that for the same reason I wouldn't take Mega Millions tickets as payment for a car.

Not to mention if the baseline prize is $1, using the language I was using above, the component of the prize that happens .1% of the time would just be worth $9,999. We could call the $1 component A, which the player always receives and the $9,999 component B, which he receives .1% of the time.