Advanced video poker question.

[Zoozie]

In short, does VP analyze tools take the 'correlation' into account when calculating EV?

For my example to become relevant we need a marginal hand in a wild game I think, so lets look at:

Deuces Wild (full pay) with a hand having 2 duece and we have the option to:

# 2 consecutive suited cards, 6-7 or higher, + deuces (3.3404255)
# 2 deuces only (3.2730188)

From my understanding getting the EV for these hands (3.3404255) and (3.2730188) are calculated from holding
the cards and iterate through all combinations and THEN do the same for the second hand. This, 'I think', is different than
from doing it simultaneously from the same deck of remaining cards.

However in the cases when the first hand makes one of the better wins (straight, straight flush, 3 of a kind) there is
sometimes a duece involved amoung these 2 cards (especially in the straight flush, and this contributed a good part to the EV).
Wouldnt we be better of if we only had held the 2 dueces since we would have gotten another duece in the first 2 cards with 1 card still to come.
Actually we are garantied 4 of a kind allready.

The consequence of this would be that optimal play can push the payout% a tiny inch more, which is still far reaching, and
therefore I am VERY likely wrong.

It is early in the morning and I hope this is not (obvious) nonsense, anyway I am used to getting a beating at this forum now

Zoozie

 

[steinhaug]

Wouldnt we be better of if we only had held the 2 dueces since we would have gotten another duece in the first 2 cards with 1 card still to come.
Actually we are garantied 4 of a kind allready.

Very interesting question indeed. I have come to learn that MG casino isn't all that random, there are patterns in the software. Some ways you can win big, some you dont. And since you are a stats freak like me I can PM you some revealing stats if you like as proof of concept.

To answer your question, I do think you are onto something. And if this works good, I am pretty sure you will get the feeling that it's working in 3/4 times - being it's pushing the payout in your favour (And it will eat it back from other players playing "expected" cards).

I almost have a feeling that the MG casino is eating far to much, and that there are scenarious where the player has an edge - making some people winning all the time.

I would atleast try this, write down your stats on it, and if it works shut down this post and do wagering,

 

[GrandMaster]

In short, does VP analyze tools take the 'correlation' into account when calculating EV?

For my example to become relevant we need a marginal hand in a wild game I think, so lets look at:

Deuces Wild (full pay) with a hand having 2 duece and we have the option to:

# 2 consecutive suited cards, 6-7 or higher, + deuces (3.3404255)
# 2 deuces only (3.2730188)

From my understanding getting the EV for these hands (3.3404255) and (3.2730188) are calculated from holding
the cards and iterate through all combinations and THEN do the same for the second hand. This, 'I think', is different than
from doing it simultaneously from the same deck of remaining cards.

However in the cases when the first hand makes one of the better wins (straight, straight flush, 3 of a kind) there is
sometimes a duece involved amoung these 2 cards (especially in the straight flush, and this contributed a good part to the EV).
Wouldnt we be better of if we only had held the 2 dueces since we would have gotten another duece in the first 2 cards with 1 card still to come.
Actually we are garantied 4 of a kind allready.

The consequence of this would be that optimal play can push the payout% a tiny inch more, which is still far reaching, and
therefore I am VERY likely wrong.

It is early in the morning and I hope this is not (obvious) nonsense, anyway I am used to getting a beating at this forum now

Zoozie

I am slightly confused by what you mean. The expected values are calculated by using all the possible combinations of the available cards. If you are playing multi-play video VP, the cards after the discard are drawn independently from separate decks, so there is no correlation. In any case, expectation is linear, so the expectation of the sum of two random variables is the sum of the expectations.

 

[Zoozie]

I am slightly confused by what you mean. The expected values are calculated by using all the possible combinations of the available cards. If you are playing multi-play video VP, the cards after the discard are drawn independently from separate decks, so there is no correlation. In any case, expectation is linear, so the expectation of the sum of two random variables is the sum of the expectations.

I am talking about single line VP, and I understand I did not make myself perfectly clear, since it is hard to explain.
(xy) means xy suited.

You have 22(67) and hold them. You are then dealt 1 card which another duece 2 giving you straight flush, and this is
best possible result. But in this situation wouldnt you rather have held 22 so you would have 222 with two cards to come?
(actually you would prefer this last situation, having the 222).
So in the some of the situations contributing to +EV for the 22(67) where you are dealt another deuce, you would actually
not have held the 67 also.


Maybe I am confused about "so the expectation of the sum of two random variables is the sum of the expectations."
Because the are not really independant as I see it, because in some situations where you get a good hand in the first situation(getting the duece for the straight flush), you would also have a good hand if you had held the two dueces only, since you now have 4 of a kind at least.

 

[caruso]

You are really looking for complications that don't exist. On a FPDW paytable, it's correct to hold consecutive suited cards with two deuces.

You have 22(67) and hold them. You are then dealt 1 card which another duece 2 giving you straight flush, and this is
best possible result. But in this situation wouldnt you rather have held 22 so you would have 222 with two cards to come?

Of course you would. If you knew a third deuce was coming, this would be the same as having three deuces in the initial deal, in which case dropping the SF and going for deuce quads would be the correct play.

If you could always predict your draw cards you'd make a lot of money. Unfortunately, you cannot.

Dancer / Paymar / Tomski et al know their stuff. Trust them.

 

[caruso]

I have come to learn that MG casino isn't all that random, there are patterns in the software.

There are no patterns in Microgaming that don't exist in whichever gambling games their cyber equivalents replicate. Whether or not the Microgaming RNG is random to the twentieth decimal place I don't know. However, it IS random enough to faithfully replicate the games to the satisfacton of any human possible requirement.

 

[cheekymonkey]

There are no patterns in Microgaming that don't exist in whichever gambling games their cyber equivalents replicate.

I think there are. I was playing the 'Pub Style' slots at Ladbrokes. I was playing a 1 coin and the machine was DEAD. You need 3 black bar wins a row to get into the feature. I wasn't getting even close - but, whenever i dropped my coin down to 50p, i kept hitting 2 black bar wins - i just knew that the 3rd black bar win was going to come, so on the 3rd spin i raised my coin to 1 again, but it wouldnt come. When i dropped back to 50p the black bar wins came in again - IMMEDIATELY, i kept repeating the process to make sure it was not just a coincidence. I finally gave in at the 50p level and took the feature the machine was so desperate to give me - AT THAT LEVEL.

 

[steinhaug]

When i dropped back to 50p the black bar wins came in again - IMMEDIATELY, i kept repeating the process to make sure it was not just a coincidence.

Havnt played this game, but it's sounds like your onto some of the patterns I'm referring to. Just like when blowing a lot on high wagering, lowering the bet will usually give you a great win on lower wagering (Compared to the lower bet ofcourse). This is by no chance random.

Or what about the retriggering which happends very often on low wagering, but not at all often on high wagering...

How comes high wagering eats more money % than low wagering? How come the jacpot often appears on x1 and not on x5 or x10?

Back to the patterns, there are truly techniques in MG yhich can be applied for better wins, and they do have to do with wagering amounts and such. The difference with MG and landbased is that we can download the Playcheck logs, which gives us superior power regarding statistics.

If it were totally random, 10.000 spins on $1 should have the same payout % as a 1000 spins on $5, but guess what... It doesnt.