johnfiske wrote:
Here in Las Vegas, many of the Spinpoker machines are being replaced
with Spinpoker Deluxe. In this version you have the same 15 card
grid but there are now 20 paylines instead of 9. Intuitively I'm
guessing the variance would decrease because as the number of lines
played increases the card positions in the grid become less
important. For example in Spinpoker there are 5 lines that pass
through the middle card in the grid which makes that position
important in 5 out of 9 hand. I'm assuming with 20 paylines, no one
position is important in such a high percentage of hands. Does my
interpretation sound correct?
I'd have to say not, at least not for the reason you suggest and not
in the manner that is likely most important to players.
Now, I realize that's a rather fuzzy response to begin with. But,
when it comes down to it, I'd suggest that the variance of this play
isn't what you're most concerned with. Most players would want to
know that in playing through $x coin-in for each of these two games,
on which one do their face the larger loss risk. That's not the same
as asking which one has the greater variance.
For example, if we were comparing $.25 3-play with $.50 3-play of a
given game, both have the same statistical variance. Variance is
expressed in betting units and isn't affected by denomination.
However, if we were to compare the risk of losing $1000 when playing
through $50,000 of coin-in on each, the $.50 play has the greater
risk. This is because the wagers are concentrated over fewer plays,
reducing the extent to which play converges on game ER.
You'll forgive the diversion here, but because in the case of
Spinpoker vs. Spinpoker Deluxe, the different amount wagered becomes a
factor in the loss potential for any given amount of coin-in. And,
irrespective of the variance statistic for each, Spinpoker Deluxe will
have a much greater loss risk for a given amount of coin-in -- not
only because there are fewer plays for that coin-in, but also because
you now have 20 hands all dependent upon the same matrix of 15 cards
rather than just 9. This translates into greater dependencies in each
play and greater risk.
If all other things were equal and say that SpinDlx hypothetically had
just 18 hands, it should be clear that in playing through any amount
of coin-in at the same denomination, you're at considerably greater
risk because any set of 18 hands played in Spinpoker is dependent upon
two independent 15 card matrices, not just one.
Now, I suspect this addresses what you, and most players, are
concerned about here. But I previously stated that I currently am not
how to calculate that actual variance statistic for SpinPoker.
Obviously I'm even less prepared to determine variance for SpinDlx.
The pure variance statistic comparison is a difficult one to make a
call from the hip on. The fact that there are likely more card
positions in the matrix that have a high number of intersecting hands
shouldn't be a factor that reduces variance vs. Spinpoker. Any
increase in the number of intersection increases dependencies in play
and should increase variance.
However, keep in mind that variance is expressed in terms of "betting
units". The fact that your total bet is divided over more lines will
be a factor that reduces variance. So, if I see this correctly, you
have two variance factors in SpinDlx that work in opposite direction.
Off the top of my head, I'm not sure which of these two would
predominate over the other in driving the direction that variance
moves in comparision to standard Spinpoker.
I'll simply say, hold onto your seat when hitting this game. (And,
hey, if you're comfortable with $.25 Spinpoker, there's a pretty good
chance that you'll be fine with nickel SpinDlx.) In any case, there's
not better teacher than experience (and a few bruises) 
- Harry