vpFREE2 Forums

VPW & MS KBJW

#1

Several months ago, I discovered what appeared to me to be errors in
the Video Poker for Winners (VPW) software when playing Multi-Strike
Kings or Better Jokers Wild (MS KBJW) with the payscale 20/7/5 for
4K/FH/FL. The software indicates this payscale has an expected return
of 100.83%. Some may consider this to not be of interest, as this
payscale for MS is not likely to be found. However, I did find it
several months ago. It may no longer exist.

For several hands with this payscale, VPW seems to use the single line
strategy (or top line strategy) for all lines including the bottom
line. For example, single line KBJW would prefer SF3 (0 gaps) over ST4
(0 gaps). Using the 6/4/2 adders for MS, which is the typical method
to determine MS strategies, the bottom line (with the 6 adder) would
prefer ST4 (0 gaps) over SF3 (0 gaps). VPW prefers SF3 (0 gaps) over
ST4 (0 gaps) on all four lines. There are many other examples for this
payscale where the VPW preference does not match the preference when
using the 6/4/2 adders.

I don't know whether VPW is more accurate and actually calculates the
best choice without using 6/4/2 adders. However, some examples such as
the one I describe here suggest VPW may be in error.

I exchanged e-mails with Bob Dancer on this subject, but he provided
only a superficial response. Do others have knowledge on whether VPW
can be trusted for this MS payscale?

BigKen

#2

bigkensd wrote:

Several months ago, I discovered what appeared to me to be errors in
the Video Poker for Winners (VPW) software when playing Multi-Strike
Kings or Better Jokers Wild (MS KBJW) with the payscale 20/7/5 for
4K/FH/FL.

I don't know whether VPW is more accurate and actually calculates the
best choice without using 6/4/2 adders. However, some examples such
as the one I describe here suggest VPW may be in error.

I exchanged e-mails with Bob Dancer on this subject, but he provided
only a superficial response. Do others have knowledge on whether VPW
can be trusted for this MS payscale?

With cursory review, I'm fully satisfied that VPW is erroneous here --
the discrepancy from the 6/4/2 calculation is of a far greater
magnitude than either the game ER or actual hit rate can account for
(the two variables that can yield a deviation -- "6/4/2" presumes a
100% ER and 50% hit rate, factoring in Free Ride appearances).

I look to give this a closer examination later.

- Harry

#3

It doesn't. It did last quite a while, though, a good 6-8 months. If
you put a substantial amount of play on the machine, chances are that
your business is no longer welcome at the casino in question as well.

The problem was that among other things, it started attracting people
who love to play with cheat sheets. There were other 100% games on the
machine, and a couple of players were pounding away, sheet in hand.
The slot department at this particular casino is a little more
attentive than usual, so this wasn't a good sign. Or maybe it was
because they were probably actually losing money on the games itself.

Unless, of course, we played it at two different casinos ;).

Generally speaking, the 6-4-2 thingie works well for me when
constructing a strategy chart.

ยทยทยท

--- In vpFREE@yahoogroups.com, "bigkensd" <kmorrissd@...> wrote:

Several months ago, I discovered what appeared to me to be errors in
the Video Poker for Winners (VPW) software when playing Multi-Strike
Kings or Better Jokers Wild (MS KBJW) with the payscale 20/7/5 for
4K/FH/FL. The software indicates this payscale has an expected return
of 100.83%. Some may consider this to not be of interest, as this
payscale for MS is not likely to be found. However, I did find it
several months ago. It may no longer exist.

#4

bigkensd wrote:

> Several months ago, I discovered what appeared to me to be errors
> in the Video Poker for Winners (VPW) software when playing
> Multi-Strike Kings or Better Jokers Wild (MS KBJW) with the
> payscale 20/7/5 for 4K/FH/FL.
>
> I don't know whether VPW is more accurate and actually calculates
> the best choice without using 6/4/2 adders. However, some examples
> such as the one I describe here suggest VPW may be in error.
>
> I exchanged e-mails with Bob Dancer on this subject, but he
> provided only a superficial response. Do others have knowledge on
> whether VPW can be trusted for this MS payscale?

Earlier, I (Harry) replied:

With cursory review, I'm fully satisfied that VPW is erroneous here -
the discrepancy from the 6/4/2 calculation is of a far greater
magnitude than either the game ER or actual hit rate can account for
(the two variables that can yield a deviation -- "6/4/2" presumes a
100% ER and 50% hit rate, factoring in Free Ride appearances).

I look to give this a closer examination later.

Ok, just gave this another look for a few more minutes. I'm not sure
what's going on with the MS KBJW strategy, but I believe this may
relate back to a previous VPW JW strategy problem identified in a post
last year: http://groups.yahoo.com/group/vpFREE/message/71990

Game strategies boil down to a comparison of calculated EV's for each
potential hold in a hand. While the EV's presented by VPW for MS 9/6
Jacks correlate to those I calculate using WinPoker, the VPW KBJW
values are wildly off.

If you play MS KBJW Level 1 and are dealt {5c 6c 7c 8s Jd}, the 3SF
hold EV is 8.027 and the 4ST EV is 8.437 -- calculated using WinPoker
(two separate methods were used to arrive at the values for strong
confidence in these values).

However, VPW reports these values as 22.48 and 19.69 respectively.
Consequently, VPW incorrectly advises holding the 3SF.

There's no way to make sense of the VPW values. Without going into
greater explanation, under no circumstance can the EV's for these
particular holds be greater than 10.

Without culling through the various games, I can't say what games
might contain a similar strategy error, but I wouldn't be surprised if
there are other JW games where the problem exists as well. (I don't
envision that any of the non-JW games are involved, however.)

Further, I don't know if the original JW strategy error reported last
year has since been corrected with a fix available via download -- if
so, it may be that the problem that has surfaced here has similarly
been corrected.

- Harry