vpFREE2 Forums

Max ER VS "other" startegies

From: Steve Jacobs <jacobs@xmission.com>
Date: Fri Feb 20, 2004 5:33 am
Subject: Re: [vpFREE] Re: Max ER VS "other" startegies
Min-Risk Strategy

In looking over the "recommended" strategy in this posting, what jumps out
at me is that KT, QT, and JT are held more often than in the max EV
strategy. For instance, JT is held above AJ, KJ, or QJ unsuited.

So, my best understanding of the whole issue is this. If I'm dealt
QsJdTd3h2c, & I'm looking to maximize my profit over the long term, with no
bankroll considerations, then I hold QJ. If I'm concerned about bankroll &
prefer to play a strategy that minimizes my "risk of ruin," then I hold the
JT suited.

Do I understand correctly?

If so, it does seem odd. Holding JT suited in this case, which takes a shot
at higher payoffs by reducing the chance of a small (at least even money)
payoff... offhand it would seem that that would INCREASE my variance & risk
of ruin.

Stuart (RandomStu)

Yes, this is right. In fact, we can test this sample
hand:

QsJdTd3h2c

Holding QJ results in the following:
16215/10037/5022/711/281/144/0/18/2/0/0 = 2.4900 EV,
4.1898 SD
Holding JT:
16215/11987/2847/711/281/204/161/18/2/3/1 = 2.4841 EV,
31.98 SD.

If we assume playing this hand repeatedly, paying,
say, 2.4 units to play (so the game is slightly
positive), we get an RoR from QJ of 7.08%, and from JT
of 95.04%.

(The risk of ruin constants are 0.01103 and 0.000212).

The min-risk strategy is defined by the following
characteristic -- it maximizes the probability that
you will reach an arbitrary bankroll size X_1 starting
from a bankroll size X_0 before going broke.

For positive games, this often means playing a tad
more conservatively and eschewing variance since you
have an edge. It's clear that if you could play VP
with ZERO variance, you would reach a bankroll goal
100% of the when playing a positive game.

But for negative games, this means playing more
aggressively. If you played a negative game with zero
variance, you'd have no chance of reaching the
bankroll target. Hence for negative games, the
"min-risk" strategy adjustment is to play more
aggressively. Like holding JT vs QJ.

Jerrod Ankenman

···

--- Stuart Resnick <sresnick2@comcast.net> wrote:

In looking over the "recommended" strategy in this
posting, what jumps out
at me is that KT, QT, and JT are held more often
than in the max EV
strategy. For instance, JT is held above AJ, KJ, or
QJ unsuited.

So, my best understanding of the whole issue is
this. If I'm dealt
QsJdTd3h2c, & I'm looking to maximize my profit over
the long term, with no
bankroll considerations, then I hold QJ. If I'm
concerned about bankroll &
prefer to play a strategy that minimizes my "risk of
ruin," then I hold the
JT suited.

Do I understand correctly?

If so, it does seem odd. Holding JT suited in this
case, which takes a shot
at higher payoffs by reducing the chance of a small
(at least even money)
payoff... offhand it would seem that that would
INCREASE my variance & risk
of ruin.

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

A word (or two or more, as is my habit :wink: about alternate strategies:

For the typical player, at denominations of $1 or less, they are an
academic (and likely fruitless) pursuit.

Tell me that you've honed your play so that in practice your errors
cost you no more than .01% (ok, I'll allow .02%) and I'll concede that
there's room that such precision play allows for play optimization
toward some goal other than "max ER". But I assert that for at least
95% of the players here (more likely 99%+), adopting alternate
strategies will frustrate and work counter to all possible play goals.

Examine the numbers cited for the example in the Alternate Strategy
FAQ discussed elsewhere on the vpFREE website:
http://members.cox.net/vpfree/FAQ_S.htm

···

------

For a $.25 9/6 Jacks progressive with a RF payout of $1300, the
greatest reduction in expected play cost between royals is from $984.3
to $976.0, a difference of about $8. Now, under ideal circumstances,
this is achieved at a nominal ER cost (.05%).

I won't argue whether this may be an efficient tradeoff and in fact
will concede that it may very well be. But, for this example, I have
my doubts that the average player, if they tackle that royal and
misses after playing a cycle through the machine, is going to go home
and crow about the $8 they theoretically saved by using alternate
strategy. Ditto in the case of the $1 player who has come home with
an expected $32 more in their pocket. (If they successfully hit, they
could give a s*** about the savings -- but that's not the focus of
alternate strategy.)

Admittedly, this example poorly illustrates the potential for
alternate strategy in some of the extreme cases -- which likely
involve pursuing plays that truly stretch the bankroll of a player and
probably involve either a much higher variance than usual play or a
higher play denomination (or both). Nonetheless, I'll shortly discuss
why such expected savings are illusory for most players.

------

Let's take a quick look at the bankroll requirement reduction detailed
first. I'm going to restate these numbers in terms of bets, to get
away from the play denomination distinction. The greatest bankroll
reduction noted is from 4516 bets to 4413 bets. I'm hard pressed to
believe that a bankroll requirement reduction of 100 bets (an amount
any Jacks player should be prepared to lose during an unfortunate
hour's play) is sufficient to evoke a huge sigh of relief from most
players.

Again, I won't dispute that under extreme play circumstances the
bankroll considerations involved may be very significant for the
precision player. But the typical player isn't likely to find any
relief here.

------

Why am I so dubious about the benefit that might be derived by the
typical player? Merely because of some of the unwieldy logistics
involved.

From the values just cited, it should be clear that they are likely
very sensitive to changes in ER -- such as would be the case if the
cost of player errors were to increase.

Now, for a precision player (again, defined as a player whose practice
errors cost .02% at the very most) I'm willing to believe that they'll
have the discipline and expertise to hone their alternate strategy
play to similar levels. But that's very unlikely to be the case for a
typical player.

This is particularly the case for the novice who's armed with an
alternate strategy, but doesn't have the understanding with which to
construct one themself under a given goal. There's no software that
I'm aware of that will alert the player to practice plays that err
when measured against an alternate strategy.

Certainly WinPoker, which doesn't permit the entry of an alternative
strategy chart, fails on this count. However, even Frugal VP - which
is notable in allowing users to tweak strategies and select them for
practice - won't flag any deviation from the alternate strategy
provided that the play is consistent with standard strategy. (I
apologize if this is a misreprestation.)

But, say a player does grasp how to construct a particular alternate
strategy and can determine a paytable that's consistent with that
strategy. When scaled where a bet unit is 5 credits, fractional
entries will be required -- something that most software doesn't
permit (neither BDWP nor FVP do). It doesn't take a rocket scientist
to rescale the chart to integer values (a single bet may now "return"
28 credits for a high pair and 22,203 credits for a RF). So, I won't
discount the possibility that a motivated player will surmount this
hurdle.

------

But, setting all this aside, here's my real take on this for the
typical player. A strongly motivated player who's reduced the cost of
their errors to .05% (a level that still represents recreational play,
albeit with more precision than most) is unlikely manage such strong
accuracy when confronted with more the more unwieldy practice just
discussed (and which more than likely applies to a subset of their
overall play). If they possessed that level of discipline, they'd be
playing standard strategy at the <.02% error cost precision otherwise
discussed (something I expect that a very small -- read as "tiny" --
fraction of players here can lay claim to). And I won't even bother
to touch in any detail on actual casino accuracy, for which all but a
handful of players are subject to greater error incidence given the
larger distractions to play.

As noted, the sensitivity of the optimized play cost and bankroll
requirements to play errors means that for most players these
advantages are likely to largely evaporate (if not, in truth, be
aggravated to larger values).

------------------------------------------------------------------

--> Bottom line, for the typical player, there's much greater
opportunity to minimize play cost and reduce bankroll requirement
simply by aggressively honing play. For the player that's subject to
a .25% error cost, simply paring that to a skilled .05% will achieve
80% of the cost between royals noted in the Jacks progressive example,
and will increase ER, rather than decrese it, in the process.

Frankly, I question the efficacy of the play of all but the very most
adept player who opts to "play around with" (I actually have a much
less attractive phrase in mind) alternate strategy concepts.

- Harry

Harry, may I borrow two dollars?.

--- In vpFREE@yahoogroups.com, "Harry Porter" <harry.porter@v...>
wrote:

A word (or two or more, as is my habit :wink: about alternate

strategies:

For the typical player, at denominations of $1 or less, they are an
academic (and likely fruitless) pursuit.

Tell me that you've honed your play so that in practice your errors
cost you no more than .01% (ok, I'll allow .02%) and I'll concede

that

there's room that such precision play allows for play optimization
toward some goal other than "max ER". But I assert that for at

least

95% of the players here (more likely 99%+), adopting alternate
strategies will frustrate and work counter to all possible play

goals.

Examine the numbers cited for the example in the Alternate Strategy
FAQ discussed elsewhere on the vpFREE website:
http://members.cox.net/vpfree/FAQ_S.htm

------

For a $.25 9/6 Jacks progressive with a RF payout of $1300, the
greatest reduction in expected play cost between royals is from

$984.3

to $976.0, a difference of about $8. Now, under ideal

circumstances,

this is achieved at a nominal ER cost (.05%).

I won't argue whether this may be an efficient tradeoff and in fact
will concede that it may very well be. But, for this example, I

have

my doubts that the average player, if they tackle that royal and
misses after playing a cycle through the machine, is going to go

home

and crow about the $8 they theoretically saved by using alternate
strategy. Ditto in the case of the $1 player who has come home

with

an expected $32 more in their pocket. (If they successfully hit,

they

could give a s*** about the savings -- but that's not the focus of
alternate strategy.)

Admittedly, this example poorly illustrates the potential for
alternate strategy in some of the extreme cases -- which likely
involve pursuing plays that truly stretch the bankroll of a player

and

probably involve either a much higher variance than usual play or a
higher play denomination (or both). Nonetheless, I'll shortly

discuss

why such expected savings are illusory for most players.

------

Let's take a quick look at the bankroll requirement reduction

detailed

first. I'm going to restate these numbers in terms of bets, to get
away from the play denomination distinction. The greatest bankroll
reduction noted is from 4516 bets to 4413 bets. I'm hard pressed

to

believe that a bankroll requirement reduction of 100 bets (an

amount

any Jacks player should be prepared to lose during an unfortunate
hour's play) is sufficient to evoke a huge sigh of relief from most
players.

Again, I won't dispute that under extreme play circumstances the
bankroll considerations involved may be very significant for the
precision player. But the typical player isn't likely to find any
relief here.

------

Why am I so dubious about the benefit that might be derived by the
typical player? Merely because of some of the unwieldy logistics
involved.

From the values just cited, it should be clear that they are likely
very sensitive to changes in ER -- such as would be the case if the
cost of player errors were to increase.

Now, for a precision player (again, defined as a player whose

practice

errors cost .02% at the very most) I'm willing to believe that

they'll

have the discipline and expertise to hone their alternate strategy
play to similar levels. But that's very unlikely to be the case

for a

typical player.

This is particularly the case for the novice who's armed with an
alternate strategy, but doesn't have the understanding with which

to

construct one themself under a given goal. There's no software

that

I'm aware of that will alert the player to practice plays that err
when measured against an alternate strategy.

Certainly WinPoker, which doesn't permit the entry of an

alternative

strategy chart, fails on this count. However, even Frugal VP -

which

is notable in allowing users to tweak strategies and select them

for

practice - won't flag any deviation from the alternate strategy
provided that the play is consistent with standard strategy. (I
apologize if this is a misreprestation.)

But, say a player does grasp how to construct a particular

alternate

strategy and can determine a paytable that's consistent with that
strategy. When scaled where a bet unit is 5 credits, fractional
entries will be required -- something that most software doesn't
permit (neither BDWP nor FVP do). It doesn't take a rocket

scientist

to rescale the chart to integer values (a single bet may

now "return"

28 credits for a high pair and 22,203 credits for a RF). So, I

won't

discount the possibility that a motivated player will surmount this
hurdle.

------

But, setting all this aside, here's my real take on this for the
typical player. A strongly motivated player who's reduced the

cost of

their errors to .05% (a level that still represents recreational

play,

albeit with more precision than most) is unlikely manage such

strong

accuracy when confronted with more the more unwieldy practice just
discussed (and which more than likely applies to a subset of their
overall play). If they possessed that level of discipline, they'd

be

playing standard strategy at the <.02% error cost precision

otherwise

···

discussed (something I expect that a very small -- read as "tiny" -

-

fraction of players here can lay claim to). And I won't even

bother

to touch in any detail on actual casino accuracy, for which all

but a

handful of players are subject to greater error incidence given the
larger distractions to play.

As noted, the sensitivity of the optimized play cost and bankroll
requirements to play errors means that for most players these
advantages are likely to largely evaporate (if not, in truth, be
aggravated to larger values).

------------------------------------------------------------------

--> Bottom line, for the typical player, there's much greater
opportunity to minimize play cost and reduce bankroll requirement
simply by aggressively honing play. For the player that's subject

to

a .25% error cost, simply paring that to a skilled .05% will

achieve

80% of the cost between royals noted in the Jacks progressive

example,

and will increase ER, rather than decrese it, in the process.

Frankly, I question the efficacy of the play of all but the very

most

adept player who opts to "play around with" (I actually have a much
less attractive phrase in mind) alternate strategy concepts.

- Harry

chiefvrwc wrote:

Harry, may I borrow two dollars?

Er, so you can stomach a cycle playing for that quarter progressive
with standard strategy rather than optmizing an alternate one??

Whatever the case, you're good for a sawbuck outright on me anytime,
chief (with no dibs on any jackpot :wink:

- H.