vpFREE2 Forums

Simulator

#1

I recently put together a simple VP simulator to look a various
scenarios and provide information about those scenarios. For example,
the following scenario evaluates 9-5 DDB. It simulates 10-10,000 hand
sessions and provides statistics for the run. It then runs that
scenario 100 times to simulate 100 different people trying this
scenario. The last two lines relate to this exercise.

···

---------------------------------------------------------------
seed = 2334555
9-5DDB 0.978639
Hands/session = 10000 with 1 total royals and 40% session wins
Total wagered = 500000 for payback of 98.238 (-881/session)
Average bet = 5, CB per session at .2% = 100
Exceeded expection 51 times and 100% 15 times
Highest payback = 103.303, lowest payback = 93.022
---------------------------------------------------------------

One interesting fact from this simulation is that 15% of the
simulations actually produced over 100% payback. This clearly
demonstrates how a higher variance game can broaden the bell curve.
Since 10 sessions could represent 5 years of play for a twice-a-year
gambler it's not hard to imagine someone thinking AP was unnecessary.
On the other hand what follows is the same number of hands for FPDW.

---------------------------------------------------------------
seed = 2334555
FPDW 1.00762
Hands/session = 10000 with 1 total royals and 60% session wins
Total wagered = 500000 for payback of 101.072 (536/session)
Average bet = 5, CB per session at .2% = 100
Exceeded expection 45 times and 100% 67 times
Highest payback = 105.029, lowest payback = 96.863
---------------------------------------------------------------

In this case a full 33% of the players would not exceed 100% payback.
Once again, it's not too hard to imagine these players might think AP
was a bunch of crap.

Clearly, for players who play far more than 100K hands every month
this is not too meaningful. However, I thought it might be
interesting to many of us who never really looked at it this way.

Finally, if anyone has a scenario they'd like to see simulated or
additional output, I'd be happy to oblige if possible (I can't do it
for every game).

Dick

#2

Great work! How about a simulation of full pay pickem with .50% and .75% cash back?

mroejacks <rgmustain@att.net> wrote: I recently put together a simple VP simulator to look a various
scenarios and provide information about those scenarios. For example,
the following scenario evaluates 9-5 DDB. It simulates 10-10,000 hand
sessions and provides statistics for the run. It then runs that
scenario 100 times to simulate 100 different people trying this
scenario. The last two lines relate to this exercise.

···

---------------------------------------------------------------
seed = 2334555
9-5DDB 0.978639
Hands/session = 10000 with 1 total royals and 40% session wins
Total wagered = 500000 for payback of 98.238 (-881/session)
Average bet = 5, CB per session at .2% = 100
Exceeded expection 51 times and 100% 15 times
Highest payback = 103.303, lowest payback = 93.022
---------------------------------------------------------------

One interesting fact from this simulation is that 15% of the
simulations actually produced over 100% payback. This clearly
demonstrates how a higher variance game can broaden the bell curve.
Since 10 sessions could represent 5 years of play for a twice-a-year
gambler it's not hard to imagine someone thinking AP was unnecessary.
On the other hand what follows is the same number of hands for FPDW.

---------------------------------------------------------------
seed = 2334555
FPDW 1.00762
Hands/session = 10000 with 1 total royals and 60% session wins
Total wagered = 500000 for payback of 101.072 (536/session)
Average bet = 5, CB per session at .2% = 100
Exceeded expection 45 times and 100% 67 times
Highest payback = 105.029, lowest payback = 96.863
---------------------------------------------------------------

In this case a full 33% of the players would not exceed 100% payback.
Once again, it's not too hard to imagine these players might think AP
was a bunch of crap.

Clearly, for players who play far more than 100K hands every month
this is not too meaningful. However, I thought it might be
interesting to many of us who never really looked at it this way.

Finally, if anyone has a scenario they'd like to see simulated or
additional output, I'd be happy to oblige if possible (I can't do it
for every game).

Dick

vpFREE Links: http://members.cox.net/vpfree/Links.htm

  SPONSORED LINKS
        Online gambling Outdoor recreation Recreation software Gambling
    
---------------------------------
  YAHOO! GROUPS LINKS

    Visit your group "vpFREE" on the web.
    
    To unsubscribe from this group, send an email to:
vpFREE-unsubscribe@yahoogroups.com
    
    Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

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

[Non-text portions of this message have been removed]

#3

Oh you probably can't do pickem with that! What was I thinking?

Donald Ross <hedonist144@yahoo.com> wrote: Great work! How about a simulation of full pay pickem with .50% and .75% cash back?

mroejacks <rgmustain@att.net> wrote: I recently put together a simple VP simulator to look a various
scenarios and provide information about those scenarios. For example,
the following scenario evaluates 9-5 DDB. It simulates 10-10,000 hand
sessions and provides statistics for the run. It then runs that
scenario 100 times to simulate 100 different people trying this
scenario. The last two lines relate to this exercise.

···

---------------------------------------------------------------
seed = 2334555
9-5DDB 0.978639
Hands/session = 10000 with 1 total royals and 40% session wins
Total wagered = 500000 for payback of 98.238 (-881/session)
Average bet = 5, CB per session at .2% = 100
Exceeded expection 51 times and 100% 15 times
Highest payback = 103.303, lowest payback = 93.022
---------------------------------------------------------------

One interesting fact from this simulation is that 15% of the
simulations actually produced over 100% payback. This clearly
demonstrates how a higher variance game can broaden the bell curve.
Since 10 sessions could represent 5 years of play for a twice-a-year
gambler it's not hard to imagine someone thinking AP was unnecessary.
On the other hand what follows is the same number of hands for FPDW.

---------------------------------------------------------------
seed = 2334555
FPDW 1.00762
Hands/session = 10000 with 1 total royals and 60% session wins
Total wagered = 500000 for payback of 101.072 (536/session)
Average bet = 5, CB per session at .2% = 100
Exceeded expection 45 times and 100% 67 times
Highest payback = 105.029, lowest payback = 96.863
---------------------------------------------------------------

In this case a full 33% of the players would not exceed 100% payback.
Once again, it's not too hard to imagine these players might think AP
was a bunch of crap.

Clearly, for players who play far more than 100K hands every month
this is not too meaningful. However, I thought it might be
interesting to many of us who never really looked at it this way.

Finally, if anyone has a scenario they'd like to see simulated or
additional output, I'd be happy to oblige if possible (I can't do it
for every game).

Dick

vpFREE Links: http://members.cox.net/vpfree/Links.htm

  SPONSORED LINKS
        Online gambling Outdoor recreation Recreation software Gambling
    
---------------------------------
  YAHOO! GROUPS LINKS

    Visit your group "vpFREE" on the web.
    
    To unsubscribe from this group, send an email to:
vpFREE-unsubscribe@yahoogroups.com
    
    Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

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

[Non-text portions of this message have been removed]

vpFREE Links: http://members.cox.net/vpfree/Links.htm

  SPONSORED LINKS
        Online gambling Outdoor recreation Recreation software Gambling
    
---------------------------------
  YAHOO! GROUPS LINKS

    Visit your group "vpFREE" on the web.
    
    To unsubscribe from this group, send an email to:
vpFREE-unsubscribe@yahoogroups.com
    
    Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

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

[Non-text portions of this message have been removed]

#4

Why not just compute the PDF? It (and the CDF computed from it) gives you eveything you
(seem to) want to know. For example, it will tell you exactly the percentage of "players" who
would end up with 100% return. PDF's can be computed with arbitrary accuracy (someone
recently posted a bt of code that even handles RoR via the PDF) and they produce much more
meaningful answers than running sets of 100-realizations ever will (for VP games). Just my
$0.02.

···

--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@a...> wrote:

I recently put together a simple VP simulator to look a various
scenarios and provide information about those scenarios. For example,
the following scenario evaluates 9-5 DDB. It simulates 10-10,000 hand
sessions and provides statistics for the run. It then runs that
scenario 100 times to simulate 100 different people trying this
scenario. The last two lines relate to this exercise.

-

#5

and .75% cash back?

This is a run of 10-10,000 hand sessions as before. I can run run any
number of hands and any number of sessions so if this is not what you
want then let me know. There's nothing in the simulation itself that
depends on the cash back, so to get the value for .75 all you need to
do is multiply this cb by 1.5. I can also change the denomination and
the number of players from 100 to any other reasonable number. That
said, here is a result for pickem.

seed = 2334555
Pickem 0.999531
Hands/session = 10000 with 0 total royals and 50% session wins
Total wagered = 500000 for payback of 100.3778 (188.9/session)
Average bet = 5, CB per session at 0.5 = 250
Exceeded expection 45 times and 100% 43 times
Highest payback = 103.5372, lowest payback = 97.077

···

--- In vpFREE@yahoogroups.com, Donald Ross <hedonist144@y...> wrote:

Great work! How about a simulation of full pay pickem with .50%

#6

Why not just compute the PDF? It (and the CDF computed from it)

gives you eveything you

(seem to) want to know. For example, it will tell you exactly the

percentage of "players" who

would end up with 100% return. PDF's can be computed with arbitrary

accuracy (someone

recently posted a bt of code that even handles RoR via the PDF) and

they produce much more

meaningful answers than running sets of 100-realizations ever will

(for VP games). Just my

$0.02.

A lot more data can be collected through a simulation. I can show the
results for multiple session so you can get a feel for how a game may
play over time. You can't really get that throught a PDF. For
example, I can provide the following breakdown for the 10 sessions of
pickem I just posted:

Hands in session = 10000, with payback of 99.89
Hands in session = 10000, with payback of 99.4
Hands in session = 10000, with payback of 96.38
Hands in session = 10000, with payback of 103.498
Hands in session = 10000, with payback of 96.16
Hands in session = 10000, with payback of 101.17
Hands in session = 10000, with payback of 98.93
Hands in session = 10000, with payback of 105.08
Hands in session = 10000, with payback of 100.76
Hands in session = 10000, with payback of 102.51

I also think a lot of folks just aren't that compfortable with
statistics. A real simulation provides a greater sense of realism.
Just my .02.

Dick

···

--- In vpFREE@yahoogroups.com, "cdfsrule" <groups.yahoo@v...> wrote:

#7

You can collect as much data as you want via simulation and will NEVER get what you can
from the PDF (well, not exactly never. I guess if you ran the simulation for an eternity you
would get back the PDF exactly).

BTW, all you are doing in the simultion is approximating the PDF. Nothing more. But it's
not a particular good approximation when you only do 100 realizations. After all, are there
really only 100 possible outcomes after 100,000 hands?

As for confusing people, the results that are reported to this is can be in the same exact
format regardless of the source (simulation or PDF).

BTW, The one thing the simulation is good for is to get "a sense of the swings over time".
But to do that, you have to look at the 100,000-hand long histories, not the final results.

···

-> --- In vpFREE@yahoogroups.com, "cdfsrule" <groups.yahoo@v...> wrote:

>
> Why not just compute the PDF?

-- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@a...> wrote:

A lot more data can be collected through a simulation. I can show the
results for multiple session so you can get a feel for how a game may
play over time. You can't really get that throught a PDF. For
example, I can provide the following breakdown for the 10 sessions of
pickem I just posted:

Hands in session = 10000, with payback of 99.89
A real simulation provides a greater sense of realism.
Just my .02.

Dick

#8

Give the guy some credit (uh....make that a lot of credit).

And, it is not just the end product that he came up with. Just think of the overall
understanding and insight that he has garnered by virtue of his going through this
exercise. We should be encouraging people to look at things that are relevant, not
discourage them.

We all look at things from different perspectives. One size does not fit all. Let's
appreciate the work done by others that can give all of us a little bit more knowledge and
understanding.

.....bl

You can collect as much data as you want via simulation and will NEVER get what you

can

from the PDF (well, not exactly never. I guess if you ran the simulation for an eternity

you

···

--- In vpFREE@yahoogroups.com, "cdfsrule" <groups.yahoo@v...> wrote:

would get back the PDF exactly).

#9

You can collect as much data as you want via simulation and will

NEVER get what you can

from the PDF (well, not exactly never. I guess if you ran the

simulation for an eternity you

would get back the PDF exactly).

Absolutely true. However, running several million hands will get you
very close. A PDF has it's place. It's very good at answering many
what if questions but it can never provide all the information
provided by a simulation, nor give a same "feel".

BTW, all you are doing in the simultion is approximating the PDF.

Nothing more. But it's

not a particular good approximation when you only do 100

realizations. After all, are there

really only 100 possible outcomes after 100,000 hands?

Wrong, for example you can give the max/min hands between RFs and
give people a feel for the potential good/bad times.

As for confusing people, the results that are reported to this is

can be in the same exact

format regardless of the source (simulation or PDF).

Wrong again, you can't report events that you can't reasonably
capture. You can give approximations but they will never provide the
same "feel" as the results of a simulation.

BTW, The one thing the simulation is good for is to get "a sense of

the swings over time".

But to do that, you have to look at the 100,000-hand long

histories, not the final results.

That's exactly what I've been stating. That's why my initial post
asked what kind of information people would like to see.

Dick

···

--- In vpFREE@yahoogroups.com, "cdfsrule" <groups.yahoo@v...> wrote: