PDA

View Full Version : Step Monte Carlo Simulation


fnord_too
08-29-2005, 10:51 AM
In MicroBob's thread I mentioned this, but figured I would start a thread since my response is burried.

I am going to whip up a little Monte Carlo to calculate EV for these things. Here is what I am thinking:

Select the level to begin the trials on.

Assume finish percentages for each step level.

Each trial runs until the player busts out of the steps and/or cashes real money.

Run 10k trials for 10K players (with same finsih expectations).

Track Net, total tournies played/tournies per trial, roi, last trial negative, max loss, total number of each level played, what ever else I can think of/you all come up with.

Maybe add the option to bail at a certain level, i.e. see how not even bothering to play below a certain level affects EV and other stats. (Wouldn't it be great if you could compine lower level entries for a higher level one?)

Comments welcome, and especially comments that involve expected finish information. I will start with the standard 1 and 2 table steps, and add mini/step ahead's if there is interest.

I should crank this out today if I don't have enough real work to keep me occupied at work.

fnord_too
08-29-2005, 02:06 PM
Something just hit me: If you win an entry to another step, does the vig on that step contribute to rake back? If so I need to track that, too, I think.

fnord_too
08-29-2005, 03:09 PM
Well, I have the one table step monte carlo written, I just need some percentages for the cashing spots at each tourney and what data you all want to see. I threw in some notional ones, but the ROI was rediculously high (not saying it is wrong, but it was in the 50+% area).

Also, if you have an average time for each level, we can figure out hourly rate (per table hour, that is).

If there is any interest in this information give me some input, else just let this thread die.

Don't worry about your estimates being off, just say what you think reasonable numbers are and if there are multiple responses I will pick a middle ground, here, I will give you a template to fill out:

One Table Original Steps
Step 1:
Advance to Step 2 percentage:

Step 2:
Advance to Step 3 percentage:
Repeat Step 2 percentage:
Loop back to Step 1 + $3 percentage:

Step 3:
Advance to Step 4 percentage:
Repeat Step 3 percentage:
Loop back to Step 2 percentage:
Loop back to Setp 1 + $3 percentage:

Step 4:
Advance to Step 5 percentage:
Repeat Step 4 percentage:
Loop back to Step 2 percentage:
Loop back to Step 1 + $8 percentage:

Step 5:
First percentage:
Second percentage:
Third percentage:
Fourth percentage:

AliasMrJones
08-29-2005, 03:17 PM
I think this is very interesting. Are you targeting the regular step or mini-step? I would think for the first few levels standard "good player" finish distributions for equivilent buy-in standard STT's would work, though this is purely speculation. How does the play improve at each step?

fnord_too
08-29-2005, 03:20 PM
Both, but for right now the one table original steps. I have never played them, so I don't know how the play is. I played some two tables, and steps 3 and 4 seemed kind of soft at the time, but I was playing early in their existance.

maddog2030
08-29-2005, 03:49 PM
I haven't played much of the steps, so I can't offer much input as to realistic finish distributions, but I am very interested in the results.

AliasMrJones
08-29-2005, 04:13 PM
[ QUOTE ]
Both, but for right now the one table original steps. I have never played them, so I don't know how the play is. I played some two tables, and steps 3 and 4 seemed kind of soft at the time, but I was playing early in their existance.

[/ QUOTE ]

I asked because in the original steps, step 1, places 1 and 2 make it to the next level, but in the mini's only 1st place gets the golden ticket. You're going to have to target one or the other as this is a pretty big difference.

Jman28
08-29-2005, 04:17 PM
[ QUOTE ]
I threw in some notional ones, but the ROI was rediculously high (not saying it is wrong, but it was in the 50+% area).

[/ QUOTE ]

This doesn't seem high to me, as you will really be playing many, many games. Do you have a way to calculate the avg # of tourneys played per STEP 1 buyin?

Newt_Buggs
08-29-2005, 04:28 PM
looks intereting, I'm definitly interested in the reults even though I don't play steps

for the rakeback, I think that it is calculated into rakeback. You can only get 20% on party now though right?