ROI Simulator / Analyzer - Free Software!

[rvg72]

There had been a bunch of discussion about buy-in drops, chances of busting based on starting roll, expectations based on "real" ROI etc so I wrote a program that analyzes all of this. Here is a screen shot:

http://www.kiosk-services.com/0805/roi.jpg

# simulations = number of times to run the sim. 1000 should be fine but obviously the more you do the more accurate the results will be.
Games Per Sim = games per sim (duh) - if you want it to analyze outcomes for sets of 4 then set this to 4 and if you want to see the expected outcome over your next 500 then set this to 500.
1st to 3rd place = input your 1st, 2nd and 3rd breakdown to use in the sim
Actual ROI is calculated after clicking on the Simuate button.
Simulated ROI is the ROI that was simulated - for any decent sample size this should equal the actual ROI. Any differences = variance due to low sample size.

Now the 3 charts...

First one shows every final ROI of your sample, the % frequency that you should expect that ROI and the cumulative total. So the 9% ROI guy above should expect a 20%+ ROI over 1000 SNG's about 1.8% of the time.

The Second chart shows the frequency of the largest buyin drops over the simulation. So, over 1000 SNG's the guy above has about a 95% chance to drop 20 or more buyins over 1000 SNG's.

The Third chart shows the lowest point of your bankroll so, if this guy started playing $55 buyins with only $550 then he has a 43.4% chance to bust out (since he has a 56.6% chance to have his worst drop be 9 buyins or less)

One of the recent threads talked about a poster who fealt he had a real ROI of between 15-20% and that his last 1350 SNG's at 6% (might have the details wrong) was an abberation. A lot of dicussions came out of that thread wondering about % possibilities that this was an aberation. Well check it with this program by entering what you think your real ROI is, set the Games per sim to 1350 and then check the All Results column to see what % of simulations finished at 6% or below.


All of this is based on 10% rake.

Hopefully this software is helpful - all I need is a place to put it so it can be downloaded. Does anyone have some FTP space they could give me? If you are worried that this steals your hole cards and sends them back to me then either don't install it, wait until someone who knows how to look for those things gives it the ok or setup your firewall to block it (it will never try to communicate to any address for any reason)

Rvg

[eastbay]

Nice work. Any chance you could turn this into a web app? I think it would by far get the most use this way.

Also, it would be cool to see probabilities associated with OOTM streaks - something people are always whin... discussing.

eastbay

[rvg72]

[ QUOTE ]
Nice work. Any chance you could turn this into a web app? I think it would by far get the most use this way.

Also, it would be cool to see probabilities associated with OOTM streaks - something people are always whin... discussing.

eastbay

[/ QUOTE ]

Thanks, I can definately add ITM and OTM streaks to this very easily - I'll do it tonight or tomorrow if I can find an hour of peace and quiet...

I wrote it in C# so I could make this an ASP.NET app with a minimal amount of work - only problem would be finding someone to host it since they'd need to be running IIS, have the framework installed etc. If there are any volunteers then I'll do it.

rvg

[bluef0x]

So the 2nd window shows only the chance that it's your LARGEST buy-in drop? Is there anyway you could have it show the chance you experience the drop, not that it's your largest? Or is there a way to derive that stat from what is shown?

This looks good, get it hosted somewhere! /images/graemlins/smile.gif

[Mr_J]

"One of the recent threads talked about a poster who fealt he had a real ROI of between 15-20% and that his last 1350 SNG's at 6% (might have the details wrong) was an abberation. A lot of dicussions came out of that thread wondering about % possibilities that this was an aberation. Well check it with this program by entering what you think your real ROI is, set the Games per sim to 1350 and then check the All Results column to see what % of simulations finished at 6% or below."

This was me, but it was ~12% over 1300 sngs where I am certain my EV is at least 15%, and most likely >20%. It'll be interesting to play around the figures though.

[rvg72]

[ QUOTE ]
So the 2nd window shows only the chance that it's your LARGEST buy-in drop? Is there anyway you could have it show the chance you experience the drop, not that it's your largest? Or is there a way to derive that stat from what is shown?

[/ QUOTE ]

Thanks, you can derive it from the chart - if you want to know the chance that you will experience a 25 buy-in drop then look at the simulation results and you can see that in 19.8% of the 1000 simulations the max buy-in drop was 24 or less so over 1000 SNG's a player with 12%-12%-12% has about an 80% chance to drop 25 or more buy-ins and about a 60% chance to drop 30 buyin's over 1000.

rvg

[rvg72]

[ QUOTE ]
This was me, but it was ~12% over 1300 sngs where I am certain my EV is at least 15%, and most likely >20%. It'll be interesting to play around the figures though.

[/ QUOTE ]

Until a find a host, I ran the numbers and gave you a 20% ROI (14/13/11.5 1st/2nd/3rd) and based on that you would have about a 5% chance to run at 12% over 1350 so not completely out of the question... Now if you are "really" a 15% ROI player (13/12.5/12) then you would have a 35% chance to run at 12% over 1350.

rvg

[suited_ace]

PM me. I have an FTP server on my computer at home, I'll set it up to serve the file until you get a decent server.

[rvg72]

OK, new version with Average Best ITM and Average Worst OOTM streaks (This time I used a player with a 14.5% ROI)

http://www.kiosk-services.com/0805/roi2.jpg

[eastbay]

[ QUOTE ]
[ QUOTE ]
Nice work. Any chance you could turn this into a web app? I think it would by far get the most use this way.

Also, it would be cool to see probabilities associated with OOTM streaks - something people are always whin... discussing.

eastbay

[/ QUOTE ]

Thanks, I can definately add ITM and OTM streaks to this very easily - I'll do it tonight or tomorrow if I can find an hour of peace and quiet...

I wrote it in C# so I could make this an ASP.NET app with a minimal amount of work - only problem would be finding someone to host it since they'd need to be running IIS, have the framework installed etc. If there are any volunteers then I'll do it.

rvg

[/ QUOTE ]

I'd be happy to host it at sitngo-analyzer.com if that works for you.

eastbay

[suited_ace]

OK, until eastbay puts it at his site, I made an FTP server here at home.

You can download the file here (ftp://roipredicter.dyndns.org/ROIPredicter.zip).

[bluef0x]

I tried running the program and it said "failed to initialize properly (0xc0000135). Click OK to terminate the applicaiton'

Anyone else getting this error?

[rvg72]

[ QUOTE ]
I tried running the program and it said "failed to initialize properly (0xc0000135). Click OK to terminate the applicaiton'

Anyone else getting this error?

[/ QUOTE ]

It does require the .NET Framework version 1.1 I believe... If anyone with .NET installed has this problem then let me know.

rvg

[bluef0x]

Oh, sorry.. I'm not on my home computer. Thanks for the program, looks great.

[rvg72]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Nice work. Any chance you could turn this into a web app? I think it would by far get the most use this way.

Also, it would be cool to see probabilities associated with OOTM streaks - something people are always whin... discussing.

eastbay

[/ QUOTE ]

Thanks, I can definately add ITM and OTM streaks to this very easily - I'll do it tonight or tomorrow if I can find an hour of peace and quiet...

I wrote it in C# so I could make this an ASP.NET app with a minimal amount of work - only problem would be finding someone to host it since they'd need to be running IIS, have the framework installed etc. If there are any volunteers then I'll do it.

rvg

[/ QUOTE ]

I'd be happy to host it at sitngo-analyzer.com if that works for you.

eastbay

[/ QUOTE ]

Sounds great - PM your e-mail and I'll send you the zip file. It is only about 10 KB.

Thanks,

rvg

[jdl22]

Nice.

Downloaded it and it seems to work. Not sure I understand the second column. Could somebody again explain what it is?

thanks.

[rvg72]

[ QUOTE ]
Nice.

Downloaded it and it seems to work. Not sure I understand the second column. Could somebody again explain what it is?

thanks.

[/ QUOTE ]

The second column represents the biggest drop in buy-ins. To use an example, let's say you are playing $20+$2's. After 900 SNG's you are up $2200 (or 100 buyins). You then lose $1100 over your next 100 SNG's. By losing $1100 you dropped 50 buy-ins from your peak of +$2200. Buy-In drop measure the biggest difference between a peak and a valley in your bank roll.

The second column shows how big variances are expected to be even if you are a top player. A 20% ROI player can expect to have a 22 buy-in drop over 1000 SNG's. A 5% ROI player can expect a 40 buy-in drop...

rvg

[tigerite]

It looks good, but please, change the title to Predictor, as that's how it's spelt /images/graemlins/wink.gif

[Burno]

This is really cool, thanks for taking the time and effort to create this.

[rvg72]

[ QUOTE ]
It looks good, but please, change the title to Predictor, as that's how it's spelt /images/graemlins/wink.gif

[/ QUOTE ]

/images/graemlins/blush.gif I noticed this last night! I'm going to change the title to something else all together because ROIPredictor isn't even accurate... A couple more minor things I need to do include:

* the first grid totals up to 100.1%
* clicking on the column headers sorts but not properly (2.8% sorts as greater than 12.8% because the 2 > 1)
* allow for < 10% rake (i.e. $200+$15)

If anyone has any other suggestions then let me know.

rvg

[jdl22]

Ok thanks, makes sense.

Excellent program btw.

[eastbay]

[ QUOTE ]
[ QUOTE ]

I wrote it in C# so I could make this an ASP.NET app with a minimal amount of work - only problem would be finding someone to host it since they'd need to be running IIS, have the framework installed etc. If there are any volunteers then I'll do it.

rvg

[/ QUOTE ]

I'd be happy to host it at sitngo-analyzer.com if that works for you.

eastbay

[/ QUOTE ]

Just wanted to follow-up to say I won't be able to host because it's a Windows only thing (which was mentioned, but I skipped over it), so other volunteers for hosting are needed.

eastbay

[ QUOTE ]

It does require the .NET Framework version 1.1 I believe... If anyone with .NET installed has this problem then let me know.

rvg

[/ QUOTE ]

Ive been trying to get this correctly installed for quite awhile now (months), could someone please give me a direct link to the download. Ive downloaded at least three different .NET's from the website and none of them seem to work.

[Nacarno]

Since you asked for suggestions... I have one. Be able to enter the number of players and payout structure for games not on Party.

[rvg72]

[ QUOTE ]
[ QUOTE ]

It does require the .NET Framework version 1.1 I believe... If anyone with .NET installed has this problem then let me know.

rvg

[/ QUOTE ]

Ive been trying to get this correctly installed for quite awhile now (months), could someone please give me a direct link to the download. Ive downloaded at least three different .NET's from the website and none of them seem to work.

[/ QUOTE ]

Here you go:

.NET Framework 1.1 (http://http://www.microsoft.com/downloads/details.aspx?FamilyID=262d25e3-f589-4842-8157-034d1e7cf3a3&DisplayLang=en)

[rvg72]

[ QUOTE ]
Since you asked for suggestions... I have one. Be able to enter the number of players and payout structure for games not on Party.

[/ QUOTE ]

Good idea - initially I am going to allow you to specify the buy-in and entry fee but if I get some time in the next little while I'll make this more open to different payout structures.

rvg

[ QUOTE ]
If anyone has any other suggestions then let me know.

[/ QUOTE ]Allow for ROI computed for 9 player SNG rather than 10 player SNG.

[ QUOTE ]
Here you go:

.NET Framework 1.1 (http://http://www.microsoft.com/downloads/details.aspx?FamilyID=262d25e3-f589-4842-8157-034d1e7cf3a3&DisplayLang=en)

[/ QUOTE ]

no worky /images/graemlins/frown.gif

[suited_ace]

[ QUOTE ]
.NET Framework 1.1 (http://www.microsoft.com/downloads/details.aspx?FamilyID=262d25e3-f589-4842-8157-034d1e7cf3a3&DisplayLang=en)

[/ QUOTE ]

FYP.

[rvg72]

I'm going to be building version 2 of this app shortly with some new functionality.

I have just recently started learning C# and am not able to do any work with it yet professionally (we use different tools right now at my company) so I'm finding little projects like this to write to get up to speed... Just wanted to say that in case anybody wonders why I'm spending this time working on an app with limited practical usage.

So, here are the changes that will be included in Version 2:

* Change the Name of the app from ROIPredicter to something that 1) is spelt correctly and 2) makes sense
* No limitations on payout structure so you could add MTT's and various STT's.
* These Payout Structure Profiles (which also include buyin and your own entered finish distribution) are saved and can be easily reloaded.
* Incorporate graphs - not sure where but I want to do some work with the graphing component...
* Minor bug fixes - first grid totals 100.1% and sorting of columns sorts alphabetically not numerically (so "2.4" > "12.1")

If anyone else has other ideas then let me know.

rvg

[rvg72]

Just in case anyone was watching this thread - version 2 is out:

Version 2 Thread (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&Number=3218857&page=1&view=c ollapsed&sb=5&o=14&fpart=1)