Variance simulation followup
 

Inspired by AgentSp's original post, i continued to improve my SnG variance simulator.

It spits out a lot of stats now, and it has built-in functionality for graphs, including export to Excel summarization of bankroll.

For stats such as OOTM, ITM and Largest downswing to become interesting, set the block size to 1. By setting the block size to 1, you effectively simulate a theoretical SnG gamers bankroll progress.

By increasing the block size you can analyze the sum of days, weeks, months and so on.

If you want to know the theoretical number of loosing weeks for an SnG player playing 50 SnG's a week, set the block size to 50 and number of iterations as you like. This will skew your ITM rate as this will now be the amount of positive result blocks.

The biggest loss/win/average is based per block.

Loosing/winning/breakeven blocks is based on the break even threshold. This way you can define what a break even block is for you.

By increasing the block size OOTM/ITM streaks also change to block based, meaning an OOTM streak of 2, with block size 50, will mean that you had to negative weeks in a row.

Please tell me if it's usable, what you like, what you don't like, and what you miss [img]/images/graemlins/smile.gif[/img]

You can download the app from here: www.improve.dk/sngcalc.rar.

My usual disclamer goes for this app too, if you fear keyloggers, adware, spyware and the likes, and you can't take my word for good, please don't use this software. I can guarantee that it's clean, no more, no less.

Please note that it requires WinRar to unpack. It does not need to be installed, it does however require the .NET Framework.

/Orca

Re: Variance simulation followup
 

Definitely fun to play with. Any chance you can allow for the 1/2/3 %s to be decimals? Right now it gives wacked out results if you don't use a whole number.

You cranked this out in like two days. Maybe it's time for me to finally learn .Net and stop doing stuff in C/win32 [img]/images/graemlins/smile.gif[/img]

Re: Variance simulation followup
 

It should support decimals, just remember to use commas, not periods (it will correct it itself though).

Regarding .NET, what are you waiting for? [img]/images/graemlins/smile.gif[/img] You can just use C++.NET if you're in love with the syntax, i love C# due to the syntax myself.

Re: Variance simulation followup
 

Nice stuff. And mentally it is nice to see that 13 OOTM is not only normal, but also expected if you play enough SnGs.

Re: Variance simulation followup
 

Aight there are a couple things I don't get.

First is the breakeven threshold. A breakeven set, to me, is one where my net profit is 0. Is there some other definition for breakeven?

If I set the block size to 1, then my losing blocks is 0. But if I set it to 50, my losing blocks is somewhere around 12%. This just seems counterintuitive. If I play 1 SNG/wk, I'm going to have a lot more losing weeks than if I play 50/week.

Are the OOTM and ITM streaks actually just negative and positive profit streaks? When I set the block size to 50, I got an OOTM streak of 5. That would mean I could have 5 straight weeks operating at a loss, not 5 weeks (250 games) where I don't cash a single time.

Re: Variance simulation followup
 

[ QUOTE ]
It should support decimals, just remember to use commas, not periods (it will correct it itself though).

[/ QUOTE ]
When I threw in some placings with decimals (14,4 9,5 and 12,6) it said my ROI is 345.55% and ITM is 100%.

Re: Variance simulation followup
 

Why do the ROI and ITM %s change with each simulation? I understand that this is just simulating games, but if I enter certain placement %s, then the ROI and ITM should be fixed.

Re: Variance simulation followup
 

[ QUOTE ]
First is the breakeven threshold. A breakeven set, to me, is one where my net profit is 0. Is there some other definition for breakeven?

[/ QUOTE ]

My thinking here is that for me, if i play 50 SnG's a day. Any day ending at around +/- 2-4 SnG's, that's breakeven. Of course, if we were to use the term literally, only +$0 would be break even [img]/images/graemlins/smile.gif[/img]

[ QUOTE ]
If I set the block size to 1, then my losing blocks is 0. But if I set it to 50, my losing blocks is somewhere around 12%. This just seems counterintuitive. If I play 1 SNG/wk, I'm going to have a lot more losing weeks than if I play 50/week.

[/ QUOTE ]

I hear ya, it's due to the break-even threshold. If you set it til 0, you'll get the number you expect. The definition of a loosing block is a block where the total loss exceeds the break even threshold * buyin. If you leave the default threshold at 4, one single block can never be considered a loosing block.

[ QUOTE ]
Are the OOTM and ITM streaks actually just negative and positive profit streaks? When I set the block size to 50, I got an OOTM streak of 5. That would mean I could have 5 straight weeks operating at a loss, not 5 weeks (250 games) where I don't cash a single time.

[/ QUOTE ]

Exactly. The OOTM/ITM are only relevant when the block size is = 1. Otherwise it says the amount of consecutive loosing weeks/months/years etc.

Re: Variance simulation followup
 

[ QUOTE ]
[ QUOTE ]
It should support decimals, just remember to use commas, not periods (it will correct it itself though).

[/ QUOTE ]
When I threw in some placings with decimals (14,4 9,5 and 12,6) it said my ROI is 345.55% and ITM is 100%.

[/ QUOTE ]

Thats weird, anyone else experiencing the same?

Re: Variance simulation followup
 

[ QUOTE ]
Why do the ROI and ITM %s change with each simulation? I understand that this is just simulating games, but if I enter certain placement %s, then the ROI and ITM should be fixed.

[/ QUOTE ]

I could calculate the theoretical ROI, and as the number of iterations went up, the ROI would near the theoretical ROI. But as you mention, this is a simulation, meaning that the results are suffering from variance, just as in real life. This also results in different graphs each time you run the simulation.