Good suggestion. Thanks! I chose to adjust 2nd and 3rd in equal absolute amounts to get to 100, and only when the total distribution was already 100.

Adjustment = 100 - (new 1st + 2nd + 3rd)
New 2nd = 2nd + Adjustment
New 3rd = 3rd + Adjustment

www.dropbox.com/s/ukpadqy24se12th/SwongSim.exe?dl=0

On second thought, let's adjust them proportionally. At least that will work out better for HU.

www.dropbox.com/s/ukpadqy24se12th/SwongSim.exe?dl=0

This last version works well, thanks

Cheers and thanks for the feedback. I also like your last suggestion and might add that in when I get to it.

Thanks! Are you planning adding some cev stuff? i would like to know given my Evroi X% what is the probability ending down(Ev) in a sample of Y numbers of spins!
Thanks again for your work!

I would like to, but unfortunately I still don't see how to accurately simulate expected chip results since it is the product of non-trivial things such as card distribution.

but you can know the standard deviation of evchips right?
What variables you need more? standard deviation and ev% expected and?
thanks!

Do we know if it's normally distributed? My sample is skewed but it's a small sample (1700 games, PT4).

It's close to normally distributed when averaged over 1700 games (30K-40K hands) because, most of the time, one wins or loses only a few hundred EV chips per tourney, and the distribution is quite symmetrical (unskewed).

Or wait, what's the skewness in your sample and how does it compare to a cash game?

I only counted the number on each side of the mean. 775 games above the mean and 945 below, over 1720 games. Not sure how that compares to cash.

Is there a way to calculate where the game is more profitable with equal chips/t (60) and rake (6%), but we don't take to consideration highest multipliers?

100 out of 100 000: x25
500 out of 100 000: x10
7500 out of 100 000: x6
19865,3 out of 100 000: x4
72019,4 out of 100 000: x2


Full tilt Jackpot SnG:
320 out of 100 000: x20
640 out of 100 000: x10
3675 out of 100 000: x6
22000 out of 100 000: x4
73325 out of 100 000: x2

Or where downswings will be lower?
I want to decide where to play if I have 60 chips/t.

Under that scenario, the expected ROI will be higher and expected downswings smaller for the first set of numbers. For comparison, you can just zero-out the buyins awarded for the top tiers. Note that if you are including rakeback in the analysis, you will need to adjust it since SwongSim calculates expected post-rakeback ROI based on the effective rake, which will be inflated when zeroing out those tiers. (new rakeback = old rakeback * old effective rake / new effective rake)

I don't believe that you can state this for sure unless you've played thousands of tourneys at both PS and FTP. And even then, the PS field of 2016 may become softer than in 2015, and the FTP field may become tougher because of regs coming from PS.

And yes, like Max Cut has said, Stars' payout structure is 'more grindable' even despite the top multiplier being x10K, because it's much rarer than FTP's x2K and FTP's structure is suffering from a deficiency of 6x tourneys.

But variance is going to be only one of the factors, the other major one being game selection (incl. playing hour selection, quitting when certain good regs are present in FTP's $50 pool and adding tables when certain regfishes are online).

Of course I talk about stable 60chips/t in theory (though I played about 4000 spins on Stars and 3000 on FTP).
I need to know, where shoud I continue to play: on 7$ at PS or 5$ at FTP.
I wondering about expected ROI and dispersion.

Thank you both, guys! So your advice is PS to play safely?

I can only tell you about expected swings, which should be relatively similar in terms of $ -- slightly bigger at FTP in terms of buyins, but playing for $5 vs playing for $7 at PS.

Announcing a brand new version of SwongSim brought to you by SmartSpin™, now with EV Chip Mode.

The fundamental use of SwongSim has been to illustrate the expected swings in spins, given a payout structure and player finish distribution. This is useful in determining things like bankroll management and perhaps which games to play.

For Spins, the win rate (finish distribution) is determined by the number of chips won on average. For this reason, we convert expected chips won per game into win rate and use that as input to SwongSim. One issue with this is knowing how reliable it is to use our reported EV chips won per game on a given sample (from PT4, for example). You see, even reported EV chips per game will have swings since it does not exclude all possible luck factors.

This new version of SwongSim includes an EV Chip mode where we can study swings in expected chips, given a specified EV chips per game and a standard deviation. With this we can ask questions like, if I have an expected chips won of 40 per game over a sample of 2000 games, how likely is it that my true EV chips per game is less than 30? To answer, we simulate 30 chips per game and see how often we end up with a 40 EV chips result (about 10% of the time, as seen in the image below, assuming a 400 chip standard deviation).


Here is the download link:

www.dropbox.com/s/ukpadqy24se12th/SwongSim.exe?dl=0

• Added a new EV chip mode to simulate swings in expected chips.
  
  

• Added option to specify the number of threads to use (from 1 up to the number of cores your CPU has minus one). This can reduce CPU load or speed up computation, depending on the setting.
  
  

I would like to thank Smart Spin for their support in bringing this new feature to SwongSim as well as more that are in the works for 2016. Check out their 2+2 staking thread:

forumserver.twoplustwo.com/222/staking-offering-stakes/smartspin-biggest-most-futuristic-spinngo-poker-team-1513469/

Thank goodness, I have the last unsponsored version saved locally

I'm glad that you've got paid now, though, you deserved this Santa's present for all your hard work.

Happy New Year!

Thanks, and best wishes for 2016! Any chance you can share your PT4 standard deviation stat, or give a tip how to make it?


As always, everyone, please let me know if you see any real problems with the new version.

Happy New Year all!



---------------------------------------------------------------------------------------
1000 simulations of 10000 games
Expected ROI (with rakeback/bonus/award): 3.5% (about 350 Buyins)
---------------------------------------------------------------------------------------
100% had an upswong greater than 2016 buyins!
---------------------------------------------------------------------------------------

Are there any real differences between the two versions except the logos that the last one has ?

And it's free.

If you wish to support my efforts and time spent in providing this at no charge, please grab the latest. Thanks!

There's the chip EV mode in the grapher, plus an MTT section under construction that does nothing so far. The chip EV -> ITM mapping was present in the last unsponsored version.

If you're paid per download, then I can't help you any more because I have the latest version too; if you're paid for every use of SwongSim, then I can keep simulations running in the background whenever I'm awake and poker clients aren't running, for a 40% share of your revenue. (J/k, I'm pretty sure you're paid for referred stakees / students only.)

Yeah Max, I got it and thanks for your efforts. Was just curious because of his comment.

Cheers 7even2euc3, happy new year!

Nothing like that; no royalty or affiliate-type remuneration based on views, clicks, downloads, sign-ups, or any other metric. Simply a small, highly-appreciated, donation in exchange for placing the logo and static splash.


On the to-do list is MTT support, mixed buy-in support, and some sort of help/manual.

Thanks again for your work,

If i understand correctly, i want to know % of days my EVchip will be negative,
All i have to do is simulate smaple with my daily volume (It should be useless to set my ITM% right?)
And after i go in draw, set the EVchip/game i think i have,
And look at the graph right?
Exemple here https://gyazo.com/230f81280007ba773219b26ab67a17f3 if my evchip/game is 55, i will be EV negative 10% of my days (85 sng/day)

How you estimate the standard deviation? big sample of many players?
How can i know my standard deviation? cant find it for chiptourney in PT4, maybe in hem2?

wow cant wait for mixed buyin support!!

Mtt would be great too!

thanks again!

Yes, assuming you use an accurate EV chips/game and assuming EV chips distribute normally, that type of simulation should give a reasonably accurate estimate. I would use a larger number of simulations in a case like that (see the image in my next post, although that is a bit of overkill).

The only thing I have to go on so far is coon74's post in the Spins thread.

forumserver.twoplustwo.com/showpost.php?p=48322344&postcount=3070

Cheers!