Random result grapher spreadsheet
 

I made this little spreadsheet to draw random results graphs based on finishing distribution. Perhaps it could be useful to help demonstrate variance to someone in a more visual manner than the ROI Simulator and a less soul crushing one than experience.

The results seem pretty much just like searching for winning players on sharkscope, but it's perhaps reassuring to see what swings can happen with no "tilt" or change in the players skill whatsoever.

I don't have anywhere to upload it hopefully this is ok.

http://www.megaupload.com/?d=6VVWLJ1Q

Re: Random result grapher spreadsheet
 

this is pretty cool. where is the calculate now button tho? i cant find it

Re: Random result grapher spreadsheet
 

nevermind, thought u actually added a "button" on the sheet.
for anyone else: tools|options|calculations

thanks again, this is very cool

Re: Random result grapher spreadsheet
 

F9

Re: Random result grapher spreadsheet
 

This is fantastic, for real.

Thank you.

Re: Random result grapher spreadsheet
 

this is awesome, ty

Re: Random result grapher spreadsheet
 

my keyboard doesnt have Function Keys....

Re: Random result grapher spreadsheet
 

This is [censored] sweet. Sticky this in the tools post, please. Nice work! Hope we can find someone to host it for real.

Re: Random result grapher spreadsheet
 

Ok, I've never used an roi simulator or anything before, so can someone just tell me what and where I'm supposed to input stuff? Do I fill in my info in the grey shaded boxes then hit f9 and simulated ROI/simulated profit is what it could be w/ variance?

Re: Random result grapher spreadsheet
 

yes, only alter the info in the grey boxes... just input the STT u play most often, and your own stats, and keep pressing F9. its been entertaining me for hours....

after u get bored with that you can fool around with other numbers, but the most interesting thing is the amount of variance and how it shows up on the graph given the same conditions.

Re: Random result grapher spreadsheet
 

This is a hell of a lot of fun. I would never think that over a 2500 game sample at 9 ROI I could run at -2 ROI. Great work OP

Re: Random result grapher spreadsheet
 

http://www.2mt.dk/sng.xls

Nice tool!

Re: Random result grapher spreadsheet
 

I don't get the finish distribution part.
What do you have to put in the grey areas?

Re: Random result grapher spreadsheet
 

percentage of 1sts, 2nds and 3rds matties.

Re: Random result grapher spreadsheet
 

Percentage of overall or of top3 only?
For example if I have played 100 sng's, I finished 10 times 1st, 10 times 2nd and 10 times 3rd.
Do I have to put 10%, 10, 0.1 or 1?
Or do I have to put 33% (because 30 ITM, and 10/30=33.3%)?

Re: Random result grapher spreadsheet
 

You put 10, 10 and 10

Re: Random result grapher spreadsheet
 

[ QUOTE ]
This is a hell of a lot of fun. I would never think that over a 2500 game sample at 9 ROI I could run at -2 ROI. Great work OP

[/ QUOTE ]

That does seem quite insane... RVGs ROI simulator says that is extremely unlikely. Using $27s and an even finish distribution it says the chance of that happening is 0% to the nearest %. Was everything definately entered correctly? I hope this was a freak occurance and not that something isn't somehow not working properly.

[ QUOTE ]
http://www.2mt.dk/sng.xls

Nice tool!

[/ QUOTE ]

Thanks for hosting!

[ QUOTE ]
Percentage of overall or of top3 only?
For example if I have played 100 sng's, I finished 10 times 1st, 10 times 2nd and 10 times 3rd.
Do I have to put 10%, 10, 0.1 or 1?
Or do I have to put 33% (because 30 ITM, and 10/30=33.3%)?

[/ QUOTE ]

You can put in the percentage of top 3, or the overall percentage, or even the actual number of finishes in each position. The actual finish distribution is then scaled based on the ROI that you put in actual ROI (keeping the ratio of 1st:2nd:3rd in the same proportion as entered).

I set it like this so that you could for example put in a 1st:2nd:3rd ratio of say 1.2:1:1.1 and then see what happens when you input different ROIs, rather than having to play about adjusting the actual finish distribution to get the ROI you want.

Re: Random result grapher spreadsheet
 

So, just to clarify: The simulated results (ROI, profit, graph) are randomly generated results that take variance into account? In other words, the results indicate one of the many (thousands? millions?) of different possible ways "I could have ran" over x # of games. Also, I assume the simulated results become closer as the sample size is larger?

Is the above correct?

Also, how do I determine the lowest/highest simulated results for ROI/Profit?

Re: Random result grapher spreadsheet
 

is there any other way to 'calculate now' since im on a mac and F9 button doesnt seem to be doing anything.

also after i changed the buyin, shoudl i not have changed the payouts (not in grey) since theyre obv different, or will that mess it up?
THANKS DAVE

Re: Random result grapher spreadsheet
 

whoaaaa. for some reason when i clicked the 'save' button it ran the calculate now function. interesting. and the first sim said a 5% roi player can run at 9% over a 5k game sample. holy crap! i want that!

Re: Random result grapher spreadsheet
 

[ QUOTE ]
So, just to clarify: The simulated results (ROI, profit, graph) are randomly generated results that take variance into account? In other words, the results indicate one of the many (thousands? millions?) of different possible ways "I could have ran" over x # of games. Also, I assume the simulated results become closer as the sample size is larger?

Is the above correct?

[/ QUOTE ]

If I understand you correctly, all that you said is accurate, except perhaps that it takes variance into account. It doesn't in anyway directly take variance into account, but just produces one single random set of tournament results based on the finish distribution and prizepool distribution provided.


[ QUOTE ]
Also, how do I determine the lowest/highest simulated results for ROI/Profit?

[/ QUOTE ]

As the spreadsheet only produces one set of results at a time, there is no way of finding lowest/highest results other than by continuously drawing new graphs and oberving.

For those kind of results you can use RVG's ROI simulator, which simulates 1000s of runs of tournaments rather than a single run as this spreadsheet does.


[ QUOTE ]
is there any other way to 'calculate now' since im on a mac and F9 button doesnt seem to be doing anything.

also after i changed the buyin, shoudl i not have changed the payouts (not in grey) since theyre obv different, or will that mess it up?
THANKS DAVE

[/ QUOTE ]

If you are wanting to simulate a normal .5/.3/.2 payout SNG you can just alter the buyin and fee and number of players and it will work out the prizes for each place automatically. You should only change the values in grey boxes.

I'm not familiar with the mac version of excel or whatever spreadsheet program you are using. I would think that somewhere in there is an option for recalculating the spreadsheet. There may be a way to add a button in the spreadsheet to make it recalculate but unforunately I don't know how [img]/images/graemlins/frown.gif[/img]

Re: Random result grapher spreadsheet
 

thanks dave. yea i have excel for mac, and its pretty much identical to the windows version. I changed the buyin and the payouts didnt adjust for some reason, so i just changed em. IT seems to have worked now, and as i said earlier, oddly when i click the save icon, it doesnt say anything about saving the file but it DOES do the calculate now function. heh.

Re: Random result grapher spreadsheet
 

I'd love to get this stickied in the FAQ if it tests out right. If you haven't played with this yet, it will make you feel good about what is hopefully "variance". I was so weirded out when I made the sample size 10000 STTs and a 13% roi, coming up with near 1000 game break even periods, and numbers were often 2% higher or lower than actual, even at this huge sample. Also makes me feel that my current month's 1200 STT HEATER may indeed be just that. Can't imagine the swings when playing above the $100 level with an roi below 5. If anyone wants an illustration of why stars needs to reduce the rake, this would be a wonderful tool.

Re: Random result grapher spreadsheet
 

This is an awsome tool to have. I will no longer be playing the $555's with no table selection thats for sure haha. Also, maybe I did something wrong but I was doing the 114's over 5k games with a 6% roi and maybe 5% of the time or so I was down after 5k games. Is that really possible? If so thats pretty damn scary.

Re: Random result grapher spreadsheet
 

Nvm I had it at 3% but that's still pretty damn scary.

Re: Random result grapher spreadsheet
 

I expanded this excelsheet with a simulating part (assuming it was open source [img]/images/graemlins/smile.gif[/img] ). It simulates a predefined number of profit graphs based on the input given. The best and worst profit graph are shown in a graph.

For example, my 22+2 results(200 simulations, ROI 8%):

http://img100.imageshack.us/img100/4...exampleui4.jpg

Download links:

Excel 97/2003
Excel 2007

These sheets contain VB code so you might need to enable macros. I deleted the initial graph because otherwise for each simulation the graph is updated which takes a lot of time.
It's still kind of slow so I wouldn't recommend to enter a large amount of simulations (200 simulations takes about a minute on my dual core).

The sheet may be expanded by showing for example largest downswing, upswing, break even strike etc. in the graph, but i guess this is rather difficult to implement. However, I'll try if there is enough interest.

Re: Random result grapher spreadsheet
 

there is interest. very cool what you did so far.

Re: Random result grapher spreadsheet
 

Im interested in that aswell [img]/images/graemlins/smile.gif[/img]

Re: Random result grapher spreadsheet
 

Please don't let Guthrie get a hold of this spreadsheet.

Re: Random result grapher spreadsheet
 

Thank you! Really helps put my mind at ease after having been through a horrible month long break even run.

Re: Random result grapher spreadsheet
 

Been working further on the simulation part, will upload a version soon. I'll try to include best run, worst run, largest down/upswing, breakeven strike, and distribution.

Concerning the last, have a look at my total profit distribution (pretty sick imo). Actual ROI is 8% at the 22+2 sng's, equal 1-2-3 distribution.

The total profit after 5000 tournaments is shown, based on 5000 simulations.

http://img514.imageshack.us/img514/5...ibutionvp2.jpg

Re: Random result grapher spreadsheet
 

Unfortunately, these programs make me wanna quit STTs altogether... though I guess 5000 isn't even that huge a sample.

Re: Random result grapher spreadsheet
 

Thanks for the mod Shrooma. Can you explain what the simulated profit and simulated ROI is over say 500 simulations? Is this the average of all simulations or is it just the last simulation that was done? Also, how can you figure out what the ROI is of the worst/best run? Thanks a lot.

Re: Random result grapher spreadsheet
 

Forgive me, as I know I'm being a nit here, but what's accounting for the variance when recalcing each time? I somehow get that the sheet is calculating different scenarios even if you play it the same way but can anyone go into it a bit more? The simulated results must be changing the finish distributions, right?

Thanks.

Re: Random result grapher spreadsheet
 

Okay, I rewrote from scratch a very fast simulation sheet supporting up to 32000 simulations of 32000 tournaments. The results shown are distribution of profit, best, worst and average runs and simulated ROI information.

Excel 2007

Excel 97/2003

Note: This program is much faster. However, 10000 simulations of 10000 tournaments took over 1 minute and 1 GB of RAM.

Re: Random result grapher spreadsheet
 

[ QUOTE ]
Forgive me, as I know I'm being a nit here, but what's accounting for the variance when recalcing each time? I somehow get that the sheet is calculating different scenarios even if you play it the same way but can anyone go into it a bit more? The simulated results must be changing the finish distributions, right?

Thanks.

[/ QUOTE ]

A randomizer accounts for the variance (based on entered ROI and finish distribution). For each tournament in each simulation the randomizer is called.
Of course, the more simulations are performed the better the simulated ROI fits the actual ROI. The main purpose of this program is to provide graphical insight into the varianced based on your own parameters.

Re: Random result grapher spreadsheet
 

very good work.

Re: Random result grapher spreadsheet
 

Thanks for improving it Shrooma they are interesting to see. I wonder if you could use the results of each simulation to produce a graph of cumulative probability of ROI/profit (preferably ROI).

Something like this that I made using the ROI and Cu columns of the output given from RVGs ROI Sim, but I guess only with one line.

http://homepages.nildram.co.uk/~walk...rofitsmall.jpg

Re: Random result grapher spreadsheet
 

this program is way too many nice. thanks.

Re: Random result grapher spreadsheet
 

[ QUOTE ]
Please don't let Guthrie get a hold of this spreadsheet.

[/ QUOTE ]

lmao