An Experiment

[sharkbyte]

Hey folks.

I am looking to run an experiment with the latest version of TA. I'm looking for someone willing to share a sample of tosses. Preferably 500 or more.

I would be looking for text or Excel files. Right outcome and left outcome is all that is necessary; and to know what dice set you used (left top/front, right top/front). All tosses do not need to be with the same set, but I would need to know which rolls go with which sets.

Also, a brief explanation of what your preferred betting strategy is.

I will take your data, run it against the latest version of TA, and provide you back data analyzing your results along with suggestions on how you might be able to maximize your tosses to your betting strategy.

Again, this is simply an experiment. I am currently only looking at my own tosses, and I want to insure the system isn't biased towards my results.

I hope one or two of you will be interested in helping me out. Just send me a PM if you are interested.

Thanks.

[acpa]

Two late to do tonight, but I'll send you an excel file that has by last two trips to Vegas.

Noah

[Mad Professor]

Hi Sharkbyte,

Noble project. Can you run toss-results that don't separate right-die/left-die results? If so, I can send you sets of ~20k rolls if you'd like.

LMK,

MP

[sharkbyte]

Hi Sharkbyte,

Noble project. Can you run toss-results that don't separate right-die/left-die results? If so, I can send you sets of ~20k rolls if you'd like.

LMK,

MP

Unfortunately, no.

The software tracks the on-axis percentages, and what faces result, to extrapolate how those results would appear using other dice sets. The only way I could use them would be to assume a random distribution in how the dice finished; which wouldn't help anyone...unless you're really a random roller and have been teasing us all this time.

[sharkbyte]

For those who might be interested what TA does is use your on-axis statistics to extrapolate results for any dice set out there. This data can then be examined to show EV advantage for some standard bet/number combinations.

It expands upon Heavy/MP/MD teachings about practicing to build consistency. While TA can break down individual sessions, its greatest benefit is the ability to combine multiple sessions and examine it as one large group. When it does this the extrapolated data is based on how well it does across all the sessions. This helps account for the ups-and-downs you experience through the normal course of play, and is doing so with long-term results. Ultimately, the more tosses you put in the better the output data will be.

Here are a couple screen shots to show the system output.



This screen shows a single session. Displaying SRR, BBS, number breakdowns, and On-Axis percentages.



The tabbed data, from the last screen. This shows the breakdown of Box numbers.



And we finish this screen with the breakout of 'group' numbers.




Now we're looking at some stats taken by looking at multiple sets. Including individual and combined On-Axis results, SRR, and BBS. Notice that 3 sessions the Left Die was On-Axis less than random, and 3 times the Right Die was less than random. This will give a good example of how we can leverage the data to still take advantage even with such variances. It also shows us that the best sets produced an SRR of 7.09 across the 7 sessions reviewed.



This shot shows us the sets producing the best EV for any of the options across the top. In this example I have selected the 6 & 8 pair.



It can also show CO options, such as Horn, World, and Lay bets, or the numbers that show the most, if you are interested in playing with a WOTCO strategy. This example shows laying the 4. The numbers down the right side show how the best set also performs with other potential lay bets. This would be a good set to lay both the 4 and the 10.



Ultimately, I want to see how a set might perform - with session rolls - using a specific betting strategy. While not yet completed, TA will ultimately be able to test customized bet strategies, and return the predicted dollar results.

I would love to hear what people think, and any suggestions there might be.

[sharkbyte]

Another point, by letting TA extrapolate dice set options, I don't have to worry about changing sets during practice - such as using a CO set and a point set. I practice with a single straight-sixes set, and let TA tell me what sets would produce the best results in any given situation. That way I can concentrate on my toss, and consistency, while practicing.

[sharkbyte]

Petey:

I'm still writing it. It's also not as easy to install as Bone Tracker. I use Microsoft Access for the front end, and I have moved to using SQL Server Express as the back end. So you would need both of these applications, as well as SQL Server Management Studio. Both SQL Server applications are free.

If you have all those, then I'd be more than happy to give you a copy once I'm done with it.

Also, when done, I plan on moving it to Access 2010 in which case I'll be able to create a version with the Access runtime. This will mean that you only need the SQL Server apps, because I'll be able to combine the front-end with the Access executable, and provide both to you.

[sharkbyte]

I am running windows 7 64bit with MS office 2010, tried to install Access 2007 32bit but it cannot find proof.cab. I am not a programmer I ma looking for something I can just load and run. Guess I will havew to wait for 2010 version because i do have access 2010.

And once I'm done, you can have a copy. It might be a bit, though. Aside from just finding time to work on it, now I've found a few new things to add since I've been processing other user's tosses.

[sharkbyte]

I assume that you are looking for data that was input into BoneTracker or a similar program which differntiates between the left and right die -

If so, I have files that I can send to you OR I can scan and save as a .pdf the actual practices session sheets which have yet to entered into the computer -

The bulk of our practice sessions are on hard copies and have as of yet to be entered into the computer -

Snakeeyes - I will take any data you have, so long as the numbers are separated. Obviously, data files are easier to load, but I can manually input some data. Not going to cut too much into my own practice time, but I can put some together. I want to make sure that I'm looking at a reasonable data set.

BTW - exactly what is " TA "

TA refers to Toss Analyzer. Its a program I originally wrote about 8 years ago to expand on what Maddog was doing with BT. It looks deeper into the numbers, including comparing all 576 possible dice sets, and shows ways to exploit your advantage in an attempt to maximize specific outcomes.

What I've been doing recently is updating TA with the last 8 years of programming knowledge, and a more in-depth understanding of what to focus on to maximize your advantage.

Feel free to send me your info to sharkbyte1@hotmail.com.