WinCraps RNG SEED Question

Would one of the WinCraps Wizards like to tell me what is wrong with this code?

I ran this autobet with BR of 1k, 2k, 3k, and 4k.
I expected the greatest win to always be the same amount, but it did not work out that way. Why not?

I found it.
In my configuration, under MISC, I had the Re-seed new games box checked.

I unchecked that box, and now I am getting the expected repeating results.

Well, I guess I don't understand this RNG yet.

I unchecked the Re-seed New Games box, and altered my code to run 5 sessions.

A session is ended by either:
- bankroll is less than $5 , OR
- 50 shooters have rolled the dice

I got the results, and then ran the Autobet again. I got the same 5 results (as expected).
I ran it 5 times, and it matched the previous results, all 5 times (as expected).

But, when I set the Autobet to run 1000 sessions, I get different results from one run to the next. Why would this happen ?

don't use RNG #3, there is something weird about it

Quote: odiousgambit

don't use RNG #3, there is something weird about it

Yeah, I saw that in the help file. I am just setting the seed number (I used 1000, and 10,000). I am not specifying a RNG. Perhaps I need to do both?

Like I said though. Using just the seed, and running a 5 session aoutobet, it behaves exactly like I would expect.
But when I run 1000 sessions, it does not duplicate wins like it should, which tells me the numbers rolled are not exactly the same

Quote: RaleighCraps

I am not specifying a RNG.

No need to leave that to chance. I always specify RNG each time, and naturally it needs to be the same one to duplicate results.

edit: although this is not something to put into an auto-bets file, but just with configuration.

I received an email which has led me to the following conclusion on my problem. I may not be correct with this explanation, but I think I am.

When I seed the RNG at the start of my autobet file, it is fixing the sequence of rolls. Let's say it is 10,000 rolls. Every dice roll will be EXACTLY the same for all 10,000 rolls.
What is wrong with my results though is, I have 2 stop criteria, number of shooters, OR when the bankroll busts. By changing the size of the bankroll as a starting parameter, I am extending one shooter's roll longer than it would have been with the smaller bankroll. Now, the next shooter will be starting their roll on a different dice roll. Even though the dice sequence is still EXACTLY the same, the start of the sequence is at a different spot.

Pretty much invalidates the comparison I had done on previous posts.

I 'think' what I need to do is pick a new seed number each time I stop the session, either due to reaching 100 shooters, or when I bust the BR.
I could do this by just incrementing the RNG by 1, so for a HyperDrive of 1000 sessions, I would seed the RNG from 10,000 to 10,999.
That would guarantee every session would have the exact same sequence of rolls up to the shortest session.
This would accomplish what I was trying to do, but at the expense of some randomness.
I would have to run this scenario many times, always starting with a different base seed number.