My new 'I have too much spare time' project

[poppo]

I don't recall whether or not you ever mentioned what game you have installed?

This is in a 5 line Gold Mountain with a SS3819.

Keep in mind this is still a short term sample. I simply used the reciprical of games won divided by games lost. In this case I have 276 won and 379 lost. Thus there has been one hit (not necesarly a win of more than wagered) in every 1.37 spins (truncated to 2 decimal places)

My 3 coin Spin Till You Win SS4198 has an even smaller sample so far. It has 57 games won and 231 lost. According to my calculations it is at a hit ratio of 1 in 4.05.

[uniman]

I don't recall whether or not you ever mentioned what game you have installed?

This is in a 5 line Gold Mountain with a SS3819.

Keep in mind this is still a short term sample. I simply used the reciprical of games won divided by games lost. In this case I have 276 won and 379 lost. Thus there has been one hit (not necesarly a win of more than wagered) in every 1.37 spins (truncated to 2 decimal places)

My 3 coin Spin Till You Win SS4198 has an even smaller sample so far. It has 57 games won and 231 lost. According to my calculations it is at a hit ratio of 1 in 4.05.

Is the hit ratio "games lost / games won" ?? 379/276 = 1.37
Shouldn't it be "total games / games won" ?  (276+379)/276 = 2.37

[StatFreak]

This is in a 5 line Gold Mountain with a SS3819.

Keep in mind this is still a short term sample. I simply used the reciprical of games won divided by games lost. In this case I have 276 won and 379 lost. Thus there has been one hit (not necesarly a win of more than wagered) in every 1.37 spins (truncated to 2 decimal places)

My 3 coin Spin Till You Win SS4198 has an even smaller sample so far. It has 57 games won and 231 lost. According to my calculations it is at a hit ratio of 1 in 4.05.

Is the hit ratio "games lost / games won" ?? 379/276 = 1.37

Shouldn't it be "total games / games won" ?  (276+379)/276 = 2.37

Correct Uniman.




<ADD> The second hit ratio should be 1 in 5.05  (231+57)/57

Poppo:
However, one doesn't need to add the wins and losses. There is a separate bookkeeping value for total games played. The correct procedure would be to change the code to read the total games played and games lost values. (In other words, substitute a loookup of the Total Games Played value for your lookup of "Games lost" and leave the math function in the code alone.

Side note: You could cheat and just add 1 to your current calculated value*, but it's not the right way to do things. The NGC would never approve.
*since total games = games lost + games won; your current value is games won/games lost; and games lost/games lost always = 1

[poppo]

Yep, you are right.   It's easy enough to fix since I already have grabbed the total games played value for use on page 1 of the display. All I need to do is change the one line in my code.  Thanks  for pointing it out.

Good think I stuck my Atoms in ZIF sockets.  

Note: I always grabbed the actual total games value from the EEPROM instead of just adding won + lost  so that I could verify I was getting the right values. I suppose I could eliminate that one read if I needed to, but I'm not running low on program space or anything.

[uniman]

That's cool, your project is really turning out nice!   

[poppo]

Thanks.  I think I am about done with this phase.

The Atom chip I am using has serial I/O (on any pins). Now I already have SAS working and can do EFT transfers to the machine from a PC. If I had a better understanding of the EFT commands, I could send some credits right from the machine (using the Atom) without the need for a PC. But from what I have seen, the EFT is not just a simple command of 'add $20'. Instead there is validation with an initial command, then a reply (that is different each time) and then another command to complete the transaction.

If I could get that working, I could add a card reader, and while not a true funds transfer, have a swipe of the card add $20 (or whatever I set it up for).

Of course after I started thinking about that aspect, I realized I could be getting the same data I am getting from the EEPROM by using SAS commands (using the Atom) and have the data in real time. The commands for meter reads are simple and the machine just sends a reply to the request. But that would tie up the serial port on the machine for SAS.

I've even been thinking of some sort of S+ 'TITO' using something like a memory stick. Pop it in, cash out (info gets written to the memory stick), go to the other machine, pop it in and add the credits. 

Did I mention I have too much spare time? 

[stayouttadabunker]

That's an awesome idea actually...
It would save live casinos millions of dollars in ticket stock and the machines wouldn't need printers...lol
You'd be a multi-millionaire! 

[poppo]

MIMO _{patent pending}

 

[stayouttadabunker]

MIMO _{patent pending}

 

You need a partner? 

[TZtech]

Poppo

Have been looking into doing something similar. On the S+ have you managed to do an actual EFT or only a legacy bonus (Andy's app does these ok to the PE+ and also SAtest ut cant seem to get a EFT transaction done)
You could write that info to a smartcard - may be more practical and easier to interface to the atom than a USB device.

Ian

[poppo]

On the S+ have you managed to do an actual EFT or only a legacy bonus ....

EFT with SAS Test.

I doubt I will pursue 'MIMO'.    But the Atom does serial (RS232) very well and easily. So whatever device, whether it be a memory stick or a smart card, would be easiest to implement if it had serial I/O. On my robotics project, I had a serial based SD card reader. It worked very well and was fairly easy to read and write from with the Atom. But deciphering the EFT validation seems like more trouble than it's worth.