Predetermined Slotmachine

I was trying to figure out how does a 95%/97%/etc. winning ratio being felt by slotmachine users?

Is it just base on the arrangement of the reelsymbols, the distribution of symbols per reel strips,

or on the machine, it tries to search for Stops that will make a hand that is 95%/97%/so on.

e.g. When the slotmachine value has $1000, and the winning ratio is 95%, the machine will search for STOPS for the
user to have a chance winning at most 95% of machine value.
machine will just arrange the stops on the slot.

*5Reel slotmachine
*app games(not real slots that gives money)

As I understand it, the 95%/97%/100.5% refers to the theoretical payback of the game averaged over a lifetime of spins. The number of paying reel combinations, amount of each payout, and theoretical frequency of each payout is determined by the game design (e.g., many small payouts with a small jackpot vs. less frequent small payouts with a larger jackpot with max coin in), with each combination of reels mapped to a certain payout.

What is not determined by the design, is the output of the random number generator (a.k.a., RNG) each time the handle is pulled or button is pushed. The RNG output maps to a certain combination of reels. The machine then stops or displays this combination, and pays accordingly. Since the RNG output is practically random, some machines may payout much more than the theoretical percentage, while others may pay much less.

Note that this is very different than "pull tab lottery" machines where the winning amounts and frequency of payouts are pre-programmed.

Quote: Ayecarumba

Since the RNG output is practically random, some machines may payout much more than the theoretical percentage, while others may pay much less.

Keep in mind, these theoreticals are based off at least a million pulls. I just did a machine here that's been up for 4 years without interuption, and it only has 240,000 pulls.

Test them by the month and indeed, you can see overpays or underpays. But stretch it out and everything should fall near theoretical. I don't think the RNG is the cause.

Quote: Face

Keep in mind, these theoreticals are based off at least a million pulls. I just did a machine here that's been up for 4 years without interuption, and it only has 240,000 pulls.

Test them by the month and indeed, you can see overpays or underpays. But stretch it out and everything should fall near theoretical. I don't think the RNG is the cause.

Interesting... that works out to an average of 164 pulls each day. If each pull is $1.25, the machine is only dropping $200, and only holding $10 per day?

Quote: Ayecarumba

Interesting... that works out to an average of 164 pulls each day. If each pull is $1.25, the machine is only dropping $200, and only holding $10 per day?

Well, that particular machine was a $100 machine. I'm sure our penny and nickel machines get many more pulls =)

Quote: Face

Keep in mind, these theoreticals are based off at least a million pulls. I just did a machine here that's been up for 4 years without interuption, and it only has 240,000 pulls.

Test them by the month and indeed, you can see overpays or underpays. But stretch it out and everything should fall near theoretical. I don't think the RNG is the cause.

It's not the RNG, it's the game design itself. Some games have higher variance (or in slot parlance, "volatility") than others. I usually do a 95% confidence interval at 1k, 10k, 100k, 1m, etc. to show the operator how fast the games converge on the theoretical. As a loose rule of thumb, games with more small awards will converge faster than games with fewer large awards. Some games expect to be within 1% after a million pulls (with 95% confidence); some are shorter or longer. Knowing that can help your floor's bankroll risk.

To put it in table games perspective, your roulette bankroll results will converge on -5.26% much faster if you play $1 on Red than if you play $1 on a single inside number. The edge is the same but the variance is much higher with the inside number bet.