Ask the Slot Expert: Another misleading pick-em bonus

Question: I have a question about Treasure Chest video poker machines.

You get to pick one of five treasure chests after hitting a four-of-a-kind. The amount you can win ranges from 140 quarters to $1199.

Does it matter which treasure chest you pick?

My belief is that it doesn't matter which one you choose. I have hit the $1199 once, as has my wife. Most of the time it is 140 quarters.

Answer: I believe that your picks should determine your winnings on any game that has a pick-em bonus. I think the game is misleading if the software running the machine uses the RNG to determine the result and what you pick doesn't matter. If the software determines the result, it should use one of the many ways to display the winning amount of a group of possible amounts without misleading the player into thinking that her picks are determining her fate.

If your picks mattered, given enough trips to the bonus, the ratios of the number of times you picked each amount to the total number of trips would each get closer and closer to 0.2 (1/5).

I don't know how many times you've gone to the treasure chest round, but your experiences seem to indicate that the software has chosen the amount you will win and your pick doesn't matter one bit.

Question: Enjoy your columns and read them regularly when they appear in the Casino Times. On occasion you are asked to field questions regarding slot payoffs and possible irregularities on the part of certain casinos. Perhaps you might find this excerpt from my own work history of interest or value.

In the mid to late 80’s I worked in the slot department of the now imploded Frontier Hotel. One might consider my position unusual in a slot department. My actual title was Database Manager. Every day we opened the little doors on the sides of all slots in a given bank of machines. Faithfully we recorded the volume of data that every machine kept religious track of. All of these numbers were then entered into the master database for that particular bank of machines. At the end of a month it was then possible to generate some very accurate reports on the exact performance of each and every machine in the casino. So it was that we stayed in the good graces of the Nevada Gaming Commission as well. After all, if you are proudly announcing to the world via placards and banners that a particular row of slots has a payback of 98 1/2% it was prudent to have the statistics to back up your claim. And so we did. Never in my employ at the Frontier did we run afoul of the Nevada Gaming Commission and that success can be attributed to the meticulous records we kept.

However, some very interesting and strange behavior patterns of certain machines became apparent after a given interval of monitoring their payoffs. Some of the slots regularly gave the player back percentages in the neighborhood of 125%! And some kept in excess of 20% as well. Not good. If we were advertising a payback of 98 1/2% then that’s what we attempted to achieve, not 82% or 124%. Furthermore, the amount of the wager made no difference to the machines. Five cent slots were just as apt to misbehave as their One Dollar brethren, so we kept a close eye on ALL of them.

Here’s how we handled the misbehaving machines. After a week’s close scrutiny verified a consistent pattern of under or over paying the customer, we put an Out of Order sign on the machine and unplugged the power supply. Next we contacted the chip maker – IGT I believe – and requested a new chip for the machine to our specifications. The problem was generally solved within the week. Happy customers, happy casino. More importantly, happy Nevada Gaming Commission. Nice bunch of fair minded fellows on the whole but not a group a sane man would want to run afoul of. They did and still do have sharp teeth.

Answer: Thanks for sharing your experiences.

Nevada gaming regulations require the collection of performance data for each machine. The meters you could see inside older games captured that data. Today I think the casino's slot accounting software captures the data and the casino can remotely query the meters maintained by each machine.

It's not unusual for a machine to pay back much more or much less than its long-term payback in the short run. Consider a session in which you hit a royal flush on a dollar 9/6 Jacks or Better video poker machine. Unless you were down $3980 when you hit, that machine paid back considerably more than 99.5% for you. The casino is in the long run, however, and your session is just a very small part of the machine's overall performance.

Another casino (Riviera, I think) had a few banks of machines that they claimed had the highest paybacks available for each game. It posted a letter from its accountants stating that the casino had ordered the highest payback possible from the manufacturer for each of the games in the banks. The claim was for long-term payback, not daily or weekly or monthly payback.

The same holds true for a bank advertised as having 98.5% payback. As long as the payback chips installed on the machines pay back 98.5% or more in the long run, that should satisfy the gaming commission, regardless of short-term deviations.

You used the words regularly and consistent, though, indicating that these situations lasted for some time. Nevertheless, did you check the performance against the confidence intervals to see if the paybacks experienced fell within the expected range for the amount of play that the machines received, either short or long term? And did their total performance fall close to their long-term payabcks?

The 1980s were the early days of computer-controlled slots. I wonder if it's possible that the chips had gotten corrupted -- maybe even intentionally.

Today the commission uses a device made by Kobetron to verify the data in chips from a machine with the data in the reference chips on deposit with the commission. The machine reports any differences between the two chips.

Back in the 1980s, the commission used an EPROM programmer to compare the chips. Ron Harris, a former Nevada Gaming Commission employee, altered the verification process so it changed the data stored on the chips. He was able to reprogram some machines so a certain series of actions on the machine would cause it to payoff.

You didn't mention anything about verifying the chips, but you did say that replacing them seemed to solve the problem. We'll never know whether these were just short-term deviations expected by randomness or gaffed chips.