Las Vegas slot machine survey update

I don't think that's too bad a "rule of thumb" for someone traveling from casino to casino... but as an actual survey hosted on your site I feel the information is just too unreliable. Unless you got a sampling size of casinos/VP/keno/slots to verify at least a couple times that it was fairly accurate, I probably wouldn't trust the information =/. I've just seen too many times in a casino if you move 3 VP/slot machines to the left it's an entirely different payback. How would you even go about selecting the VP/keno? Would it be a spot check? Would you check ALL of the VP/keno in the place and make an average of some sort?

When it comes to slots, it's sad they can get away with not letting you know what you're playing.

Quote: Romes

I don't think that's too bad a "rule of thumb" for someone traveling from casino to casino... but as an actual survey hosted on your site I feel the information is just too unreliable. Unless you got a sampling size of casinos/VP/keno/slots to verify at least a couple times that it was fairly accurate, I probably wouldn't trust the information =/. I've just seen too many times in a casino if you move 3 VP/slot machines to the left it's an entirely different payback. How would you even go about selecting the VP/keno? Would it be a spot check? Would you check ALL of the VP/keno in the place and make an average of some sort?

I would do random checks. For example, five machines in each casino. I would make no effort to sniff out the best machines.

Quote:

When it comes to slots, it's sad they can get away with not letting you know what you're playing.

I agree 100%. I applaud the UK law where the return percentage has to be disclosed on the machine itself or in the help screens.

Quote: Wizard

I would do random checks. For example, five machines in each casino. I would make no effort to sniff out the best machines.

The results would be more statistically reliable if the sample size were determined by the number of slots in each joint. Checking five in "The Cromwell" might be overkill while five in the Venetian could be way too few to produce reliable results.

The total number of machines in each joint is published somewhere, right?

Quote: Ayecarumba

The results would be more statistically reliable if the sample size were determined by the number of slots in each joint. Checking five in "The Cromwell" might be overkill while five in the Venetian could be way too few to produce reliable results.

This could make for a textbook statistics problem. Yes, I think the Gaming Control Board publishes the number of slots per casino. Maybe I could do something like 1 per 1,000, with a minimum of 5.

To me, for "reliable" information this would fall in to a divide and conquer strategy. I would break them down to The Strip, Downtown, then North, South, East, and West... possibly doing the other bigger city destinations if your survey were to reach that far. I think the numbers, much like their table games and VP will change pending these locations. Mid strip, tough to find 9/6 JoB... Drive 15 minutes to Boulder Highway.. tough not to find 9/6 JoB.

After that I would do a set percentage of total machines (as you're suggesting). The problem is it's such a beast of a feat to tackle... Without putting in a crazy amount of effort the data is unreliable, but the problem then becomes you need to put in a crazy amount of effort =P.

Just get Axel to pay his minions $50 per casino to do all the scouting for you =).

Yea, that's it.

Publish the Wolf of Vegas survey!

I likes it precious....

Quote: Wizard

This could make for a textbook statistics problem. Yes, I think the Gaming Control Board publishes the number of slots per casino. Maybe I could do something like 1 per 1,000, with a minimum of 5.

I don't think any casino in Nevada has more than 5000 machines so the 1 per 1000 seems superfluous..

First off, I love this idea! Perhaps with pics and machine numbers sent the entire forum could help out with this? And lastly, I think it would probably need to be more like 1 sample per 300ish slots with a minimum of 5 samples with a note of preferring different slot banks for samples instead of 5 in a row. Just bouncing some ideas. Maybe some simple excel table could be used for sending samples if that idea takes off?