Originally Posted by
mickeycrimm
Mission, we have an anomaly on the Harley's. It has to do with the frequency of catching the mega free spins. I've collected a lot of stats on the HD's. Here are some of the frequencies I came up with:
1. Frequency of Bonus/Bonus landing on the 1st and 3rd reels is 13.
2. Frequency of mini symbol landing on 5th reel is 17
3. Frequency of maxi symbol landing on 5th reel is 52
4. Frequency of mega symbol landing on 5th reel is 92
Using my stats I initially put the frequency of catching the mini freespins at 221, maxi 661, mega 1157. My actual results on the machines were very close on the mini and maxi but not even close on the mega. I was averaging over 2400 spins per mega hit. And everyone else was having the same result. Then someone put up the par sheet numbers on twitter:
mini = 220
maxi = 675
mega = 2553
So I had come up with accurate numbers for the mini and maxi frequencies but the same methodology didn't work for the mega. It's an anomaly I can't explain.
A Binomial Distribution would be sufficient to determine the probability of your results v. what was on the PAR sheets. How many spins did you sample? You can get into some Variance with a frequency of 1 in 92, especially if it's actually supposed to be less likely than that. Looks like it should be about 1 in 13 and 1 in 196 for Mega based on the PAR sheets, so we can start there.
Let's say that your sample size is 20,000 spins, we can then use a BInomial Distribution calculator:
http://vassarstats.net/binomialX.html
Okay, so we say 20,000 spins. You got it at a quip of 1 in 92, so we want to see 217 total in 20,000 spins. The actual probability is supposed to be 1 in 196 to see that on the fifth reel.
<0.000001
Yup, so you definitely shouldn't have seen that many. WAY outside the third standard deviation.
Conclusion
I can only reach three possible conclusions to solve the anomaly. First, I would be interested if the PAR Sheets contain the frequency of what the Mega landing on the fifth reel is supposed to be.
1.) Your sample size was less than 20,000, in which case, the Mega results could be skewed. If your sample size was MUCH smaller than 20,000, then something like that could reasonably be that skewed.
2.) Somehow, the Mega symbol is MORE likely to hit if the two Bonus symbols don't hit.
-Notice that the Mega symbol was showing up more frequently, but the probability of hitting the Mega Free Games came out pretty close.
3.) You just had a sample WAY outside the third standard deviation.