Statistics Question: Confidence

Mechanics, discrete, statistics, ...
Post Reply
Posts: 6
Joined: Sun Jul 15, 2012 11:55 pm

Statistics Question: Confidence

Post by renosis » Thu Oct 02, 2014 12:51 am

I need to figure something out to settle an argument. I am a layman when it comes to Mathematics / Statistics. I have solved over 100 Euler Problems, so I don't think I am a complete moron.

I play this game sometimes called Warframe. Warframe has this character called the "Stalker" who sometimes spawns on a level to kill players, he is kind of like a Bounty Hunter. He has a 3% *known* chance of appearing on any level. This number was released by the developers. The Stalker drops special game items if you are able to kill him.

Some players have the belief that if you are using low level weapons and equipment, the "Stalker" is more likely to attack you. There is no confirmation or denial of this from the Game Developers. This is where the problem lies. I believe this is just "Confirmation Bias" on the part of the players who believe this true, because once you hear this rumor, and then you see the "Stalker", the "Stalker" appearance becomes more memorable because it was a difficult fight and the player probably lost due to the low level weapons.

What I want are some numbers, so I can prove that any one player is not going to see him enough times and log it to get a statistically valid proof of this. A player rarely ever see this character. To even skew it in favor of the believers, i could say that at best players see him 2 times a day, and the game has only been out a year, so a player, who has seen the "Stalker" a lot could have seen him MAYBE 730 times.

How many times would a player have to see him with both low level and high level weapons to even give themselves confidence of lets say 75% that they are right?
What kind of sample size would I need to prove this true or untrue with a p > 0.05 I suppose?
How can I figure this out?

Post Reply