Does causation imply correlation?

Yes, obviously. 2 33.33%
No, obviously. 3 50.00%
Ummmmm. . . 1 16.67%
Voters: 6. You may not vote on this poll

League of Legends, Statistics, and You.

Comment below rating threshold, click here to show it.

Tetraplasm

This user has referred a friend to League of Legends, click for more information

Senior Member

07-24-2011

This may SEEM like an off-topic discussion, but it does in fact relate in general to League of Legends.

Many people have (over a course of a long time, not all at once) posted polls asking things like "Should eve be nerfed?" or "do these kayle changes suck?" or "is leona super underpowered and annoying for wearing high heels?"

Surprisingly (or not, to those like me who are taking statistics in school/have taken) the results come back into the favor of the opinion that the OP posts with (unless the OP has a ridiculous opinion like Twitch is super OP or something).

When the data comes back from these polls saying "99% of players HATE RIOT'S GUTS for SLIGHTLY CHANGING A CHAMPION!!!" You may think "Oh, geez, I guess EVERYONE hates Riot." (fake poll, I just made it up for the sake of the topic)

Wrong. Not necessarily.

People who are currently angry/dissatisfied are more likely to go take a poll that bashes what they are angry at. Also, the sample size of the poll is extremely important. Take 99% of the poll's answers that voted they hate Riot - it might be something like 500 people. 500 people out of the hundreds of thousands, if not millions, of people who play.

The same can go in a positive aspect. Say there is a poll that asks if you loved the winter map and desperately want it back, results come back 80% of people want it. Well the poll hasn't adequately covered all the players in league of legends, now, has it? It hasn't even probably gotten all the players on the forums alone to answer it.


All this to say, the poll feature is cute, but don't take it for the word of God/whatever diety you love. For this reason, Riot tends to patch things based on internal testing and their methods rather than being the community's *****.

TL;DR Before you QQ about some poll, look at the statistical facts of it. If you don't understand statistics (believe me, if you think you do and haven't taken stats class, you probably don't know as much as you should. I thought I knew a lot, boy I was wrong, lol.) then you really shouldn't be analyzing polls, now, should you?

Thank you for your time.

EDIT: Lol, totally typo'd the poll. It should be "Does correlation imply causation." D'oh.