"Correlation does not mean causation" is a theme taught to me by my statistics teacher in highschool.
It is equally likely that the team that loses are more likely to verbally abuse their teammates
Hence this is misleading
Edit: reworded the post to be more "correct"
You are correct that correlation != causation, but it can be assumed that this statistic was computed from a population. That is, this statement comes straight from the data and is not an inference about future data.
Yea, RiotDerivative is right. Not that I would be able to verify his claims. Perhaps they will change when you take his AntiDerivative and discover his true form?
I'm a stuffy statistician, dont worry about it. i'm finishing up my degree at UoM and yeah, you're definitely right in saying that because they come from the same population isn't a case of correlation vs causation. Being that you just ran the variables against each other within the same pop there is no inference or extrapolation happening. Simply an analysis.
On a side note, as a statistician myself I recently took Calculus 4: Multivariate & Vector calc and am wondering how applicable this stuff is in real world use. I'm assuming that there are programs and everything already in place and that the need to use spherical/polar/cylindrical coordinates in triple integrals isn't ever practiced. How much modeling do you do that is anything beyond possibly calculus 2?
It is actually pretty relevant. Anytime you deal with distributions containing multiple variables, you will be dealing with multiple integrals. I see it everywhere especially in Bayesian statistics. If you do statistical computing, one of the main responsibilities is "optimization." When you have multiple variables in an analysis or simulation, all of the data points form a surface or multidimensional geometric figure that you must optimize on. This is common in statistical computing and machine learning.
The only time I have seen polar coordinates was in seeing a proof on creating random normally distributed values using only uniform random numbers. That was pretty cool.
>> I must be wasting my time studying Sociology because I couldn't follow that at all.
Sociology has a lot of uses in games and in general!
Quote:
Originally Posted by Robtard
When you're looking at data with tons of variables and you want to check out just a few for "optimizing", would you just do partial derivatives to pull those variables out while leaving the rest constant to drop off and then at that point you can be left with something in R3 or greater for analysis? that stuff isn't so bad as far as finding critical points and whathaveyou but don't you typically just jam it into statistical software (such as the open source R) to get what you're looking for?
I guess that i'm just struggling with that age old idea of "I'm never gonna use this stuff when I'm in the work place"
Also, do you still run hypothesis tests? I feel like i've been doing them forever and hope they're worth it.
What you are saying is basically correct. You could optimize for a particular variable or parameter by doing partial derivatives. If you take mathematical statistics (usually junior/senior level), you will learn this. The same is true for statistical models that have a lot of parameters. You can study the behavior of one or more parameters by integrating out the ones that are not of interest.
Something like R would be used to actually do the optimization, but it is good to know how it works under the hood. For example, what if this complicated surface is not differentiable? Then, R won't give a very reliable result. If you don't know that, then you might report a result that is incorrect.
I still use hypothesis tests for certain things to convince myself, or someone else, that I am correct ;-). They are more powerful for small samples though (less than millions).
I must be wasting my time studying Sociology because I couldn't follow that at all.
Definitely not a waste--but as you go deeper in social science fields you'll find that having strong stats knowledge will put you far ahead of your colleagues and really make you stand out! Because social science relies on fuzzier measures, knowing your tools (e.g., statistics, psychometrics, etc.) super-well is a major asset
Summoners who verbally abuse their team lose 16% more games.
"Correlation does not mean causation" is a theme taught to me by my statistics teacher in highschool.
It is equally likely that the team that loses are more likely to verbally abuse their teammates
Hence this is misleading
Edit: reworded the post to be more "correct"
"Correlation does not mean causation" is a theme taught to me by my statistics teacher in highschool.
It is equally likely that the team that loses are more likely to verbally abuse their teammates
Hence this is misleading
Edit: reworded the post to be more "correct"
On a side note, as a statistician myself I recently took Calculus 4: Multivariate & Vector calc and am wondering how applicable this stuff is in real world use. I'm assuming that there are programs and everything already in place and that the need to use spherical/polar/cylindrical coordinates in triple integrals isn't ever practiced. How much modeling do you do that is anything beyond possibly calculus 2?
The only time I have seen polar coordinates was in seeing a proof on creating random normally distributed values using only uniform random numbers. That was pretty cool.
Sociology has a lot of uses in games and in general!
I guess that i'm just struggling with that age old idea of "I'm never gonna use this stuff when I'm in the work place"
Also, do you still run hypothesis tests? I feel like i've been doing them forever and hope they're worth it.
Something like R would be used to actually do the optimization, but it is good to know how it works under the hood. For example, what if this complicated surface is not differentiable? Then, R won't give a very reliable result. If you don't know that, then you might report a result that is incorrect.
I still use hypothesis tests for certain things to convince myself, or someone else, that I am correct ;-). They are more powerful for small samples though (less than millions).