I think I have found the URL but so far can only access json files of cases that are presented to me in the tribunal, before the verdict is decided, do you know if they have made it this way recently or am I using the wrong URL?

The reason I am asking is because I am interested in performing a similair task and it would be rather inefficient if only 20 samples could be collected a day.

I never found anything like that. I know that my code still works.

:3
I was using the right url but hadn't spelled it correctly.
One last question:
would you mind sharing (creating a paste in pastebin) the code you used to get the amount of games in a particular case because that is stored in a javascript inside the html of the reform card and the method I am using currently is bottlenecking the rest of the code severely.
I understand if you don't want to and all the information you have given is still appreciated.

Awesome stuff!

Hey for anyone interested, I wrote some software to do basically the same thing as OP, shoot me a PM if you want it.

Also, have you tried analyzing how often writing in all caps results in a punishment? I'm too lazy to do the statistics and writeup, but based on the data I have, there is definitely a strong correlation.

Very good read. Im a bit curious what you are using to pull the information.

Id be interested in stats on tribunal users if that is possible. For example if there are any users that have a significantly low accuracy that is outside the normal.

Sir, it's time you give advice to another trying to follow in your footsteps.

http://na.leagueoflegends.com/board/....php?t=2959928

so it was YOU who ran this before!!!

Your agreement section was most interesting to me, since when I ran my data, there clearly was dip in majority (controversial) pardons.

Is this a change in behavior since august? If it is, it may actually lend credence to my theory that people are leaning more and more toward punishes (and my theory that its due to justice rating+%correct)

IF this is true, it restores some of my faith in humanity, people aren't actually getting on their horses when they see someone say noob/bad... They might actually just want justice rating.

oh? how so?

below is stats. You can see that majority pardon is only 1.5% of pardons. Where as majority punish is 43% of punishes.

More importantly, anyone know how I can have it output it into more code format? I hate how hte data looks when I output it currently.

===============
Pardoned cases by agreement:
majority : 47 1.5
strong : 2335 74.39
overwhelming : 757 24.12
Total : 3139 100
===============
Punished cases by agreement:
majority : 4875 43.08
strong : 3814 33.71
overwhelming : 2626 23.21
Total : 11315 100