why is it hard to report/detect inting (i mean those obvious cases) but its easy to detect bad chat?
Because a chat filter is... well, just a chat filter. Basically, if someone uses a flagged word (fuck, shit, ZT language such as slurs) it's marked and the context is weighed lightly (ex. saying "Fuck me" is acceptable, "Fuck you" is not.) Realistically it takes less than a few minutes to fully and accurately review a chat log.
Intentional feeding and griefing is a much harder thing to detect just due to the sheer amount of variables. Let's say there's a 0/12/7 Jax on the enemy team.
- How often do they play Jax?
- Are they experienced in this particular matchup?
- He's got 253 CS by 27 minutes, is he split pushing or refusing to group?
- Did the jungler gank/camp him?
- Did he say anything in chat?
- When did they last play League?
- What rank are they?
- His jungler came up and he went in with him, is Jax only comfortable when teammates are around?
These are just some examples of the thousands of variables that come into account when reviewing someone for intentional feeding and griefing. Basically, you have a word checker (Did the person use the word "fuck" in a negative way?) versus checking someone's playstyle, mouse patterns, damage dealt to champions/towers/objectives, CS, ward score... everything.
:p