Troll Tally

I had an Annie take ghost and smite to support top lane and an Udyr take cleanse and ignite to jungle. None of this was really addressed going in and tbh I should have just dodged. It all started when a support with a bard-themed name got mad because our top laner took Bard (which he turned out to be decent with). He mentioned something about trolling but I passed it off as most of the League community is sarcastic and I looked up his Annie stats which were pretty good. I figured he was going to supp bot and was having a laugh at not changing his summoner spells til the last minute (turns out I was wrong). The whole game was a mess. I ended up going 5/0/3 despite the Udyr running straight to the opposing mid laner and griefing followed by a bot lane grief and a terrible AP item build.

This circumstance has prompted me to conduct a troll tally: a list that says if my team had a troll or the opposing team had a troll (or both). I need your help community. To keep it scientific, I will have to define what it means to troll. I won't include simply raging as this happens from time to time, but I should include excessive raging (but I need to define excessive raging). What do you consider excessive raging in logical terms? Also intentional feeding will will be a must. So that just leaves excessive raging (still needing to be defined) and griefing. Can you think of anything else I should include? A lot is said that trolls will balance out i.e. I will have as many trolls on my team as the teams I play against. But I have heard this theory debunked as games are generated randomly which does not necessarily mean trolls will balance out. All in all, they don't have that much of an effect on climbing, and these results will only be of games that I have played. Although I don't provoke other players and pretty much keep quiet, this will not be a sufficient sampling. It's just out of curiosity. I will post my results at the end of the season if anybody is interested or wants to join in and record the data, too. Thanks for your help :)