Actually, I've had plenty of games where I went positive and lost. Conversely, I actually recognize my mistakes and work to improve on my mistakes rather than lament my misfortune on the forums to people who don't give a shit. I lost 9 games in a row and dropped back to gold 4 during the comcast issue - where I personally played 2 bad games and had a couple of games with leavers on my side - yet here I am again in gold three just a few games after being demoted. The difference between you and me is that I know I'm not a pro and so I don't spend my entire league game criticizing my teammates and instead spend it critiquing my own play.
Pretty much this. Enough people have shown evidence of ELO hell not existing, and also, y'know, statistics and facts from Riot/analyzers, to pretty much sadly shake one's head at anyone blaming their persistent rating on other people. Even on Reddit and a lot of the league forums 'ELO hell' is used sarcastically more than not now. Your faulty understanding of the chance of a troll on your team vs. their team is further indicative of you not trying to understand shit and just continuing to try and belittle other people instead of thinking critically for 30 seconds.
Out of the ~1000 normal games and ~50 ranked games across 3 seasons that I've played, I don't think I've seen more than 20 hard-core trolls like you're describing. I don't have people in my games constantly raging for no reason, or just shitting the bed for no reason, and there's been plenty of games where I've come from way, way, way behind and won to even get negative when we're down by 15 kills 10 minutes in. Hell, even 4v5's aren't worth letting your 'game on' gland suicide and just moping your way to another loss, plenty of 4v5 games go to the 4.
In theory, you and I are playing the same game. Hell, I would be surprised if we hadn't had the same people in some of our matches (considering how many normals I've played, and how many baddies are in normals). Yet my game experience is nearly 100% opposite to yours. Considering the player population we interact with is statistically similar, what is the variable? Now, because of my approach, knowledge, and fairly minor mechanical skill to back that knowledge up, I've climbed from bronze (season 1) to getting placed into Silver 1 in season 3, and gone from being ~40 wins below losses in normal to being 40 wins ABOVE losses in a matter of a couple hundred games. This is all due to shit >I< did, not shit the (9 * 1050) other personalities in my games did.
You don't want to take responsibility for your actions, you don't want to man up and really look at your own approach, skills, knowledge, and interaction with players in an objective way. If you were actually good, that wouldn't matter (plenty of asshats get up to Diamond 1 and even Challenger), but... you are not good. And you will continue to not be good for as long as you see fit to blame other people for every intricacy of the game. Tons of people do this in various ways in real-life situations and interactions, and it's bad there too.
Edit:
They wouldn't place in bronze to begin with if that was the case. Besides which I'm quite sure normal queue mmr doesn't bleed into ranked.
Normal MMR does have SOME impact on your first couple of placement matches. You apparently have to play 10 normal matches before you can do your ranked matches, because the game doesn't want to default you to a set number for some reason (seems dumb since 10 matches is fairly irrelevant anyway). Also, the 'test' accounts Riot just gave most of the Pro gamers to analyse various things with MMR/Player Behavior were just started at level 30 with no normals played and auto-placed in to Bronze I believe (the pros keep talking about bronze players in those games, not sure if it's just trash talk or legit commentary).