Champion points too closely tied to how well the team did

I've noticed that my ranking and CP does not seem to be as closely tied to my actual gameplay as I would like. I am still new-ish to the game, experienced enough to know what should be done, newb enough to sometimes screw it up in the heat of the moment; I play unranked but I'm probably mid to upper bronze level. One thing I always do is after a death, or major team fight, I sit down and ask myself what happened, what I did wrong, and if there was anything I could have done better to prevent a death. For this reason I'm painfully aware of why I'm still only bronze, I see plenty of times where I could have positioned better or forgot to use an item that was of CD; or used an item in a hopeless battle that wasted it. However, my point is that I think I have a good idea of when I play well, and when I'm playing foolishly.

The problem is that my ranking after the game doesn't seem to be tied to that as well as I would like. Sometimes I'm playing late at night and not on my game and making a few mistakes; nothing horrible but not doing my best, and yet I'll get an A+. Other times I'll be playing perfectly, or as perfectly as my limited skill ever plays, and get a B- or even C.

The problem seems to be that my CP ranking is too closely tied to the team as a whole. For example I may be playing rakka and were winning bot lane against the other champs, and even surviving multiple jungle ganks without any help from our jungler. I may in fact be a large part of how we survive ganks because i"m on my game. Meanwhile the rest of the team may be loosing horrible and and the rest of the enemy team snowballing, such that by the time we get close to team fights there is an unquestionable advantage for the other team. My team get's butchered in a few team fights, we surrender, I end up with a C ranking. However, I was on my game, we did great in bot. Yes during team fights I didn't manage to save many people, but that's because some fed assassin was killing them as fast as they showed up; all I could do was try to save a few and not get assassinated myself. However, because my heal totals were lower then the average raka I get a bad ranking.

Least you think I'm complaining about not getting high enough rankings, it's actually the opposite that I hate worse. I may have a great carry on bot who is just killing everything, but i'm playing poorly. I misposition and nearly get killed and my carry manages to burn down the person attacking me to save me. I miss my e when I try to root anyone and I never seem to use my global ult when it get's the best advantage, and otherwise am not on my game that night. However, since my carry is snowballing the enemy team I get credit for not dying when he save me, and for healing him after he butchers the enemy team alone. I may walk away with a relatively high ranking, say A+, despite knowing that I didn't really make my presence known throughout the game. This is worst in my mind, since the idea of using the rankings to tell how well your doing is defeated if you can't trust them.

I understand the difficulty of making this work perfectly, and I suspect that supports are some of the hardest to rank, since so much of our contribution to the team depends on interaction with, and thus skill of, our carries as well. However, I think there are some small things that could be done to try to skill the champion ranking to the champ, and not the team's, performance. The basic idea that when your team is loosing lower the expected values (kills, damage, healing etc) that you require to reach a ranking. When your team is snowballing increase it. Basically if my team is winning then the enemy team is under leveled and geared, has less armor, and I have less reason to be afraid of being killed if I go all in; so of course I can safely get a few more kills on my lighter defended foes. If my team is leading and I'm up against people with a level advantage getting kills should be a far more impressive sign of skill!

1. the easy solution is look at the final kill/level/death totals of you and your enemy team at the end of the game. If your team has a massive level disparity relative to the enemy team at end of game use that to adjust the expected values to reach a ranking

2. more complex example would be look at how those not close to you do against a champ, and compare it to those who are close to you. If someone is regularly and consistently dieing horrible when not near me, but he dies far less often when near me that suggests I'm good enough to save an obvious feeder. It may be that this person is still feeding to much, which puts us at enough of a disadvantage to cause us to loose horrible, but if I can limit the harm a bad player does that may be a sign I'm playing well even though my team is loosing. Of course this one may be much harder to implement.

I would love to see at least the first suggestion be implemented. Tying my ranking to how well the team does takes away the biggest advantage/fun of having character rankings in my boat. I want to be able to say, yeah my team bombed horrible, but look I did a decent job trying to salvage everything despite that. Or alternatively, yes we won, but clearly I was carried by better teammates my ranking was bad, lets figure out what I did wrong to fix my next game.