Hearthpwn

Quote from Psy_Kik >>

If you peak at 6k it's obviously pretty hard to notice this - and I am telling you, if you are peaking at 6k then you are making a shitload of mistakes in terms of optimal play, every single game.

 Gotta be said, it's hard to spot mistakes when you're pretty much half dead by T6, which means basically one bad attack order later and you are dead (the 15 dam cap is basically nonexistent due to people quitting early - perils of rank floors). This metagame is really heavily hero-based, I'm finding - I've been hitting 7k previously, but currently swinging between 5.9 and then a couple of 6th places and I'm back down at 5.6k (seriously annoying how one 6th wipes out multiple 3rds). Quite a few bad heroes, and with the early power of weaver or just the opps hitting synergy early it's easy to just lose the entire earlygame when your HP does nothing for multiple turns.

I do struggle with how much the metagame shifts between 'games over by T8' and 'can still be out 5th on T13' between games based on what's in, though. Adapting to metagame shifts quickly in BGs has never been my forte, if I'm honest. Probably averaging about 4th or so (not usually quite hitting 5k before I bounce back), but the variance is absurd at the moment. Just not much you can do when it's stuff like Patches/patchwerk/afkay etc that just don't really work well, then you don't hit a weaver.

Weaver does feel like a crutch that I overrely on, though. Just lets you handle early-mid with ease if you hit one.

With that said, BGs is not rigged, and anyone claiming it is should show statistical rigour.

Quote from RandomNoob >>

It doesn't even matter if it's not rigged as long it feels rigged.

And there are indeed rigged elements in Hearthstone. Some games in solo adventures have rigged card draw (drawing the correct card to satisfy the story telling). 

Also take in consideration that Zephrys the Great is a card they have been working on since early Hearthstone (as devs said) and a big business like Blizzard would not allow a card out of 135 from the set to be that demanding unless the code behind can be reused to force that ideal 50 % win-rate that keeps the customers.

 Using Zeph as an example is bad. Many situations where the perfect cards he offers are not perfect, even within the very limited selection he can offer. Doesn't give transform removal for deathrattles under about 7/7s, for example, and can't handle multiple things to deal with. Opp has a secret, a weapon, and a board? Time for a bad value option, ST removal and a boardwipe, no flare or ooze. It gets even worse when you make bad plays in order to increase the odds of a good Zeph (via remaining mana manipulation) and then he just... whiffs.

Zeph's logic is nowhere near as powerful as people make it out to be. Can almost always see a lethal (there's a few times he's missed it for me, though, because the logic doesn't take ongoing effects from quests into account), and will sometimes give you a good card.

On the actual topic, I've been stuck at 4k in one meta due to how heavily weighted the good heroes were (no pass), whilst my rating is usually significantly higher. My rating usually notably shifts when there's a meta change, which will be down to skill level relative to the lobby. Both of these act completely contrary to your hypothesis. I will say that there's extremely small gains in higher finishes relative to losses when you lose - one 7th loses you more than you gain in 2nd, one sixth usually loses about 60 whilst a third gives 30-40ish, but that's how the mode goes. If you're not averaging above fourth, you're not climbing.