What % chance do you have against Magnus?

The ELO system is designed in a way that a ~120 rating difference should indicate a 2-1 favorite or "twice as strong"  (calculations are approximate, probably 1-5% error)  These algorithms make it so exponential gains in skill are measured in a linear way.
ELO performance calculator: http://www.paxmans.net/performance_calc.php
+120 [2-1]
+191 [3-1] (In theory, a 1191 should beat a 1000 in 3/4 games, on avg)
+241 [4-1]
+280 [5-1] 
+311 [6-1] (In theory a 1311 should beat a 1000 6/7 games, on avg) 
+338 [7-1]
+361 [8-1]
+381 [9-1] (A 1381 should beat a 1000 90% of the time, in theory) 
+511 [19-1] (A 1511 should beat a 1000 95% of the time, in theory)
+676 [49-1] (A 1676 should beat a 1000 98% of the time, in theory) (aka 49x stronger)

Every 120 pts you gain, you should be 2x stronger than before (in theory).  A 220 should be 2x stronger than a 100 (don't laugh) a 340 should be 2x stronger than a 220 and 4x stronger than a 100, a 460 should be 8x stronger than a 100, etc.   I can't find out what the floor is for a FIDE rating, I know USCF goes down to 100, I think.  
If a 100 FIDE player is possible, and let's say Magnus floats around 2850, he is 2750 points higher, divided by 120.5 means he is approximately (2^22.816) 7.3 million times stronger than a 100. 
So if ratings started at 100, and moved exponentially in accordance with skill, magnus would be approximately 730 MILLION FIDE. 
Just to give us more perspective and to see how much we suck compared to Magnus.
Magnus vs You 
FIDE ratings  (These are approximations intended for entertainment purposes)

If you are 2730 rated (thanks for reading my thread btw) You are expected to score 1/3 vs the WC
If you are 2610 rated you are expected to score 1/5 vs Magnus, aka he is 4x stronger than you.  
If you are 2490 rated you are expected to score 1/9. He is approximately 8x stronger.
2370 1/17, 16x stronger
2250 32x stronger
2130 64x stronger
2010 128x The WC is 128x stronger than you (theoretically)
1890 256x
1770 512x
1650 1024x
1530 2048x (It might take more than 2049 games to beat him though ) 
1410 4096x
1290 8192x 
1170 16,384x 
1050 32,768x 
930 65,536x
810 131,072x  So you're saying theres a chance!  (probably not) 
100 7.3M 

This thread is just for fun but if someone has any major corrections, please chime in.  I also don't actually know what the FIDE rating floor is.

There's definitely a flaw with your definition of "twice as strong": being twice as strong as someone does not mean that you would score 2/3 versus them over a large enough sample. It means that the difference between you and them is the same as the difference between them and a beginner. In other words, you have to score 90% or more against somebody to have any claim to being twice as strong as them.

What's your definition of "twice as strong"? I don't get it.

Hmm I still don't follow.  This seems to make sense:
An 1120 beats a 1000 2-1 (in theory)
A 1240 beats a 1000 4-1 (in theory)  The 1240 is 2x as good as the 1120 at beating lower rated players.  

An 1120 is expected to lose 4 games to 1 vs a 1360
A 1240 is expected to lose 2 games to 1 vs a 1360

The 1120 is half as good as the 1240 at beating higher rated players

If I beat someone 66% of the time then I'm 2 times stronger than him???? what??

Now YOU are phrasing it wrong..  This doesn't apply.  The argument would have to sound more like this: Person A is TWICE as strong as person B at winning races against X competition.   Therefore person A will win against X twice as often as person B.  There are more factors than just speed in a race, endurance (stamina) 

Geez, what a weird definition.... 

My chances are obviously much less than 1%, and if I do, there's a really good chance it's not because I'm better than him, but here are the most possible causes for me beating him if I do happen to win:

1. He was bribed to lose.

2. He died.

3. Some sort of major, yet sudden health problem occurred.

4. He got really bad sleep and completely overlooked a piece or two.

5. His brain was taken over by aliens.

....

48362. I figured out which part of his brain is most essential to playing chess, and I teleported that portion of his brain into the core of the Earth, so he played the equivalent to that of a player with a 700 Chess.com rating.

48363. A dog startled him and he accidentally tipped over his king.

48364. I am better than him.

48365. His brain randomly got cluttered with a wave of thoughts about cat videos, and he couldn't stop thinking about cat videos, and it was so bad, that he started to hallucinate cat videos, and I looked like an upright cat to him, and the chess pieces looked like cats, and the pawns looked like little balls of yarn, and the chess board was a litter box, and he just stared at these cat hallucinations, then he lost on time.

48366. He is addicted to HotHands, and he ran out, so he couldn't concentrate at all, and he fell for an easily avoidable mate in 2.

48367. He had the sudden urge to solve a Rubix cube as soon as possible.

Differences in chess skill is a lot harder to quantitatively measure because there are so many factors involved in skill (along with variance).  Person A could be a stronger overall player than Person B, but worse at tactical calculations.   Chess involves calculation, memorization, creativity, speed and arguably mental resilience, determination, focus, competitiveness, susceptibility to mistakes, susceptibility to variance in mood and mental clarity are all important factors as well.  It is difficult to put a quantitative measure on chess skill, so I am attempting to use this fun (but over simplified) performance based formula.  I realize this cannot be perfect, or anywhere close to it, but I am still attempting this fun exercise. I figure this will get the ball rolling and maybe we will all figure it out together  

Huh me vs magnus? I'll probably win if he doesn't have a queen and a rook

And finally this can never be truly accurate anyway, because certain players will overperform or underperform against certain style differences.  Person A could excel mostly in tactical positions, and therefore will have a worse chance to beat Strong player X who's biggest strength is also tactics (reducing imbalances).  If Person B is the same rating or "strength" as Person A, but he is an opening nut, and a positional guru, while weaker in tactics, he may actually have better chances to score an upset against Strong player X than person A.

Oooh, now I get it.

If I'm 120 points higher, then in any number of games, my number of victories will be twice the number of my opponent's victories. Therefore, I'm "twice as strong".

Makes sense, makes sense....

 Also another way it can be argued, and possibly more legitimate.   If player A has  66% expectancy vs player X and player B has a 33% expectancy vs player X, is player A twice as good as Player B?  Maybe this is the more accurate way to do it.  
This would mean that 240 pts means 2x in "strength" not 120.

Ok I think my math was all wrong.  But I want to figure this out.  
Is this how it should be done??
Person A 880 rating, 33% expectancy vs 1000
Person B 1120 rating, 66% expectancy vs 1000
Person B is 2x as strong as person A?

I need to go for a drink...

very fun

I'm trying to remember what American player it was who said his greatest chess blunder was waking up Bobby Fischer!

Bisguier. Last round of the US Championship and Fischer's nodded off, if he loses on time then Bisguier takes the title. Fortunately for Fischer he had a sporting opponent and once he'd stretched and yawned it was game over.