What's funny is, IMO, it's more likely for a random move generator to beat Carlsen than an experienced low rated player.
This is because a low rated player has internalized certain rules and methods that will make the likelihood of playing certain moves zero... and sometimes the best move is a 0% move... but a random move generator will always have a possibility of playing the best move.
As for having an off day, it's really hard to play more than 400 points below your rating. It requires something extreme that essentially takes away the person's will. For example being very sick or in a lot of pain.
If Carlsen is delirious and blunders enough material, sure you'd win, but that's not really a chess game 
If Usain Bolt were having a medical emergency could I win a 100m sprint? Well yeah... but that's not really a race.
And now someone might show me an OTB game where the upset is more than 400 points.
Sure sure, not every 400+ point upset involves a heart attack... but I would argue that every 400+ point upset (that isn't due to inaccurate ratings) involves a significant lack of will power from the higher rated player.
So if Magnus is just having a slightly off day, the beginner is hyper analyzing every move, and there is plenty of time for said "beginner" to fully blunder check, would it be possible? Could I even?
Calculation is a very poor tool for understanding the position. Players really only do a lot of calculation in 2 situation:
1) There are a lot of forcing moves and they have to be sure they're not losing material
2) They don't understand the position and as a last resort try to calculate their way into some understanding
#2 is very ineffective. There are simply too many possible positions. It's why early chess computers were weaker than GMs even when they could calculate millions of positions per second. If all the moves you calculate are garbage, then calculation doesn't matter.
If I were to play Carlsen (or any GM) there might be a moment where I can trade into an endgame (for example). But is that endgame good for me? Is it a draw? Sometimes, no amount of calculation will tell me the answer. I could sit there and calculate for 24 hours, it wouldn't matter... but if I'd learned the answer via study and experience (like the GM has) then I'd know almost instantly, no calculation needed.
In other words even if you calculate 1 hour on every move, you'll know less about the position than Carlsen does after a few seconds. This is why GMs can play 50 people at once and win every game.
So if Magnus is just having a slightly off day, the beginner is hyper analyzing every move, and there is plenty of time for said "beginner" to fully blunder check, would it be possible? Could I even?
If Magnus has an off day, he still analyzes deeper and more precisely in 10 seconds than you in an hour.
Even posting the question is a tell: it shows that you brutally underestimate the skill difference between a 800-rated player and a super GM.
So if Magnus is just having a slightly off day, the beginner is hyper analyzing every move, and there is plenty of time for said "beginner" to fully blunder check, would it be possible? Could I even?
Calculation is a very poor tool for understanding the position. Players really only do a lot of calculation in 2 situation:
1) There are a lot of forcing moves and they have to be sure they're not losing material
2) They don't understand the position and as a last resort try to calculate their way into some understanding
The exact opposite of this is true. Grandmasters do a lot of calculations (and I mean a lot) before every move. That is what they spend their time on. They calculate more variations in a boring, flat position than an average player in a full month.
I know I'm posting a lot but...
One time I played in a simul against a 2600 GM. I saced my queen for 3 minor pieces... and I'd had this position in front of me on an engine within the last week. In other words it was an incredibly messy tactical position that I had a non-zero amount of experience in.
He paused at my board, but less than 10 seconds. I don't know how he saw through all those tactics so quickly. His next few moves were not as good as stockfish, but they were good enough to get a better position and then beat me.
Which is to say... my first example was how calculation wont help me understand a whole endgame... but also, even in positions where calculation IS effective, they're still monstrously superior. You'll calculate a lot of superfluous garbage in situations where they'll lock on to the critical line and so one quick calculation and they're already done analyzing.
So if Magnus is just having a slightly off day, the beginner is hyper analyzing every move, and there is plenty of time for said "beginner" to fully blunder check, would it be possible? Could I even?
Calculation is a very poor tool for understanding the position. Players really only do a lot of calculation in 2 situation:
1) There are a lot of forcing moves and they have to be sure they're not losing material
2) They don't understand the position and as a last resort try to calculate their way into some understanding
The exact opposite of this is true. Grandmasters do a lot of calculations (and I mean a lot) before every move. That is what they spend their time on. They calculate more variations in a boring, flat position than an average player in a full month.
There was some psychology study that revealed the rating that calculates the most is 2000-2100.
Beginners calculate less because they're slow.
GMs calculate less because they're more efficient.
If you think GMs are strong because they calculate a lot then you fundamentally misunderstand how chess works.
You'll calculate a lot of superfluous garbage in situations where they'll lock on to the critical line and so one quick calculation and they're already done analyzing.
No. How could they know in any position what the "critical line" is? It is only critical in hindsight.
Instead, they calculate a lot of superfluous garbage extremely quickly and efficiently. In a slow game, they do it before every move.
So if Magnus is just having a slightly off day, the beginner is hyper analyzing every move, and there is plenty of time for said "beginner" to fully blunder check, would it be possible? Could I even?
Calculation is a very poor tool for understanding the position. Players really only do a lot of calculation in 2 situation:
1) There are a lot of forcing moves and they have to be sure they're not losing material
2) They don't understand the position and as a last resort try to calculate their way into some understanding
The exact opposite of this is true. Grandmasters do a lot of calculations (and I mean a lot) before every move. That is what they spend their time on. They calculate more variations in a boring, flat position than an average player in a full month.
There was some psychology study that revealed the rating that calculates the most is 2000-2100.
Beginners calculate less because they're slow.
GMs calculate less because they're more efficient.
If you think GMs are strong because they calculate a lot then you fundamentally misunderstand how chess works.
You could not be more wrong even if you tried. Good job.
No. How could they know in any position what the "critical line" is?
The same way I do when I play weaker players. I know a lot of patterns and I prioritize piece activity. This means certain candidate moves I can dismiss without having to calculate a single move.
Instead, they calculate a lot of superfluous garbage extremely quickly and efficiently.
LOL
Llama36: just watch any GM analyze their game in the post mortem. They sometimes do it on camera for the fans. They show that they calculated every line to the end.
@OP: There is a formula to calculate the probability.
https://wismuth.com/elo/calculator.html#rating1=700&rating2=2800
I guess it is theoretically possible, but in reality it doesn't look too good for the 700 player. 
According to the calculator, even a 2300 rated player has only a chance of 0.8% to beat Magnus in a single game.
Llama36: just watch any GM analyze their game in the post mortem. They sometimes do it on camera for the fans. They show that they calculated every line to the end.
N-no... in post-game interviews they'll generally give a 4-6 move variation and render an evaluation.
Maybe it's a language issue... maybe you don't mean "calculate" the same way I do. For example in my own games, I might decide a move is bad because I know certain patterns... ok, like the 5 move smothered mate for example. I can "see" that and avoid it without calculating the individual moves the same way an adult reads a word without looking at individual letters. But maybe you would call that calculation. Maybe we're using the word differently.
Llama36: just watch any GM analyze their game in the post mortem. They sometimes do it on camera for the fans. They show that they calculated every line to the end.
N-no... in post-game interviews they'll generally give a 4-6 move variation and render an evaluation.
Maybe it's a language issue... maybe you don't mean "calculate" the same way I do. For example in my own games, I might decide a move is bad because I know certain patterns... ok, like the 5 move smothered mate for example. I can "see" that and avoid it without calculating the individual moves the same way an adult reads a word without looking at individual letters. But maybe you would call that calculation. Maybe we're using the word differently.
We are not using the word differently.
Of course, if you notice a pattern and you know that some move is good or bad, but that happens only occasionally. In most positions there are many plausible candidate moves and many implausible ones.
What do you think GMs are doing when they play a single slow game for 5 hours? What are they thinking about? Their lineup in their NBA Fantasy League game? No. They calculate a lot, they calculate even seemingly crazy moves, because they do not want to get crushed by an unlikely-looking tactical blow.
But sure, in positions where calculation is useful (tactics or certain endgames) the GM will calculate (and show in the interview) ridiculously long lines.
Calculation is a tool. In some positions it's useful, and in some positions it's not. Experienced players don't waste energy calculating garbage, they're very efficient at dismissing bad moves, and spend their energy on important moves.
But sure, in positions where calculation is useful (tactics or certain endgames) the GM will calculate (and show in the interview) ridiculously long lines.
Calculation is a tool. In some positions it's useful, and in some positions it's not. Experienced players don't waste energy calculating garbage, they're very efficient at dismissing bad moves, and spend their energy on important moves.
Okay, I see what you are saying, and I disagree. That would suggest that they play their moves in a few seconds in non-tactical positions, but in a slow game they don't. They usually think for at least 5-10 minutes on every move (unless they have only 1 legal move), calculating a lot, much faster and deeper than any 2100 rated player.
Llama36: just watch any GM analyze their game in the post mortem. They sometimes do it on camera for the fans. They show that they calculated every line to the end.
N-no... in post-game interviews they'll generally give a 4-6 move variation and render an evaluation.
Maybe it's a language issue... maybe you don't mean "calculate" the same way I do. For example in my own games, I might decide a move is bad because I know certain patterns... ok, like the 5 move smothered mate for example. I can "see" that and avoid it without calculating the individual moves the same way an adult reads a word without looking at individual letters. But maybe you would call that calculation. Maybe we're using the word differently.
We are not using the word differently.
Of course, if you notice a pattern and you know that some move is good or bad, but that happens only occasionally. In most positions there are many plausible candidate moves and many implausible ones.
What do you think GMs are doing when they play a single slow game for 5 hours? What are they thinking about? Their lineup in their NBA Fantasy League game? No. They calculate a lot, they calculate even seemingly crazy moves, because they do not want to get crushed by an unlikely-looking tactical blow.
Well for example, I might calculate a 2-3 move sequence where I open the c file but my opponent can use c4 as an outpost for their knight. If it's not a situation I immediately understand then I'll spend some time trying to figure out if that's ok.
So, without calculating, I imagine my rook on c1 their knight on c4. Can I play around the knight (ignore it). Can I sacrifice my rook for the knight? How many moves will it take my opponent to get there and can I use those moves to make something else happen? Which squares / pieces will be vulnerable when my opponent is making this maneuver? If I get 3 free moves in a row, how can I pressure those squares / pieces? If we both improve our pieces in a logical way, then what is my way to generate play? What is my opponent's? Do I have at least as much play? etc.
Some positions are calm and have no tactics... so move-by-move calculation is pointless. So I sit there thinking about things like this.
Okay I know this is a bit out there but bear with me. Obviously on any given day in rapid/blitz/etc. Magnus could probably wipe the floor with me. Also I am not necessarily talking about me specifically but in general relatively lower ranked players who have a decent grasp on the game as not to be a complete beginner.
In theory, is it possible to beat Magnus Carlsen as a 700 chess.com ranked player? Like if given classical time controls and actually playing each move carefully taking as much time as necessary to avoid blundering. I know everybody blunders once in a while even GMs and I've also seen a few times on this site where somebody low ranked gets lucky and beats someone significantly higher in rank than them. So if Magnus is just having a slightly off day, the beginner is hyper analyzing every move, and there is plenty of time for said "beginner" to fully blunder check, would it be possible? Could I even?