Looks like pound for pound the ant wins eh.

Actually an elephant sized ant couldn't exist - it would asphyxiate itself.

Elephant sized ants are suicidal, TIL.

It also wouldn't be able to move.

Yes, an inconvenient side effect of being dead.

Ok then, how about an elephant vs a swarm of ants with total ant mass equal to one elephant mass.

Google calculations give me ~1.5 million ants.

"Giant ants were the terror of the movie Them! (1954).  Rajesh and Howard realize giant ants would be a cool new method of transportation.  But Sheldon Cooper is right:  unfortunately physics determines that giant ants cannot exist on our planet as we know it."

Spock made a brief appearance in the movie. If giant ants aren't possible, they must have shrunk the people.

An elephant is not much the fighting animal... it probably wouldn't recognize the danger in time. 

also, 1.5 million ants are way more flexible than an elephant! There are just so many ways they can split and swarm... how will he block 10,000 of them coming up his nostril? while 20,000 more are busy tickling an armpit, 5000 busy bothering the end of his digestive canal, and 15,000 more crawling up an ear? This could drive anyone crazy!

Piece value in chess: 9 for a queen, 10 for a couple rooks, 12 for four minor pieces - in total, 31 pawns. 

Now, how would a setup of King + usual pieces, standing on their first line, fare against an array of King + 31 pawns where the black king stands on e8 and the pawns occupy the whole of the 7th, 6th, 5th and 4th ranks except for, say, d4? 

I remind you that each pawn can make a double jump on its first move!! Monstrous indeed :-)

I'd take that against a 2700 player any day of the week - hey! Material is even and he even gets white :-) he should be happy (true that I'm no 1300, but I promise to patzer him a couple pawns at some point...)

I meant that their legs would break from the weight. 

only if the 2700 player intentionally loses otherwise odds are the same as a snowball in hell.

And those odds would be...?

Actually this may be playable or even winnable for the one with the pieces...

I think about sacrificing a bunch of pieces on one side (like, the a-b file area), then penetrating with a Queen and a rook (maybe two, if two are left) to just checkmate the king... the pawns are slow and won't do much damage - but it depends how long it takes to actually arrange the pieces for these sacs... an interesting question after all

Rhapsody in Blue is always nice.  Better with an orchestra, of course.  One of those few "concertos" that gets played solo.

I think one of the real problems with this is that a GM can see much further than a 1300, which is why they make better positional moves.  The 1300 could even make a move that looks perfect but the GM can quickly see a 13 turn tactic that destroys them.

I think the easiest answer is to say that the 1300 has a chance to beat the GM but its implausible.  Sure, the GM could die a few turns into it, or the 1300 could be Bobby Fischer who never played a rated game since getting to 1300.  Maybe the 1300 is a hustler who purposely keeps their rating low.  Maybe the GM is the hustler who purposely loses the first game and then bets a huge sum of money on the second?  Oh, maybe the 1300 is a super intelligent alien chess genius and it's their first chess game?

I think one of the problems is that the gm sees much farther than the amateur, understands positional chess much better, doesn't blunder near as much and the blunders are of a much more "minor" nature, knows the basic mating patterns, knows basic endgames, knows typical middlegame plans, and understands what good and bad pieces are, knows what good and bad pawn structures are and how they relate to the pieces are...But, that is just about it.

I agree completely erikido23 and everything you just stated is why it would never happen.

Anytime actually. It depends on the game they are playing. If the 1300 rated player is a professional athlete in any sports like basketball, tennis, badminton, table tennis, wrestling, mma, golf, etc and plays the 2700 rated master in his profession instead of chess, the 2700 rated master is beaten everytime. Just try to imagine a 1300 rated professional boxer fights a 2700 rated master in a boxing ring.! Lights out!

Okay, so here's an extension question. What rating *would* you need to be to have a cat's chance in hell against a 2700 GM in a one-off game?

Ah, that's easy - 2300. Definitely.

... although:

http://www.youtube.com/watch?v=3nYwTWycVSM

http://www.chessgames.com/perl/chessgame?gid=1606556

Agreed, and add to that the inability of the 1300 to effectively capitalize on any mistakes the 2700 did make along the way.

Remember, ratings measure results, not "strength."  The maximum deviation of 350-400 points is the outer limit of probability for any result other than the higher rated player wins. 

That's not based on some theoretical supposition, it's a statistical fact.  Beyond that rating difference - assuming as we must that both ratings are reasonably accurate - the chance of the lower rated player not losing approaches zero.  The only reason it isn't actually zero is due more to mathematical theory than chess practice.