What is average depth human is searching in online chess?

Fezzik i just returned from country chess youuth cup U16 and i was 2nd... I played with player positioned one and in one position i had 3 best moves and only 3 , I have watched depth 20 for all of them and couldn`t find winning position for me with all best moves for both sides ... I have josen best variant and it was draw at the end at 89th move...

P.S : Game time was !h per player ...

Yes, we'll have to tweak the equation to account for proof of anything being consumed and for the preservatives and chemicals included in crappy beers, plus other chemical consumed by the player-- plus their cumulative effect.  I need a beer just typing this stuff.

 Well, to be honest, your incredible Fide rating speaks in favor of you..

But 15-20 moves really is a lot, and I think you can imagine we are quite sceptical about it, since only you know whether it is true..

The complexity of the board determines the depth in which can be seen.

For example:

if I have a rook and a pawn, and you have just a rook, I can see 20 moves ahead.

If we each have 3 or 4 pawns, and a couple of minor pieces I can probably see 4 or 5 moves ahead.

If the game just started and we are on move 1, I can only see 2 or 3 moves ahead.

Case in point, a fun puzzle where  club players I showed this to (barely 1600-1700 rated) figured the whole idea out (25+ moves (not ply, sorry) of calculation! Gasp!!!)  in a matter of a few minutes.

White to move. Can he win this?

That's more like 80 ply although I lost count

The king is in the square, so I look only one move ahead. Once the first task is accomplished, I'll look forward two or three moves as I execute the second task. Then perhaps two moves ahead executing the promotion and then mate.

Assuming White makes the only correct first move, Black can resign.

*edit*

On second thought, White has a bit of work to do. I may need to look four or five, maybe even ten moves ahead.

Easy got it up in 1 minute 

The idea here I guess is: First take the a pawn and then go and take the pawn on f3 and then all the way back to a6 and force the pawn on f4 to move and then all the way back again and take it and after f4 the pawn on e4 marches to victory.

Took me 48 moves to checkmate Hiarcs 12: that's 96 ply, but I did not calculate more than two or three moves. Black's choices are nil; White's plan is clear cut. Very little calculation is necessary.

Glad people found that easy ... but the point is ... can you say you really saw 48 moves ahead to a won game for White?

I could argue "yes, indeed!" ... but really?

I hope this gives the OP an example of  how  "depth" can mean very different things :)

In Shivsky's position, you need to see one geometric figure (the square of the pawn) to realize that White is not losing. Then you need a clear plan involving key squares on the g-file and on the a- and b-files involving zugzwang to understand White's win. You also need to understand breakthrough. It is an elementary pawn endgame, but not one requiring calculation.

Yes, I see two dozen or more moves ahead, but I calculate very few.

in here with analysis board you could sort of brute force and play out the game to mate, which isnt possible with OTB chess. with analysis board you can play against yourself till you find the move that you want to use. actually till mate if you take the time.... this isn't possible in OTB unless you have some sort of blind chess ability to play to mate against yourself with virtually all moves.

I think it's impossible to thoroughly calculate more than 5 or 6 ply ahead in complicated positions which you have not seen before.  Even if each player only has two reasonable candidates for their move, calculating 6 ply means looking at 64 possibilities.  Sure, some moves will be basically forced, but other moves have 3 or more reasonable possibilities.  Good players will look much deeper along some of these lines, but they cannot look deeper at all the reasonable lines.