Is it cheating if I use a real chess board during a live game?

Brilliant!

Yes. Touching or moving pieces on the 2nd board before entering a move on the screen is easily seen as outside assistance- (analysis).

A side issue is - what of “touching” pieces on the screen via a mouse, picking them up and hovering over squares and deciding not to move that piece? Returning it to it’s original square and making another move. 
Analysis- yes. Absolutely. OK to practice in Live ? Yes. Why ? Because the practice can not be prevented as chess interfaces are written. A double standard ? Clearly

But the reality is, they would never know!

Engine play, and searching for engine use while logged on (ICC used to do this, they may still) is easy.

Books are a little more dangerous to make accusations of.  Some non-GMs know some lines extremely deep.  For example, I know the following line, and did not consult a book or database to enter it here.

But how will they EVER know if you have a 3-D board out?  They won't!  So you are safe there!  It isn't like you are playing the blindfold world championship!

Chess.com and FIDE have absolutely nothing to do with one another.

With a few differences, chess.com is closest to USCF, not FIDE.  USCF has no rule about writing moves after they are made.  You can write it before or after, but you cannot write a move not played - that is deemed note-taking.  So, in essence, you write a move, you are stuck making that move or you can get a director call against you.

But just to prove Chess.com is nothing like FIDE, in my 3rd or 4th OTB tournament ever, 3rd round, I had White.  Black's flag fell, and I tried to claim win because Black could get his pawn to h2, king to h1, my king to f2 and mate with the Bishop on the long diagonal.

I was White, had a light-squared Bishop against a knight and h-pawn.

In FIDE, that would be a win for White.  In USCF, that's a draw.

Well, guess what bucko?  That's a draw here too!

Perhaps you are correct, and it is difficult to figure out. But if someone plays longer live games and is around 800 rated and suddenly starts calculating on a 1 500+ level, I think it would raise suspicion in the long run.

Perhaps they wouldn't know that he is using a second board to cheat that way, but the algorithm might still conclude that his overall play is not achieved normally.

As for your knowledge about lines, that is a different matter, but I would think that knowing some 20+ move line for some person rated 2000 is entirely possible. But if you suddenly start knowing that amount of opening theory in all of your live games, it would probably become more suspicious in the long run.

But opening knowledge might be more of a moot point and it could be one of the more difficult things for cheat detection to figure out (I am only guessing here however).

On the other hand, using a second board is something that might affect calculation significantly, and I feel that cheat detection would figure it out in the end. I could be mistaken, of course.

I don't see why, since they wouldn't be engine moves and the player would presumably win more games, improve and the new rating would reflect the improvement.

I mean, I had an opponent that I've played twice, and he even had losing positions in both of our games (or at least 1 losing and one worse position, both games he won in the end by the way) and he got banned afterwards. To me he didn't look like a cheater, I never suspected a thing. He seemed to be playing normal moves, I had good positions in both games, he made one not so noticeable tactical blunder, which I missed, and I never felt under pressure in a sense that I often feel when playing against someone who plays perfectly.

They are able to figure out not so obvious cheaters. If that wasn't the case, they wouldn't be able to catch some titled players as well. These titled players know a fair amount about chess, and those caught cheating most likely can cheat in a more refined way than most run of the mill engine users ever could.

Now of course, detection is not perfect, there are some that slip through the cracks here and there.

But that being said, someone rated 800, that starts calculating way better all of the sudden and starts rapidly raising his rating would probably become suspicious. Perhaps they wouldn't figure out what exactly he is doing, but they might think that he is, for instance, turning the engine on from time to time, or something like that.

I agree that it is less straightforward to figure this out, but I think there is a good chance the algorithm would spot it in the long run. Of course, one could only guess.

At what point is it deemed as being too many lines for a 2000 to know over 20 moves deep?

I can think of at least 4 lines I have known beyond 20 moves, and many others into the mid-to-late teens.

The lines I know more than 20 moves deep would be:

The French Winawer line shown a few posts ago.

Certain lines of the King's Indian Mar Del Plata

The 16.Rb1 line of the Botvinnik Semi Slav all the way to the perpetual check by the White Knight

Basically the entire Ivanchuk - Shirov game from 1996 in the Semi-Slav that showed 23...d4 was a huge error.

Now I do not play the semi-slav any more.  Haven't in 20+ years.  But back in 2000 I won a game over the board in the 16.Rb1 line after White played Nf7 instead of going for the perpetual.  Unfortunately, the guy I beat passed in 2014.

Other openings I know 15 to 20 moves deep include:

French MacCutchen w/ 6.Bd2 (main line)

Bronstein Gambit (The line in the Saemisch KID where Black gets BBPP for Q.)

Exchange QGD lines

Closed Sicilian

Etc.

Well, if I specifically avoid lines I do not know (Grunfeld, Dragon, Marshall Gambit, etc), and only play what I already know, I could make it look like I know too much for my rating as my opponent is always the one going out of theory.

As I've said, opening knowledge is a different thing, and it is probably less obvious. What is too much opening knowledge for your level, and enough to get someone like you banned? I admit that I have no idea. I am not sure if cheat detection would be able to differentiate your opening play from an opening play of a super GM. I am more inclined to believe that it would, but I am really not sure.
 
That is an interesting question.

I guess we must be looking at different guidelines then. Staff, including MGleason and Martin have said the use of a second board is ok, or "fine". There is no rule against it. It doesn't violate any fair play policy. It's not oustide assistance, so that's the end of it. There is nothing they can do, other than repeat for you, again, that as a long as you use it for strictly relaying exact game moves only, then there is no problem. 

They no longer encourage OR discourage it, because it's a non issue. It's not for you to decide if someone does or does not have any advantage by relaying moves to a real board. It's up to the player. Just like it's up to the player to decide if he or she wants a white and green board online vs. a brown and tan one. It's player preference, and is not outside assistance. 

If you are convinced that using a real board (in the manner described here) is "highly discouraged" in the guidelines maybe you could point to exactly where that's stated. Because I don't see it anywhere. 

I agree.

It is not too honest though.

Better don't use a real board as supplementing tool.

WOW this has become big

Zhss

No rule exists. Past language of the fair play policies made a few suggestions to avoid the practice in Live play. Now…. the issue is completely avoided by saying nothing - except to state -

• In Live Chess, no outside assistance OF ANY KIND is permitted.

It leaves Staff free to make any interpretations that they deem fit for individual cases. 

Yes, it seems to be a contradiction. In my own opinion, picking up a piece with the mouse and hovering it is not a contradiction, however, because it makes the 2-D experience more intelligible to our minds. It makes it more like a 3-D visual experience. Some people have difficulty with flat 2-D boards and this practice redresses the balance, so it's perfectly admissible.

Picking up a piece via the mouse and hovering over a square obviously helps with analysis and clearly is of outside assistance. 
The player is better able to visualize the position 1/2 ply in advance. He can observe the position as it would stand;  better enabling a visualization of the next possible move by the opponent. How many times have we done this and quickly realized oops - a mistake.
Clearly a slight advantage in analysis over OTB play. Each player can make the practice so no advantage is gained per say in that regard. 
Online the practice can not be stopped - hence it can’t be regulated against. I imagine it would be possible to program the interface differently such that if the mouse clicks on the piece (touch move) it must be moved. However- the practice is so ingrained it will never be changed. I think there is little question it violates the touch move rule that must be followed OTB.

I think it evens up the difference between 2-d and 3-d boards for those having difficulty with 2 d. I don't have difficulty but a lot of of ppl do, therefore it's ok for them to do it.

 First of all the 2d chess board is an advantage over a 3d chessboard.  Its easier to see patterns on.>>

For you. Others find it very difficult. People are not all the same.