R database players scared of playing real chess?

Okay. I appreciate your effort.

1. Does the database limit my candidate moves? It might for a 1300 player, although I doubt it. It certainly does not for me. Quite the opposite. In any given opening position, there are normally no more than three moves that I will consider when playing OTB. Using databases during correspondence play (make no mistake, "Online Chess" on Chess.com is correspondence), opens my thinking to more possibilities. Databases increase the number of my candidate moves.

2. Absolutely. Considering my opponent's responses is part of every phase of the chess game. This process is just as true while playing in a bar while drinking, playing in an OTB tournament, or playing online at any time control. 

Consider a couple of specific cases:

a. In a typical correspondence game on this site, I will use the Game Explorer to check the candidate moves and the scoring percentage. I assume that my opponent is doing the same. The research begins while exploring several moves deep. Looking through high scoring moves, there is often a move that initially looks good, but there is one less popular move that changes the percentage dramatically. Why? Sometimes that's because an old line has been refuted. Part of the game of correspondence is found in the effort to anticipatehow one's opponent is using the Explorer, and use it deeper to steer the game into a line that is more favorable for my side.

Once I get an advantage, there is still much play. At some point, every game will leave the databases. Having used the database to steer the game to a favorable position, I also have given myself a position that is not typical of those I usually play. Having worked harder in the opening because I deviated from my standard repertoire, I now must work harder in an unfamiliar middle game.

b. I offer a game that I played via postcard in 1997-1998. I had a growing collection of books, but no databases. My books included MCO and two volumes of ECO. I also bought one issue of Chess Informant. I was playing the English Opening and Reti as White in most of my OTB games and I had an opening monograph on the opening in addition to ECO, volume A and MCO. I played through several of the English Opening and Reti games inChess Informant 64 while waiting for my opponent's moves to arrive in the mail.

This game caught my interest:

Ribli,Z - Sherzer,A [A12]
Magyarorszag 64/7, 1995

1.Nf3 Nf6 2.g3 d5 3.Bg2 Bf5 4.d3 e6 5.0–0 Be7 6.c4 c6 7.b3 0–0 8.Bb2 Nbd7 9.Nbd2 a5 10.a3 h6 11.Qc2 Bh7 12.Bc3 b5 13.cxb5 cxb5 14.Qb2 Qb6 15.b4 axb4 16.axb4 Rfc8 17.Rxa8 Rxa8 18.Nb3 Bd6 19.Ra1 Re8 20.Bd4 Qb8 21.Bc5 e5 22.Ra6 Nxc5 23.bxc5 Bc7 24.d4 e4 25.Ne5 Bxe5 26.dxe5 Nd7 27.Qd4 Nxe5 28.Qxd5 Nc4 29.Bh3 e3 30.f3 Rd8 31.Bd7 Qc7 32.Ra8 Bc2 33.Bf5 1–0

In particular, White's move 14 suggested a strategic idea that was new for me: using the queen to support rook action on the a- and b-files, and using her to support action of the bishop aloong the long diagonal. My game never reached this position. Nonetheless, the growth of my underdtanding of the strategic ideas helped guide me to victory in this game.

I discuss this game in more detail at http://chessskill.blogspot.com/2011/03/playing-by-book.html. My opponent was a well-known chess promoter (he founded New York's Chess in the Schools program). I do have quite a few blog posts concerning database use. Feel free to poke around and read some that I have not linked to here.

3. Of course, using an engine is good preparation for OTB. I use engines for training. However, rules do matter. Using engines to analyze games in progress is against the rules on this site. I do not do it.

Using engines to prepare a repertoire, however, is not only useful, it is something that I find essential. Again, I find a specific instance useful to illustrate. In 2008, through a few strokes of luck I qualified to play in my city's championship--a two player match of four games. I was a strong B Class USCF player at the time (~1740). My opponent, the reigning City Champion was a FIDE Master, but no longer at his peak rating. He had been in the mid-2400s in his youth, but had dropped to 2250.

As luck would have it, he had a fairly limited opening repertoire. He played the Advance Variation against the French. He played the Dutch Defense as Black against queen's pawn openings. I learned a new line against the Advance variation, and I spent a lot of time working on the Raphael variation against the Dutch. In my preparation (I had about three weeks), I managed to draw HIARCS 10 in a fifteen minute game. In the match, I lost the first game badly. It was played on Thursday night. Games two and three were on Saturday. A friend of mine (an expert) and I spend all morning Friday using engines, Chess Informants, and database refining my line against the Dutch. 

I had Black in game two, and played a long game that was almost drawn. I earned the FIDE master's respect. I earned a draw in game three. There was no game four, because my opponent had the 2 1/2 points needed to win the match.

The games are posted at  http://www.spokanechessclub.org/gamecorner.htm.

On some sites, and in many correspondence associations (ICCF for instance) engines are allowed during correspondence play. I have played this sort of chess, sometimes called "advanced chess". I played on another site that encourages it. Yesterday, at the OTB tournament I am playing in this weekend, I was discussing this form of chess with friend who played some ICCF events. We both quit because of the added effort needed. He said it was like having a second job.

It is possible to play engine assisted correspondence chess (on Chess.com, these games must be unrated and by mutual agreement). For me, the level of database research needed to play well in correspondence chess on this site makes the game more satisfying. Engine assisted correspondence chess is less appealing. Nonetheless, that too, could help one's OTB game if done well.

I usually have ten to twenty games going at once. I play here, on another site, and am currently using two iOS apps, and Chessimo on Facebook. The last of these does not allow database use, and I don't use databases in those games. In my games where databases are allowed, I use them selectively. I do not reference them at all in 1/3 - 1/2 of my games, use them sparingly in a few, and do deep research in a few.

In my highest rated win on this site, my opponent walked into a known trap that I had studied for less than two hours in Starting Out: The King's Indian two years before the game. I then had to convert the advantage of an exchange, which I did by steering the game rapidly to an ending where I could return the exchange to go into a favorable pawn ending. I did not use databases in that game. My memory served me well.