Questions for 2000+ rated Daily/Correspondence players

1. Do you use any openings databases besides or in addition to the chess.com masters database?
2. Do you have any opening books or courses that you feel have helped significantly?
3. How many of your games are won and lost in the endgame?
4. On average how long do you spend analyzing a position before making a move?
5. What is the longest and shortest amount of time you have spent on a move?
6. How many candidate moves do you consider?
7. How many moves ahead do you analyze before evaluating the candidate move?
8. How many opponent responses do you typically consider for each candidate move?
9. Do you analyze weird looking moves in case there is a hidden tactic a few moves ahead?
10. If there are two opening moves in the database with conflicting data, e.g. one has a higher winrate but also a higher loserate, or one has a slightly higher winrate but is played less often, what are the things that you consider in trying to assess which move is best
11. When playing a lower rated opponent, do you try to play very solid even if that means settling for a draw, or do you play aggressive to go all out for the win?
12. Any other tips on becoming a strong correspondence player?

Thanks in advance for your words of wisdom.

@1
1) Yes
2) Yes
3) Most
4) As long as allowed
5) Days
6) 3
7) depends on the position
8) 3
9) Yes
10) Analyse
11) Always avoid losing
12) "In correspondence he who goes to bed later wins"
- Fritz Baumbach, 11th ICCF World Champion

1. Yes

2.  Yes

3. Depands

4. Depands of many things

5.  Same

9. Yes

11. If player dont see mistake move I m playing with more mistake because looking for his more deep mistake and making good moves after.  If is solid player than game is pretty solid.  Waiting for mistake 

1. Which one(s)?
2. Which books/courses helped you most?
7. In a very complex middlegame how many moves ahead would you look, and how do you copy with the combinatorial explosion of having about 3 possible responses to every move each player makes?
10. Could you elaborate?

@4
1 Everything available
2 None in particular
7 It depends on the position and on the available time. At 50 days per 10 moves it is possible to spend 40 days on move 10 or move 11. The first moves may go at 1 day/move.
10 If there are 2 possible replies, then analyse deeply so as to find the truth.
Winrate or loserate, or played more or less say nothing. The move played less often and with the highest loserate and the lowest winrate may well be the best move. The truth comes from analysis.

When I played it consistently:

1) I used the Mega Database and the Correspondence Database.

2) I used the books and courses I based my OTB repertoire around.  The idea being to help hone how to play the resulting middlegames iny OTB events.  If your goal is to be a better correspondence player, you may want to go a different route.

3) That I cannot recall.  In ICCF, you can claim a table base win/draw when you reach a certain point (I do not recall if chess.com allows you to use the table bases in their daily games ... Would need to check on that).

4) until I was satisfied with my selection.

5) 5 min and 3 days (I did not play anything longer than 3 days/move).

6) depends on the position

7) depends on the position

8) depends on the position

9) depends on the position

10) with my repertoire, I usually have 2 ways to play in the opening.  For example, against the Caro-Kann, I will play the exchange or the advance.  I pick which based on what I feel like playing.

11) I play solid and look for a mistake.  I found that many lower rated players do not take advantage of the opening books and will often make a serious mistake in the opening.

12) identify what you want to use correspondence for.  If you want to use it to improve your OTB play, then stick close to your normal repertoire and see how best to play the middlegame positions.  If you just want to get better at correspondence itself, find ways to keep it interesting.

I don't play daily here, at least not seriously, but I have a few times (on past accounts) and just wanted to say...

In terms of candidate moves, I think what matters a lot more is having many candidates for your opponent. Once you've done some analysis, and you're reasonably satisfied with a line, you should go back over it, 1 move at a time, and on each move try to improve your opponent's play. Doing this over and over and over (so, similar to @tygxc 's #12 I suppose) is what leads to strong play, IMO.

Of course you should have multiple candidate moves yourself, but I feel like that's a given.

And of course some positions don't require long analysis... and the first half the of the game can be more research (looking up games) but anyway...

@7
"the first half the of the game can be more research (looking up games)"
++ Correspondence games can be lost or won in the opening.
If you follow some path that leads to trouble, you will suffer for months with a bad position.
Over the board you can get out of it, but in correspondence there is no mercy.

Questions one and two are covered in some detail at http://chessskill.blogspot.com/search/label/correspondence

I create a record of each game in a database called "In progress". I analyze there. Within ChessBase, I have ChessBase Mega 2020, PowerBook 2020, all Chess Informants, ChessBase Opening Encyclopedia, ICCF database, and others. The articles at the link offer some clues as to how I use these. Answers to other questions can be discerned, too.

I have not gathered data, but I would have expected more games to reach the endgame than the number that have.

I've moved instantly (sometimes to my regret), but 2 hours per move is typical with some moves getting a full weekend. I recall one game in the French that sent me to Chess Informant and I cranked through fairly quickly every game ever published there that had reached our position (about 120 games). I looked at the whole games. Even after we departed from book, some games that I examined that weekend guided my play. I won the game.

For this one I wanted to say... 

If you reach a big shift in ideas (like going into an endgame) without forcing it, then that's a good place to stop. Another way to say it is... if you reach an endgame via forcing moves, then ok, the strength of those forcing moves may depend on whether you can win or draw that endgame... but if you reach it via non-forcing moves, then you should probably do many more runs though the analysis that lead to that point, because the strength of those moves will depend on the improvements you can find there, not on the endgame.

Big shifts in ideas tend to happen when the pawn structure changes and to a lesser extent whenever pieces are traded.

If you're very satisfied with the analysis that led up to a certain point, then sure, you can keep going, but as you might imagine, it's hard to be extremely confident of a 30 move long analysis

1. Mainly openingtree.com and my own opening repertoire

2. For the Caro-Kann I used Gotham Chess course, which was great when I just started playing CK. Not so much for more experienced players

3. Going by Insights (Daily games since 1 Jan '22): 45% of my games are decided in the endgame (+ 36  -18 =10)

4 and 5. Impossible to say. My guess would be < 1 minutes for familiar positions, around 2-5 minutes for most moves, 15-20 minutes for challenging positions. In exceptional cases 45-60 minutes.

6. Varies. Usually 3-4. But there are some situations where I have only one move i consider and just do calculation to check if anything is wrong with the move. In most cases I'm comparing 3 or 4 candidate moves.

7.  This is less straight forward, more of an iterative process. I first do a quick scan for each move, playing the most obvious moves or until I find move for my opponent that creates a bad position for me. The I start exploring the best line to see if I missed a way for my opponent to punish me and for the other lines I try to find a way to improve on the current line.

This is assuming I have time and energy to do so and the position is relevant enough. otherwise I'll just trust my instinct. Which goes for all my answers, time constraints ofte prevent me from doing a full analysis.

8. As many as I need to find a response i don't (really) like or until I'm fairly sure I'm good.

9 No. Not sure what you mean by weird move. I see no reason to check out a weird move. I do sometimes check out obvious sacrifices they can play

10. With conflicting lines between databases or within one database two moves that look equally good, I just play out the top move for each 4 moves in a row and decide whether I like the position or not

11. Against lower rated players I play more aggressive, assuming the pressure will cause them to make mistakes. 

12. I have nothing

How about you, @KevinOSh ?  What are your answers?

No chess.com does not allow tablebases or engine use.

I had a quick look at corr database and there is a quote from Ginger GM saying nowadays you can use engines in correspondence games. Definitely not the case on chess.com. I don't think I understand the point in playing chess when both players can use engines. To me it seems like one engine playing against another, not a real competition between humans.

I joined ICCF to get access to their databases. I played two games, both against the same opponent. The win was very satisfying. Assuming my opponent, like I, was employing an engine, I labored hard to create an unbalanced position that seemed equal for many moves, but that seemed to offer winning prospects in the late-middle game or endgame. 

True cyborg chess is something more than merely engine vs. engine. Playing this way, however, was even more time consuming than regular correspondence chess. I see how it can be appealing, but I don't want to play this way on a regular basis.

Here's my win with annotations I created during the game:

Engine use I remember.  I thought they had added the ability to use table bases a while back, but I could be mistaken (it has been a while since I played daily games with anyone but friends).

Regarding centaur chess:  it is not just "engine vs engine".  In fact, if you try to just play the top engine moves in an ICCF game, you are likely to get crushed.  The whole purpose is to drive the engine to find the strongest possible path.  Many times, the horizon effect leaves a position that looks reasonable early on in a completely lost position later.  If I recall, the USCF correspondence chess does not allow engines either (not sure how they check that), but they do allow opening books/databases and table bases.

Yes, USCF and chessdotcom forbid engines, including tablebases. Naturally, there is always the fear that your opponent might not be observing these rules.

Detection can be difficult, but there are ample resources. I got many of my suspicious opponents banned during play. Enough that I suspect my own rating became inflated.

I played correspondence continuously from 1996 (postcard) to 2018. Postcard gave way to email early this century and I started playing on websites in 2003. There was overlap when I played both postcard and email, and again when I played both email and websites.

The ICCF games were played on their website.

I rarely play correspondence these days. I put a lot of effort and time into it when I do and I've been losing interest, partly because I'm weary of my constant suspicions.

Firstly, we are both well under 2000 rating so no chess masters but of course you are correct in thinking that our answers still have value. So thanks for taking the time to reply and here are mine:

1. Yes
2. I am a big fan of Marc Esserman's book
3. I think it is a minority of my games, but this is due to playing a lot of lower rated players who lose an exchange or allow a fork etc.
4. I guess somewhere between 5 and 10 minutes on average but it varies a lot depending on the complexity of the position
5. Shortest is a few seconds. Longest is probably a little over an hour. I make a gut decision if I evaluate two moves to be about as good as each other and additional analysis is unclear. However after the game the engine usually tells me that the two moves are not even nearly as good as each other, so that's a weakness that I have.
6. Usually two or three. Sometimes its just a simple recapture which i forced so just one. Occasionally 4, 5 or 6 moves all seem logical, and in those positions it is so much harder to decide
7. Sometimes just a couple of moves. But if it is unclear I sometimes analyze up to 10 full moves ahead. But this can lead to getting lost in a dozens and dozens of lengthy variations, so there comes a point when its necessary to stop and make a call.
8. At least two. The more the better. I am still occasionally surprised by a move my opponent makes, but much less often than when I was like 1000.
9. No, but I probably should
10. I usually end up playing the move that best fits my style of play and my understanding of general chess principles. If there are two moves and one has slightly a higher winrate but I don't understand the move at all, its probably a bad sign.
11. I play pretty similarly to how I play against stronger players. With opening databases available its often impossible to get a significant advantage in the opening but as long as it goes into the middlegame about even then there are plenty of chances to find an advantage later on in the game
12. If only I knew, I would be a better player

@12

"I don't think I understand the point in playing chess when both players can use engines. To me it seems like one engine playing against another, not a real competition between humans."
++ It is a real competition between humans, using all available resources.
"6. In ICCF event games, players must decide their own moves. Players are permitted
to consult prior to those decisions with any publicly available source of information
including chess engines (computer programs), books, DVDs, game archive databases,
endgame tablebases, etc. TEAM: In addition, acceptable behavior includes
consultation between/among players of a team including a team captain about positions
in active games in their team event. No other consultation with another person
concerning analysis of an active position is allowed in either a team or individual event. "
https://webfiles.iccf.com/rules/2022/ICCF%20Rules%20update%20for%201-1-2022%20-%20correction%201-25-22.pdf 
The point is, that it is impossible to impose and patrol a prohibition.

To illustrate what I mean by "weird looking moves" an example is move 17 in this game:

Hah, that reminds me of a tactic I found because I was in a difficult position and it seemed like none of my moves worked, so half in a daze, I just started trying almost random legal moves... and I accidentally found a knight sac that looked pretty interesting. It required me to sac a few more pieces, but surprisingly all the lines worked out for me.

It wasn't as fancy as the queen sac in that game, but it reminds me of it because at first it looks completely stupid... but each move carries a big threat. In that position move 17 didn't but each move after that threatens mate.

1. Just the chess.com database.

2. The only opening book I read many years ago was I think 'Play the French', but it might have been a different one I can't remember. The rest of my limited repetoir came from my coach years ago.

3. In my last number of games I seemed to have a winning or much better position before the endgame. Unless I was playing against people who would get banned in the middle of the game.

4. I would often spend many hours over a move.

5. I don't know exactly. Longest would be many hours over a few days. Shortest would be instantly for a forced move.

6. Depends.

7. As long as it takes until the position can be evaluated (there are no more tactics).

8. Depends.

9. Yes absolutely. In fact I made sure I would have at least a cursory glance at every single possible move I could make (on the first move of calculations).

10. I didn't pay much attention to winrate. I wanted to find the move that would lead to positions I was comfortable with. Sometimes that would be a move that's not played often, and sometimes a move not in the database at all. Also it's better to play a move you understand that's not in the database than one that's most popular that goes over your head.

11. My openings are such that I generally play the same responses every time. They will often lead to drawish positions but that's all I'm working with. Outside of openings daily is such that playing aggressively or not didn't really come into it. There's the time to try and find the most accurate moves objectively. I only play blitz nowadays and in that I find that if I can just focus on keeping it solid and not forcing things lower rated players will make a serious error at some point. Sometimes deep into the endgame, but the error will generally come in the end.

12. Only play a few games at a time. Maybe join a good votechess club where strong players are there. Learn how to think and analyze. You'd be surprised how many good players and better blitz players than me don't know how to think. They just play moves which look good and have good instincts. Analyzing if you don't know what you're doing means looking at the features of a position (Good Knights vs bad Bishops etc.) and drawing up plans for both you and your opponent based on them, it will also take into account calculations of course. Maybe Jeremy Silman's books are good for learning to analyze, but I know a lot of people don't like them.