martes, 17 de octubre de 2017

Review: Chess software

I have been wanting to make a review on the software I use when I study chess but never got around it.

As I reviewed John Nunn's Chess Puzzle Book I realized that being able to talk about this chess software is quite useful for the reviews, so I decided to set aside laziness and go for it!

In this review I will single out what I do in chess and which software best suits my needs. However I would like to highlight two things before we begin:

  1. I have a open source operating system in my desktop (so, not windows, not apple). Many software won't work on it even if I were willing to use it.
  2. Given the  choice, I try to go for the open-source solutions. I am kind of stubborn and follow that principle quite a lot.

Playing on-line:

I play on-line in a server called Free Internet Chess Server. As the name states it is free and it is enough for me. It is like an ICC from ten years ago (maybe ICC still has the same interface, I wouldn't know) and it does not take me long to find games there. My rating fluctuates between 1900 and 2000 generally (I am fide 2200).

I'm sure there are better sites there, but this is the one I know and I'm able to use given my software requirements.

To play on it I use Xboard (Winboard for Windows users). On my current Operating System (Linux Mint) I only had to download Xboard from the repositories and issue the following command on a terminal:

xboard -ics -icshost

Xboard may not be the best software ever, but to play on-line you do not need much. A board and a way to seek opponents. Xboard gives me that and makes it easy for me to play. Windows users may not be able to relate, but being able to set things easy is a big plus.

Database handling:

My main reason to handle game databases is to manage my opening repertoire. I like to put the lines in the books on a database so it is easier to maintain and improve. There are two applications there that do what I need:

  • SCID: This application is for the desktop computer and you may think of it as the open source version of Chessbase. Maybe it is uglier but I can do whatever I need to do with my files: I can create variations, comment on the game, etc. It also has a mode that I love called 'Opening Training' where it will make the moves for one side randomly from the opening file you have chosen and wait until you insert the move of the other side. This is a great way to study openings. It is very similar to what you can do in but it will use your files!
  • Scid on the go: This application is for android and is a derivative version of SCID. This android version is able to read databases, search positions and names and create games. I use it to introduce the moves from the book while on the bed before I go to sleep. Afterwards I put them in the computer through dropbox. It has the option to hide the notation of the game so you can train the openings there too (but it won't pick the move at random this time, but the main move of the game you are at). It is not as great as in SCID but it does the trick when you are studying in the subway. There are two things I do not like about this app: (1) it does not allow you to introduce comments on the notation and (2) when the game has a lot of variations and comments it turns slow and somewhat unresponsive when 'clicking' on a move. Furthermore, it won't allow you to change the settings of the engine.
Analyze games:

This is a different section because in the normal world there are both Chessbase and Fritz. As I do not play against the computer the only reason for me to use Fritz (or Houdini or whatever) would be to analyze a game. My engine of preference is of course Stockfish which fits the bill perfectly as it is not only free but open source, and it it the best engine overall (including commercial engines) by elo (or very close to it that you would not see the difference anyway), and I use it in the following apps.
  • SCID: On the desktop computer I use SCID for analyzing too. It allows me to do whatever I want as noted earlier. 
  • Droidfish: On the cellphone I use Droidfish. I just recently found out about its hidden potential (previously I only thought of it as an interface to play the computer). It allows you to add comments (which is very important for me as i not only introduce the lines from opening books but also create tactical and positional problems from them) and you can change the engine settings. I use that to check the positional exercises I create from an opening. An exercise is good if the first line of the engine is at least +0.2 better than the second one, so when I find a position which seems interesting (generally the ones in diagrams or before a ! move) I load the engine showing two lines and there I know if the position is a good exercise. Sadly Droidfish does not seem to be able to manage databases, so once I have finished analyzing the game/exercise and I have added all the comments, I share the game with Scid on the go where I am able to store it in the correct database. Being able to do it swiftly is one of the great discoveries I had recently.
Train tactics:

I train tactics in Chess Tempo. The site is free to train but some functionality is only available for paying members.  I do not pay so I have no idea if the paying section is worth it, but the free version is great. You can solve tactics slowly or in a mode called blitz (which takes into account the time you use to solve it). There are like 100,000 exercises to solve with a wide variety of different levels. The exercises you are presented with depend on your rating so you won't get something out of your league (too easy or too difficult).

I have been using the site for two month now and I have slowly climbed to 2130 elo in the standard mode. At this level I have to spend like 20 minutes per exercise to get them right. That means that those exercises I'm solving are harder on average than John Nunn's Chess Puzzle Book. 

There are other sites available (, and I guess they would be good as well. I'm trying to spend 30 minutes each day solving tactics on-line (which sadly now is like two exercises, as I take sooo long to solve them) and my hypothesis is that it should raise my tactical awareness, specially when done without time taking into account. The blitz mode may be more fun but you do not invest as much in those problems and the improvement comes from the energy invested.

In particular ChessTempo (I cannot say for other sites) will show you exercises where only one move wins (a computer evaluation bigger than 1.75). The exercises are different from normal tactical books as they are not chosen to fulfill any criteria on beauty but just to have one good move available. It takes time to get used to it but all the positions are from real games, so you will be very close to real game scenarios (except for the fact that you know there is only one good move and that good move is winning).

To me this site has make puzzle books more or less obsolete . Note that this does not mean I won't be reading puzzle books in the future or I'm giving away the ones I own. I will continue to use puzzle books and I will buy more, but that's because I like the feel of studying with a book, writing in pencil what you have thought so you  can come back a year later and see if you have improved, and specially that if the author is good he gives you something special in the solution or in the selection of exercises.

But when you are studying tactics for tactics' sake, then you may as well go to one of this sites, you do not need anything else, as those are places for the day to day training. The site will bring you puzzles at your level and your rating will be updated after each puzzle, so you will see if you are improving or not.

Chess diags is a little app for android for solving mates. It starts with mates in 1 and goes all the way to mates in 10. There are like 2,000 mates and the potential for anyone to create their own set.

The difficulty is varied, there are a lot of trivial mates, but others really forced me to set the position in an engine.

As far as I can see, the only problem with this app are: (1) It is a data dump of a lot of mates, so some are good, some are bad, it really does not care if you are a queen up, you just have to mate; (2) it does not have an engine that might help you when you are stuck.

SCID on the go as training tactic is feasible if you have a database of tactical problems as you use the training mode in the same way as you did for your openings. You make a move and if it is the correct one SCID will answer it with the main line of the variation until there are no more moves.

Reading books:

To read books I use Forward Chess. It is an app for android and iOS that a lot of publishers use to sell their electronic books.

The application is not bad (although there are tons of things I would it to do that it does not but SCID on the Go does) and you do not really have a choice there, if you want an electronic book from one of its publishers you have to have this app (which is free).

I like it mainly for opening books so I can study them in the subway but I have bought normal books there too (Chess Structures for example) and it is useful to have the engine running at your fingertips.

Importing a diagram to your phone:

The problems with physical puzzle books is that often you are left wondering why your move did not work. Recently I found an app called ChessOcr which will let you read the diagram with the camera of the phone and send it to the app of your choice for studying that diagram (I send them to Droidfish).

The app is great and has saved me tons of time by not forcing me to set-up the board on Droidfish, but you will have to pay for it :( But if you are like me and wonder why your intended move is bad and find yourself setting up the position by hand in an engine a lot, maybe you could consider buying it!

3 comentarios:

  1. The rating at Tactics Training on chess tempo might be overestimated. I start yesterday, after reading your review,and I am already 100 points over my FIDE rating

  2. What you report is very odd. I got two friends to use the site. One has 2000 elo and she is hovering around 1850 in CT while the other is 2150 and he has recently crossed the 2000 mark (with a clear upward trend). Both are from the tactical type of player. I myself am at 2200 elo and it is taking me forever to raise it (I have around 2160 now on CT).

    Although I see myself crossing the 2200 barrier on CT soon (usually my performance is higher) I may be the atypical as I spend a lot more time than average on the problems (a lot more than on a real game) so it is only logical that I inflate this rating.

    Anyway there are some reasons you may find your rating on CT higher than on FIDE:

    1. You are specially great at tactics and FIDE ratings measure your overall performance.
    2. You live on a place where fide ratings are low (it happens in Spain that some regions have lower elos than others for the same strength, so it is only natural to assume this happens between countries even more).
    3. Your elo is too high or two low compared to ours and in that range the elo of the problems is different for some reason (so say a 2200 problem is indeed 2200 FIDE, but a 1600 problem is instead 1500 FIDE).

    Whatever the reason this should not be of great concern as the important thing about this elo is not the absolute value but the variation. It helps you see if you are improving or not. It helps you too if you have some friends as you can have a healthy rivalry to improve.

  3. Thanks for answering. The 3 reasons you mentioned seems logical. Point 3 looks possible, since I have lower elo, but i.e I solved problem 1044322 in seconds, and if I remembered correctly its rating is about 1650+. This puzzle might have one asterisk in one of the Yusupov's orange books.
    Do you (or your friends) have tried It would be interesting to know if your rating is similar there.
    Keep up the good work on your reviews!