Tuirgin

A Scheming Workflow—Swimming the Aquarium

by Tuirgin on May.17, 2009, under Chess

One of my bigger game problems is that I have lacked any consistent method for analyzing and playing my games. I might eyeball a game from my cell phone and make a move if one is obvious—if it isn’t I’ll wait until I can get to a computer. At the computer I’ll work positions out on the analysis board. If I’m really invested in the game I’ll get it into SCID, Aquarium, or Chess Assistant so that I can utilize my databases of games and opening references. Occasionally I’ll make private notes of general ideas, but my analysis of the board is really rather arbitrary and unorganized, and I rarely bother to make notes.

This, of course, is a mistake. I’ve tried a variety of methods to improve my approach to games, but it’s only this weekend that I’ve found a method that really works for me—it involves SCID and Aquarium. In this workflow SCID is optional, but I like the convenience of it. I use the Xfcc feature to play my SchemingMind games directly from SCID which is my daily workhorse chess database. With a game loaded in SCID I can copy the PGN of the game to the clipboard and then paste it as a new game in Aquarium where I’ll do the bulk of my position analysis. (For those of you who have suspicious minds, I hereby affirm upon my future grandchildren’s children’s best friend’s heads that I only do engine analysis of my games after they are finished. While the game is in progress I restrict myself to perfectly legal references—databases and opening books. Cheating isn’t fun and absolutely no one benefits from it.)

SCID is perfectly capable for exploring variations, but Aquarium has a few features to recommend itself for use. I recently began using Jeroen’s New Opening Book for Rybka Aquarium. I’d suggest reading the article just linked for the full details on the opening book. One of the less obvious things you can do with it is to create a new tree configuration which blends the Jeroen_Book configuration with the Annotations configuration. Once done you have both the excellent details of the Jeroen Noomen opening book plus the text commentary provided with Aquarium’s Annotations tree.

Jeroen Opening Book Tree with Annotations

During openings I use Jeroen’s opening book and the annotations to explore possible responses and to try to get a basic understanding of what is happening, what is the logic behind the opening. And throughout the game Aquarium makes it very easy to keep track of the variations I’ve run through and refuted. I find documenting my work on games much easier to work with this way. Once I determine my next move, I return to SCID and make the move, write my comments, and synchronize with the server. If I want to keep private game notes on the SM server, I’ll have to load it in a browser—there’s no way to make private notes with Xfcc as far as I can tell. The benefit of having private notes on the server is that I’ll have access to them no matter what computer I’m at.

After my games are finished, I move them from my Active Games database to my Finished Games database, and depending on how interesting the game was, I’ll either have Aquarium do a fast blunder check or I’ll let it do a full game analysis. Once the analysis is done, I’ll run through the game again, using Aquarium’s other analysis features to explore particularly interesting positions. The one important thing I have yet to add to this workflow is my own analysis prior to the engine analysis. But hopefully now that I’m keeping notes on my games, I’ll find it easier to analyze afterward.

One more thing I’ll mention about the Jeroen opening book—it comes with an i-book (no, it has nothing to do with the cult of Apple), which is an update on enriched documentation features of the past. The i-book provides a classifier, notes on some of the more interesting openings, and a few other odds and ends. Aquarium has a learning curve because it is providing new solutions to problems with rather stale solutions. So far I’m impressed. It is certainly making my life easier and more interesting.

:,

23 Comments for this entry

  • chesstiger
    May 18th, 2009 on 4:33 am

    What i mis here is you setting the position up on a board and analyse the position without moving the pieces, just writting up variations.

    After you have done that you may check the variations you have written down by moving the pieces to see if what you saw thru your mind’s eye (board) is correct.

    Reply
  • Austin Lockwood
    May 18th, 2009 on 5:41 pm

    Hi Christopher,

    I wonder if it’s worth sending an email to Convekta suggesting that they add XFCC support to Aquarium?

    Now that SCID and ChessBase have support, it could be sold to them as the latest ‘must have’ thing ;-)

    Your comment on cheating is interesting, and that’s my greatest concern about promoting this techology. We all know that cheats will cheat whatever; it’s more to do with impressions – telling people that they can view their games in ChessBase often raises an eyebrow. I got ChessBase to disable ‘engine kibitzing’ for SM games, which should answer some of the critics.

    Cheers, Austin

    Reply
    • Tuirgin
      May 18th, 2009 on 8:34 pm

      I’d definitely like the feature, that’s for sure. Aquarium is becoming one of my favorite apps with it’s publishing goodies and general usefulness. It’s DB features are still pretty rudimentary. But as CA is at end-of-life a lot of us are hoping to see ChessOK focus on Aquarium all the more.

      There is a thread over on Chess.com where someone mentions Xfcc support as a desired addition and someone else jumped down his throat, accusing him of wanting to cheat. It makes no sense to me, since it’s quite easy to cheat regardless of Xfcc. Xfcc is a convenience feature. People that are cheaters will go out of their way to cheat. And it doesn’t take much to download a PGN and load it up for an engine to analyze. You don’t even need the PGN, just grab the FEN and analyze the position. Cheating is easy and there’s no way to eliminate it except through building a culture of respect and honor. SchemingMind seems like it is pretty well situated there.

      Maybe a different tack is educating people about the benefits of Xfcc rather than confronting the cheating perceptions head-on.

      Reply
    • Tuirgin
      May 21st, 2009 on 6:29 pm

      After your comment here I went over to the Rybka forums where the developers hang out and posted a request for Xfcc support. One of the developers commented back:

      We decided to study this subject deeper to determine whether to implement this or not. I’ll report the results here.

      The thread to watch is here: http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?tid=11038

      I can’t help but think that the more people that chime in wanting Xfcc the better. Whoever is reading this and interested in Xfcc spreading, specifically to ChessOK/Convekta products, please go leave a comment and let others know to do the same.

      Thanks!

      Reply
  • tommyg
    May 19th, 2009 on 12:04 am

    I think chess.com has the cheating thing well controlled. I have read many threads where the staff are pretty confident they can spot it. And to be honest if there really were that many cheaters I would never win any games!

    Reply
    • Tuirgin
      May 19th, 2009 on 12:11 am

      I’m not too concerned about cheaters. Someone who is cheating while playing at my level needs to cheat better! It’s a legitimate concern at the higher levels of chess. But it’s an entirely different issue from Xfcc.

      I’m more bothered by the disappearing opponents. As I noted in our game, I’ve won 3 (2 different people) games on SM and 6 (5 different people) on Chess.com because my opponent just disappeared. Wins don’t come often for me, and I’d like to enjoy them. There’s no pleasure in winning on time.

      Reply
  • Austin Lockwood
    May 19th, 2009 on 4:46 am

    Hi Christopher,

    Unfortunately the problem of vanishing members is a problem everywhere. For an individual player, the best you can do is to check the records of your opponents before starting any games; personally I don’t mind so much if someone makes a single move and then never comes back because not much has been lost, but to spend days analysing a mid-game, only to have it abandoned is extremely frustrating, I agree. One possible solution would be to restrict the number of games new players can play until they’ve built up a certain level of ‘trust’ – it wouldn’t get rid of the problem completely, but it might help a little.

    Tommy, I’d be very skeptical about anyone who claims that they can reliably detect engine influence in correspondence games; I’ve never met anyone yet who can consistantly distinguish between high level pre and post engine era correspondence games.

    Austin

    Reply
    • Tuirgin
      May 19th, 2009 on 1:26 pm

      How would you even begin to test for cheating? I realize that some have put a great deal of time into detecting “signatures” from some of the more popular engines, but to prove it? That would be difficult and very time consuming. But there will be tell-tale signs. If I suddenly started beating players rated in the 1800s-2000s consistently you’d have to be suspicious of me. Or if I played slop for half a game and then suddenly started playing with mechanical precision. Someone who is letting engines make their decisions is going to be pretty consistent unless they are only using them for specific moves. It’s a shame people get so creative at cheating rather than just working on their game. BUT… I haven’t run into it. For me it’s just a theoretical problem.

      Reply
  • tommyg
    May 19th, 2009 on 1:06 pm

    Austin,

    I could be skeptical or I could look at the results at the website based on the games I and others have played and conclude that either chess.com has it fairly under control or the majority of the people are NOT cheating. Most likely it is both.

    Reply
  • Austin Lockwood
    May 19th, 2009 on 1:32 pm

    Hi Tommy,

    I think it’s true to say that the huge majority of people on most sites don’t cheat; also, the people who do cheat very quicky rise to the top of the ratings, so unless you’re playing at that level yourself, you’re likely to avoid them.

    Austin

    Reply
  • Keoki010
    May 19th, 2009 on 2:27 pm

    Tuirgin could you tell me how you get the annotation text to show at the bottom of the tree display. I have tried clicking on seperate annotation but gives me one line and then chops it. If you want just email me the configuration. Thanks keoki010

    Reply
    • Tuirgin
      May 19th, 2009 on 2:34 pm

      If I’m understanding correctly what you see is a ‘+’ then one line of annotation. The ‘+’ indicates that it’s an entry that can be expanded, so just double click on the ‘+’ and you should see everything.

      I’ve noticed both in the Annotations tree in Aquarium and in the Openings reference from ChessAssistant that some of the annotations seem to be cut off after a certain point. I’ve noticed this several times, where it will indicate that it will show a line and then stops without showing it, or even that the texts ends mid-sentence. I don’t know if I’m missing something, or if the text annotations were taken from another source and were abridged (rather abruptly) if they got too long.

      Let me know if this doesn’t help and I’ll try to help figure out what’s going on.

      Reply
  • Keoki010
    May 19th, 2009 on 3:39 pm

    Yes, that is the answer. Thanks a lot! I made myself some other problems now. Have to try and fix them. Again thanks.

    Reply
  • paradox
    May 26th, 2009 on 11:33 am

    Hi!

    If I may add some comments about Scids functions you may not have noted yet. I’ve no problem with Aquarium (except that it requires several GB of runtime environment ;) but just to point you to some really cool thing you may have missed…

    The tree view of Scid is pretty similar to the one you show in your screen shot also containing similar information. We are a bit more conservative with using space, ie. we only list the Scid-enhanced ECO instead of the full name, however.

    If you want to have textual annotations for your openings you will definitely want to check out the Masks features from the tree windows menu. Masks allow to annotate either the current position by some textual remark or add commentary to a given move. We show only one line of them (again to save space) and you get the full text by hovering with the mouse over the line or the position annotation. Note that you can have several masks and you can fill the masks once and lay it above any of your databases. Ie. you’ll not only have the annotation in the DB where you added it but once you opened the mask you’ll be able to see it in any other base as well.

    Now, masks work by positions and not by opening lines. This means that you have a possibility to handle transpositions seemlessly. To me, Scids masks are one of the most powerful annotation tools within the opening. I may add that besides textual comments you’ve the possibility to mark moves by symbols (two per move are possible), by NAG codes (!, ? etc.) and by colours.

    Also note, that the current cvs version (ie. the next version of Scid that will be released) also offers an alternative view to the masks called “Mask map”. This is a tree view of your current mask and thus also shows you actual lines of play. This tree can also contain NAG and annotation symbols as well as your comments.

    As you mention SchemingMind: you may know the game explorer and its extensive annotations. Scid also offers seemless integration of GE by means of a special SchemingMind plugin I did for the CC features. Using either the menu or the g hotkey will open the GE in your browser, that’s the simple thing. But you’ll also have the G shortcut which will fetch the annotation right from the GE and add it either to your game as game commentary at the current position or as a position comment to your currently opened mask (if you opened any).

    Concerning private game notes on the server: you’re right they are not possible via Xfcc, otherwise you’d have them. Sorry for this inconvenience. Maybe one could suggest it to Martin Bennedikt to add for a future release. Still, I’d then have a logical problem on how to distinguish private and public commentary.

    To make long things short: you should definitely check out the Masks feature and probably contact me in the Club concerning the plugin ;)

    cu Paradox

    Reply
    • Tuirgin
      May 26th, 2009 on 11:55 am

      Thanks for commenting. I certainly haven’t turned my back on SCID. I’m a bit of a software junkie and will continue playing with anything that doesn’t just completely piss me off. SCID definitely hasn’t pissed me off. I can’t say the same for ChessBase the times I’ve been able to play with it.

      I do have the plugin you developed and it is convenient. Thanks.

      I’ve looked a bit at the masks feature, but only a little bit. I’m just starting to learn about the use of trees and position databases in chess software, so it’s a concept I’m not completely comfortable with, yet. Has anyone written up a good introduction to using these features with SCID? I know SCID has a lot of features I haven’t used. Most of it seems to be documented, but documented more for people who already know what they want to do with the features and are familiar with the functions they provide and I’m not one of those. I’m tech savvy, but very new to chess technology.

      Along these lines, the benefit with some of the features I’ve worked with in Aquarium is that I’m able to make use of already-provided annotations. I’m just learning, am not making a lot of my own annotations and commentary, yet, so am using notes provided with the software to learn basics of various openings, etc. As I learn more I may find that SCID gives me greater flexibility than Aquarium.

      At any rate, I remain very impressed and happy with the culture present at both SchemingMind and with the SCID devs. Thanks for all your work!

      Reply
  • paradox
    May 27th, 2009 on 10:36 am

    Hi!

    First of all, I had no doubt about your usage of Scid ;) I just wanted to point you at a lesser known and pretty new feature that is very closely related to your workflow and I just guessed you might have missed it. I think it is a great feature especially if you want to collect information from various sources and combine it in one single spot.

    Concerning Scids documentation: it is in the process of being rewritten entirely together with a tutorial and I have to say, not all parts are as extensive as the CC parts where I really tried to cover everyting in detail (at least in the english version). Help on this project from our users would be greatly appreciated.

    You face a major problems if you want to write decent end user centred documentation. The first one: programmers do not like to write docs. (I admit, I hate it myself and only do it to offer the users some help.) Secondly: usually the people who do the coding write pretty bad docs for an end user. You know, I know the concepts and algorithms of Scids CC part by heart. For me some things are that plain obvious, I just don’t loose a word about them. Its like not seeing the forest sake of all the trees. Still you might not even understand from a users perspective what’s going on there. And the third one is NLS. For a free time project doing some 14 languages is really some kind of efford…

    Combining all this: it would be really great if we could find some users who would help us writing decent docs. I’d even go so far as to offer you support and anwering for your questions even in detail if YOU write the docs for it with your own words from your user centric perspective. (Ok, you’ll get support without writing docs as well, but a lot of others would gain if you do some write up ;)

    If you want to, we could easily start up at Scids tree/Mask features. You’d learn a lot about them on the way maybe Pascal (who did the coding of that part) would get very valuable input for improvement and at the end this would be a really good chapter in the docs. (Yes IMHO english version is sufficient. If someone wants catalan or whatever, well (s)he can translate it :)

    cu paradox

    Reply
  • tommyg
    May 28th, 2009 on 4:46 am

    You are almost making me want to give aquarium another try! I just got so frustrated with it. I think we share the same less then favorable opinion of chessbase products. :)

    I will follow closely but I am thinking that the main reason I was not overly enamored with the Aquarium GUI compared to the Shredder GUI was that (and this sound silly) the Aquarium GUI almost did so much of the work that I felt too passive even though there was so much to learn about the program. The Shredder GUI is much more utilitarian in nature so I was more involved right away. I also remember that that playing games in Aquarium was weird for me as well. And I also think these things are such a personal choice. Different interfaces make different people happy.

    All that being said I am watching you work your way through Aquarium with very curious eyes! :)

    Reply
    • Tuirgin
      June 2nd, 2009 on 6:03 pm

      I saw your comments but didn’t have time to leave a reply at the time. Then life got in the way and I forgot about them… sorry about that.

      Aquarium takes some getting used to, but I like it quite a bit. When analyzing you indicate how many long the variations should be, and how many plies at the end of the variation are untrusted (and therefore dropped). It also focuses on the part of the game where moves would make a difference. So, if you get 2/3rds into the game and it’s already significantly in one side’s favor, you aren’t going to see a lot of analysis of the endgame.

      It has a very convenient online 6 man tablebase. It’s pretty quick. If you want your engines to use tablebases then you have to have your own install, though.

      And analysis is stored in trees, I think, so that given the same position in another game you can access the analysis for it. I think it does this — don’t quote me on it.

      I still don’t play against the computer much. No matter the handicap I get clobbered. Playing “real-time” I just don’t take the time to analyze well.

      Reply
  • tommyg
    May 28th, 2009 on 4:48 am

    Oh and the funny thing is after getting beat up by my chess coach in our last lesson I am actually getting back to old school. He wants me to have less time at the screen and much more time at the board itself. And he isn’t old school, he is only 20 years old!!

    Reply

Leave a Reply

Looking for something?

Use the form below to search the site:

Still not finding what you're looking for? Drop a comment on a post or contact us so we can take care of it!

Visit our friends!

A few highly recommended friends...

Archives

All entries, chronologically...