New Chess Improvement Plans

After my less than satisfactory tournament results at the British this week I've decided I need to restock and come up with a coherent and efficient training plan.

If anyone cares to give some imput that will be great.

1) My first idea is JUST to concentrate on ICS alone now as far as studying is concerned. I'm the kind of person who loves to dip in and out of books, starting one, then dropping it for another. Since my harddrive crash and the loss of the first 5 chapters of Yusupov's Build Up Your Chess I've lost heart in studying that again as I feel compelled to redo the 5 chapters I've already done so I have a complete database and I really don't want to, so for now, I'm going to leave them alone completely.

So my chess study is ICS, ICS and more ICS, so hopefully I can actually make some progress and march beyond the first month.

I will start the course again, from the beginning as soon as I've finished this tournament and got essential non chess chores out of the way.

2) I think the next crucial step is analysing my own games in detail. At the moment I just run them through Rybka and spend 10 minutes or so seeing where I went wrong, and then store them in a database. I think I need to do much more than this, annotate them thoroghly with my own thoughts etc and highlight the key parts of the game, what I did right, and where I could have improved. I think this is a key component.

3) Opening Study, I'm actually thinking of wholeheartedly adopting the ICS reportoire modules as well, starting with learning the basics of the 1.d4 reportoire, and then branching into the Scandanavian and Dutch defenses, leaving my 1.....e5 and against 1.d4 the Queens Gambit Declined Orthadox variation until I've finished the course and want to start studying independently. I think playing the ICS openings will attach me further to the course, reinforcing my commitment to it and my investment in it. At over £300 for both modules, it's not a cheap course so I want to get the maximum value from it.

4) NO MORE INTERNET CHESS!! From now on I'm banned from playing online, I simply don't see anything productive in it. Rather than waste hours each week playing pointless 15 minute games I am instead going to join a second chess club in my area, who play on a different day to my own club and play in their internal competitions for graded games with proper time controls. The best way to improve I think is playing REAL chess, over the board, for grading, under league and tournament conditions. That way I will improve in terms of chess practical experience, psychology, getting used to playing pressure games etc.

5) Finally I need to focus on my physical health. As a former cancer patient with chronic lung damage I've noticed that as I've neglected myself the past couple weeks I've felt really wheezy and out of condition. I've even been sweating and panting just from walking around etc. I'm not overweight (well flabby but only 147 lbs) but because of my lungs I'm really feeling drained now after just 7 straight days of tournament chess. I've been unable to work for the past decade since a bone marrow transplant so this week by my standards has been quite tiring. I feel I really need to start eating more healthy and getting outside in the fresh air, lots of walks, and gentle exercise as physically I've felt a wreck this week and I'm sure that doesn't help my play (to say nothing of my quality of life and life expectancy!)

So anyway, that's my plan in a nutshell. All my non ICS books will now be put back on the shelves in the lounge and from now on my training is ICS, ICS and ICS, along with time spent analysing my own games, according to the ICS method of course.

I'm actually excited by the prospect!

You should put 5) in first position.

Getting healthy and well rested with good sleep, worth by themselves 100 or 200 ELO points, if not more !

For the other points, it's your business, everyone is different at studying, if there was a miracle recipe, we would all try it...

Personnaly i decided not to subscribe to opening course, will see.

I like your ideas. ICS has enough material to be worthy of someone's complete focus - especially if you have a BCF anywhere under 175. Stay healthly and good luck.

Cheers guys, HangingKing I definitely agree I need to work on my health more. My physical condition is pretty poor at the moment and I really felt it during this tournament.

Getting physically fitter is a must.

Bobbyblunder I agree that when properly studied ICS is definitely more than enough to focus on.

I've also sat down and properly analysed one of my games for the first time today. My analysis previously consisted of running it through Rybka, playing through the variations for 15 minutes and then archiving it in a nice database.

Tonight I spent over 3 hours analysing a single game putting in all of my thoughts and paying special attention to where Rybka recommended alternative moves, trying to really see why.

I gained a lot from this process, in the particular game, my first tournament game where I hung my queen to lose the temptation was to ignore it as obviously I knew why I lost.

But I learnt some great things. I saw a couple of ways to really improve my position at various places in the game and punish him positionally for his mistakes. I learned I placed a piece inaccurately at the beginning of the game and how it affected my position thereafter. I saw how I missed opportunities to seize control of the centre, and also how I had the chance to simplify and exchange into a winning endgame.

I also spent an hour going over the correct opening moves for the game and feel like it was a very rewarding night's work,

I'm going to try and spend another few hours tomorrow night analysing the second tournament game even though I know I played poorly.

In doing this I've actually come to realise that game analysis itself is long and hard work, equal to studying any book, and ongoing so I'm definitely just sticking to ICS and own game analysis for now.

That's plenty of work...............

I agree pick something and stick with it.

I never put a game into Rybka until I have first worked through it on my own.

I don't think the internet games are bad, only blitz games. I have had some fantastic slow (30 30 or slower) games on the internet. It is when I play fast on the internet that my game suffers.

For me ICS is a supplement to playing slow games and going over them first on my own and then with a coach (and usually Fritz but really computer dependence is very very bad for the improving player and too many improvers use it as a crutch in my opinion as evidenced by the people posting 'Fritz gives it a +0.2' and such at ICC during games).

I'm still on my chess break, though, and don't care much about the game right now.

They are not supposed to use an engine during games ?!

Or maybe you talk about retransmitted GM events on chessfm, you can follow on a board at ICC.
My opinion is these people you talk about, are not improvers, they just want to be bright in society and tell you which is the best move (i.e. the one given by the computer), most of them don't even understand the move they propose.

Anyway for having analysed games through engines, a +0.2 at my level means nothing ! I even analysed GM games, and it means nothing, it is too few, at GM level maybe 0.5 start meaning something, and at my level at least 0.9 is the beginning.

What's sure, is when you make a mistake (or you opponent makes one) then the score change violently, this is what is important, detecting mistakes.
And once you found the mistake you can search (without the computer) a better move, and if there is none, look backward where you started diverging from an healthy game.

I agree with you, blitz is only for fun when you have some time to kill within the day but not sure you can stay. However, slow games on internet are good, but i wonder if people are as serious as they are in face to face play.
According to what Bilbo says, if found the tournament more difficult than online play, despite the fact he was among people of the same grade.

I don't know, maybe in september after school holidays, i will join a local chessclub to see by myself, the problem is i don't want to do all competitions, since i do not have the time, and they may refuse me...

Hey hangingking.

regarding finding these tournaments more difficult, I don't compare them to online play, to me that's not real chess. I simply played badly in these tournaments I think in large part through non chess factors, like poor healthy etc. I'm not comparing my results to online games but rather my club games and previous tournaments, (I came 2nd in the Torbay U100 last November).

The competition wasn't tougher here than there, in fact a lot of the same guys played, but I just played shockingly poorly!

Still I've not shrunk from my losses. Rather than throw any evidence of these losses away I have instead analysed them.

So far I've analysed three of my 7 games played, all losses and even though my play was crap I feel I've learnt plenty of pointers.


One area I disagree with you guys though is in using computers for analysis.

For me that is the starting point. If I lose badly in a match because I don't understand what's going on I see very little value in then going over it again unassisted as I havn't suddenly developed new skills to analyse properly this time around as opposed to during the actual game!

Instead I run it through Rybka then analyse it using Rybka's analysis.

So I still spend several hours going over each game but rather than exploring blind paths and wrong ideas as I would if trying to analyse myself I devote my time to trying to understand WHY Rybka recommends the moves it does.

In other words I spend my time on learning to play the right way and no time fooling around trying variations that are completely irelevent to the position at hand.

I really believe this is a better method for lower level players.

I need the computer to tell me the right moves so I can then try and understand why they are the right moves.

Spending time on the board playing more wrong moves of my own ideas is simply a waste of time in my opinion.

I feel like I've learnt more about chess in this past weekend analysing three games with Rybka than I have in maybe the past two or three months, and what's better is unlike book or ICS study, this is DIRECTLY relevant material I'm learning as it comes from my OWN games.

I really think unless you already have a full positional understanding its simply a waste of time and effort to analyse a game yourself without a computer first telling exactly what you should be analysing.

It's like digging for treasure. One way would be to randomly dig where you think something valuable might be buried, just making your selection based on essentially guesswork with very little understanding of the history of the area and any idea of what's under the ground.

A far better way would be if you had access to maps and historical documents revealing already the likely locations of treasure and even better sophisticated tools that can allow you to search an area with geo-imaging so you can see exactly what's under there before you start your hardwork of digging!

That's going to be my motto in chess. I use the computer to find the best moves, and then I spend my analysis time working out why they are better than my moves.

Trying to work out better moves than I found in the game on my own is just pissing in the wind imo.

Hmm, here is my opinion on this:

I disagree because i tried your method and it worked for me at the beginning, i mean it helps me reach 1600. But then, i found myself stagnating for a long time, something was missing. So i tried something else.

I think trying to understand deeply computer moves is not easy at all, because they often do moves that are justified only by tactical advantage 10 moves later, and i'm not sure we have the calculation ability to see beyond these moves.

I tried myself to play with engine help for example having 2 boards, 1 set to play against a lowered engine and another (mimic) analysing with rybka : one every five moves for example just play the rybka move, and the other four you are on your own.
I realized soon, i was unable to understand the finality of the rybka move in the time given to me, and carry out the line chosen in secret by the computer.

This is why courses like ICS comes into purpose, they give plans and strategy in order to build your own lines and not count on the miraculous move, because the miraculous move gives sometime a big advantage, but often has also drawbacks that could be disastrous if you are not aware of them. For example, if the miraculous move is a sacrifice, what to do next ? That not so simple.

This is the reason why you have to keep distance with lines proposed by engines, because they involve a lot of knowledge, to be able to reproduce them in your everyday game.

You should rather, use engines to evaluate your moves. One thing is to detect big mistakes, that's very easy. Second thing is to try to think again on the game by yourself, choose a move and evaluate it with the engine.
That's a completely different method, than just "looking at the solution" !
So of course it's not like hypothetical treasure quest because you can use the engine as an oracle (able to tell you : your move is good or you move is bad).

Unroll the game and stop it when there is a mistake. Go backward 1, 2 or 3 moves and try to find a line different from the one you chosen in the played game, and only then perfom the move on the board and evaluate it with the computer (just evaluate, not analysing at this point). If the result is OK fine you have learned something, if not, just try again... At the end if you still dry on the position, look at the line proposed by the computer.
But i think, immediately looking at the computer solution, will not help you that much improving.

Also i realized this method makes me now more confident in my moves.

I'm actually pretty confident in the engine I use (Rybka 3 Human).

I think it gives variations prety similar to what a human would play. I've set the material value to .035 so I get a decent amount of analysis but am not overwhelmed.


To give you an example of some of Rybka's selected best moves and how I understand them.....

In a couple of games Rybka pointed out opportunitys that I had to attack my opponent's pawn centre, and playing through their line choice I could immediately see the benefits of wiping out White's pawn centre when given the chance.

Other examples were moves like Re1 criticised for competing with black for an open file and reading between the lines it's clear that it because White has more space and thus doesn't want to exchange rooks.

Another common thing I tend to do in QGD positions as Black is to put one knight on e4 and the other supporting it on f6. I saw Rybka consistently doesn't like this idea and I've realised in analysing with it now that it's because the knights far from supporting each other are duplicating each other's function and getting tangled up with each other.

I always thought the Nf6 upheld the e4 knight but instead what its actually doing is taking away its best retreat square, just as the e4 knight takes away the knight on f6 best attacking square.

I saw I missed a couple chances in one game to get a rook to the back rank early in a game.

I think its more important to spend time actually analysing these good moves than it is to come up with move ideas on my own.


My point is this, its true a compute enginge, even Rybka's positional understanding, is different from, and maybe not objectively as good as a human grandmaster, but its still MUCH MUCH better than mine at my current 1600 level.

If I spend time analysing a game blindly so to speak, trying to come up with ideas of of my own to pursue and spend 45 mins on a line that Rybka then shows instantly fails because of either hidden tactics or in reading between the lines because it involves bad positional choices then that's 45 mins of analysis wasted.

I would rather know the best moves first and spend that 45 minutes figuring out why they are the best moves.

I remember BDK saying something similar when he was doing the ICS Consequences exercises from month 1.

He found he was spending far more time, and thus remembering more clearly, his own faulty analysis that he spent a considerable amount of time and effort on in trying to answer those questions than he was the actual ICS solutions.

For me this is a bad idea. When you are are analysing moves and moving pieces around you are reinforcing those thoughts and ideas into your mind. If your analysing bad moves based on limited understanding you are actually reinforcing negative ideas into your play.

I think its a case of having to unlearn.

I reckon it will take me another couple of years before I am competent enough positionally and have the evaluation skills to look at my own games unassisted. In the meantime I will focus on Rybka's moves and seek to analyse and understand them rather than my own moves that cause me to lose games to 1400 rated players!

Rybka's way may not be the way an elite GM would play a position, but its a lot better than how I would.

And in my experince, its not very often that Rybka recommends a move because of a 12 move sacrifice idea, it's a suprisingly positionally minded engine, especially on its human setting.

We have a long thread on similar stuff here.

Of course computers will win against human, but that doesn't make them best instructors, especially with strategy, as the human will explain the position, not just spit out variations.

As long as you spend time on games that's fine, to each his own. It just makes me a bit surprised when someone doesn't trust themselves enough to go through a game on their own, but I used to be like that too so certainly understand it. I wouldn't even want to do postmortems with people at a tournament for such reasons, as I didn't have enough confidence in myself that I would see, say, or hear anything useful. Talk about stupidity, and missed opportunity on my part! (that is main point of this whole post, so you can ignore the rest ).

I'm not much better now that I was then, I just had a shift in attitude (and got a good coach). For one, typically the practical consequences in the game reveal a lot. So, for example on move 10 I didn't see why X was a positional blunder, but 20 moves later it became clear why. It doesn't take a GM level understanding to see why the move 10 move made the position cramped in the rest of the game 20, this despite the fact that it might take a GM to see right at move 10 that it was bad. Later moves have a way of showing you what is good or bad in a game, the things you worry about at move 20 reveal weaknesses in your thinking at move 10. Hence, trust the game to help reveal the evaluation of earlier moves.

Plus, the more you do it on your own the more you will be confident in real games at tournaments, not insecure because of dependence on the Rybka nipple.

Key is to talk to chess peers after the game, go over it, hash over ideas, see what they were thinking, play around with pieces, refute each other. The more exposure to real people (including yourself), interaction, conversation about positions, with nobody involved allowed access to the computer nipple, the better. Then the computer later, if you must, will reveal some weaknesses of what everyone said, and that is a nice final evaluation of the position, but not necessarily superseding what you found in the conversation.

But I'm not dogmatic about this. If it seems to help, and keeps someone into chess, and they don't have access to people and coaches to talk about chess, or people at tournaments don't want to do postmortems (as is often the case with lower-rated players in tournaments for some reason) then why not. I don't think it is a coincidence, however, that the best players eschew a computer-first approach even for their most novice players. They all say to use the computer last, as a check on your thinking. It took me a while to believe this.

Going over a game unassisted is one thing. Going over a game on your own, and then with a good coach, in a real human conversation, is another. There is nothing better for memory formation (in my experience anyway) than what I get in conversation with a human, especially after I have gone over the position on my own and formed an opinion. Then they come along and crush it with great ideas (and some variations if it is tactical), and explain the general principle, what to look out for, etc..

That is very different than staring at some puzzle for 10 minutes, talking to nobody, and remembering some incorrect thinking about the puzzle. For puzzles, in particular, I try to spend at least as much time on the solution as the problem itself. If I had a coach to talk about each puzzle, that would probably be a wonderful way to counter the 'memory of the wrong answer' effect, but it seems a better use of coach time to go over my actual games.

I was in LA once and saw a bunch of people going through a tactics book together, arguing, fighting, hashing out possible solutions, refuting each other. No computer of course, no nipple. That's the ideal in my mind. Too bad it is so hard to find. The thing to do is create that conversation with yourself. Refute yourself, fight yourself, find problems with what you just thought or said. Imagine Bobby Fischer is sitting across from you. What might he say? Of course you won't be as good as Him, or a group of friends, but it is a great way to go at a game. You can always grab the Rybka teat when you are done.

To repeat, of course Rybka is good, and can even find moves for which we can give a positional justification. Focusing on that sort of misses the larger point I think.

The irony is that at that other thread Bilbo was taking the human side against Rybka, saying brilliant and insightful things like:
"There are many occasions in chess where objectively the best move in a purely analytical sense is not the best practical move to play."

As for internet versus in-person, I have never understood people who say internet games aren't "real chess." If your rating improves 400 points at ICC, you have gotten better. There is a rating system at ICC, different from FIDE of course, but you will earn a rating based on performance in slow games, and play people at a similar rating so, on average, you should have a decent game of chess.

What I do is set up the game on the real board, and only look at the computer at ICC to make my move and see what move my opponent made. Some great long battles I have had, that are as real as many of the horrible games I've played in in-person tournaments.

Perhaps it is a question of what motivates you. I know many people aren't motivated to concentrate at ICC because in their mind it doesn't feel as real. I don't have that problem. Indeed my best game of chess ever was a 45/45 game at ICC.

Hey BDK, I agree I was defending the ICS team's ability to come up with ideas on their own that may prove as viable practical alternatives to Rybka's choices, but the difference is I don't trust MYSELF to do it

I am now offically rated, and a 1600 level player, and that means that I fall well well short of excellent positional understanding.

I hope to improve but going over my own games it fair to say that even if Rybka, (which its specially human oriented play style) makes the odd move which is positionally not what a GM would play it will still play a far better positional game of chess than I can!

It's been working for me tremendously so far since adopting this strategy.

I won a nice game tonight at the club using exactly the stuff I had learned from Rybka over the weekend.

I fought far more aggressively for the centre, and when I had the chance to play e5 as black (it was a Colle), one move before white was about to play e4 I went for it and it paid off.

I equalised in the centre almost immediately and went on to win the game.

If I just sit down at the board with one of my games unassisted I feel like I'm just wasting my time in all honesty. I might spend an hour exploring some ideas that are entirely irelevent to the position at hand and then when I see Rybka's analysis I only glance at the correct moves, all of my focus, moving the pieces around, calculations etc, in sum, all my effort, has been invested studying inferior lines.

You wouldn't do that in any other discipline!

Far better to get the right moves and decent analysis from the beginng and then devote all your time and energy on trying to work out exactly why Rybka chose the lines it did.

I still put as much effort into analysing the game as before, I'm just studying decent lines as opposed to my own faulty ideas.

It's working for me at least, and even if the entire chess world deplores my method I will bhappily be a lone wolf in the wilderness, a maverick!

I should also say that I believe, quite confidently that Rybka understands positional chess, and will play moves in accordance to the needs of a position far more accurately and consistently than any player under 2000, which for me means it beats studying with a stronger, but still low rated player.

If I had both an 1800 looking unassisted over my game and Rybka I would trust Rybka more in a vast majority of cases to show me the best moves.

Be happy, we could have studied GO...

Do you do math homework by struggling with the questions first, or just look up the answer? Looking up the answer before any struggle would be a huge mistake, a squandering of an opportunity to learn. When solving tactical problems, I don't just look at the answer. I struggle with it, try to figure it out on my own first. Typically ten minutes for a single position. This is key!

That initial exploration of a game provides a set of hypotheses about the position, the hypothesis being that X is the best plan, Y is the best move, Z will be my opponent's best response. Such exploration of a full game provides an entire narrative arc, a story about what happened, what should have happened, what could have happened. Then, after struggling to generate these ideas, I will find some outside source better than me to test my ideas. This way, my understanding/knowledge/confidence in my skills increases, my ability to have a conversation about chess, my ability to literally generate my own ideas about a position, improves. Then, if my hypotheses are refuted, I get strong feedback in the form of negative reinforcement (oh crap here's what I missed and here's how I need to revise my approach), or even better positive reinforcement (wow my ideas were right, I should remember in positions like these here are some good ideas). Because the ideas are mine, because I generated them, I am more likely to remember the lesson learned. This crucial process of idea proliferation and then refinement via reinforcement doesn't happen with the nipple-first method.

Also, you didn't respond to a key point I made in response to your lack of confidence, that at move 10 while it is hard to see why something was bad, by move 20 it is revealed. So, having a full game there is very helpful for evaluating positions earlier in the game. As I said, a GM might see why X was bad at move 10, but even us lowly patzers can see why it was bad when we have the whole game there to show us the refutation.

It would be interesting to see if there is even one coach out there that would recommend firing up Rybka before at least a cursory examination/interrogation of the game, a struggle between player and game. Perhaps for internet blitz games, where you just want to find tactical problems.

While the Rybka-first method is obviously suboptimal from a chess-improvement standpoint, if it is the only way you like to study your games, then it is the best way for you! Hell, if I studied optimally all the time, I'd probably get sick of chess.

As I said, I think the computer is both useless and indispensable for postmortems. It isn't an either-or thing, but a matter of emphasis. Plus, if you don't have access to other people (that's one thing you didn't seem to mention that was a main point of my post), if you can't have a good conversation about chess, then Rybka is a good interlocutor.

I agree with your maths analagy BDK but you're missing the point a bit, I've already studied the positions, I played the game!

I'm not talking about crappy internet sessions all my games are over the board, properly graded games with classical time controls, actually mostly 40 moves in 1 hour 30 mins with another 30 mins to finish, so I've typically pored over the board for around 3 hours per game on average at least.

I wouldn't study GM games with an engine first as I havn't already spent time studying it, but my practice essentially is my study and analysis of my own thoughts, then I analyse it with Rybka to see how it would have played differently to me.

None of us have infinite time and I don't want to be spending more than say 6 hours a week, on any single of my own games. So I play it, which is 3 or 4 hours, then I will study it for another 2 or 3 hours, make detailed notes and file it away, reviewing all my game notes periodically.

I think its a wonderfully productive way to study, I wish I had always done this as it makes the whole process efficient and rewarding to me.

It's like when Jeremy Silman brought his Complete Endgame Course based on the idea of only providing rating appropriate material in each section rather than divide by chapters on king and pawn, rook and pawn etc, which simply overwhelmed.

I remember the very first chess book I started to read was Capablanca's Fundamentals, which was so ridicuosly over the head of a beginning player it was absurd. In the very first chapter he's talking about how to mate with Queen and King versus Rook and King, at a time when I still havn't heard of en passant yet or the names of any openings other than the Ruy Lopez, and have never even learnt how to properly mate with a king and rook against a lone king.

I have another endgame book which starts with basic mates as chapter 1 and includes the mate with king, knight and bishop vs king, ON CHAPTER 1! No wonder I never progresses.

I just feel the same with aimless self analysis when I'm a low level player. It's just trial and error and I'd rather actually spend time on trying to understand why the computer analysis is good rather than try and come up with more failed ideas myself.

I honestly feel I can jump a couple hundred elo points over the next year with this method.....I just don't want to be spending hours on guesswork self analysis, and reinforcing my poor ideas by moving pieces around on a real chessboard, it's just reinforcing bad habits.

I should say though that as I increase in positional understanding I WILL do my own analysis more and more.

Right now at 1600 I simply don't have more than the most rudimentary positional skills and so am likely to wonder off on paths to nowhere if I self analyse for hours before using an engine to aid me.

Yes, if your games are four hours then that is a good chunk of thinking already, and probably you want to avoid burnout.

Do you have people you can talk about chess with?

Not really, I mean I belong to a local club but we don't really socialise outside of club night and there are no really strong players to discuss things with anyway.

The higest players in our club are just under 2000 and 1900 respectively.

To be honest I don't think their understanding of positional play would be any better than Rybka's either.

I think in Rybka I have found my perfect partner. I will slowly build up my own positional understanding through the ICS course and from my own games and the meantime I will seek to understand my chess engines annotations.

Of course it gives a few strange moves occaisonly but if you are devoting you study time to actually trying to understand Rybka's annotations rather than looking at your own ideas you'll be able to understand the logic behind its ideas and why it selected those lines.

It worked brilliantly for me yesterday. I saw from analysing my tournament games that I missed opportunities to wrest control in the centre and was adament that in this game I wouldn't make that mistake and I played a lot more aggressively than I would in normally in the opening, opening up the game and playing e5 before my opponent could get e4 in.

It worked a treat. I also learned that I place my knights wrongly especially as black, often having one of f6 supporting another on e4.

Messing around with my own ideas I never took the time to try and understand why Rybka didn't like this plan but now as I'm soley concentrating on trying to work out the logic behind Rybka's moves, rather than thinking up flawed plans of my own, (I come up wth enough of them in my actual games!) I am spending my time correcting my errors rather than trying to think up more.

I actually think this will make me stronger than any player at my club within a couple years. Screw convential chess wisdom, I'm going to do it my way

Thinking it won't help to talk to people at your club is a big mistake, unless you beat them all every game. In any case, it sounds like you could use a good coach. Just going over games with Rybka is sort of sad.

I did try and get a coach, I asked Jack Rudd, an international master who lives in the same county as me but he was too busy to take on coaching.

He looked over a couple of my games however that had been puzzling me, mainly because Rybka's suggestions were baffling me in one game in particular but he explained that it was correct, basically an opponent had lauched a pawn storm against my king and so I started bringing pieces back (both knights) to defend.

The computer didn't like that at all and actually recommended moving the existing knight away from my kingside, leaving it knightless in the face a pawn charge!

But Jack said that in such situations the pieces only get in the way and allow the pawns to attack with more forces as they must flee each advance.

But it confirmed to me that in the vast majority of case Rybka will play if not the absolute best continuation for every move, still a far better set of continuations over a series of games than any strong player at my club would.

The great thing though is that as Rybka has no real commentary you still have to really analyse yourself and think about the position to determine why Rybka is making the choices it is.

So I don't analyse my games in any less detail than you, I just focus on learning the right continuations as opposed to wasting time on my own guesswork.

It's a MUCH better way to analyse games, you have to trust me on this!

I felt I've learnt more in the past week doing this than I have in the past couple months.

Actually I should clarify my thoughts a bit more to avoid misunderstanding.

I'm not saying you should never analyse a position, like ICS exercises, or following grandmaster games etc, here you should absolutely try and solve the problems on your own here.

I'm talking about you own games, after you have played them.

As I no longer play online chess and only games with full time controls my practice lasts at least 3 hours or so per game, so that is my own work, analysis, plans etc.

After the game, I fire Rybka right up and my post game analysis will consist in me seeing why Rybka rejected certain move choices and why it recommended others.

So I have already analysed the game for 3 hours when I was playing it.

I'm not talking at all about no longer working on your own to solve test positions, or trying to play through master games etc, on the contrary you won't learn anything unless you thoroughly absorb yourself into the game or position and really seek to evaluate and analyse.

But once you've completed your own analysis, fire up the computer and make sure you spend at least as much time on trying to comprehend all the subtleties and reasons behind your engines move choices.