Fromper's 2009 improvement thread

[Sammy39337]

I have come across an interesting article by Dan Heisman: the 10 Biggest Roadblocks to Improvement One of them is: Worrying too much about losing and ratings.

[Fromper]

Originally Posted by Sammy39337

I have come across an interesting article by Dan Heisman: the 10 Biggest Roadblocks to Improvement One of them is: Worrying too much about losing and ratings.

That's not as big an issue for me as it sometimes sounds in these threads. I'm really not obsessed with ratings. They're just a tool to measure my progress over the long term, so I talk about them here because I'm talking about long term goals. And while they occasionally annoy me with the short term statistical changes in the wrong direction (down), it's not like I base my actions on them.

For instance, I mentioned that I started playing at a club that meets on Wednesdays after losing my last job and having more free time. At the time I first went there, I considered the possibility that I'd lose rating points playing there, because I'd been playing poorly since returning from my two month break from playing. So I thought I could avoid losing rating points by practicing at home some more before starting at that club. But I realized that what I learned from playing there would make up for any lost rating points, so I chose not to worry about the rating.

Now being a little too timid during games, due to fear of losing, is sometimes a problem for me. But as I mentioned, I think McDonald's statement in that book has given me something to think about with regard to that, and it should inspire me to play more fearlessly going forward.

--Fromper

[Octal]

There are a few other quotes from that book that I thought would be good to keep in mind:
"General principles are there to aid your chess understanding, not control it."

"Black must always be willing to play ugly looking moves in the Dutch." After all, 1 ... f5 is about as ugly as you can get.

And the other one went something along the lines of:
"If you are contemplating a pawn-grab, or sacrifice, many players will spend a long time contemplating it, conclude it is unclear, and then decide on a safer continuation. When in this situation, play the move immediately, or use your time finding a different continuation."
I couldn't find the previous quote in the book again (I only skimmed it when searching for it), although it was something similar.

[ketchuplover]

1...f5 is beautiful!

[Perseus]

Originally Posted by Octal

"Black must always be willing to play ugly looking moves in the Dutch." After all, 1 ... f5 is about as ugly as you can get.

They must have missed 1. ... f6; which is playable actually.

Originally Posted by ketchuplover

1...f5 is beautiful!

It certainly does have it's moments

[Fromper]

So I went to the Saturday club today and lost to the top player, who I seem to play as black every single tournament. I was getting bored with playing the Tarrasch against him the last three times, and I'm up to the Stonewall chapter in the book on the Dutch, so I tried playing that in this game. He did a good job of taking advantage of the typical weaknesses in the Stonewall (black's bad bishop, hole on e5, etc) and stopped my king side attack from going anywhere. This is why the guy is rated almost 400 points above me.

But I also think the Stonewall isn't really my style. I'm looking forward to reading more on the Classical and Leningrad, as those might suit me better. That's really why I bought the book.

In my second game, I beat a guy rated just slightly below me, so I scored 50% and should gain a few rating points from these games. The final two rounds of this tourney will be next Saturday. Not many people showed up this time, so I can guess who I might play, and there really are no easy opponents for me. Everyone in the tournament is at least 1500!

So Abba did a good job in his 2009 goals thread of breaking down specific goals besides just one rating number, and also breaking down what he's doing to get there. I've decided to do something similar.

My goals for the year:

1. 1800 USCF rating. Preferably in time for the Miami Open in September, as stated in the first post in this thread. This is the ambitious long term goal. The rest will be more realistic and shorter term, I think.

2. Win or draw against someone over 1900. Obviously, this will happen if I come anywhere close to that first goal. I've gotten 4 wins and 2 draws in my life against opponents in the 1800's, but I've lost every game against opponents over 1900. It's about time to break that trend.

3. Improve my time management. I just re-read one of Dan Heisman's Novice Nook columns about this. I thought about time management more during my games today than usual, and I think it helped. The key for me is that I spend too much time thinking on non-critical moves. Cutting that out should really help me avoid getting into time trouble in the endgame.

Hmm... there are so many specific areas of the game that I want to improve on that I'm not sure if I should keep going with goals this way. I could be here all day. So instead, let's move on to study goals that will be used to achieve these goals.

At the start of this thread, I did what I usually do - listed various things I want to work on in order to improve. I've tried setting specific study goals in the past ("I must do this many tactical puzzles, this much reading, and play this many games every week"), and it always falls apart due to lack of realism. But the more general "I should study more" method that I've been using isn't quite strict enough to motivate me. So this time, I'm trying a cross between the two.

Basically, I'm listing what my ideal study for any week would be, while knowing that there are going to be very few weeks when I'll actually follow this plan perfectly. Why have the plan then? Because on weeks when I do a decent amount of play and study, this will remind me to balance what areas I'm focusing on, instead of just spending a ton of time on one thing that week. But on the other hand, I know that this is just a guideline, not a strict plan, so I'm not going to get frustrated and give up if I can't stick with it. Knowing my own personality, I think this "half and half" attitude is probably the key to realistically sticking with a study plan.

That said, here's what I'd ideally like to do every week to improve my chess:

1. Play at least 2 slow USCF rated games per week. I actually played 83 total games last year, so I averaged almost that much. But there are two additional weekly clubs that I should be able to go to regularly now that I'm at my new job. So between the 3 clubs, I could play 3-4 games every week if I actually make it to them all, plus additional games on weekends with big tournaments. Also note that by "slow", I mean G/70 or slower. I wish the clubs had all G/120 or slower, but I have to go with what's available.

2. Play at least 3 other games per week. This includes internet games, blitz, etc. I just think that the more I play, the more I'll improve, and a good balance of slow, serious games with some faster, casual games will help.

3. Study my own games - All of them!!! Ok, maybe not skittles games at the club where we're not keeping score, but all the tournament and internet games, anyway. This is where I think the most real learning comes from in Chess.

4. Do at least 50 tactics puzzles per week. Obviously, if I'm doing really tough ones, that number will go down, and if I'm doing easy ones, the number will go up. I think having a number to shoot for instead of saying how much time I'll spend will motivate me better.

5. Read at least one chapter of a Chess book every week. Right now, that's the book on the Dutch, along with the two endgame books that I'm studying. It'll change as I go. But just making sure to stop saying "I should read some of these books" and actually doing it will certainly help.

6. Play through at least 3 master games per week. If the chapter that I read for goal #5 happens to include 3 master games, that does count.

7. "Other" study. I've got that deep calculation exercise I'd like to try a couple of times, but it's time consuming, so it'll be once in a while as I have time. As I mentioned, I'm also considering hiring a coach, which will probably be too expensive to be every week. So having an "other" category in my study plan to cover these sorts of things seems like a good idea.

Speaking of hiring a coach, anyone have recommendations? I don't know of any locally who are close enough to be really convenient. I know NM Dan Heisman coaches via the phone/internet, and I like his teaching style from reading his columns, so I was thinking of contacting him. Anyone have other recommendations for people who coach remotely? How much should I expect to pay?

--Fromper

[Fromper]

I just realized that the obvious followup here is "How am I doing this week?". I just have to figure out if I'm using calendar weeks or what. I'm thinking a week in my study plan should be the weekend, plus the following week days. That way, I have the weekend to get the bulk of my study for the week done up front, and the rest is just motivating myself to do "one more thing" after work one or two nights during the week.

So here's the short form of the study plan:

1. Play at least 2 slow USCF rated games per week.
2. Play at least 3 other games per week.
3. Study my own games - All of them!!!
4. Do at least 50 tactics puzzles per week.
5. Read at least one chapter of a Chess book every week.
6. Play through at least 3 master games per week.

Last week (ending Friday), I accomplished all but #2 on that list. I played two games at the Saturday club, looked them over on Sunday, started using Chess Tempo for puzzles and did maybe 40, so close enough, and I read a couple of chapters in the book on the Dutch, which includes plenty of complete master games.

So far this week, I've accomplished goal #1 by playing at the local club, and I did a few more puzzles at Chess Tempo afterwards. I also analyzed my win at the club with my opponent, though I still need to go over the loss. I've got no plans for Sunday, so I should have plenty of time to study and play.

Overall, I'm going through a very highly motivated phase right now. We'll see how long it lasts.

--Fromper

[chessdad]

I'm currently coaching my son to play chess, and his study plan is:

1. 10-20 tactics puzzles per day
2. 5 or 6 FICS games per week (usually 15 min time controls)
3. No openings, except for "try 1. ... c5 and improvise from there"
4. Sometimes I point out one or two interesting things from a game, but we don't analyse them move by move.
5. Endgame exercises against the computer.

His FICS rating went from 1250 in October to 1500 right now. Most of his wins/losses are still determined by tactics, so improving tactics skills remains the most important objective. I try to teach by discovery. Let him try different things, and get a feel for what works, and what doesn't. For instance, last week I set up an endgame on the computer, with a rook and two pawns against a rook, and let him try to figure it out (taking back allowed). Half an hour and 20 variations later, he couldn't do it, so we stopped, and tried again the next day. After 3-4 days, he started to figure it out, and checkmated the computer in 48 moves with only one takeback. Not only has he learned the right way to do it, but he has also learned how not to do it, and why it doesn't work.

I'm no expert, but I believe this method is better than taking advice from books which are usually correct, but not always. If you always follow the "proper" advice, I think you risk missing the really creative exceptions to the rule. Besides, reading books takes time away from actually practicing, which is more important (IMHO). Nobody learns to ride the bike by reading a book.

So, instead of replaying all the moves from a master level game, try finding a position where you think one player has a big advantage, set that position up on the computer, and try to convert it to a win. If successful, find another position with a smaller advantage.

I also don't believe that long games are all that useful unless somebody already has a fairly high rating. A common recommendation is to play slow and take your time to check all simple tactics before touching the piece. I think that's rubbish. Instead, I believe one should be able to spot all simple tactics more or less instantaneously. If you can't do that, play less games, and spend more time on exercises. Of course, once you can do simple tactics quickly, it's okay to spend more time on deeper variations.

Of course, this is all just my own opinion, but based on my son's progress, it seems to work.

[Fromper]

Chessdad,

Thanks for your perspective, but I have to disagree with you about skipping book study and master games. While kids have the enthusiasm to keep trying the same things over and over until they get it right, that's definitely not the most efficient way to learn, and most adults just don't have that kind of patience. How old is your son, anyway?

While I do agree that many adults (myself included in the past) put too much focus on books in their chess study, and not enough emphasis on actual practice, that's no reason to eliminate chess books altogether. I've realized that I need to play more and study less, but I can also continue to learn from reading books and studying master games.

Let's use your example of trying to figure out the endgame of two pawns and a rook vs a lone rook. I like the idea of letting him try to figure it out for himself first. But once he spent at least half an hour on it and was ready to give up, that's when I'd pull out an endgame book and say "Let's see how the masters say to do it". Read through the technique, then let him try it again using what he's learned. Have him practice it a few more times, checking the book again if he gets it wrong, until he can do it right every time. Not only will this cut the learning time down greatly, but it also addresses the fact that eventually, there will be tougher techniques that he just won't be able to figure out for himself without help.

As for openings and tactics, I do agree that at low level, you should skip opening study (though making sure he knows opening principles is important) and focus heavily on tactics.

Personally, I'm already at a higher level than your son, where many of my tournament games aren't decided entirely on tactics. So while I am doing some tactics study, I definitely need to focus on other areas of the game more heavily than him.

In case there's any confusion, your son's 1500 FICS rating at fast time controls just doesn't even compare to my 1540 USCF tournament rating. First of all, internet ratings are always highly inflated compared to over the board ("OTB") tournament ratings. Second, his rating is based on fast time controls. At slower time controls, the quality of play increases dramatically, and if your son is only playing fast games, he's never faced that. If he were to show up to an adult USCF tournament right now and play slow games against opponents rated 1400-1600, there's no doubt in my mind that he'd probably lose every game.

Needless to say, I completely disagree with playing only quick games and no slow games. Staring at the board for extended periods of time to try and find the best move without moving the pieces IS Chess! On the other hand, I made the mistake in the past of playing only slow games, which is why I'm making a point of including some faster games and relatively fast master game study in my study plan today. I need to learn to spot things faster, and this is how to improve on that.

As for playing through master games, the point isn't to try and memorize exactly what they played. It's to get used to seeing the types of play that show up over and over in high level games. Eventually, the player will start to mimic some of the ideas, if not the actual moves, and that should raise the level of their own game.

But don't take my word for it. Most of my study plan is based on what I read in Dan Heisman's Novice Nook column at chesscafe.com. He's a master who has coached plenty of students and specializes in helping beginner and intermediate level players. Besides the opening principles article linked above, I'd also recommend you check out this article called "The Four Homeworks".

--Fromper

[chessdad]

I agree that slow games are useful, but only after somebody is good enough to spot 99% of all simple tactics right away, and uses their time to think of more complicated and deeper variations. A simple test is to use Fritz (or similar program) to perform a blunder check after playing a game to spot any places where a tactic was missed. If there's still missed tactics, spent more time on practicing those, otherwise spent more time on positional evaluations and long term planning. For example, one of the teachers (rated 2000) at my son's chess club was talking about some subtle strategic decision at move 20 of one of his own games. When I ran his game through Fritz's blunder check, it showed a tactical blunder at move 15 that both players had overlooked, and which changed the evaluation more than 20 moves of subtle strategic planning combined. The tactic wasn't even very deep or complicated.

I've read some of Dan Heisman's columns. One of his recommendations is to pick a move, and then spent some extra time double checking it doesn't fail because of basic tactics. I think that's a bad strategy, since you're only doing this for a single move. Instead, this looking for tactics should be done for every move in every variation that you consider. This is impossible if you can't do it quickly. Therefore, learn to see tactics quickly and automatically, then check all variations. When you have decided on a move, there's no need to double check it. Just play it right away.

As far as the rook+pawn ending, we did go over the first attempts, and look at the points where he got stuck (for instance by getting his king separated from the pawns), I then showed what the computer would have played instead. I do try to keep advice and rules to a minimum, though. He's 6, by the way.