Gotta Get that Feedback!
"How can I best improve my play?"
"Does doing X help me improve?"
"My goal is to be an expert. How can I best get there?"
These are very typical of many questions I get on our monthly Chess.com TV show Q&A with Coach Heisman (the first Friday of each month at 5-6:30 ET).
Obviously there are lots of answers (and material) on this subject, but one thing is clear. In all learning of complex subjects there must be a triangle present:
- Get feedback
Let's consider how one learns in school as an example:
1. Learn/study - this comes from teacher lectures and reading homework
2. Practice/play - this come from homework problems, lab work, answering questions in class, taking quizzes and tests, going to the blackboard to do problems
3. Get feedback - this comes from teacher answers to your questions, remarks about going to the blackboard, reviewed homework, and marking and reviewing quizzes and test. Even parents and peers can provide feedback on your work.
Imagine what school would be like if we got rid of #3, Feedback. You would operate in a vacuum. You could have all kinds of misconceptions about the subjects but no one would ever know and there would be little way for you to find out and be corrected. You would just go on making the same mistakes over and over.
For example, if you were learning addition and you added wrong, such as thinking that 5+5 = 55, it would be difficult to find out. No one would tell you about your problem, and you would never see your incorrect answers marked in your quizzes and tests. Not even in the answer section (a passive form of feedback).
Hardly anyone would do well under such circumstances. Yet many, if not most inexperienced players believe they can do just that with respect to chess. They either want to lock themselves in a closet and read chess books ("I am not going to play in tournaments until I am strong!") or just want to play and study.
One thing that makes the "play and study" seem effective is that it can work for a while when you're playing strength is fairly low. Many players go from 800 to 1400 or so without a tremendous amount of feedback.
When I mention feedback as an important part of learning, I often hear objections like "Rubinstein did not have it" (I highly doubt it, but even if so, it would be an incredibly exception, not a great model for most!) and "Fischer did it all by himself!". Sure, and young Bobby's daily trips to NM John Collins' home after school each day, where he played and analyzed with other strong Collins' students like GM Bill Lombardy, we can ignore that, right?
Feedback can come in many different forms. You don't have to hire a professional instructor such as myself. Obviously someone you see often, like a family member or friend, is ideal. So make friends with some strong players. Go to a local chess club and be proactively friendly. Ask the strong players if they will analyze your games with you and, even if they will not, they will at least usually let you sit in with their post-mortems with the other strong players.
When I wrote the Novice Nook "My Top Tips for Chess Improvement", the #1 tip was "Hang out with good players": Go to as many tournaments and club meetings as you can. Review your games afterwards with opponents and strong players. Talk chess, eat with other players, let them criticize your play and knowledge. One great skill in improvement is the willingness to accept criticism and the ability to see how you can apply it to future play and study to "subtract negatives" (e.g. remove misconceptions) and add positives.
A lesser but important form of feedback can be from engines and books. Review your games with engines for identification of critical errors. And after every game, fast or slow, review the opening with a book, database, or engine and answer the critical question "If I played an opponent who made the exact same moves against me sometime in the future, where would I differ my play to improve it?" Don't make the same opening mistakes over and over; use feedback to make each mistake only once, if possible.
Add consistent and good feedback to your plans and see how it affects your improvement .
Someone on the show asked me about the TCEC computer chess tournament. He asked specifically about learning from the openings and, indeed you can, although at last for the later rounds all the openings are "fed" to the computers so that they will play the same lines with white and black against each opponent once double (and more) round robin play commences in the later rounds.
At the moment, Komodo is running away with the semi-final, although it's only loss was to perrenial rival Stockfish. Two seasons ago Komodo edged Stockfish, but last season Stockfish beat Komodo easily. Look for those two to likely make the final again.
One person on Twitter insisted to me that Shredder was by far the best program. He was a little out of date but I had no luck convincing him by throwing facts his way. For example, the current CCRL "pure" rating list has Komodo first, Stockfish second, and Shredder 21st! His Shredder was beating his Stockfish, but I could not get him to admit that his version of Stockfish was way out of date and not nearly as strong as the current one. C'est la vie
Hope to see you next show! Get your questions ready...