Frequency of occurrence
Slowly the big picture of a whole chess game emerges.
- 250 scenarios for the opening
- 50-150 scenarios for the middlegame
- 50-150 scenarios for the endgame
- 50-150 scenarios for tactics
with one week per scenario it takes somewhere between 7.5 and 13.5 year to master the game to a certain degree. This leads to the question: with which scenarios to start? With the scenarios that have the highest frequency of occurrence. The scenarios that happen in each and every game.
Every game:
Middlegame scenarios
- battle for the center
- development
- pawn structure
- piece exchanges
- activate your pieces
- bury your opponents pieces
Tactical scenarios
Not every game:
Openings
Endgame scenarios
Interesting conjecture regarding the time required to identify and absorb all those logical scenarios.
ReplyDeleteAnders Ericsson’s research regarding university violin students (popularized in Malcolm Gladwell’s book Outliers: The Story of Success) indicated that the average solitary practice time was about 7,500 (NOT 10,000) hours. He estimated that the very best students (seniors who were expected to eventually play in the world’s most renowned orchestras) were only about half way to their best performance levels and that for them to reach their highest skill level (around age 30) would require 20,000-25,000 practice hours in total.
I am always skeptical of “averages” on average.
Assume the two end points (a low of 400 weeks and a high of 700 weeks) are accurate enough for government work, and it takes an average of 7,500 hours to reach the point of half way up to the highest level (Magnus Carlsen at 2882 Elo). [The average USCF rating is 1500 (≈1400 Elo).] Assuming that the range (lowest Elo [1400] to highest Elo [2882]) is representative, with a normal bell curve distribution (it’s actually shaped more like a Rayleigh curve, but who cares?!?), the median Elo rating would be 2141, which is USCF “Expert” (USCF 2000-2200) or slightly below FIDE Candidate Master (2200-2300).
Note the following regarding Russian categories:
The level of 1st category was more or less equivalent to a current classification of 2000 Elo. After that one could become a Candidate Master, which corresponded to a current Elo of around 2200. As for the Master category, it was 2400, maybe more.
[Recall that mister Lasker’s 200-hour curriculum specified that the goal (if reached) would end up somewhat near to the skill of a first category player—someone who if given any odds at all would surely win. That’s not really true for that skill level, but that’s a story for another day.]
Mashing all the maths together gives a raw (rough) estimate of how much effort would be required.
On the lower end, 7500 hours spread over 400 weeks works out to 19 hours per week. On the higher end, 7500 hours spread over 700 weeks works out to 11 hours per week for almost twice as long on the calendar (13.5 years). Converted to hours per day (5 days per week): 2-4 hours per day.
Leaving aside the difficulty of consistently doing that for 7.5-13.5 years, there is at least one other more serious problem.
University violin students follow a curriculum that has been established over hundreds of years. All of the “logical scenarios” (musical scores) and how to best play them (technique) are known and standardized. this is one of the critical requirements for Ericsson’s deliberate practice regimen.
Unfortunately, in spite of all the chess training material and training programs, there is no standard curriculum for advancing skill to the first category level [ignoring mister Lasker’s 200-hour “program”]. Your 23+ years pursuing this goal is indicative of the great difficulty in finding and learning standard scenarios (if such things exist, which I believe DO exist).
I’m not trying to be “Debbie Downer.” I’m perfectly happy with learning one thing at a time, until I run out of time. (At 76, I suspect I will run out of time a l-o-n-g time before I reach the point of knowing sufficient logical scenarios (presuming such a collection is available somewhere). Such is life: It Is What It Is [IIWII].
One man’s logical scenario is another man’s heuristic, maxim, general principle or “rule.” (I can think of so many “exceptions” to THAT “rule”!)
In the meantime, onward to the skillful siege of the castle of knowledge (and don’t forget to bring the Trébuchet)!
What I wanted to express is some sort of celebration about the fact that I had the feeling for the first time that I mapped out the full area of scenarios of a game.
ReplyDeleteI'm not too serious about numbers. The first 23 years were taken by experimenting with the method. When the method was clear, it took me another 1.5 year for mapping out the area of scenarios. The next step, which is this post about, is to prioritize the scenarios.
7.5 years to become a master, 13.5 years to become a (super?)grandmaster. Ceteris paribus. My book "Grandmaster at 75" is on the line.
Congratulations on formulating “the full area of scenarios of a game”! I think that is an achievement within the chess context similar to the mapping of the human genome.
ReplyDeleteThe 6th Edition of the Encyclopedia of Chess Combinations contains 3155 problems, cataloged into 10 categories (as previously noted in a recent comment):
1. Annihilation of Defense
2. Blockade
3. Clearance
4. Deflection
5. Discovered attack
6. Pinning
7. Demolition of pawn structure
8. Decoy
9. Interference
10. Double attack
Does this reference work provide complete coverage of the 50-150 tactical scenarios? Does it provide coverage of at least some of the 50-150 middlegame scenarios as well? Or, are scenarios something else entirely, such as heuristics, maxims, general principles, “rules”?
Do you have a complete list of books/videos that cover all scenarios? If so, please share the list!
I’m eagerly awaiting the publication of “Grandmaster at 75"!
Since "areas of scenarios" is ultimately vague, covering the whole chess area is not so difficult. Opening, middlegame and endgame, and you are already there.
ReplyDeleteI compared that with my own games, and noticed a few holes in my bucket. With the aid of a course about an opening, it is not so difficult to estimate how many scenarios you need to know as a bare minimum. I counted it for the quickstarter of the Nimzo Indian and came to 50. Since I play 5 openings, I extrapolate that to 5x50=250.
To estimate tactics wasn't that difficult either. There are about 50 different themes at Chesstempo (I didn't count them). The fact that you can combine them into 3155 different combinations is irrelevant. There must be 50 to the power N different combinations, I would think. Where N depends on the current planet constellation and your digestion. I looked at a few blitz games of Hikaru Nakamura to get an idea of the tactics you need to know and how well you need to know them.
I looked at the Trébuchet lately, and noticed that there are about six scenarios to adopt to go for a win or a draw. To a certain degree, these scenarios are position independent. I compare that with the "100 endgames you must know" course from Chessable and made an estimate.
A game ends with mate or gain of wood to prevent mate or and endgame. That is covered by tactics and endgames.
The preparation for mate is covered by Vukovic and the PoPLoAFun system.
This was all already known. What was not known, was the the middlegame stuff. See the list in the post above. With the aid of the courses I could estimate that.
After identifying the areas, the next questions are of course: which scenarios precise, why do you need a week to master just one scenario and can't that be done faster?
ReplyDeleteI just found another example of the “box canyon” (ravine, ambush, whatever) that involves a Knight rather than the Queen or Bishop. In all cases, the defender is constrained by Function and by the restriction of the surrounding terrain to a narrow channel. An attack deflecting the defender (usually the King) away then allows the attacker to take the previously defended piece.
ReplyDeleteFEN: r2q4/ppp1pkbp/4bnp1/3n4/3PNB2/5P2/PPP3PP/R2Q1RK1 w - - 0 1
Puzzle #67jHV from the lichess.org game ChessHobbyPlayer (2357) vs HAMOUMRafik (2290), a 10+0 Rapid game. The associated puzzle themes are: Crushing; Deflection; Middlegame; Short puzzle. Note that both players are above average playing strength AND the game ends on the 15th move, which implies that neither player would have been in time trouble.
1. d4 Nf6 2. Nf3 g6 3. Nc3 Bg7 4. e4 d6 5. e5 dxe5 6. Nxe5 O-O 7. Bc4 B(b)d7 8. Nxf7 Rxf7 9. O-O Nb6 10. Bxf7+ Kxf7 11. Bf4 Bg4 12. f3 Be6 [Entering the “box canyon”!] 13. Ne4 [Headed for the ambush point] 13… N(b)d5 [not “SEEing” the terrain problem] 14. Ng5+ Kg8 [Certainly NOT 15… Kf8, allowing a royal fork] 15. Nxe6 Resigns [being a piece down without compensation.]
I realized that this is a good example of the benefit of SEEing both positive forms and negative spaces. The attacking White Knight and the deflection of the Black King away from defending the BBe6 are the positive forms. The negative space is the restriction to a single square from which the Black King can defend BBe6. The “box canyon” is formed by the BPe7 and BNf6.
The essential insight is that we often pay little or no attention to the background/frame of the “big picture” because we are so focused on the foreground/primary item of interest. This holds true whether we are just training or actually playing a game.
Perhaps too much of our tactical training is focused on SEEing only the positive forms (foreground) with little or no attention paid to the negative spaces (background). When we can SEE both positive forms and negative spaces, we will be capable of SEEing the entire picture (the “essence” of the position).
Note that this is NOT synonymous with the “eagle’s eye view.”
Correction to the game score:
ReplyDelete7. Bc4 B(b)d7
should be:
7. Bc4 N(b)d7
I hear what you say, but it doesn't trigger something within me. Usually that takes some time.
DeleteI know these "types of positions" due to a study of 100 different sacrifices on f7. When d6 is played, e6 is weakened. Making it vulnerable to a sac on f7. Usually there are a few scenarios that can make this work:
1. when you can mate the king
2. when you can force a royal fork
3. when you can trap the queen
4. when you can force access to c7 where you can fork rook and queen
5. when you can force access to c7 where you can trap the rook after luring away the queen from c7
Now you found a 6th scenario, when there is a piece put on e6
This post is about frequency of occurrence. And working on special cases of a sac on f7 lacs a high frequency. Which doesn't mean your effort is not appreciated.
That's why I focus on the scenarios around the conquest of the center. But I haven't found a beginning yet. So when I become quiet or so, you know what I'm working on. Help is welcome. The difficulty is to find an entrance point.
Not a problem.
ReplyDeleteWhat area(s) of the conquest of the center can we help you with?
Out of curiosity, I looked at Chess Tempo’s Game Database to see what White players considered a ‘playable’ (perhaps winning?) opening. I was somewhat surprised to find that every possible legal move had been played multiple times. Even more surprising was the winning stats for White!
ReplyDeleteThe 5 openings with the highest “winning” chances for White:
1. a4 – 158 games [W57%-D12.7%-B30.4%]
1. h3 – 298 games [W54%-D14.4%-B32.6%]
1. Na3 - 38 games [W52.6%-D13.2%-B34.2%]
1. Nh3 - 98 games [W51%-D18.4%-B30.6%]
1. a3 – 1598 games [W50.1%-D15.1%-B34.8%]
The average rating for all games was 2400+.
The frequency of occurrence does hint that these openings might not be the best ways to begin the game for White, in spite of the glowing winning stats.
How does one go about figuring out a logical scenario for things like this?!?
IM John Watson certainly seems vindicated in declaring the “rules independence” of modern chess.