Playing folks 200 points better than you

Even if it took me a few posts to get it

A strategy to consider when facing a much stronger opponent

1. Maintain a strong pawn structure and don't hang a single pawn. You will need an equal amount in the endgame.

2. Trade pieces, not pawns, as soon as possible. The stronger player's attacking chances will diminish as each piece is removed from the board.

3. Study the heck out of K+R+P vs. K+R+P and K+P vs. K+P endgames. Promote one of the foot soldiers.

The way I see it, the much stronger player has a tactical strength that will crush the weaker player and will rarely hang a piece or fall for a simple tactic like a fork, skewer or pin. By trading away the minor pieces, the queen and one rook - well - your chances have dramatically increased. The queen is worth 9 pawns. Make the stronger player work for the win!

---

"Chess is 99% tactics".

I have more opportunities to apply tactics against opponents of +/- 100 strength. I'm getting more tactical practice and will get stronger over time. Playing someone 200 points stronger doesn't work for me. I wouldn't get anything out of a match against someone 400 points stronger than me on a regular basis except a major headache. I'd quit the game.

I do train with someone 600 points stronger than myself. I'll play normal chess - not the crazy sounding trade down strategy - and learn from his comments.

I rather enjoy the challenge. I recently beat someone that was rated over 600 points higher than me so I say that it is a good thing to play someone better than you. It is a great confidence booster and a great lesson in chess. Go for it!

I suppose you have read Dan Heismans recommendations for choosing your training game opponents?

The best way to improve is to play is to play stronger players-  One of my only negative experiences on this website was a game with a player rated about the same as I at that time. I won after blundering then came up with what I thought were clever moves- I later noticed he played only player lower then him. He'll never improve because is hated losing and his victories were cheap shots at weak oppenents.Best victories are because of thoughful and clever chess. No one likes losing but if you look at the rating loss verses a much stronger opponent, it is often quit small- take the risk, play your best, ask questions, learn - you will soon be up there. (coming from someone who is about 1400....I hope what a said is still meaningful.)

1. I never heard those stats before. It would be intersting to see those stats on a graph with a curve. Going from 50% to 1% would represent an exponential curve - pretty sure. Is the curve really that steep? My first guess would be that it would not be a curve, but a straight line going upwards at a 45 degree angle. A class A player would more than crush, but absolutely destroy a class C player? Wow!!

2. I feel it's necessary to win on occasion to improve. With my elementary students I will let them beat me from time to time so that they gian experience in the process of winning, i.e., building a decisive material/positional advantage, mating net and checkmate. I guide the new beginner through this winning process on every move, and as they improve I will slowly remove the training wheels.

In fact, the USCF rating systems are based on this 400 point diff. curve.

Which is why when a stronger player (>400 points) wins a tourney game, he gets close to zero points for that win ... the "expectation E" was already said to be very close to 0.999 so a tournament win (1.0 point) doesn't give you much of a difference to gain rating points for.

musikamole

it's based on a normal distribution, of the form e^(-(x^2)) with some constants thrown in.

stwils

since you like so much playing people above you, why would you avoid playing people below you? they like that experience, too. if you cant enjoy the game, enjoy helping your opponent improve.

A bell curve?

I believe the curve shown here at chess.com for rated competition is in the shape of a curve. It makes sense that there's a pile of people in the center and a few on either extreme, but how do you explain the winning chances going from 50% to less than 1% between, i.e., Class C vs. Class A?

It makes more sense for this extreme outcome to be true when the players are on opposite ends of the bell curve, i.e., Novice vs. GM. I never took a math class only devoted to probability and statistics, so my understanding is quite limited.

"In probability theory and statistics, the normal distribution, or Gaussian distribution, is an absolutely continuous probability distribution with zero cumulants  of all orders above two. The graph of the associated probability density function is “bell”-shaped, with peak at the mean, and is known as the Gaussian function or bell curve" - wiki  

I'll take a stab at it.

Your question: How do the winning chances shift from 50% to 1% between C and A player?

Well, your question is based on the existing concept of what a C and a A player rating means.

Let's define those FIRST.

An A player can be considered a 1800-2000 rated player by USCF (or any rating system) standards.

The C player is considered a 1400-1600 player (there's our 400 pt diff).

Let's call these players A and C for simplicity.

Now when they compete against each other,  the system adjusts their post-game rating based on whether they "performed upto" their expected result or not.

So Expectation E such that E(A wins) = 0.95 - 0.99 and logically, P(C wins) = 0.01- 0.05  (variation thrown in based on the choice of rating system we follow)

So the basis here is a rule that maps this Probability/Expectation of result to the rating difference.

An approximation of the expected result (lots of crazy math here) is basically that a 400-500 point rating diff. means the stronger player wins 95-99% of the time. 

So if the A player wins, no biggie. Expected Result equals Actual Result.

If that doesn't occur and if C wins ... there's a disturbance in the force and the "rating system" will make the C player "grow" in rating by a computed value X and drop the A player's by a similar value.

In other words,  A and C have a "playing strength" and the system attempts to merely "figure out + track whatever the heck it is". Over a period of time (100s of games), assuming A and C don't ever get better or worse in chess playing strength,  the system's rating of these players will match very closely to their true strength.

So your question is actually a little cyclical ... a player can only be called A and another C IF the "winning chances drop between 50% to 1%"  as compared to C playing another C.

Put differently, if your future self came back in a time machine and  beat you over 9 times out of 10, chances are, he's grown 400-500 points stronger than you as far as the USCF rating system is concerned.

This 400-500 point spread is merely chosen by the same system that defines what an A and C player are rated along the bell curve.  It could have been "any" number to define this "95-99%" spread.

Update:

I'd like to add that the skill differential in this game is gigantic .. quite possibly the widest of any game in the world (i.e. skill of the novice  going up to the skill of the World Champion) ... so it makes sense for the "scales" to be somewhat logarithmic/exponentially graded ... also explains why it gets harder and HARDER to climb from class D to C and then to B and then to A.

Think of it this way => A  C player knows 10 times  as much as a D but a B player may know 100 times more than a C.  

For my own USCF rating, I made my E to D to C climb in 1 year, took 2.5 to climb to B and am struggling ... 2 years and counting and still haven't reached A. ... when I play a player who is truly 500 points lower or higher ... those  95-99% statistics sure feel real to me :)

Wow! A B player may know 100 times more than a C player. That alone gives me a headache.  I wish you success in hitting class A.

My training partner has a USCF rating in the 1800's - Class A. I can't imagine what the difference is in playing strength between Class A and Master (1800 to 2200).

1000 times greater?

Those 100x multipliers were just meant to advertise how exponential the curve gets ... I'm not sure it's accurate *grin.  But you get the idea!

Shivsky:

"Another way of looking at it is to keep raising the bar until you are winning NO MORE than 35-40% of your games and no LESS than 20%.

 Those are the waters you will "need" to swim better in if you want to get stronger without ever getting complacent."

And this is true of things other than chess, also.  There's an educational theory out there by Lev Vygotsky, where the goal of the instructor is to give tasks to the students that are just beyond what their current ability is, but not completely out of reach.  This is known as the ZPD or Zone of Proximal Development.  If you can put yourself there, you'll learn a lot faster than if you only practice what you're already good at or try to leap too far beyond what you can do.

Additionally, as you do this, you increase the "ability you already have" which then allows you to move into the next ZPD -- this process is called scaffolding.

You've got to have an opening repertoire, solid endgame, and excellent tactics before playing someone 200 points better than you will improve your chess.

Otherwise, it's like...."Hmmm, I lost. Better work on my tactics."

how to make a sharp sword? they need to be first baptized by fire, then hit by the hammer repeatedly. in short, you gotta have you're @$$ kicked like everybody else before you learn to play. personally, i like playing with the strong raters. i dont want to play with those who dont finish their games. such a waste of my time.

With all due respect, I think you misunderstood my comment because you agree with me in your disagreement. Look:

1. If you don't have an opening repertoire, then you can't improve it if you play against someone 200 points better than you.

2. If your tactics suck, then you won't survive long enough against someone 200 points better than you to understand how to improve in your endgame!

3. As far as learning from your tactical shortcomings, this is why I said playing against someone 200 points better results in, "Better work on my tactics."

4. Your second paragraph doesn't address the particulars of playing against a 200 point rating difference.

If you don't have a repetoire and facing someone 200 above you, then you need to learn 'on the job'. The higher player is more likely to play main line theory for longer than someone 200 below. So get a few references and look up the theory! This way you will learn something about openings while playing stronger players.