The word “unofficial” while used, is hardly accurate. For these ratings,
are based on the same results that the future “official” ratings will be,
with the exceptions of the generally small effect of re-rates. In terms
of accuracy, for especially the very active U1200 or so player, due to
the increased K factor in the rating calculation, the difference between
the “published” and the current may very well be hundreds of rating
points.
I direct or help at many events in which players are passionate about “climbing
the ladder” from U300, to U600. to U900 to 900+ sections. This accomplishment,
more than prizes, cash or trophies, fuels their passion. The current ratings quite
frequently leave the monthly published in the dust.
Further, not all youngsters progress at the same rate. Some, even some of those
quite active, have remained U300 for years. Yet, they continue to play because in
lower formatted playing groups, at least they have a shot to be competitive, even
occasionally, win. Current ratings ensures that those players too advanced do not
compete with them.
Love of ratings is a very misguided passion. Passion for doing things right leads to excellence. Passion for improvement leads to excellence. Understanding what is improvement and what is not leads to excellence.
Agreed, Tim.
Or What if an event is rated DURING a tournament? (Not all tournaments are consecutive rounds on consecutive days.) What if an event has qualifications and a limited number of spots - do you keep changing who has the last few spots up to the day of the event?
OOPS - and NOW I see you also wrote this - GMTA.
Seems like a bad reason to go out of your way to introduce inconsistency just to tick them off.
Or, consider a large tournament with a Friday night round and Saturday RE-ENTRIES. Would a re-entry have to enter at a new Saturday morning updated rating - potentially in a different class?
There are probably an endless number of nightmare scenarios that can be created around using unofficial ‘most recent’ ratings.
The National Scholastic already have a policy for which published ratings they use. Events that begin before the 10th of the month use the previous month’s rating list, otherwise they use the current month’s rating list
SuperNationals V starts on April 5th, 2013, so it is already scheduled to use the March list, which will have a cutoff of February 15th, or 49 days before the event begins. That should be plenty of time for teams to decide their rosters, make travel arrangements, etc.
You can always come up with bad scenarios when common sense is thrown out the window.
The greater problem is tournament directors without common sense, which is the basis of a whole
lot of your scenarios. And if that is the basis, the problem lies elsewhere.
First, If you would read my original post, certainly I am not suggesting this approach for major events such as the SuperNationals. But, small local events, in which the staff capability
matches the local need. Second, perhaps with more experience
in running scholastic events, you will understand the relationship
between ratings and passion, and the great benefit USCF offers, thus
through its ratings service.
First, if you would read the post that you quoted, you would see that someone else (not me) made the comment about SuperNationals – it was made by Boyd (Eastside).
Second, I have experience running scholastic events dating back to 1970, and have likely run hundreds of such events. How many more do you think I’d have to run to have the amount of experience you think I require?
Perhaps, Rob, especially because of my successful teach experience, I have more relevant experience than you.
Your argument that we should stress ratings because kids get excited about them makes me think that we should feed kids nothing but candy because they like that too.
Frankly, if candy truly gets a player to study chess a little longer, provide better notation,
etc, then it might be a good motivation technique to use. Whatever withing reason that
fuels the passion. (however, for the most part, if the spark is not internal, good luck
with fueling the passion).
Secondly, both players, and their supporting parents quite often need a tracking mechanism, which ratings are, to measure progress. Trophies won, quite often at
in-school tournaments, and small local clubs quite often do not accomplish this, for
being “king of my hill” may or may not be relevant in terms of comparing to a larger
setting. And, sometimes, motivating these kings of their hill to higher accomplishments
is difficult because of their current satisfaction levels.
When a chess player learns, and grows and improves over time, there are certain patterns that have been documented by chess coaches and by those who understand the rating system.
First, a player who improves will not necessarily win more; If one is 800 strength, and improves to 1200 strength, one is still highly likely losing to 1600s.
Second, such a player may simply be losing later in the game. The quality of their game will improve, but now instead of losing with a tactical error on move 20, they are losing an inferior endgame at move 45 - a situation that they weren’t capable of reaching previously.
Third, there is a tendency for players to “jump up” in rating; because of the situations described above, as a player learns their rating will tend to be stable, or perhaps even to decrease slightly (due to confusion) as they try to implement the new skills they are learning. Once these new skills are learned, the player will “suddenly” win games that they did not win before, and the rating will jump.
Now, let’s consider the above learning process in the context of Rob Jones’ suggestion - that we should play up the rating. We quickly see that such an approach is pedagogically irresponsible to the student, since they will be misled into thinking that they are not improving when they are, and that they are improving when they are not (since the rating jump comes at the END of improvement.)
These are things that good, long-term, reputable coaches know. We may disagree on many things, but we understand that Ratings are a TOOL to help us discern overall improvement over time, ratings are not a goal in and of themselves, and to hold them up as such to the student is inappropriate.
It appears that Mr. Jones’ mantra is short-term “marketing” (rather than branding) irrespective of the damage it may do to a student and their long-term ability to enjoy chess.
An old friend from Northwestern once lamented that he would win only when his opponent would blunder. I replied that no one can win unless the opponent makes a mistake. In turn, he responded that he would win because his opponent dropped his queen, walked into a mate, or allowed a family fork, whereas I would win because my opponent had a backward pawn, or less space, or a bad bishop.
That friend was rated 300 points lower than me. Now he’s 200 points higher than me.
Ratings are one indicator of improvement. Unfortunately, they can be deceptive, too. A fluky win may cause a rating spike. Subsequent results may cause the rating bump to disappear. While we look at ratings, the concern is for trends. Ratings graphs can give you some information, but not all about the player’s level of improvement.
Besides rating, I also look at individual games of the player. One thing I look at is who is the highest rated player that the student has defeated. Then we look at the quality of that game and similar wins. We look for patterns of good moves and mistakes in both the wins and the losses. This qualitative assessment can be more important than the quantitative assessment based on rating. As the player improves, good games multiply and lead to a sort of “crtitical mass” that is an indicator of an upcoming rating spurt. A player who stops learning or gets lazy will start showing patterns of sloppiness, impulsive moves. If it happens in one game, not a big problem, but poor behaviors tend to fester and spread to all phases of the game, leading to rating drops. Not just young players, but all players should look at their games objectively to find out what is going on, talk to a strong player or coach, and try to correct their play. Poor results and rating drops are motivation killers for many people because they focus too much on their rating.
Ratings under 1000 are so volatile that it is very difficult to say whether a player rated at 400 is really all that much stronger than a player rated at 300. Because of the K values in the rating system, one or two games can lead to big surges up and down. A rating can go up because you are “bunny bashing” and not playing strong enough opponents to give a challenge. What I do not like to see in scholastic tournaments is when a player gets slotted into a rating group section and get stuck playing the same players. I would prefer my students to play up as much as they feel comfortable doing. Sometimes they have to be pushed and take a good drubbing to find out where they really are and see what it takes to beat higher rated players.
Kids, some very strong ones I might add, have told me that the 1000 mark is an important psychological barrier for them. After they break 1000 or 1200, it is time to wean them off the fast time controls and prepare them for adult styles of play. They are more at ease and ready to work harder after they hit this level.
In this response, I choose to respond only to those of material, rather than subjective, and have deleted the
subjective.
Now in regard to his points: Of course a person who improves will not necessarily win more. This is
due to the fact that quite a few tournaments, and most in the DFW area, are sectioned according to
ratings groups. A 1-3 performance in a higher section may, in fact, be a stronger performance than
a 3-1 performance in a lower section. Certainly, most of the coaches I know do encourage their
kids to “play up” when they are eligible to do so. There are other factors as well. For example,
a youngster who earns a 1400 rating on primarily G/30, may truly struggle at say, G/90 or longer
time controls, as they have not developed the necessary end game skills to compete with those
of similar ratings, who have been playing at these longer time controls. I know of no coach who
encourages their kids to take their opponents rating into consideration in how they play the board.
The idea that a player who improves from 800 to 1200 is still likely to loose to a player rated 1600
is a no-brainer. Of course. But, the ability of the 1200 compete with the 1600 is far stronger.
And certainly, victory is more within the realm of possibility. There ARE significant differences between players rated 300, 600, 900, 1200 etc. Most chess educators I believe recognize this fact.
Each becomes their own benchmark towards progress. Labeling lower rated youngsters abilities in
a negative way, by way of comparison with players who are stronger, rather than celebrating each
step upward in a positive manner, accomplishes little positive.
Further, as students learn new tactics, openings, etc, and try to implement them into games, they
quite often, will suffer the effects of error, and thus ratings loss. The successful coach is there for
them in these times, carefully explaining that the path “up” is not always an upward diagonal.
Sometimes little bumps are part of the learning process. All err, the difference is the reaction afterward.
Those who improve are more apt to learn how they can improve, those who stay stagnant generally are
those who mope or whose passion for the game is not as great.
Last, ratings improvement hopefully does not happen at the end of a learning process, but is continual
throughout this process. Playing frequently, and playing with principles is a key to this process.
If you graph the ratings improvement of most players prior to their peak rating, the graphs will (very roughly) resemble a rising flight of stairs. There will be flat areas, then spikes of increase over shorter periods than the flat areas cover. These flat areas can be thought of as times of apprenticeship, where the apprentice is acquiring and assimilating tools in their chess toolbox. They’re learning how to beat a certain level of player consistently. If this player plays and studies frequently, there will come a point where the tools are no longer clumsy and mechanical to the player, but instead flow naturally as part of their thought processes during a game.
That epiphany is what I often describe to parents as “The Click”. It’s the time when players are really able to use those tools effectively as a natural part of their games. Those periods right after “The Click” are when you see those sharp spikes in the rating graph. This player will rise quickly, but then level off. That next, higher plateau is when the player runs up against stiffer opposition, who have all the tools this player has, and maybe more. Now, the player must go through another period of apprenticeship, where he has to learn some new and advanced tricks. The process then repeats itself.
You may have heard of players talking about the law of diminishing returns. As you study and improve, those plateaus can go on longer, and the spikes are shorter. You often have to put in more time and effort in the plateaus for less and less rating increase in the spikes. Past a certain point, most players just lose interest in putting in that much effort for what they see as too little gain, which is often when they stop playing.
It is very rare to find a player whose ratings graph resembles a smooth hyperbolic curve over the start of their chess career. It is even rarer to find a player whose ratings graph resembles a smooth parabolic curve over the same period. These are generally prodigious players, who really are the exceptions that prove the “rule” of incremental ratings improvement.
What Boyd points out concerning the process of learning chunks of information and improving as steps upward with periodic plateaus has been studied and documented in many fields. I have looked at ratings charts of hundreds of promising juniors and noted this progression. In only a few cases was there a huge spike , or what I call a “rocket,” of ratings improvement. At a certain point there is always a slowing down or plateauing of the rating until another surge occurs. If the player is satisfied or stops pushing himself, the rating curve stabilizes in a narrow rating range or degrades over time.
A particular interesting thing happens to some students who start out with a high initial rating and then do a “U” , that is have a drop in rating, when they get a coach. While the player is learning new things, breaking bad habits, or adapting to the style of play and openings the coach suggests, he or she may experience a rating drop until the new information gels. Then the rating goes up again. This can occur when the player is forced to drop junky openings that work against kids for a more solid repertoire, for example. Or when the player is asked to play for more tactics and sacrifices. Many times ratings drop when the players transitions from Game 30 tournaments to events with slower time controls.
As has been pointed out in a number of sources, but especially in a book by Malcom Gladwell, expertise in any field takes about 10,000 hours of study and directed practice. This has been shown with music virtuosos, chess players, scientists, and other masters of their arts. Prodigies are few, and even these work very hard over a number of years to excel. There are no quick fixes or short cuts to improvement. It takes dedication, motivation, and hard work to be successful at whatever you want to do, especially if you aim to be one of the tops in your field. Sometimes you have to calm a parent down when their little angel does not become a GM by suppertime, in spite of their expectations.
I cannot tell you how much I appreciate this post. Thank you.
such a curve without a “dip down” is rare. What does cause
issue with some parents, primarily, is this. As a player improves, esp, out of the provisional arena, then to 1000, and
then 1200+, the K factors change. I have had parents wondering why the next “jump” from say 1100 to 1800 is so
difficult when their kid went from 400 to 1100 in three months.
Those at all familiar with ratings understand this is the
difference between swimming across a pond, and swimming
across the English Channel.
Kids, some very strong ones I might add, have told me that the 1000 mark is an important psychological barrier for them. After they break 1000 or 1200, it is time to wean them off the fast time controls and prepare them for adult styles of play. They are more at ease and ready to work harder after they hit this level.
Above quote from: tmagchesspgh
I think a better way to phrase this is that:
a. It is better to ADD, not “wean them off” faster time controls,
such as G/30, and prepare them for longer styles of play
such as G/90, 30 sec increment, for example. Add is far
better than “wean” as the number of longer time control
tournaments, and the number of available games, is far less
generally than the short time control formats. Certainly, most would agree that four monthly G/30 tournaments, with
one longer time control event is the ideal, as opposed to
the one longer time control event alone. Each game when
the proper focus and principles are used, can easily be a
chess lesson unto itself.
b. The reason for youngsters rated below 1000 NOT to play in the longer regular events,
is that most of the sections in these events are designed for established and skilled
players. To force a skilled player to play rounds in a slow time control event with
players basically lacking the skills to compete, is simply rude. Neither they, or their
opponent generally gain much whatsoever from this experience. Esp with youngsters
who play poorly (by comparison), diminishing the value of the event for all concerned.
(For example, I have seen parents put their kids rated U800 into U2000 sections in which
to play!!–not as big a deal for the shorter time control events, but a real drag in longer)
I haven’t gone and double checked, but IIRC, Elo also documented this in “The Rating of Chessplayers Past and Present.” It’s a very well known phenomenon. As I said in my initial post, good, experienced instructors know this.
First, I would disagree with using that kind of external motivation (and to what end?) but secondly my comment was about giving candy to kids simply because they loved candy - without regard to the impact it has on their health - simply because they love candy.
BTW, I recall a promising student who, in his second year of scholastic tournaments had a sudden drop off in performance. For two months we were trying to figure out why until my wife discovered that the boy - who was not allowed candy at home - was sneaking off to the tournament concession stands to buy and consume a couple of chocolate bards prior to the first round. Once we stopped this process, his prior promising results immediately returned.
While ratings are “A” tool they are not “THE ONLY” tool to measure improvement. A good coach can watch the quality of play and see what is working and what is not over several games. Moreover, ratings have their faults. As noted, improvement comes in pieces and only when enough pieces are in place does “coordinated improvement” along with winning and a ratings jump occur -thus ratings increase at the end-part of the improvement process. Secondly, as a player continues to improve they will plateau and often even get a little worse. If one uses ratings as the primary tool one will always be behind the process and will - without a deeper understanding - misunderstand the process. This is something good coaches know, which is precisely why they do not stress ratings.
As we note above, improvement leads to changes in results - but not necessarily immediately. Ratings change ONLY WITH changes in results. Thus, while learning may be a continuous process (but not necessarily continuously increasing/ improving of strength process) the calculation of ratings is definitely discontinuous, being reliant upon individual game results, which are also discontinuous. Hence your statement is necessarily incorrect.
I was thinking of players whose rating improvement one could graph to illustrate the point of how unusual continued, uninterrupted ratings improvement is.
If you look at the ratings history of Sam Shankland, you will see an approximate hyperbolic curve at the beginning of his career. Even with the younger Shankland, though, you can see the occasional dips and plateaus that correspond with his “apprenticeship” periods that I mentioned earlier. However, his ratings graph on the whole is amazing to look at.
Perhaps the best example of a parabolic ratings graph can be found with Bobby Fischer. Unfortunately, I don’t know of any place that has his complete USCF rating history available. What I do know is that his first rating, in May 1956, was around 1700. By the May 1957 rating list, he was around 2250. Such an increase by itself is remarkable. When you factor in that he won the first of four consecutive US titles later in 1957, you don’t really need to see a graph to understand how quickly his rating must have gone up.
