Hi, I am looking for help for a claim I’m making in my forthcoming book. I want to know if someone rated 1050 USCF could defeat 90% of the youth who compete in scholastic chess tournaments. I would like someone to answer who has access to recent rating statistics.
Specifically, I am trying to verify a claim by Joseph L. Bell based on his research on 2004 rating data: “Bell recommended that Chess merit badge counselors be rated at least 1050 USCF. Bell estimated that someone with that chess rating could defeat about 90% of the youth who play in scholastic chess tournaments.” In this quote from my forthcoming book, Bell refers to Bell, J. (2011). Recommendations for chess merit badge counselors. Retrieved from meritbadge.org/wiki/index.php/Im … elors1.pdf.
If you go to the PDF link, Bell used 2004 ratings. I would like to know if the ratings distribution today (2015; or 2014 if 2015 is not available) would support the claim of a 1050 player being able to defeat 90% of scholastic chess tournament participants. If not, what would the percentage be instead? Maybe 80%? Please help! Thanks very much.
Well, 90% of USCF-rated scholastic players, or all scholastic players? That’s a big difference. In my scholastic tournaments (1997-1999), there was a section for rated players and a section for unrated players. In an unrated section, I went 5-0 and won a free USCF membership. When I played in my first USCF tournament a couple months later, I got my first rating of 1137. So in that unrated tournament (3 sections), out of at least sixty kids, I was probably the only one who was (maybe) 1050 strength.
MSA puts 90 percentile for juniors (which I think is defined as U21) at 1388 which is significantly higher than 1050, though I’m sure it will tumble a few points if you got rid of 18, 19, and 20 year olds.
I’m going to go a different direction. Knowledge of chess is different from playing chess well. I can think of at least one good NTD that I’d have full confidence in as a merit badge counselor, but who has only a three-digit rating. I can think of multiple CTDs, LTDs and SrTDs in the same boat.
The only section that possibly deals with the playing strength of the counselor is the one involving going over the Scout’s games. Even that can be looked at as more of a score-keeping verification than as a teaching session (granted, I often do some analysis as well in that section even though that doesn’t seem to be required).
You might see if its possible for Mike Nolan to run some sort of an analysis that shows what the initial rating was for players below age X (take a pick). One could then argue that if anything initial ratings should be at or higher than the general population for that age, since they would be likely to have more chess training than the typical player who has never played in a rated event. (Coaches/parents would likely teach a novice something prior to their first rated competition, and such teaching would be likely to occur at an equal or greater rate than someone who never plays in a rated event.)
Then you could compare the 1050 rating against that initial rating and test the statistical reliability of your assertion.
Thank you Jeff and Kevin for your responses. As I wrote in my first post, “Specifically, I am trying to verify a claim by Joseph L. Bell based on his research on 2004 rating data.” So Jeff might be right about tournament directors rated 999 and below being qualified to counsel the Chess merit badge and Kevin has an interesting analysis idea. However, I was asking a more limited question about the veracity of Bell’s claim. Kevin, Mike Nolan emailed me some statistics which were helpful for my analyzing what Bell wrote. Again, thanks Jeff and Kevin for taking the time to respond! I appreciate it.
I dare say that scholastic chess has changed since 2004. The rating of the very top players and the number of advanced kids have both increased at a national level. There were about 50 masters under age 21 a decade age. Now there are 100 rated 2257 and higher, and likely 150 masters total. The number of youth masters (defined under age 21) has tripled in a decade!
Yes, my analysis doesn’t properly address the stated question, which addresses the 90th percentile of the youth who compete in scholastic tournaments. However, the growth in competitive young chess players is not limited to those who are destined to reaching 2200. We only must ask if the overall number of scholastic tournament players has also increased to match the growth of competitive players (whether you use “masters” or “1050 rated”). I am pretty sure the number of USCF scholastic members has increased in the last decade, but it has not tripled.
My hypothesis states that the 90th percentile in 2004 is weaker than the 90th percentile today. That would mean the claim by Mr. Bell would stand a better chance of being true in 2004 than 2015.
Hi, Michael! Thanks for your response. My understanding (from a PM from Mike Nolan) is that being in the 90th percentile is not the same as winning 90% of games. Bell’s claim was that a 1050 player would win 90% of games against scholastic chess players. Maybe Mike Nolan will respond here in the Forum to clarify the difference between 90th percentile and 90% winning rate. He might also be able to answer your question about the 90th percentile change from 2004 to 2014.
OK, I misread the original question, so here is a tweak of my original method that should serve your purpose (with a few provisos).
If Mike can provide the average post-event rating of all scholastic participants (possibly below a given age) during a year (2014 for example) then you can calculate the winning expectancy between that average and 1050. That winning expectancy is an estimate of the percentage you describe.
Your statement is then: “A player rated 1050 would have an estimated score of XX% against scholastic participants (under age X) who played during 2014.”
“Could defeat 90% of the youth” is a different statement from “could win 90% of the games against the youth”. I can defeat pretty much any 1800 but would only take about 70-75% of the games (with a 30% draw rate that would mean winning about 55-60% and losing about 10-15%). I can defeat 2200s (admittedly not often, but I have done so). For a 1050 to win 90% of the time you’d need the field to average around 500-600. For a 1050 to be able to defeat 90% of the players, you’d need 90% of the field to be under, say, 1300.
It sounds like you are focusing on winning 90% of the games, so you must have some additional information that gives that interpretation as opposed to being able to win (at least some of the time) against 90% of the players.
To be in the 90th percentile among junior players, a young player will be rated at about 1385 today in 2015. I am not sure what rating would hit the 90th percentile in 2004, but I am willing to guess it is a bit lower than that. There are quite a few juniors, over 4000 of them, who are above 1385 in USCF rating. This pool of stronger young players grows every day.
I have had a number of students in the 500 to 1600+ range in USCF rating. Their ratings are pretty volatile. When they are above 1400, they handle players, both adult and junior, below 1200 easily and can score against 1600+ players. The players who were at around the 1050 mark did very well against players who were 200 or more points below them, but really struggled to beat 1200+ rated players. While a 1000 rating seems to be a threshold rating that many juniors strive for and consider a turning point toward “serious” chess, it seems to me, based on observation of the performances of my students, that a 1200 rating is the line where the young player starts to be more effective against higher rated players while dealing with his lower rated peers efficiently. Those performance observations are based not only on the statistics but also of the quality of their play and conceptual understanding, which are qualitative measurements. A player who is 1050 rated has too many holes in his play because of a lack of knowledge and mechanical deficiencies in his ability to analyze positions. He still has tendencies toward impulsiveness, of moving too quickly. A player above 1200, and especially above the 1400 mark which is slightly above the 90th percentile, is less impulsive and is beginning to approach the game with greater depth because his tactical and planning skills have increased. Because of these factors, I have a hard time believing that a 2004 or a 2015 junior player with a 1050 rating would be able to beat 90% of the scholastic community who is USCF rated on a consistent basis. He would be doing well to be scoring 50% of the time given the range of juniors who are available to play. That is what the stats of my students in the 1000 to 1100 rating show.
Of course, there are 1050 players and 1050 players. By that I mean there are players who have earned the rating by playing a broad range of rated players versus those who have that rating by playing “bunnies.” A player who plays only players rated 750 and below might have a 1050 rating but it is “soft.” If he is not playing higher rated players, is playing only lower rated players who hang pieces and fall for back rank mates frequently, and is not playing in tournaments with adults, his rating is hollow as the pool of players he is competing against is not robust. I have seen a number of players who have such “soft” ratings who are sheltered and do not venture out of the safety of their niche.
BTW, given what counselors are expected to teach the scouts (see link in OP), and the requirements necessary to attain the merit badge, an instructor who is below 1000 rated can still teach the group. The skills being taught do not require a master or expert. Part of the program also requires the person trying to earn the badge to understand aspects of directing a tournament. There are a number of low rated tournament directors who would be effective in showing the kids how to play and how to direct simple tournaments for their peers.
I have found that lower rated instructors sometimes do a better job of teaching because they are closer to the subjects and understand the struggles of the newbie better because of their own experience. A higher rated player might take too much for granted. It is easy for a person who is lower rated to be anxious that he will make mistakes while instructing kids. The kids really don’t mind or look on him as being poor. They know he is trying and are energized and motivated when they see that they can solve a problem that the instructor struggles with. It shows them that they are improving rapidly.
Thanks for the recent posts! The original Bell (2011) quote, from the PDF link I gave in the original posting, is that a 1050 player “would likely win a game of chess against 90% of youth who play in scholastic tournaments.” I may have misinterpreted by paraphrasing Bell here as meaning that a 1050 player “could defeat about 90% of the youth that play in scholastic chess tournaments.” When I read the PDF, I thought “win a game” meant “consistently win” not “win once in a long time.” I appreciate everyone’s posts and think that this part of my book will definitely benefit from what I’ve started thinking about because of this Forum.
Well, can the merit badge counselor mate with king & rook vs king, and does s/he really understand KP vs K? The rest should be a breeze for the 800-rated adult.
I recently had the pleasure of assisting my friend Mark Frank at a Merit Badge workshop. Lots of fun! I started with White Kg6 Rf7, Black Kg8 (White to move), introduced Zugzwang (comparing king vs king in checkers)…
Let’s say Joe Blow could defeat 90% of the scholastic players who have played in at least one tournament during the past year.
What if this same Joe Blow visited each scholastic tournament held during the past year? Would he be able to defeat, on the average, 90% of the players present at each such tournament? I doubt it very much. It would probably be more like 60%, or worse. At each tournament, the more active players would, by definition, be present in greater numbers than the less active. And the more active players would likely be stronger than the less active.
So you need to define your hypothesis a bit more precisely.
is to win 90% of the time against opponent “X”, it generally implies that you are to be 350-400 points above someone. The median junior rating is under 600 (meaning 50th percentile is like 573 or something like that)
a 1088 could win 90% of their games against a median child, and probably even 90% again against 60% of the players. I dare say that at the really low ratings (where the so-called k-factor or variability from one result is higher), this may not hold as true, but if you took every child, from 1-100, (worst to best), and counted how many games out of 100 a hypothetical 1088 scout player would win it would look like this.
Anyway, that might be a more rigorous method. The question is does the huge number of 99%'s early on make up for the bottom 10%? I’m not sure but it probably falls just shy.
His whole premise is off though. The chess taught in the MB is minimal really. Teaching the terms is c3ertainyl not based on chess skills, solving mate in one problems is pretty simple stuff. the requirement to play three games, take notation and go over the games with the MB counselor does not imply any great analysis on the part of the counselor. Indeed when I’ve taught the MB - and I’ve had some 50 to 70 in the last two years - the students are most frequently basic novices. The notation is barely sufficient to lay out the game - if it is that good.
BSA MB work is more designed to make someone familiar with an area and not proficient in it. Hence the MB for architecture, aviation, medicine, dentistry, nuclear engineering don’t attempt to create licensed professionals but rather may spark an interest in that young person for further study. That is why there are some 110 MB with only 13 specific ones required for Eagle (it used to be 12 specific ones). There are 21 MB required for Eagle in total and the other eight are elective. The most I’ve ever had from one of my scouts was around 55.
So one doesn’t need great chess skills to be the counselor for this MB. One ahs to have a familiarity with the subject matter. One does have to 1) enjoy working with young men 2) understanding something about scouting and 3) have some teaching skills. The last is really important if this is done in a group setting.