Reforming the NM title to avoid corrupt gaming of the system

All Things Chess
68

I think it can be both @relyea. There can certainly be instances of players conspiring to help the almost-master (or almost-expert) pick up a few points, even if only tacit…“let’s play one of our small closed events among friends and colleagues…nudge nudge.” Also, players can play well timed games as a house player and take their pick of opponent as an assistant TD, for instance.

And again, to clarify, I am relating to a specific instance. Perhaps it isn’t generalizable as effectively a case study, though it seems this sort of thing isn’t uncommon. It’s certainly a thing that can happen and that I request shouldn’t be dismissed. It’s an instance of an affiliate and (assistant) TD abusing their power, but perhaps in ways where there is plausible deniability or simply no explicit rule against it, hence my search for some change. So it’s not chiefly about the minutiae of expected ratings gain calculations, but maybe how this can be blended with artificial arrangements to boost one’s rating, particularly when they’re close to that new threshold.

Keep in mind that I’m referring to solely getting to that new level, not across hundreds of points where one might attempt to win tiny incremental points. I agree that this would be incredibly difficult, indeed a sisyphian task. I’m referring to those instances of hitting new levels when the player is very frustrated but close. For instance, they may have the inherent strength to achieve 2250 but are stuck at 2175. Because, regardless of strength at one’s best, it’s damn hard to reach new heights like 2000 or 2200, and I think a big reason is consistency. This probably explains why many IMs never become GMs. They may be able to play like a GM at their best but need to do it over dozens of games to hit 2500, apart from a few stellar arrangements.

Now, I similarly have issues with the FIDE closed tournaments where they hand pick weaker titled players as point-bleeding punching bags, also an artificial scenario designed to reach new heights on the international scale. I just don’t like unfair, anti-competitive situations like this, so it’s all in the same vein and maybe that helps to explain where I’m coming from. If I ever get to IM – and I can beat them sometimes, but similarly not consistently enough – I’d want to work my butt off for it. I think we should all be held to the same rigorous standards.

Something concrete that I want to broach is the concept of (casual) arranged matches between two individuals. I think I’ve mentioned here that I think it’s unnecessary to have these matches, which could otherwise be casual, be rated, and that this should only be to determine champions – regardless of what level, e.g., club, state, etc. Should these matches be allowed? Why do these individuals need to be removed from the broader rating pool to compete for points only among themselves? Can’t they just play each other in the club? This is simply a more concentrated version of insiders playing each other, where these effectively tacit (or not) agreements for the almost-titled player to hit their plateau occur. Because why not? They all have a stake in it. And people cheat. This is just another level…and often not really “cheating” because, well, it’s allowed, or the ambiguity based on unclear guidelines may allow sketchy situations to glide by the ethics committee, for instance. So basic measures, like banning such arrangements, seem sensible. You guys are talking about wonky rating calculation changes, but how about the sorts of changes re: arranged and insider matches I’m referring to – has anything been done? It doesn’t seem so.

I don’t know exactly where the debate is now on the closed FIDE title hunting tournaments are. Maybe my views here align with a broader movement against these.

69

As noted upthread, the expected performance formula only looks at the difference in ratings.

So a 600 player is expected to have the same performance against a 400 player as a 1600 player would have against a 1400 player or a 2600 player would have against a 2400 player.

I don’t think the expected performance formula has changed much since Elo’s days, and I doubt he had access to millions of games back in the 1950’s like we have in the current database to attempt to incorporate actual ratings instead of relative differences. It’s a simplification, but probably a reasonable one.

Moreover, the rest of the formula, especially the bonus formula (which US Chess has but FIDE does not) probably causes the ratings to match up with the expected performance curve over time (within limits as Tom Doan noted upthread.

Here are the graphs I generated back in 2018 showing expected vs actual performance in regular-rated games by players at 3 skill levels: 400-1000, 1400-1999 and 2000-2999.

Until you get to experts and above, the graphs are very similar, and at relatively small ratings differences all the graphs pretty much sit on top of each other. And in all 3 graphs, the lower rated player does a bit better than expected, for the reasons Tom Doan outlined upthread.

800×800 9.66 KB

70

Low-rated GM’s are only useful for achieving the count of opponent GM’s. (And foreign-flagged 2420 GM’s have been popular forever for that reason). It doesn’t help with actually achieving the norm since the opponents must still average 2380 (so 2420 is barely useful) and the performance has to be 2600+ (where low rated opponents force a very high score).

If you have any evidence (or very strong suspicion) that someone is throwing games or submitting unplayed games to help someone achieve a higher rating, you should forward specific information to ethics or TDCC or ratings. No one is going to act on vague “I’ve heard things”.

71

For better or worse, USChess has long determined that having as many games as possible that are played, be played in a rated fashion. I assume that a huge portion of the reason is that we think that the results of games, any games, is predictive and we want our rating system to be as accurate as possible.

[Delete 5834 words of blather related to rating floors.]

The reason we have match restrictions, and we have many, many more restrictions on matches than any other type of event, is not to deter fraud, but because the more concentrated the opponents, the less predictive the results generally.

Re insider tournaments, events where players are disincentivized from succeeding over the board are covered under the fraud provisions. And it doesn’t matter if my three buddies who happen to be floored masters and I get together to play a quad at my house with the understanding that I’ll win all my games or if an out-of-the-money GM approaches me before the ninth round of the World Open offering to lose for $5000, it’s already covered by the various Codes and will not always get caught, but is still difficult to get away with, and I’ve seen nothing above that would fix that.

All of Mr. Bennett’s suggestions above appear to have the dubious benefit of making things much more difficult for players who are playing “fair”, but do absolutely nothing not provided for in the current rules to forestall dishonest players.

Regarding FIDE norm round robins, I point to Mr. Doan’s answer above and note that like some of the other situations above, it is much more difficult to execute than Mr. Bennett seems to assume, which is to say there just aren’t a ton of GM norms being created in events with 2430-rated GMs.

And few of us can consistently play up to our peak performances.

1 Like
72

That’s the Bernoulli Theorem: The relative frequency of success in a sequence of Bernoulli trials approaches the probability of success as the number of trials increases towards infinity.

And the more games you have, the more ‘anomalous’ results you’ll see, regardless of whether they’re caused by mental errors, tiredness, or just a player eager to get his last round over with (or lock up first place) and head home so he takes a draw in a favorable position. I’ve had a player resign against me in a position where I felt I was dead lost.

73

I’m not sure if we’re agreeing or disagreeing, but I’ll point out that the ‘anomalous’ results aren’t really all that anomalous. I mean it is the same players who are most prone to mental errors, or getting tired, or playing tournaments far enough from home that they’ll want to take a quick(er) draw to get on the road.

In other words, players who have a tendency to fight fiercely for every half point are likely to have better results and consequently higher ratings than those eager for a quick handshake at any opportunity.

US Chess ID 12498726. US Chess National Tournament Director, FIDE and ICCF International Arbiter

75

People’s safety has to be in first regard when holding a tournament. Why are you guys so immersed with Safe Play/Safe Sport. I’m totally in touch w/ ground on this. I witnessed this at a tournament(s). So much for being grounded in reality.
By the way, the tournament that I referred to was a State Championship. Any thing goes huh (who would have thought).

78

I grew up in Delaware, and we had a little bit of the “Alaska” problem. I remember it was a sensation when an Expert moved to Wilmington. I had gone to college long before any NM moved or achieved the title there.

It was somewhat easier to gain rating points by going up to Philadelphia, but they were tough there. I always gained a lot when I went to NYC. But we couldn’t afford to go there much, even though it was just a 2 1/2 hour drive on the turnpike. I remember asking our high school principal if we could be subsidized to go up to the National High School championship, which we might have done OK in (we went undefeated in every high school match for 3 years that I was there) but he said no, all his resources went for our bad football team.

I only got Expert and then most the progress to NM myself by playing in big places. There are local rating pools, groups of people who play each other with limited outside contact, and their ratings could drift higher or lower than the national average for a given playing strength. In every case I know of, those local pools are tougher than the big cities, where masters and GMs are adding their points into the system and you can get some of them. I doubt that someone in Alaska can make NM in that rating pool, no matter how good they get. And if they do get NM that way and it isn’t obviously fishy, they might be a real threat for some higher titles if they spent some times in big city chess.

The talk of subsidizing someone to make NM is nice. But where would the money come from? There’s not a lot of money to be made from the NM title even for the title holder. But apparently the NM title continues to have some importance in the minds of some people, and I’m not aware that scandals have discredited it.

I can think of a lot more troublesome things going on in the world of chess now, like the questions that are now unavoidable about the major online platform on which lots of money prizes are awarded. I wouldn’t want to make it even more difficult for someone in a small town or a distant state to dream of winning an amateur level chess title.

1 Like
79

Elevation in status, we are now the name of a problem! :laughing:

Joking aside, it is hard to say for sure. There is only on GM from Alaska, and he did just that: moved away to play ‘big city chess’.

At the last tournament where players from outside came to play in Alaska this past year (there was only one tournament last year), several players came from Illinois. The strongest one was rated 1800 ish.

There was a five way tie for first place. One was the strong player from Illinois, three were unrated Alaskans, and one provisional alaskan rated about 1600. Seems we did okay for having no tournament experience in the past 17 years. (We being Alaskans, not me personally, I came in 16th.)

I am not sure if that proves your theory, or breaks it, but it is interesting.

80

It has been a long time since I have posted. Life, age, and being extremely busy promoting chess has indeed taken its toll. However, I must agree with the very great and
astute NTD Jeff W. in this:
As far as TDs gaming the system goes, such gaming in the past had resulted in retraction of the NM titles, correction of the ratings and suspensions applied to the TDs. So I also know of people that have done it, but they have not gotten away with it
)))))))))))))))
Many years ago now, at a major event I was directing, A certain discredited TD came
to me and asked why US Chess was “persecuting” him. As in removing more than a
half dozen tournaments in which he personally played in defeating players lower than
the 400 pt match threshold, or who did not have enough games played to qualify for
match play, in supposed regular tournament pairings. Strangely, I did review these events when I heard about this from various sources, and certainly did agree with
US Chess that tournaments which have more byes than actual games played certainly
do create attention, at the very least. And this particular individual, a TD, was “inching”
towards becoming an NM which each submission of these events. For as a teacher
myself, I do understand the importance of such titles in charging higher teaching fees.
– there can be a $$$ incentive to submit these illicit bogus events.
And what looks like a skumk, smells like a skunk probably is ot a cat with a white stripe.
My thoughts,
Rob Jones/Senior TD

81

I am aware of three situations in the past 20 years where multiple events have been deleted from the system because of concerns over their validity. There were a few others that were a bit iffy but allowed to remain in the system.

Nearly every month we have to delete at least one event because it was submitted more than once, usually because two different people at the club submitted it.

A few years ago I tried running a query for events with an unusually large number of players with byes or unpaired rounds. It was difficult to sort out the ones that were suspicious from the ones that were just non-Swiss events like a ladder. There were some scholastic events where players would go into the media center over their lunch hour and play whoever was available, with the results submitted at the end of the week. This resulted in some very odd looking crosstables, but there was no manipulation going on, just a variable pool of players from day to day.

82

Just change the requirement from 2200 USCF to 2200 FIDE for USCF NM title. How can you be a NM master without attending regional and larger tournaments?

  • FIDE enforces real time controls for their standard rating
  • All the feeding goes away for the < 1800 USCF group as they will not have FIDE ratings

It also doesn’t make sense that all of USCF titles are norm based except this one. USCF should concentrate on the amateur ( < 1800) titles and leave the big ones to FIDE (which still has it’s problems for “special” tournaments for IM + GM titles).

Want to identify a fake USCF Expert or NM?

  • FIDE ratings > 100 points from their USCF rating
  • All their norms and rating points are from short time controls events local events
83

The title of National Master predates the norms-based titles by, oh, about 80 years.

FWIW, OLM also is not norms-based, nor is Senior Master (which is separate from Life Senior Master-the norms-based title.)

84

Not too many things from the 40’s that couldn’t stand a tweak or two.

USCF Life Master is kind of like a who cares with the widely recognized FIDE Master/IM titles. I doubt most of the USCF members in their 40’s or younger would even know it exists.

For the OLM title, just rename it Stockfish master.

If i’m looking for a master for coaching and see a USCF NM title, am i getting the equivalent of FIDE CM or USCF Expert? Sucks for the folks who earned their title the old-fashioned way that their hard work has an asterisk.

There is reason in the larger tournaments the top two sections are FIDE rated.

85

Having more than one section of an event be FIDE rated is a relatively new trend, here’s what I see since 2005 for events that had one FIDE rated section and events that had multiple FIDE rated sections.

What’s interesting about that is I recall a discussion a decade or so ago about how FIDE was making it hard for US organizers to run FIDE rated events and people were saying the number of them would decline. Nope!

year one mult
---- --- ----
2005 102    2
2006 101    3
2007 101    4
2008 125    6
2009 142    7
2010 171    6
2011 215   19
2012 230   21
2013 200   22
2014 229   13
2015 188   28
2016 176   31
2017 176   44
2018 211   43
2019 249   69
2020  77   20
2021 118   73
2022 204  125
2023 258  206
2024 143  111

As someone whose US Chess rating peaked at 1651, NMs and FMs are levels I can only admire.

1 Like