Reforming the NM title to avoid corrupt gaming of the system

All Things Chess
21

Having a geographic requirement seems troublesome, there are players who for one reason or another cannot easily travel out of state or afford to.

Looking at a fixed number of games prior to the awarding of the NM title almost certainly will include some but not all of the games from the first such event. Similarly, looking at a fixed number of events will result in a variable number of games, which will complicate analysis.

I don’t know if posting the list of players who earned the NM title in any given year creates any issues, I would prefer that be cleared with the office.

Also, keep in mind that any FIDE rated players in their first US Chess events will bring in their FIDE rating as part of the initialization of their rating. That’s true even for Grandmasters, and the NM title isn’t awarded until the player’s rating is established.

1 Like
22

Yes, I think I’ve come to agree to some degree about a potential geographic requirement, @nolan, esp. if the financially burdensome component can’t be readily overcome via, for instance, grants (or old school “patron” type support for the artist) for near-master players. Still, I think this is worth looking into as a possibility with U.S. Chess Trust. And I did mention the open tournament requirement – outside of one’s own org that they run or work for. What do you think?

Actually, come to think of it, even without drilling into the numbers too much, I look back at the data you’ve provided:

Looking only at wins, only one had a best win against a player under 1700, another 9 of those NM’s had their highest wins against 2000 players in the 180 days prior to earning the NM title,

Another 18 had their highest wins against 2100 players in the 180 days prior to earning the NM title.

Now, a best win over 6 months against a 1700? That must be pretty overtly “gaming the system.” I don’t see how that can be justly, by any objective measure, obtaining the title, clearly suggesting the need for reform to avoid such scenarios – or again, at the very least, having ethics action taken. Unless there were really no way this player could have played a 1700+ over six months, but it’s highly dubious to me (and perhaps blunderous to allow it). I think this case should be inspected and am not sure how that was approved. The only explanation I can imagine is that it was permissible by the rules, which don’t apparently cover this sort of scenario. Or in a very small state, but still, I think 1700+ could be readily found, no?

Further, you also mention nine new masters who, over six months, played 2000s highest. That’s where digging a bit would be interesting; something uncouth may be found there. Six months is a long time. So maybe they played a couple vs. 2000s and it may be found that similar arrangements to those ones I mentioned had occurred, where they generally played much lower rated. Worth revisiting by the board or a committee in my view to ensure nothing corrupt occured there.

To address @alaskalinuxuser on how “the numbers don’t lie,” there may indeed actually be some unfortunate truths in there – even if, to be fair, the vast majority of titles were legitimately obtained. But let’s say 5, 10 corrupt cases…should those be allowed?

It seems that you’ve largely ruled out the main points that have been addressed for reform of acquisition of the NM title. So I wonder whether you see any need for reform in any manner whatsoever when it comes to this topic? How about the insider scenarios I’ve mentioned where one solely plays tournaments run by the organization that they run (or work for)? I’d argue that at the very least, such scenarios should raise red flags, and are worth reviewing to see if any “gaming” occurred there. Because as long as there are loopholes, some will undoubtedly try to obtain the title by any means necessary and exploit them. Not sure if this would be formal changes or internal (kind of like legislative vs. executive action, to use a rough analogy). And again, floating the open tournament requirement as something worth pondering.

1 Like
23

You have to keep in mind that not every US Chess member lives in a major metropolitan chess hotbed.

When one lives in a smaller city (say, under 500,000), there is likely to be only one active chess organizer, two if you’re lucky. And it may be 100 miles or more to the nearest city with an active chess population.

If one happens to be one of the best players in that city, there aren’t many opportunities to play higher rated players without traveling or someone running events designed to draw strong players from other cities. (Organizing and directing are separate skills, and strong players might not be interested in those tasks.)

I agree that we do need to monitor for abuse of the ratings system, including those chasing Expert or Master titles, but other forms of ratings abuse may be more common. I would discourage adopting rules that might seem inflexible to situations that not all strong players face.

3 Likes
24

It is true that as a matter of practice the Executive Board has exercised responsibility over titles and ratings. However, under the Articles of Incorporation, the authority of the Executive Board is limited by the “mandates of the Board of Delegates,” except to the extent that those mandates contravene applicable law, the Articles and the Bylaws. The Delegates also have general policymaking power as the legislative body of the corporation (the use of the term “among other things” in the Articles implies that the list of express powers is not exclusive). Accordingly, if the Delegates so desired they could intervene in this matter.

I am not expressing an opinion one way or another as to whether the Delegates should do so in this case. However, there should be no confusion that they indeed have that authority.

1 Like
25

Indeed. And I’m aware that this can absolutely look like metropolitan-area player elitism. Regardless, however it may look, that’s fine; I just want the system to be fair – as do all of us here I think. And, while I’m all for equity when it comes to rural areas (broadband access, for instance), that doesn’t mean that just because someone is from a rural area and becomes a master by beating a bunch of low rated players they should get a total pass (and I don’t think you’re saying that). I think here we arrive at the issue of intentional vs. unintentional ways in which someone can obtain the master title by playing mostly much weaker players (i.e., corrupt vs. happenstance.) Ultimately, there has to be a balance between competitiveness and feasibility/access. But I think, with some of the points you’re raising, we risk sliding too far on the feasibility end of that spectrum, leaving the competitiveness end out to dry. Which is specifically why I suggest some measure to insert more competitiveness for the aspiring master, providing a boost on the feasibility end. Well, maybe we take a “carrot” rather than “stick” approach and, without (hypothetically – not backing down in this regard completely!) necessarily changing the requirements for NM, to provide resources to those who lack access to world (err…national?) class play. Because I do think we all, including the players in remote areas, want to see the new master achieve their title by playing robust competition.

By the way, I think we can also flip this on its head: while it seems desirable to help players in rural areas face stiffer competition, there should be ZERO EXCUSE for players in metropolitan areas with bountiful high rated players to get their title by playing far weaker competition, or only insiders of their org. The latter case should certainly at least be flagged and may indicate unfair play (to be reviewed). This may, actually, get at the crux of what I’m after. I think taking the stick approach with metropolitan candidates and carrot approach with rural ones may be a productive way to go. It all sounds very wonky, but this rural vs. urban issue is a real one, and highly complex. (It’s almost beginning to feel like a sweepstakes or recycling thing… 5¢ in some states, 10¢ in others.) And again, as for how these changes would be implemented, it may be a case of bylaws reform / official action by delegates, or it may be a softer approach in terms of the ratings committee being more vigilant about looking out for this stuff before challenges need to be made; however, delineating clear guidelines to avoid ambiguity would be in everyone’s interest, lest we create more problems than solutions. Would an ADM still be the way to have the issue raised, for these kinds of points to be considered, and for reasonable actions, if needed, to be taken? (Maybe I have the presidential immunity case in mind, but clearing up any ambiguity – in an expedient manner – does seem to be the way to go!)

And, for those with connections to U.S. Chess Trust, based on this intriguing discussion (hopefully we’re all learning something and have something new to chew on), I propose, if not already in effect, some form of grant for players who meet this exact requirement: say, 2150-2200 in an area where they don’t have easy access to a pool of experts and masters. Does that seem like a reasonable approach @nolan @alaskalinuxuser @ChessEntrepreneur. How this would be done, I don’t know, but hopefully someone in the know, if the idea seems reasonable, could get this in motion.

26

Thanks for the addition @ChessEntrepreneur.

I want to give you the chance to insert your opinion, though (or more facts), if you’d be so inclined! :grin:

27

I mis-explained the data, for which I apologize. The row I was citing was draws only, the player in question (a FM from India with a 2324 FIDE rating at the time of his first US Chess event and with only a handful of US Chess events) did have a draw against a 1692/24 player in the event where the NM was awarded, but also had a win against a 2232 player and win against a 2305 player. (And the 1692 player had a post event in the mid 1800’s.)

So this player was master-strength in FIDE events before ever playing in any US Chess events, but US Chess doesn’t award the NM title until the player’s US Chess rating is considered ‘established’, ie, a ‘game count’ > 25. A FM would start out with 15 ‘games played’ when the rating is initialized from FIDE data.

28

While I am glad to be part of the discussion, I do think it is important to note that I am not anywhere near a contender for NM, and I am a very new USCF member, as well as a club level TD. I say this not to think less of myself, but to say I may lack the experience in both competitive play and in tournament directing to make sound judgement on this matter. So, take it all with a grain of salt, so to speak.

I do agree, to some extent, that one cannot be a National Master if one has never played nationally.

I also see that it may become a stumbling block for those in my location, as I litterally meet chess players weekly or at least every other week, who tell me that they cannot afford the $45 USCF membership fee, and this stops them from signing up to play in my local tournaments. How much more then, for someone to pay 20+ times that to go to another state to play? I don’t know.

I also, being new to USCF, don’t understand where all this magic money comes from to pay for people to go and play in other tournaments. Perhaps I need to shake this magic money tree for 50 vouchers to sign up new players who can’t afford to become USCF members. :grinning:

However, I do ultimately agree that one cannot be a National Master, without at least playing nationally.

29

There are quite a few cities in the USA, like New York or Chicago, where one may play against highly rated players who all live within a few miles of the tournament site.

Does that mean those players have not ‘played nationally’?

Let’s suppose we have a 2100 player living in a city where all the other players are rated 1800. If in an event he plays 10 1800 players and wins all 10 games, his rating will go up around 21 points. If he scores 9.5, his rating will go up around 19 points. If he scores 9.0, his rating will go up around 10 points. If he scores 8.5, his rating will go up about 1 point. If he scores 8.0, his rating will go down about 10 points. If he scores 7.5, his rating will go down around 19 points.

So the player will have to win consistently (and probably play about 50 games) to earn a 2200 rating against that pool of 1800 players. A few slipups and 2200 will be an elusive goal.

(Data from the ratings estimator, so these aren’t intended as precise calculations.)

30

I think that’s 29 on the 10-0.

Note also that in the “corrupt” small pool, the willing and eager lower rated opponents will lose more points than the would-be master would gain (because the higher ratings have a lower K factor).

This is a “solution” in search of a problem.

31

That’s totally fair. Thanks for bringing it back to the notion of national level play (as you write, at least to some extent) to become an NM. And affordability/accessibility is certainly an issue. Which is why, based on this robust discussion with a lot of informed folks, as I was hoping for, perhaps the way to go is support for rural/small state players so that they can indeed play vs. very strong competition (again, not to be metro biased or something – I did AmeriCorps in the Missouri Ozarks btw and am very interested in rural issues – but this is just the reality with the big fish small pond scenario." Not only would this make it more fair for everyone else, but the aspiring master could have a more meaningful experience, become stronger, and feel great about their title once they earn it fair and square.

1 Like
32

That’s fair. I don’t know about the whole NY area not being “national” literal meaning there; I mean that level of competition, and NY is among the strongest. That’s the main point, strong competition, as fair and square as possible.

But I take your point about the difficulty in beating a bunch of lower rated @nolan, esp. if much lower, and I get that the risk of losing points is high as well. Still, I imagine some players have gotten the title this way? If so, the hypothetical wouldn’t really matter but rather an assessment of the result.

And again, I’d return to my point of “insider matches” for lack of a better term, which did raise my awareness of the issue in the first place. Not sure if you’ve addressed this point yet – if not, what do you think? I think this is effectively an explicit cheating issue, whereas playing a bunch of lower rated is likely unintentional, maybe not the deliberate gaming scenario, and again I’d return to my positive solution based approach here of offering opportunities so that even such accidental scenarios don’t need to happen because the remote player would have more opportunities. And I want people to be able to gain the title, just to do so competitively. Does that make sense?

33

I think it depends how we define corrupt @wintdoan, since those willing accomplices wouldn’t mind losing a few points if the idea were to help their friend or colleague push through. And keep in mind that we may only be talking about the final dozen or two points, which are the toughest; this is where a microscope might be best put to use.

These issues I raise, to clarify, are based on the (what seemed to be obviously unjust) scenario that was brought to my attention last year and concerned me in the first place. And frankly I’d like to know whether this is in fact widespread or not (which is beyond the headline data, and which I’ve heard about from the first couple of people I mentioned this to) to determine whether or not action needs to be taken. Who knows, maybe not.

But I do think that regardless, something interesting that has arisen in this discussion is about rural / small state players and not necessarily punishing those who fairly obtained the title, but helping those who haven’t yet to have access to face a high level of competition to get stronger and have a real shot at obtaining the NM title in the first place – based on the difficulty, as you write, of racking up points incrementally (not to mention the undesirability of doing for a variety of reasons brought up). Though, again, I don’t know how common this actually is in practice beyond hypotheticals and a few anecdotes so I’d genuinely like to know.

34

If the anecdotes are to be believed, there was probably a lot more finagling of ratings in the past than there is now. (Prior to 2005 there was no record that I could find of ratings or floor changes made manually by the office, so lots of opportunity for mischief existed and even justifiable updates were not logged for subsequent review.)

The match rules limit gains or losses from a single match to 50 points, with additional limits of 100 points over 180 days and 200 points over 3 years, and the gains and losses are tracked separately, so someone who has a 30 point gain from a match, followed by a 20 point loss from a match can still only gain another 70 points via matches within the 180 day period.

I’ve argued for stricter rules for crossing major class boundaries in the past, but there really isn’t a lot of hard evidence of a current serious problem here.

35

Thanks again for the data. Is a 50 point gain from a single match a non-issue? Again, if someone is looking for a boost within the last ~50 points they might get a little help from their friends, and this would more than do the trick. Not sure why such matches are even necessarily if it’s not a formal championship or something. Can’t they just play each other off the books?

Certainly, though, it does seem that more corroboration would be needed to take action on some of the points I’ve mentioned.

36

Mr. David Bennet,
When a rule book is a couple hundred pages long, gaming the system is expected. To have the rules challenged continues the game from going extinct. Chess itself in the mechanical game is finite. It has already been proven that white has only three opening placements that speak of only one win and two draws.

A long time ago I proved that one could win a five or six round Swiss by just ordering byes. Of course the rule book has evolved, however to contest the game on the mechanics of time travel and being in tune with rules regarding triple occurrence of position and picking up clock the game can get incredibly interesting.

When looking at the longer games of Magnus Carlsen, in the first 25 moves mate should already be decided. It doesn’t need to go to an end game that has a dead zone or neutral zone. The game itself evokes debate on the world of the day. Wouldn’t it be considered gaming to skip winning in 25 moves and instead play longer to reset the challenge just be the end game?

37

I think that Mr. Bennett is searching for a solution for something that is likely not even a minimal problem. The changes of the rating system to 1) rate events tournament by tournament rather than game by game and 2) to effectively make the minimum rating change .01 points instead of 1 point have made it incredibly difficult to increase one’s rating by playing much lower rated players. He didn’t mention the practice attempted at least briefly by many players in the 2100s of playing house games against players rated about 800 who received a bye, but this ptactice is unlikely to be fruitful now, or at least so slow as to be beyond anyone’s patience.

What is difficult here is that, as Mr. Aigner suggests, we have three “Master” things which are not remotely equivalent, only one of which impresses the chess public at large. We have the NM title, the LM title, and the Master class. The OLM title isn’t germane to this discussion. Simply put, while the first two can be very much cherished by the title holder, it is only the third that impresses other players. We’re all a little befuddled and many scoff to see a player rated 2185 on the wallchart as NM John Jones, much less when Jones is rated 2030. This is the reason why many players stop playing when they reach 2200 and only coach.

Finally, the phrase “National Master” reflects the national nature of our rating system. There is enough “cross-pollenation” of the system that a 1600 player in Fargo will have about he same results against his neighbors as he would against similarly rated opponents in Orlando. Mr. Howard will find, for example, that a few players temporarily stationed at Fort Wainwright will stop by his tournaments, or someone on vacation will stop by, or a few of his players will decide to go to Nevada and play the National Open, or visit their grandkids in Boston and drop by the Metrowest club a couple of weeks. Enough of this happens that there are very few areas that are not well calibrated with the rest of the rating system. So National Masters are not necessarily those who have played nationally, but those who would play at the Master level anywhere in the nation.

2 Likes
38

@relyea thanks for your feedback. To be clear, are you really suggesting that (1) there isn’t a problem at all with TDs and affiliates abusing their power to create artificially arranged matches and that (2) smaller pools of players, which aren’t broadly exposed to the general playing pool, doesn’t create any deviation in ratings? Now for those in a small town or state who don’t have a choice, i understand, and therefore encourage ways to help them get exposed to that broader rating pool, but for those who have the option and yet choose to engage in anti-competitive behavior to avoid the broader rating pool, that’s a problem. Is there really no there there? It sounds a little too simple and dismissive.

40

Some solutions are overreactions in search of problems. If you inadvertently dropped you phone in a puddle and need to dry it out then a simple way is to do so is to put it in an oven at 500 degrees to dry it out quickly, but that is a very sub-optimal solution (unless you don’t mind destroying your phone).

Alex was not saying there is no problem. There are, however, existing tools to deal with it that do not cause other serious difficulties. There have been cases of abuse by TDs and affiliates that have been dealt with using various levels of harshness. There is no overriding need to do something else to replace the mechanisms that can already be used.

Many isolated pools end up being underrated because they’ve improved from their initial ratings and are playing other underrated players. Cases of overrated isolated pools (see Claude Bloodgood) are identifiable and can be dealt with using existing tools.

Major changes to handle edge cases are generally overreactions.

41

@jwiewel How effective are the mechanisms in place to deal with such nuanced scenarios like (again, the example I have in mind) players of the same org or friends choosing to play each other and therefore intentionally shrinking the pool? Now, is there a clear rule against this sort of thing? Doesn’t seem to be. And the players can argue that they simply want to get in more practice and that it’s convenient to play each other – and they can probably point to the fact (correct me if wrong) that there are no rules against this. Clearing up any ambiguity would be helpful, creating standards for players, TDs, affiliates, and ultimately to whatever committee (ethics, rating, TD) the issue is brought. I think an extreme scenario, which really isn’t that much different, is one on one matches. Do they really need to be rated? Maybe if it’s for a state or club title, but otherwise, based on the skewing effect this will likely have (advertently exiting the broader rating pool, which is the point I’m arriving at), why do it? Do these small group / insider matches really need to be rated? Why not just have them be casual unrated games, as is customary for this sort of arrangement.

And I understand your exaggeration, but the phone isn’t being placed in the oven. Instead, it’s simply waterproofed to begin with! Not too tricky and a reasonable measure, beneficial to all.