FIDE ratings and juniors

[Andrew Bak]

Pinched from another thread:

OK. Let me give you an example. You are a 1900 rated player, but really are 2100 strength. So you gain on average 3.9 points per game. You play 50 games in one month, gaining 50 x 3.9 = 195. You are now rated 2095.

However, if you had played 25 games in one month, you would have gained 25 x 3.9 = 97.5 rating points. Next month, you are rated 1998. So now you are under rated by 102. You play another 25 games at 2100 and gain 25 x 2.1 = 52.5. So you are now rated 2050.

Both players started as 1900 players. Both players played 50 games at 2100 strength. But one played them in one month and is rated 2095, the other played them over two months and is rated 2050!

This seems to create a problem for juniors. If they play tournaments when they are young, they are saddled with a very low FIDE rating that will take a long time for them to rectify as they grow up and improve. I'm thinking that this has a negative affect on their development as they will not be able to easily play against stronger players at congresses, particularly in the early rounds and they will not be able to play on high boards in the 4NCL until their rating catches up.

In addition, the monthly FIDE rating system makes it even hard for juniors' grades to "catch up" to their actual level. Very few people play 25 FIDE rated games a year, never mind a month!

Is there a case of allowing juniors to have a permanent k factor of 30 until they reach 18 years old or obtain a rating of (say) 2400 so their rating more accurately matches their ability?

[Roger de Coverly]

and they will not be able to play on high boards in the 4NCL until their rating catches up.

That's no longer true. The 4NCL will allow board orders to be set using converted ECF grades. Where Juniors have played most of their chess against other juniors, the ECF grades can be inflated.

Adults with modest ratings have started to complain that playing in e2e4 tournaments is a good way of keeping their ratings modest with a number of Juniors whose FIDE rating lags their ECF one.

Early to say perhaps, but is the International system going to get seriously distorted because of different national policies on how early in their development, players are brought into the system?

[Christopher Kreuzer]

Adults with modest ratings have started to complain that playing in e2e4 tournaments is a good way of keeping their ratings modest with a number of Juniors whose FIDE rating lags their ECF one.

I believe I was one such who raised this. It is not just juniors, though (and this takes things off topic). Adults with modest (possibly under-rated, possibly not) ratings playing each other end up consolidating those modest ratings. My view is that anyone who thinks their FIDE rating is low and could be improved, needs to play up a section. The gains in terms of FIDE rating are higher there. I am considering playing in the upcoming e2e4 Sunningdale tournament, and am considering playing in the Open section, partly due to this, and partly because I need to get experience against players graded considerably higher than me in preparation for a forthcoming tournament abroad.

I played in the Open at the Hinckley e2e4 tournament back in August 2012 for the same reasons, and scored 50% (3.5/7, ECF grading performance of 191 and gained about 20 FIDE rating points). I also managed to score 50% in the Imperial College Chess Congress Open (2.5/5) in February, but don't have the calculations to hand relating to that. Those were relatively small congresses. My experience in the massive FIDE-rated Open at the London Chess Classic was quite different (3.5/9) as I shed a bucketload of FIDE rating points there. I think the only other FIDE-rated chess I've played has been in the 4NCL (division 3) and the tournament I played in Kazakhstan. In both cases, it was a bit hit-and-miss as to whether my opponent would have a rating or not.

Quite what to conclude from all this, I'm not sure, but that experience over the past year could be compared to that of juniors and what sort of tournaments they should play in. if anyone can also conclude from this what tournaments I should play in, that would be great!

[Roger de Coverly]

I played in the Open at the Hinckley e2e4 tournament back in August 2012 for the same reasons, and scored 50% (3.5/7, ECF grading performance of 191 and gained about 20 FIDE rating points).

Provided you are good enough not to become stranded at the bottom of the tournament, that might be a good strategy. It should avoid the worst of the under rated players if you are usually going to be in the bottom half of seeded pairings.

[Christopher Kreuzer]

I played in the Open at the Hinckley e2e4 tournament back in August 2012 for the same reasons, and scored 50% (3.5/7, ECF grading performance of 191 and gained about 20 FIDE rating points).

Provided you are good enough not to become stranded at the bottom of the tournament, that might be a good strategy. It should avoid the worst of the under rated players if you are usually going to be in the bottom half of seeded pairings.

Stranded at the bottom of the tournament is what happened to me at the London Chess Classic! I'm not sure whether this larger grouping to get stranded in is a feature of larger tournaments (I think it might be), or whether I just played badly (probably a bit of both). I do think a win in a smaller tournament will send you up to meet higher rated players faster than in a large tournament, where a win may just mean you meet an in-form player coming down the other way.

[MartinCarpenter]

The thing with being stuck with with high/low grades certainly not just a problem for juniors either - basically if you're not playing much and your first FIDE rated tournament is unusally bad/good it can take an awfully long time to get rid of it.

You only need to look at some of the FIDE ratings of the Yorkshire players in the third division of the 4NCL to find some quite good examples. Fundamentally of course, accurately grading people who aren't playing much is an impossible exercise

In fact looking at those FIDE ratings makes me worry a bit about people getting ratings based off them by playing in that competition. They're really not aligned with playing strength very well.

[Christopher Kreuzer]

One possibility would be to have the k-factor for the current rating period be based on the degree of improvement in the previous rating period (until a certain rating is reached). That way if you are improving a lot, the k-factor can be boosted to reflect that. If the improvement slows down, the k-factor would go back down. Simply having a high k-factor for the first 10 or 30 games isn't really enough.

[Greg Breed]

I got my FIDE rating a number of years ago when I was lower graded in ECF terms. My FIDE rating hasn't changed much over the years because a lot of my opponents in tournaments haven't had a FIDE rating themselves. In fact I was doing well against them and badly against those who did have a rating so my rating did not improve.
As it turns out I think my ECF grade is now spiralling down to meet my FIDE rating anyway!

[Roger de Coverly]

I'm not sure whether this larger grouping to get stranded in is a feature of larger tournaments (I think it might be), or whether I just played badly (probably a bit of both).

The London Chess Classic Open is unusual for a British tournament in that there is no parallel event for lower rated players. At the British Championship Congress, lower graded players avoid the Major Open and play the 5 day morning or afternoon events instead. Similarly with Hastings. At the e2e4 events, there's always a second section with a rating cutoff in the 1850 to 2050 range. All this affects the pairings, making pairings against rated 1600 players much less likely, quite simply because they aren't in the tournament.

[Leonard Barden]

This seems to create a problem for juniors. If they play tournaments when they are young, they are saddled with a very low FIDE rating that will take a long time for them to rectify as they grow up and improve. I'm thinking that this has a negative affect on their development as they will not be able to easily play against stronger players at congresses, particularly in the early rounds

In addition, the monthly FIDE rating system makes it even hard for juniors' grades to "catch up" to their actual level. Very few people play 25 FIDE rated games a year, never mind a month!

Is there a case of allowing juniors to have a permanent k factor of 30 until they reach 18 years old or obtain a rating of (say) 2400 so their rating more accurately matches their ability?

the small number of K25 games coupled with drastically lower Fide start ratings is a serious handicap for younger juniors entering the Fide system at 1800 or less.

I don't see the slightest hint that the ECF is asking, as it should, for the start K factor for juniors to rise to 32 or 40 for a significant number of games.

In my opinion this is something for which Stewart with his contacts in Fide, the International Director and the Fide delegate should all be actively campaigning. None of them are.


The above is what I wrote in a similar discussion way back in October 2010. Nobody took any notice.

In my opinion the low ratings start and low K factor has already handicapped players like Franklin and the retired Williams and is currently handicapping Sanders and Wadsworth who are our best prospects.

I'm thankful that back in the 1970s the Fide rating floor was 2200 and our players were not then saddled with baby ratings. If it had been as is now the English Chess Explosion would have developed much more slowly.

The start ratings in the younger age groups in world and European youth events show that the Russians and ex-Soviets start their best talents off in Fide events where they would not be underrated.

I felt in 2010 that the BCF/ECF attitude to this major problem for our chess future was totally amateurish and I still think so.

[Christopher Kreuzer]

I felt in 2010 that the BCF/ECF attitude to this major problem for our chess future was totally amateurish and I still think so.

You make a very good case. What is the best way to lobby the ECF to take action? I don't think posting here is necessarily the best way. I presume lobbying is/has been done by other routes as well? Has it ever appeared on any of the ECF agendas?

You mention some juniors specifically. Matthew Wadsworth, (Isaac?) Sanders, Sam Franklin and Peter Williams. I know Yang-Fang Zhou rose in rating very quickly (if you look at his FIDE rating graph, the jump is very noticeable), and attained the IM title rapidly. What went right there? In that age group (U18, around 17 years old, though some might be over 18 now), the next highest-rated junior is James Holland, who I saw playing in an evening league match earlier this week. That prompted me to look up previous results, and reminded me that he got an IM norm at the British last year. I think his FIDE rating is lagging a long way behind his ECF grade. Are there lots of juniors in that position? How many have IM norms or the potential to achieve that (Peter Roberson has two IM norms)?

The useful links relating to norms are:

http://ratings.fide.com/title_norms.pht ... ountry=ENG
http://ratings.fide.com/title_norms.pht ... ountry=ENG
http://ratings.fide.com/title_norms.pht ... ountry=ENG
http://ratings.fide.com/title_norms.pht ... ountry=ENG

Not strictly related to ratings, I know. But norms and titles are the other measure along with ratings.

The former juniors or juniors with IM norms (I looked at anyone born in the late 1980s or 1990s) appear to be: Peter Roberson, Peter Poobalasingam, James Holland, Sam Franklin, Callum Kilpatrick, Chris Dorrington, Martin Brown. Those with GM norms are Callum Kilpatrick and Thomas Rendle. I didn't include anyone born before 1985.

[Roger de Coverly]

I think his FIDE rating is lagging a long way behind his ECF grade.

The published ECF grade has an artificial enhancement included in it. For their real performance you need to look at one of their opponents in the earlier grading period and see what they were recorded as.

If they have continued to improve, you get the benefit of their improvement. If not, you don't get the benefit of the enhancement.

[Christopher Kreuzer]

I think his FIDE rating is lagging a long way behind his ECF grade.

The published ECF grade has an artificial enhancement included in it. For their real performance you need to look at one of their opponents in the earlier grading period and see what they were recorded as.

If they have continued to improve, you get the benefit of their improvement. If not, you don't get the benefit of the enhancement.

Could you expand on that with an example? It should be possible to link to the ECF grading pages in question. I need to go back to the post above and add in some examples of other juniors or recent juniors with IM or GM norms.

[Leonard Barden]

Yang-Fang Zhou rose in rating very quickly (if you look at his FIDE rating graph, the jump is very noticeable), and attained the IM title rapidly. What went right there?

Yang-Fan succeeded in winning a couple of tournaments with very high percentages, gaining around 100 Fide points in each case.

[Sean Hewitt]

I suggested the following to FIDE last year

Players aged under 18 should have a k factor of 30 until they reach 2300.

I'm waiting for a reply!