2nd GM norm for Daniel Fernandez

[Stewart Reuben]

No, they couldn't. That requires more information than the FIDE rating system gives (you'd need the actual probabilities of a win, draw and loss, not just the expected score).

I am not sure I follow that Jack. Of course when Elo made his statement they used to average the ratings of the opponents, both for title norm requirements and calculating rating changes. Now it is only done for the former.
As a first approximation, it would surely be sufficient to choose a 2500 rated player against a field of 2400 needing a score of 7/9. You might have to calculate the scenario for 7 wins 2 losses, 6 wins 2 draws 1 loss, 5 wins 4 draws. You can look at the profile of the player on the FIDE site to find out their result history.
against a field of 2450 needing a score of 6.5/9.
against a field of 2480 needing a score of 6/9.
Make the assumption that all opponents are within 100 rating points of the average rating.
It is only an approximation, so assuming 2400 and 6 wins, 2 draws, 1 loss would be fine.

I never asked Arpad Elo, but I assume he chose to average the ratings of the opponents to determine the rating change was because the system was designed for Title Norms, not Rating changes. I got Vishy's help to get FIDE to agree to change to rating each game separately.

David Welch made the interesting suggesting that the norm requirement for each game be determined separately and the total added together. That is logical. But if the variation in rating of opponents is wide but still within 400 points, it becomes easier to get a norm, perhaps as much as half a point less. I don't understand why.

[Roger de Coverly]

Your expected score against a particular set of opponents works on a game-by-game basis, whereas a title norm works on the average rating of the whole field.

Didn't they use to do ratings by average field as well? So if you played 6*2450 and 3*2050, the expected score is based on playing 9*2317. Although mostly modernised, the Elo system does after all predate the wide availability of even handheld electronic calculators. A possible reform would be to rewrite the Norm regulations to be based on performance ratings. You might have to include a minimum standard of opposition to prevent a Norm seeker gaining a Norm by a near 100% performance against a low rated field. Using the "rule of 400" could work, so that 6/9 against 6*2450 and 3*2050 computes as (6*2450 + 3* 2050 + 3*400) /9 = 2450

[Roger de Coverly]

I never asked Arpad Elo, but I assume he chose to average the ratings of the opponents to determine the rating change was because the system was designed for Title Norms, not Rating changes.

I imagine also that both Elo and Clarke were designing systems that could be calculated using the office technology of the day. So adding up a series of numbers, dividing by the total number of games gives you the Clarke result, with the extra steps of looking up a result in a table and adding or subtracting it to or from the previous rating giving the Elo result. Having to borrow a Colossus or derivative from GCHQ or a University to do the BCF grading was presumably not on the cards back in the 1950s.

[NickFaulks]

I am not sure I follow that Jack.

I do, and Jack is quite right. The part of Elo's groundbreaking work which addresses probabilities ignores the possibility of draws, and is as a result embarrassing nonsense.

[Stewart Reuben]

In the early days of the FIDE Rating System, Mrs Elo used to do some of the arithmetic at the kitchen table. There wouldn't have been anything as advanced as a calculator. I am not sure even a slide rule was used. An abacus would have been more useful anyway. I am quite certain no computer entered into consideration.

The BCF System for congresses was described as being done by HG Felce n the back of an envelope. That was all changed by the number of events exploding post 1965.

[LawrenceCooper]

A last round win to finish on 6.5/9 in third place with a 2574 performance. Sadly no norm but a useful rating gain.

[Mick Norris]

A last round win to finish on 6.5/9 in third place with a 2574 performance. Sadly no norm but a useful rating gain.

Yes, Dan should be up to 2485 as a result, if Chess results calculation is correct of a 13.4 rating gain

[Mick Norris]

Dan now in Oporto
results

[John McKenna]

Canadian GM Kevin Spraggett (2542) and German GM Vitaly Kunin (2576) in the lists.

Kunin is the man to watch out for - he recently won the secondary event at the Capa Memorial in Cuba -

http://en.chessbase.com/post/capablanca ... -and-kunin

[Alan Walton]

Well considering Kunin lost to a 2118 in round 1, definitely worth watching

[Peter D Williams]

Morning friends
Lets hope Daniel does not switch federation we lost Ynojosa Aponte, Felix Jose http://ratings.fide.com/card.phtml?event=3900258 because his father had to go back to Venezuela to look for work.

Right time for a nice cup of tea

[Alan Walton]

Morning friends
Lets hope Daniel does not switch federation we lost Ynojosa Aponte, Felix Jose http://ratings.fide.com/card.phtml?event=3900258 because his father had to go back to Venezuela to look for work.

Right time for a nice cup of tea

Considering he has just paid (unless the federation paid for it) a significant amount of money to change from Singapore to England, I suspect he wouldn't want to pay that again

Also you have to remember he started his junior career in England back in 2003-2006, before his family moved to Singapore

So Peter I think you are not really comparing a comparable situation

[LawrenceCooper]

Morning friends
Lets hope Daniel does not switch federation we lost Ynojosa Aponte, Felix Jose http://ratings.fide.com/card.phtml?event=3900258 because his father had to go back to Venezuela to look for work.

Right time for a nice cup of tea

Daniel transferred from Singapore to England this year so hopefully he's here to stay.

It was a shame to lose Felix, I liked the family (his sister was also a very decent player) and felt his friendly rivalry with Yang-Fan was good for both of them. Whilst no longer representing England I'm pleased that he's carried on playing, made his IM title and has played in Olympiads for Venezuela.

[Peter D Williams]

I remember felix sister she was a steady player hard to beat if i remember right. Yang'Fan is almost 3 years older than Felix so those 2 would never have played in age group events http://ratings.fide.com/card.phtml?event=468584 . Maybe it helped Felix has he is ranked Number 3 in Venezuela so would be picked for plenty of top chess events for Venezuel. I have some film of Felix playing chess here in UK

Lets hope Daniel is our next GM

Well time to check the tortoises.

[LawrenceCooper]

I remember felix sister she was a steady player hard to beat if i remember right. Yang'Fan is almost 3 years older than Felix so those 2 would never have played in age group events http://ratings.fide.com/card.phtml?event=468584 . Maybe it helped Felix has he is ranked Number 3 in Venezuela so would be picked for plenty of top chess events for Venezuel. I have some film of Felix playing chess here in UK

Lets hope Daniel is our next GM

Well time to check the tortoises.

Yes, different age groups but they were often on similar scores and so friendly rivalry existed although they both wanted the other to do well.