24 April 2013

Of Rating Pools

Before moving on to the next part of my Pattaya Analysis, here is the first of two articles on ratings pools and ratings deflation for the benefit of readers who may wish to have more information on these subjects.

The FIDE Ratings system is based on statistical concepts expounded by Professor Arpad Elo. The rating of a player is not an absolute measure of the playing strength. It is instead a relative measure of playing strength compared to the playing strength other players in the rating system. The FIDE Ratings List compiles the ratings of players from member countries worldwide. The integrity of relative measure requires that there be sufficient rated games between players from different parts of the world.

However, there are cases when a group of players play primarily among themselves such that their ratings change only relative to each other but goes out of sync compared to players in the wider system. For example, we could have Malaysian players playing rated games primarily with each other, say they play only Selangor Open, National Closed and one or two more local rated events with 90% Malaysian participation. This situation creates what is referred to as a separate rating pool, which of course may be out of sync with the wider global ‘ocean’.

For illustration, let’s say we have


rtg
rtg diff
Fadzil
2128

42
Sumant
2086
10
Jianwen
2076
48
Shazwan
2028
4
Yit San
2024
18
Yit Ho
2006


The ratings difference predicts that Fadzil will score 62% against the other players in the group if they play many games against each other and if their relative difference in playing strength remains the same. Yit Ho will score 41% against the other players.

So if they only play among themselves for 5 years and all of actually them improve but their relative strength differences don’t change, they would find that their ratings remain the same.  An analogy could be that they all go to Shaolin temple and train together for 5 years and only spar against each other. Sumant may feel that he has not improved as his score against the other players remain basically the same. The same goes for the other players. They are in the same pool which is disconnected from the outside world, so to speak.

After 5 years on intense training and getting pass the 36th Chamber of Shaolin Temple, they reintegrate themselves into the outside world. So they go and play many open events in Europe for two years. Since they as a group have actually improved compared to the ‘world standard’, they should see their ratings go up progressively until it is reflective of their playing strength based on the ‘world standard’. The new ratings could be as follows:

Fadzil
2522

42
Sumant
2480
10
Jianwen
2470
48
Shazwan
2422
4
Yit San
2418
18
Yit Ho
2400


However the relative rating difference between the 6 Malaysian players remain the same. Thus for the 5 years while in Shaolin (but where their sparring games are rated by FIDE), they would be under-rated as their ratings don’t change although they are improving relative to the outside world. When they start playing in Europe, their ratings would start creeping up towards their new strength based on the ‘world standard’. Thus their small pool would be integrated with the world ocean.

 ‘World Standard’
It is accepted that there are various such pools all over the world, especially at the lower end of the ratings scale. FIDE would not be too concerned so long as the upper regions i.e. ratings upwards of 2500 are still generally reflective of the ‘world standard’. It is also generally accepted the biggest pool comprise of players playing the Open Events in Europe. The ratings of these players are therefore the ‘world standard’. It is in comparison to this ‘world standard’ when considering the question of under/over ‘ratedness’.

Some years ago, there were discussions locally on whether Malaysian players were generally underrated. In the last 10 to 15 years since 1998 when young Malaysian players started going to the World Age-Group in numbers, and since 2004 with the first edition of the revived (Dato’ Arthur Tan) Malaysian Open, we have more and more opportunities for our players to compare ourselves with players of other countries.

The evidence so far does not appear to support the argument that Malaysians are underrated. If our Malaysian players were truly under-rated, then we should also see Malaysian players make a net gain of rating points in the Malaysian Open or other local events with foreign participants, or when they go out to play in foreign countries, like the just concluded Bangkok Open. This does not seem to be the case so far.

The just concluded Bangkok Open, I dare say, has set a new standard for international open tournaments in this region. The Open Section had over 200 participants, coming from numerous Asian and European countries. There was a good distribution of players with ratings ranging from 1600 to 2600+, and there were strong young players, seasoned master players and strong GMs. Such an event give the participants a great opportunity to compare themselves with players from other countries, of different ages and of playing strengths.

The next part of my Pattaya Analysis will consider this aspect of the ‘different ratings scales’ in different countries.

No comments:

Post a Comment