It's clear from my follow-ups that I like the ELO system, but didn't start the thread to claim that it was better than the current ranking system. I pointed out problems with both. I'm looking for both accuracy and fairness. Players shouldn't have to play to maintain ranking positions or rating points - if it is known that they haven't lost playing strength, their inactivity alone shouldn't cause their ranking or rating to go down.
Table Tennis isn't the only sport with very bad pairing policies. I remember a number of years ago, Serena Williams had left the game of tennis for a while. By the time she returned, her world ranking had dropped considerably due to inactivity. Serena came back and entered a tournament. The organizers strictly followed the seeding rules and didn't give her a seed due to her ranking. The problem was that her fitness and skills hadn't deteriorated - she was still one of (1-5 or so) the best players in the world. Because the pairing rules were based on seedings, Serena had to play a top seed in the first round! That wasn't fair to either of the players and screwed up other pairings as well. Pairings in sports such as Table Tennis and Tennis should be used to try and ensure that the two best players meet in the finals. Best players should be defined as of the start of the tournament. Defining best by adding other factors (playing and placing in more events than stronger players than you) hurts the purpose and makes tournaments worse. Seeding and pairing rules need to be flexible. If some player would normally be seeded #3, but it is known that they have a very sore ankle, have the committee use their handicapping skills (or hire people with those skills) to decide how many seeds to drop the player.
I get the point about the current system being designed to encourage participation - but at what cost?
Having Ding Ning and Liu Shiwen play each other in a qualifying match is absurd.
I'm not sure you get it since you haven't proposed a system that solves the problem yet you continually castigate the current system despite the fact that it has so far solved the problem it was designed to solve. You can't get accuracy and fairness if you pretend that your definitions are not in conflict with what is being prioritized.
I prefer to measure performance, not strength. And I have zero problem with players not been seeded according to strength since performance tends to approximate strength. The problem with measuring strength is that if you do so, and people keep their strength measurements without actively playing, it encourages them not to play unless they want to. Even the ITF/ATP/WTA have minimum participation requirements in addition to performance based systems.
The rules were followed for Serena, but if sometimes, players get injury protected rankings for a few tournaments in tennis. I think the issue back then which may have been fixed now was that pregnancy or motherhood was not a protected rankings event, and it might be now. But even then, complaining that you just want your own rules rather than a system does have its risks, doesn't it? What if Serena played badly? Did the system then get it right? It can't be all about one player like Ma Long - there are many times the system gets it right (Pitchford over CNT Xue Fei) and then people just ignore that.
Right now, there is a mild controversy in that Federer is being seeded #2 over Nadal because of grass performance. Again, there is no perfect system. The most important thing is to remember what is given priority and to maintain the system. If you complain about the system without consistently referring to the goals of the system, it isn't clear you are really dealing with the constraints inherent in the problem.