Different countries ranking systems

Does there exist a perfect ranking/rating system yet? The answer is NO (by all means prove otherwise).

There is no doubt people can take advantage of any rating system. So? Some folks bragging about how they sandbagged their ratings and won some little prize money ($50 to $200) only to make others loathe them more. In general, the US rating system serves its purpose and of course if USTTA can improve it then it is even better. Will and when that day come remain to be seen.

Back to the OP, there are a lot videos of folks 2000-2300 USTTA rating and I myself can approximate very much which rating I possibly belong to. Similarly, people from different countries can show videos of different ratings from people of their own countries and we can do the same self evaluation. However, the only way to get a true USTTA rating is to participate in many USTTA rating events without any hidden agenda. Many people are doing it (only a small population abuse it) and their ratings truly represent their TT skill level.

Even under idealistic participation by a player, the USATT match rating system is flawed by natural deflation causing elements thus ratings do not necessarily reflect a player's real skill level at a point in time.

Very true. It only gives an estimation how good a player is. But I do not yet know a system can dynamically reflect a true player' skill.

Loopadoop said:

The USA match rating system is flawed causing too many players to avoid tournaments. A better rating system would be divisions or classes. .

Click to expand...

I agree completely. In other sports if conditions are far from optimal, you just participate. In the USATT if you are slightly ill, there is not enough space to play like you normally would, the floor is slippery or whatever, your national ranking can plummet in one event. It makes no sense to play. In every other sport I have participated in, you just play. If I am an "A" class racer and have a horrible day, I am still an "A" class racer at the end of the day. The way the current system is now in table tennis, it does not encourage participation. It is very easy to lose a ton of rating points and it is quite hard to get them back. The re-adjustment process is also somewhat arbitrary.

I think having divisions would encourage participation. You could have some points system to advance divisions. Once you achieve a certain division you cannot drop down a division for life unless you petition to be moved to a lower division. This would prevent sandbagging which is a very big problem for people rated 1900-2300. I personally know several people currently who lower their ratings just to win some prize money here and there. They also get better odds to gamble if they claim to be 1900...

Loopadoop said:

The USA match rating system is flawed causing too many players to avoid tournaments. A better rating system would be divisions or classes. .

Click to expand...

There are certainly a lot of flaws. I lot of the ways to fix them result in the same people missing, manipulating, or cherry picking events. There is a way to game every control measure as long as there are those motivated to maximize their chances to win an event.

Simply trying your best to compete at ones level or a level above that is not enough for them. No fixing that crowd, no software update patch available.

lightspin said:

I agree completely. In other sports if conditions are far from optimal, you just participate. In the USATT if you are slightly ill, there is not enough space to play like you normally would, the floor is slippery or whatever, your national ranking can plummet in one event. It makes no sense to play. In every other sport I have participated in, you just play. If I am an "A" class racer and have a horrible day, I am still an "A" class racer at the end of the day. The way the current system is now in table tennis, it does not encourage participation. It is very easy to lose a ton of rating points and it is quite hard to get them back. The re-adjustment process is also somewhat arbitrary.

I think having divisions would encourage participation. You could have some points system to advance divisions. Once you achieve a certain division you cannot drop down a division for life unless you petition to be moved to a lower division. This would prevent sandbagging which is a very big problem for people rated 1900-2300. I personally know several people currently who lower their ratings just to win some prize money here and there. They also get better odds to gamble if they claim to be 1900...

Click to expand...

I have a nagging suspicion that the same sandbaggers will game any other system you'd come up with. Also - in the division system you are proposing you eventually end up with a bunch of A division players who actually belong in Div C (because they are no longer even remotely close to Div A standard of play but somehow are still around).

To be honest I'm still not sure what people are finding wrong with current USATT rating system: if it prevents you from participating in tournaments, perhaps you should look in the mirror and ask yourself whether you are assigning too much significance to a number that tries to track your strength as a hobbyist TT player. Your livelihood does not depend on it.

A division players in C division have no effect unless they are winning money and trophies then they should be elevated.

I know lots of players who avoid USATT tournaments to protect their ratings. Especially coaches as they age. Players don't want to waste money and ratings points.

Loopadoop said:

The USATT match rating system is easy to manipulate for your benefit.

a Division A player shouldn't be able to go down to Division E very easily or in a short period of time. If he does and is dominating, his status should be elevated accordingly.

Click to expand...

Well, let's just say all these under-whatever events make it easier to manipulate than a division system.

I could see why the USATT stays with the current system as it works pretty well for the geography and population. Most people could only make it to 1 or 2 tournaments a year due to travel time and expense. Being able to play in multiple events at a tournament becomes an important factor.

Long story short, the very reason a Div A player plays in Div E is due to the measures implemented to discourage campers. After the reform, points are kept only for a year after the expiration of membership. Tons of players have to start from square one because of that. This severely breaks the curve.

Now, we have people constantly complaining over the meaning of membership because they get to play just 1 to 2 events within their divisions and most get eliminated just after 1 match.

Loopadoop said:

Even under idealistic participation by a player, the USATT match rating system is flawed by natural deflation causing elements thus ratings do not necessarily reflect a player's real skill level at a point in time.

Click to expand...

Every Elo-type rating suffers that problem. Division system that uses Elo-type rating suffers that, too. To be honest, all rating systems suffer that. Changes in rules, equipment, styles etc. occur over time. People have much better access to resources and training conditions nowadays. Players in the same spot from the 90's would likely have trouble staying in that spot now.

We have people camping in the upper end of Div B that are capable of entering Div A on purpose because they get eliminated in 1st round otherwise. It also sounds better on paper if they finish in the top in Div B, especially for those who coach.

zeio said After the reform, points are kept only for a year after the expiration of membership.

Division points history should be kept just like they had membership.

The USATT match rating system hasn't worked for a long time. Membership has not increased compared to other sport hobbies. Less than a 1,000 member increase in the past 20-30 years. In my area, club and tournament participation is at an all time low. Ratings are very inaccurate. Players have to go out of state to increase their ratings.

Pickleball just started at one recreation center has more organized participation than TT.

Loopadoop said:

A division players in C division have no effect unless they are winning money and trophies then they should be elevated.

I know lots of players who avoid USATT tournaments to protect their ratings. Especially coaches as they age. Players don't want to waste money and ratings points.

Click to expand...

Your first sentence is the essence of Korean Amateur system of sandbagging. The have divisions, no numerical rating. You do not move up a division in a city class, unless you win your division in the city tourney. Club owners and players are motivated to keep their players way under-classified, so when the team event happens (it is always a handicap) the most sandbagging team has the edge. However, since over time EVERY CLUB is doing this, what it created is a system of divisions that are totally stacked in the top end of each division. Since everyone is doing it, on one level it is fair, but not in absolute terms nor according to my personal ethics.

I entered the Korean system in my city north of Seoul at Div 5. I NEVER won a city or regional tourney in Singles, ever. Yet, as my play level rose, I pressured coach to classify me at the appropriate level I was at as I grew. I played Div 4 when I was strong Div 4. I got to Div 3 when I was middle of the pack for that division. I got 3rd place in Div 3 in city and started playing Div 2 city. I COULD have stayed Div 4 and won some matches straight up without handicap vs Div 1 o Div 2 players, but I wouldn't have any of that. I wanted to be known for the level of player I was, even if I never won as you need to defeat many players 2-3 levels better than you once you reach round of 32 or 16.

In USA, I know many players proud of their rating (ME TOO) and once they get a level, stop doing sanctioned tourneys (NOT ME). It is laughable.

What is even more laughable are Asian parents' attitudes towards coaches. They expect a coach to compete and show the coach is way better than everyone. These parents do not want to see weakness in a coach. So, if a coach does tourneys and doesn't win, these parents lose respect for the coach. SO funny. Once a high level player stops training and does coaching full time, they drop 1-3 levels. No way to play like they did before consistently.

What is even funnier is the high level coaches do NOT want to publicly rally with the "Wrong" player either. Think, if an older, very NON-Athletic looking dude with Pips sticks his bat out there and does an off the bounce block for a winner and the American-Chinese parents of their kids being trained by that coach see it... it is OVER, parents will not respect that coach. I have seen players "asked" to not play in the club for these reasons.

Winning tourneys while coaching has NOTHING to do with being an effective coach (although knowing what it takes to win and having done it is important). Parents are part of the culture problem too.

Anyone name me a control measure to stop sandbagging in USA and I will show you a way(s) for the sandbagger to beat it.

It has already been though of, discussed, tossed around and the sandbaggers will still sandbag.

The only way to really stop the sandbagger is to ban them from playing. However effective that is, that will take away from the bottom line of an already stressed club or organization. That solution is not going to be used very much if at all.

Also, the reasons for the deflationary pressure on USATT ratings has been widely discussed. The ratings system itself still does its job - rank players relative, but many factors skew it.


Overall, players are getting better in every class. What playing level used to be 2000 just 10 years ago... many 1700-1800 level players do that stuff and play that level.

Another factor is many highly skilled players enter, and it is difficult to properly classify them right away.


Another nagging factor is some places have some really skilled players who are 1700 level, but somehow, are classified as under 1000, MANY of these in an area. So, when players do tourneys in this area, you have the natural loss of a 1600 player to the 1700 skilled 823 rated player and you get what you get over time.

tropical said:

2100-2200 : https://www.youtube.com/watch?v=yEBWZ4qOzYo

Click to expand...

In France, around 1800 points I mean

Loopadoop said:

zeio said After the reform, points are kept only for a year after the expiration of membership.

Division points history should be kept just like they had membership.

The USATT match rating system hasn't worked for a long time. Membership has not increased compared to other sport hobbies. Less than a 1,000 member increase in the past 20-30 years. In my area, club and tournament participation is at an all time low. Ratings are very inaccurate. Players have to go out of state to increase their ratings.

Pickleball just started at one recreation center has more organized participation than TT.

Click to expand...

That's how they used to do it. Points accrued stayed forever. That's what led to campers and the upward flow became even more difficult the higher you went, hence the reform. You don't play for a year? You disappear on the ranking list. You don't renew your membership? Your points get tossed out a year later. Now what? These campers wreck havoc in lower divisions instead.

Could USATT do something about the system? Of course. Is the division system the answer? Not necessarily.

The ratings are prone to be inaccurate when your members are spread so far away from each another that small groups of players compete for points among themselves. This won't change even if you opt for the accumulative rating system with a retention period.

Der_Echte said:

...

The only way to really stop the sandbagger is to ban them from playing. However effective that is, that will take away from the bottom line of an already stressed club or organization. That solution is not going to be used very much if at all.

...

Click to expand...

I know at least one club that does this (banning sandbaggers from tournaments) - not in US, though. As expected, plenty of anguish and complaints from banned players - although there don't seem to be a lot of them.

When I was in Korea, we had a few good tt athletes in our association. I would invite them to play in the region north of Seoul. Every time I brought our best players, two would win their singles division. They didn't live north of Seoul and that greatly upset some top players and coaches.

They all complained and my club coach who always registered them for tourneys was asked to not have me bring those players north of Seoul for regional tourneys...

Guess what I did? Yes, I brought them... with a chip on their shoulders for one last shot at glory.

The tourney director, instead of outright banning them, reclassified the 1 to 2 divisions higher. This made them give up 3 points handicap to equally skilled players. Of course that made the deck stacked against them.

Of course one of them won his Div 1 anyway. The other 2 were rated Div 0 (ZERO !!!) and Div -1 (MINUS FREEKING ONE). This was unheard of, but we went out with a bang.

All the other clubs had double digit underclassified players spread in all levels. At the top level, playing field was level... for the sandbaggers.

Sent from my SM-N950U using Tapatalk

This is not table tennis, but I think it's interesting to compare. I have attended/helped organise an annual competition for another sport for many years. While there is an individual competition with divisions (which does see sandbagging) the more prestigious competition (with a bigger trophy and prize) is the overall club trophy, which is awarded to the club with the most points of all their competitors combined according to a system. It works out that a even very strong competitor, who could win the individual comp, will add the same to the club score as only two or three intermediates or beginners.

What this means in practice, is that the club that brings the most people, usually wins, or at least has a big advantage. So in the lead-up to the comp, there is a big campaign by the clubs to increase participation, get new members to enter, and help train all the members. As every point counts, getting a beginner to a reasonable intermediate standard is more effective than just preparing the best few competitors. I think it's mostly a good system, though it favours the bigger clubs.

I guess what I'm saying is, if people are going to game every rating system, then if you design the rating system to favour bigger teams, what you get is people persuading every friend and relative to come along, assuring them that even if they only get a handful of points it still helps them win, and shower them in advice and equipment. The clubs also usually subsidise entry, which is like $10. It's good for the sport.

hillwalker said:

I guess what I'm saying is, if people are going to game every rating system, then if you design the rating system to favour bigger teams, what you get is people persuading every friend and relative to come along, assuring them that even if they only get a handful of points it still helps them win, and shower them in advice and equipment. The clubs also usually subsidise entry, which is like $10. It's good for the sport.

Click to expand...

The people at USA Hobby Table Tennis Coalition on Facebook can design a class or division system that can't be gamed. If they start a new USA association it will set the standard for the rest of the world.

In Estonia we have non-ELO ranking system. It is introduced from Russian scientist work more than 15 years ago. It is very precise – players end up competition close their initial seed. Lets say +/- 3 places. Drawback is that calculation is not easy to understand or done by-heart .

Some key points. Competitions do not have bonus points (only Estonian Championships, which do not affect at nearly at all). So same formula – no tricks, whatever competition you play. No politically giving more weight to some competition. I found it rather good that ranking formula does not include ‘motivation’ component, which can be abused. If we want it for seeding, then lets stick to max exactness.

Each match has value. If players ranking points difference is 0-2 points. Value for winner is +2 and -2 for looser. If difference is 3-13, value is +/-1. If value is 0, which means players stronger player does not gain any points if wins. If weaker player wins it gets a bit more value pints than difference of player rankings. At the end of competition player absolute value points are summed and this becomes “weight value”. It shows relatively how much player plays. Higher the weight, more exact the ranking is. Weight is reduced every month. So we avoid devaluation. So we reduce weight, not ranking points. Reduction of ranking points and possible bonus points would make ranking less exact.

Positive and negative value points and weight is put in formula and ranking points change is calculated and added/ substracted to/from previous ranking. Formula takes account weight so if weight is high it gives less ranking points (positive or negative). So good player who accidentally loses, gets less minus than this who plays less. If new player comes it rises more quickly to it right place as it gets more ranking points due weight is low.
Formula also takes account value for un-real wins and does not award weaker player linearily. Let say if at the end of competition you wins and losses value totally is +2 (non-absolute value), you will get around ranking point change +1. If weakest in ranking wins strongest in ranking, having value points +100, it gets only 10 ranking points. So, function is having kind of parabolic curve, not linear dependency.

For foreign players we calculate rating for seeding purposes, but do not include it Estonian ranking. Referees can access these values. For Estonian players at the same time we do not calculate rating when competing with foreigners. Foreigners initial ranking is more wrong which does not support accuracy of the ranking. Foreigners also tend to leave earlier and will not play last matches which would affect ranking accuracy negatively. Referees can also access values for players which are not in visible ranking (zero weight) and usually by prospectus referee can decide how to seed those players.

We have over 600 players in men’s active ranking and over 140 woman (visible, when weight is over 0 or played within half year). Over 9000 players in database from which around 2000 played last year at least once in ranking tournament. National association allows on Saturday’s only few big ranking competitions and on Sunday it is free for unlimited local competitions which distance is more that 100km each other. This helped us to increase number of players taking part on over the country and organizers have maximum possible participants (usually over 100 participants and each player gets 6-8 matches). Sundays facilitate local competition and where more hobby players and children can take part close to home. Negotiations with Association genrate sometmes a little bad blood regarding Saturdays, but it is worth after everything is settled.

Bigger countries might have problem that one part of country does not play often with other part and may happen kind of clusters, but I suppose this can happen with other systems too. We have same ranking system for leagues (series of team competitions). No ranking for doubles competitions or double maches in teams.

http://www.lauatennis.ee/web/node/353

I didn't know this type of ranking that incorporates uncertainty about the player's skill made it into table tennis. That's good, definitely much faster to update towards your true level upon joining.

My 2c. A peculiar flaw of the USATT is that the range of beginner to intermediate/advanced ratings is ridiculously stretched compared to the range of advanced to elite/professional ratings.

From what I can see of 2100/2200/2300 players over videos, it feels like Kanak Jha should have 3500 points if not much more (he's had really good wins recently, I don't know his exact level on the world stage). The best non CNT could maybe sit at 4200 or so.