US Ratings to UK

[Creek]

Are you sure about the percent share and ratings in the US?
From all I heard, USATT is German TTR + ~350-400

With 1850 TTR I'm in the top 1,5% in Germany, but to be in top 1,5% in the US I would need 2400-2500 USATT according to your list, which would not be TTR + 400 in my case, but + 600.

Either the +350 is wrong or your list 😅

 

[Grumms]

Are you sure about the percent share and ratings in the US?
From all I heard, USATT is German TTR + ~350-400

With 1850 TTR I'm in the top 1,5% in Germany, but to be in top 1,5% in the US I would need 2400-2500 USATT according to your list, which would not be TTR + 400 in my case, but + 600.

Either the +350 is wrong or your list 😅

I'm not sure, I just quoted the pyramid from the reddit thread I quoted.

2 different systems in 2 different environment can't have a perfect correspondance, especially near both ends. The +350 can be OK for 98% of the people, but not for you.

 

[Tony's Table Tennis]

From a certain level, USATT rating should be told by tenth instead of hundredth (2610, 2620, etc..., with a 2680 player being way stronger than a 2600).

you are so right
because after say 2600, the gap of 10 or 20 is actually probably worth 50 to 100.
The question is, from what level, the point difference is worth differently....

I really like your comparison - UK1600 to 2500 is all USATT2700, haha
UK has a 1000 point range, while USATT is maybe only 100. so 10 to 1 ratio

as Nextlevel has said, there isn't enough high level players in USA to make that 2600 to 3000 more accurate....
But does UK have enough to make the high level accurate?

 

[NextLevel]

you are so right
because after say 2600, the gap of 10 or 20 is actually probably worth 50 to 100.
The question is, from what level, the point difference is worth differently....

I really like your comparison - UK1600 to 2500 is all USATT2700, haha
UK has a 1000 point range, while USATT is maybe only 100. so 10 to 1 ratio

as Nextlevel has said, there isn't enough high level players in USA to make that 2600 to 3000 more accurate....
But does UK have enough to make the high level accurate?

Nowhere in the world has enough high level players to make those ratings extremely precise except maybe French, Swedish and German league - they are generally accurate though. The thing is that you would have to use them in the ITTF tour. There was an ELO component of ITTF rankings for a while before they switched to the new system. That said, German TTR should be more interesting as a proxy.

 

[Tony's Table Tennis]

Nowhere in the world has enough high level players to make those ratings extremely precise except maybe French, Swedish and German league - they are generally accurate though. The thing is that you would have to use them in the ITTF tour. There was an ELO component of ITTF rankings for a while before they switched to the new system. That said, German TTR should be more interesting as a proxy.

Well based on German TTR, then USATT low end is too high

So like my earlier question, its either low end too high or high end too low.

TTR does take in WTT matches, so that is indeed interesting. Going to make my job easier, since most of the guys im sending to Europe all have WTT history

 

[NextLevel]

Well based on German TTR, then USATT low end is too high

So like my earlier question, its either low end too high or high end too low.

TTR does take in WTT, so that is indeed interesting. Going to make my job easier, since most of the guys im sending to Europe all have WTT history

Not sure what you mean by USATT low end is too high. It's not a perfectly linear relationship and there are definitely enough matches at the lower levels for ratings accuracy. It is the higher end where we don't see 2700 and 2800 players play each other often and that is true just about everywhere in the world except for maybe the top league systems so that isn't going to resolve anytime soon.

 

[Tony's Table Tennis]

Not sure what you mean by USATT low end is too high.

grumms shared chart is a good example
Post 25

https://www.tabletennisdaily.com/forum/topics/us-ratings-to-uk.35097/post-478495

 

[vvk1]

grumms shared chart is a good example
Post 25

https://www.tabletennisdaily.com/forum/topics/us-ratings-to-uk.35097/post-478495

Did you not read my exchange with Grumps above? He completely misunderstood the contents of the TTE ratings data dump and consequently the charts/conclusions/etc are meaningless.

 

[Grumms]

Did you not read my exchange with Grumps above? He completely misunderstood the contents of the TTE ratings data dump and consequently the charts/conclusions/etc are meaningless.

You did not replied to my question. Does players have different ratings. A -17y has a -17 rating and a senior rating, and a vet rating ? That seems right :

So my UK chart is wrong, since I counted several times the same players. I should do it only for adult players. But the way to rate player in UK is really flawed, because most players above 13y will play against adult (at least that's the case in France in the team leagues).

 

[Grumms]

I re-did my chart only with UK senior ratings :

Since there are only 1000 adults rated in the UK, the statistical relevance is limited. In France, 50.000 adults are rated.

 

[Tony's Table Tennis]

Did you not read my exchange with Grumps above? He completely misunderstood the contents of the TTE ratings data dump and consequently the charts/conclusions/etc are meaningless.

Im not too nother on the count
But range, looks right
The updated one too, 1900-2600 for 2700 usatt.

There is no doubt usatt from say 2600 onwards, every 10-30 points is almost a different level

 

[Grumms]

Yes, for the high end of the rating, you'll never find accurate correspondance. But it's the most useless ratings for us, since we'll never touch, by any means, those ratings.

 

[Tony's Table Tennis]

Yes, for the high end of the rating, you'll never find accurate correspondance. But it's the most useless ratings for us, since we'll never touch, by any means, those ratings.

yep,
I get you.
I have U17 TPE national players and higher, a U15 Indian national player, probably even U19 Korean national players,
that I will be nominating for European leagues

so its good that I can have some form of idea on TTR/USATT,
obviously, with time, I can just learn and focus on TTR more and ignore USATT all together (USATT was my first familiarity for level vs rating)

 

[Grumms]

The rating approach is somehow very limited, even within a country. The density of players can completely change the rating distribution. For example, in France, a 1500 in Paris is way stronger than a 1500 in Toulouse. This is caused by one thing : the players are rated according to the player whose they play against. A county level player will only play his county mates, then their rating will allow to evaluate them in their county alone. It's the same for Chess and the ELO rating.