PDA

View Full Version : Returning players and ratings.



lost
04-10-2009, 09:20 PM
I have just returned to chess and want to know how do you calculate your rating based on tournament results when you play an unrated player.

Could Bill Gletsos let all of us know how to calculate our rating with an unrated player.

lost

Bill Gletsos
04-10-2009, 09:51 PM
I have just returned to chess and want to know how do you calculate your rating based on tournament results when you play an unrated player.

Could Bill Gletsos let all of us know how to calculate our rating with an unrated player.

lostIf the "unrated" opponent is listed as xg on the master file list then the system will use the internal rating it has for the player.
If the "unrated" opponent is not on the master list or is listed as "unr" or "new" then the rating used will be based on their performance during the entire rating period.

As such you cannot calculate it.

lost
04-10-2009, 10:35 PM
Ok thank you Bill.

lost

Saragossa
04-10-2009, 11:20 PM
To avoid starting another thread: What is the FIDE to ACF conversion? Like if I beat someone who only has a FIDE rating does the change in my rating go completely according to that? Or is there a formula?

Garrett
05-10-2009, 08:25 AM
Far out how many points did Damien Norris shed from the zonal ?

his ACF rating is 1620

Damien was always a great player and from memory always well over 2000.

I'd hate to be playing him in his next tourney at that rating.

cheers
Garrett.

Garvinator
05-10-2009, 10:46 AM
Far out how many points did Damien Norris shed from the zonal ?His ACF rating is 1620
Could have been worse. He could have lost another 21 and been able to play in the Australian Minor ;)

Watto
05-10-2009, 10:53 AM
Far out how many points did Damien Norris shed from the zonal ?

his ACF rating is 1620

Damien was always a great player and from memory always well over 2000.

I'd hate to be playing him in his next tourney at that rating.

cheers
Garrett.
He shed 448 points.

Garrett
05-10-2009, 01:08 PM
okay thanks

448, that sounds harsh.

Damien is a great bloke and I think it would be devastating to shed that many points from one tournament (edit - I don't know for fact it was one tournament).

Garvinator
05-10-2009, 01:24 PM
okay thanks

448, that sounds harsh.

Damien is a great bloke and I think it would be devastating to shed that many points from one tournament (edit - I don't know for fact it was one tournament).
Norris's results from the Oceania Zonal:


Rd. SNo Name Rtg Res.
1 60 Stahnke Alexander 1795 s 0
2 62 Ford Daniel 1792 w
3 52 Juszczynski Mathew 1914 s 0
4 56 Rachmadi Herman 1842 w 0
5 72 Cigelj David 1238 s 1
6 61 Egan Bill 1795 w 1
7 49 Jule Alexandra 1928 s 0
8 63 Searle Zachary 1791 s 0
9 70 Lapitan Daniel 1287 w 0

Which means he performed at around 1550 or so for the tournament. Considering that he has played nine games in the tournament, his current rating, while seemingly harsh, is based on his most recent data and 1620 does seem fair for the results he has recorded. There is nothing in these results to indicate that he performed anywhere near his old rating of 2142.

The better way to look at it is that his past rating is from so long ago that this is a new rating based on his current results. If he plays more and performs way above 1600, then is rating will shoot up just as much.

Garrett
05-10-2009, 01:48 PM
If he plays more and performs way above 1600, then is rating will shoot up just as much.

Are you sure that is the case ?

His old rating may (I don't know) have had a question mark or two.

Now that the rating has no question mark is it possible to gain 442 points if he performed at 2200 for a single 9 round tournament ?

cheers
Garrett.

Kevin Bonham
05-10-2009, 01:57 PM
I'm pretty sure the ratings Garvin shows above for Norris's opponents are actually FIDE ratings where the player has a FIDE rating. Try these (noting that they are ACF ratings for the previous period not this one) :


1 Alexander Stahnke 1635 0.0
2 Daniel Ford 1619 0.5
3 Mathew Juszczynski 1872 0.0
4 Herman Rachmadi 1569 0.0
5 David Cigelj 1238 1.0
6 Bill Egan 1497 1.0
7 Alexandra Jule [WIM] 1826 0.0
8 Zachary Searle 1746 0.0
9 Daniel Lapitan 1287 0.0

On this basis his TPR was only 1374 though it is a little higher when performances in the period by his opponents are taken into account, as Glicko does.


Now that the rating has no question mark is it possible to gain 442 points if he performed at 2200 for a single 9 round tournament ?

No but a couple of such tournaments in the same ratings period would more or less do it.

I will shortly move the Norris discussions into the main returning players thread or merge the two threads.

Vlad
05-10-2009, 05:10 PM
I'm pretty sure the ratings Garvin shows above for Norris's opponents are actually FIDE ratings where the player has a FIDE rating. Try these (noting that they are ACF ratings for the previous period not this one) .

If you want to calculate his true performance rating you need to take ratings for the current period. This is exactly what Glicko does.

Garvinator
05-10-2009, 05:40 PM
On this basis his TPR was only 1374 though it is a little higher when performances in the period by his opponents are taken into account, as Glicko does.Well, in this case his performance was even worse that my estimate. Certainly nothing to suggest that his current rating deserves to be anything like 2100.

Kevin Bonham
05-10-2009, 05:55 PM
If you want to calculate his true performance rating you need to take ratings for the current period.

I meant that they are for the period in which the Zonal was played (the ratings that were current at the time the event was processed), rather than being the ratings that are current now.

Vlad
05-10-2009, 06:28 PM
I meant that they are for the period in which the Zonal was played (the ratings that were current at the time the event was processed), rather than being the ratings that are current now.

When I say current I mean September 2009. When Glicko does calculations it eventually uses ratings that will be the output. We already know what the output is so one can just use it.

Bill Gletsos
05-10-2009, 06:45 PM
When I say current I mean September 2009. When Glicko does calculations it eventually uses ratings that will be the output.That isnt correct.
When doing a rating period the rating system uses an intermediate rating for a players opponents rating based on the opponents performance during the entire rating period but that intermediate rating can vary a fair bit from the opponents published September rating.