Tantrix Masters Tournament - Elo Rating Calculations

The Elo ratings implied by this tournament were found as follows. Since it was the first tournament, ratings were iterated until the calculation made 'pre--tournament' rating = post-tournament rating, using twice the lobby ranking for each player as a starting point for the iterations. Please note that because of the limited amount of rating information at this stage, these ratings will be re-calculated from all available tournament game information both before and after the World Championships this year.

(23 August 1999) The pre-WTC recalculation has now been done and revised ratings based on results in this tournament only but with average opponent ratings based on the recalculated ratings for the other players can be found on the overall rating list. This original calculation has been retained for reference, so that you can see how ratings will be calculated in future years but the figures are no longer valid.

For a key to what the columns mean, see below the rating lists. For an overall rating list after all of the tournaments now played, go to the Current Overall Elo Rating List. Click on one or more of the links at the bottom of the page for further explanation.

Official Tantrix Tournament Elo Ratings (G at least 6)
 
Posn Name Nat. G AOR % TP Diff. Rating Lead Nickname Tournt
 
1 Alex Kowalczyk AUS 8 1754 64.4% 208 1963 74 AlexKid 1
2 Jason AUS 8 1763 58.8% 125 1889 16 jzyhk 4
3 Zenon Kowalczyk AUS 8 1765 57.5% 107 1873 7 ZAZZA 2
4 Jérôme Papillon FRA 8 1740 58.8% 125 1865 75 papillon 3
5 Britta Steude NZL 8 1701 56.3% 89 1790 11 Bdot 7
6 Sandrine Papillon FRA 10 1713 54.7% 66 1780 97 didi 5
7 Steven Trezise GBR 8 1762 44.4% -80 1683 24 steven2 6
8 Julia Schwarz GER 10 1647 50.8% 12 1659 27 jade 10
9 Tessa Kerr NZL 8 1720 43.8% -89 1632 70 Sara 8
10 Peter Petrecz HUN 10 1700 40.3% -138 1562 43 Pepe 9
11 Dave Dyer USA 10 1596 44.5% -78 1519 11 ddyer 12
12 Laszlo Harsanyi HUN 10 1651 40.0% -143 1508   leslie1 11
 
Unofficial Tantrix Tournament Elo Ratings (G less than 6)
 
Based on 2-3 games against just one player each so probably VERY inaccurate.
 
  Name Nat. G AOR % TP Diff. Rating   Nickname  
 
  Laurent Berguin FRA 2 1659 43.3% -95 1565   lolo  
  Fred Trivett USA 2 1508 50.0% 0 1509   PizzaMan  
  Jesus USA 2 1519 20.0% -476 1043   cyberjesus  
 

Key:

G is a measure of the number of games played. Count 1 for every game played in an all-play-all. In a knockout (including the qualifying this time) where M games are played in a match, count only SQRT (M+1), rounded to the nearest integer, to avoid giving too high a weighting to games played against the same opponent all at once. This gives 2 for 2 or 3 game matches. Average opponent ratings and % TP scores are adjusted accordingly too.

NB. A rating will be "official" if G is 15 or more. For this year, it will be official (if less accurate!) if G is at least 3 plus the number of months of 1999 completed so far.

AOR is the average rating of the player's opponents, obtained for the first tournament by iteration to give a set of self-consistent ratings. If this is lower than for others, it may mean the player themselves did well so the % TP is high - all the ratings are consistent with each other so such a player will not be at a disadvantage.

% TP is the % of TPs scored out of the maximum possible (i.e. out of 20 per game) in all games played in the tournament.

Diff. is the rating above or below the average opponent rating implied by the Elo system for that TP % - click on the links at the bottom of the page if you want more explanation.

Rating is the player's rating based on the first tournament, equal to the average opponent rating plus Diff.

Lead is the number of rating points the player is ahead of the next player.

Tournt is the player's actual finishing position in the tournament. This still stands - the ratings are NOT supposed to replace the tournament positions.

Further information:

The following links contain more information. Those marked with *'s are fairly simple (I hope!), the others are aimed at the statistically-minded who want to check out the mathematical theory behind the method and look for any flaws - I know there are some of you! :-)

* Simple interpretation of the ratings *

* Why do rating and tournament positions differ? *

* A table of Diff. v % TP *

How the difference table was calculated

Description of the theory and method

* Tournament categories *

* Tantrix Master titles *

OVERALL ELO RATINGS

MASTERS TOURNAMENT RESULTS

BACK TO TANTRIX TOURNAMENTS RESULTS INDEX PAGE

The date and time are :

Friday, 25-Jul-2008 03:05:15 GMT
Friday, 25-Jul-2008 04:05:15 BST (local)

This file was last modified on Monday, 10-Mar-2003 21:17:37 GMT

1306