JavaPairing Home Page

Welcome to the
JavaPairing
home page!

available translations

Download and install JavaPairing now!

get Java™

Donations? email contact

use a Simple version on-line for smartphone and tablet!

Simple version of JavaPairing for all browsers

This is a tool to manage a chess tournament, team or individual, i.e. player registration, round generation, results, ranking & crosstable, report to FED and WEB site generation. Implemented engines are: swiss Dutch FIDE, swiss Dubov, swiss Simple, swiss Perfect Colours, Amalfi Rating, round robin

how to install browse the JavaPairing user manual get it PDF

The distribution can run as is on any platform, including Windows, Linux and Macintosh, only needed is Java support (JRE 1.6 or more).
Expand the '.zip' file to your hard disk, maintaining the subfolder structure.
Execute the 'JavaPairing.jar' file in the 'dist' subfolder. By default the program speaks English, but local languages are easily handled by substituting strings at runtime. You have to supply the 'JavaPairing_xx.properties' file with translations where xx is the language code (see 'i18n' folder for details). Pack your language file into the .jar and start. 
Italian, French and German translations are already available! If available, a non native language may be invoked specifying xx and YY as parameters at startup, where YY is the country code (i.e. to see Italian outside of Italy, specify 'it IT').

Download JavaPairing now! get Java™

Donations? email contact

improvements / bug fix log after rev. 1.0

rev. 2.9 - September 2015

032 topic: improvement; general

    - fixed bug rounding Elo variation
    - fixed bug generating FIDE report (half point bye)
    - fixed bug reading FIDE report (half point bye)

rev. 2.8 - September 2014

031 topic: improvement; engine

- Improved the Swiss Dutch Engine for computational complexity when in a big group even the "split according the color due" workaround does not work
- Improved the TRF FIDE import when in the same round more BYE than one happen (the last one read is kept if odd. Others are assumed to be half-point-bye or full-point-bye. Warning! This may cause a checker error!)
- Improved the checker and the cross table generation as stated at previous point

030 topic: improvement; general

- Updated Elo FIDE calculations to rules dated 1st july 2014. K=40/20/10

rev. 2.7 - November 2013

	*** 'SWISS DUTCH' PAIRING ENGINE ENDORSED BY FIDE !! ****

rev. 2.6 - June 2013

029 topic: improvement; general

- Improved the Swiss Dutch Engine (aligned to Istanbul 2012. bug C.1, X and Z)
- Improved read/write FIDE report (aligned to Istanbul 2012. forfeit with colour)
- Improved cross table and statistics
- Added import of teams from a .csv file (the team name shall be on the previous field of that containing the player names and on its own record; a filter has to be prepared regarding the player record. Data is loaded by specifying '*' in the 'Load player from database' field)
- Now it is possile to blank the player's name in insert results of a team tournament (i.e. to put 2-0f, no board)
- Improved FIDE report import
- It is now possible to batch check FIDE reports, specifying the file name as a startup parameter.
- It is possible to import player's data up to 5 files at a time.

rev. 2.5 - February 2013
the endorsment procedure restarts

028 topic: improvement; general

- Developed the checker to test a FIDE report against the Dutch engine (checkbox in the 'open tournament' window)
- Improved the Swiss Dutch Engine (aligned to Istanbul 2012)
- Computational complexity better handled
- Removed auto acceleration of the swiss (new check box in the setup window)
- Fixed a bug in ranking with assignment of 1/2 point to the BYE
- Updated lower FIDE Elo limit to 1000

rev. 2.4 May 2012

n.b.: the endorsment procedure is suspended waiting for B.3 rule explanation by FIDE

027 topic: improvement; general

- Improved the Swiss Dutch Engine
- Result 0-0 is now allowed. For adjourned games use 1/2-1/2
- Updated K=30 to estimate the tranche FIDE for unrated players
- Improved mouse click on the tables

rev. 2.3 March 2012

026 topic: improvement; engine

- Improved the Swiss Dutch Engine for extreme situations.
- Added 'Verbosity Level' to the 'explain' window (the text is also saved to the 'log.txt' file, coded "UTF-8". It may be needed to specify this coding when opening the file with a text editor or a wordprocessor to see up and down floater constrains represented by up and down arrows)
- Reporting score brackets theese simbols are used:
      w  = colour due white
      b  = colour due black
      w? = colour due mild white but can change (rule A7.e)
      b? = colour due mild black but can change (rule A7.e)
      W  = colour due absolute white
      B  = colour due absolute black
      W? = colour due absolute white but can change (rule A7.d)
      B? = colour due absolute black but can change (rule A7.d)
      .  = no colour preference
      -  = no floater constrain
      up arrow   = upfloater constrain   (rule B5 then B6)
      down arrow = downfloater constrain (rule B5 then B6)

rev. 2.2 November 2011

025 topic: improvement; general

- Added a tournament simulator with parametrization. To enter digit 'simulation' in the field 'Load player from database' of the 'Registration' page
- fixed Buchholz Kallithea bug (two or more unplayed games)

024 topic: improvement; engine

- Fixed some bugs of the swiss Dutch and Dubov engines 
- Added improvements to avoid factorial complexity. Anyway, it is not excluded the need to abort the engines in extreme situations where calculation takes a very long time. If this occur, please try to pair the round with a different engine

rev. 2.1 August 2011

023 topic: improvement; engine

- Swiss Dutch algorithm has been improved to prevent factorial complexity in a large group with unbalanced colours (sample from 'Città di Cordenons 2011'; round 2; group 1; x=2; p=18)
- bug fix Dutch pairing algorithm ('sample from Mario Held 14_5'; round 4; group 5); when x=0 strict remove floats first, then the strict flag. 
- bug fix colour assignment
- bug fix final order pairs
- improved colour due explanation

022 topic: improvement; general

- Thanks to Georges Marchal [gmarchal (at) scarlet.be] added the French translation of the program!
- Code updated for NetBeans IDE 7.0
- Elo FIDE calculations updated to rules valid from 1st July 2011
- Buchholz calculation updated to Kallithea Congress 2009
- can read (import) FIDE report

021 topic: improvement; output

- Improved printer pagination of current round and player cards. Remember to use the 'send to browser' button and do 'print preview' on that! If this call does not work in your PC, directly open the file 'temp.html'.

rev. 2.0 January 2011

020 topic: improvement; general

- A dialog has been added at the startup to quickly allow to select individual tournament/team tournament/open existing file.
- Added player cards.
- Added 'Weighted Boards' tie-break for team tournaments. Individual results are multiplied by a board coefficient (1.9 - 1.7 - 1.5 - 1.3 ... 0.1) then added together.
- Improved graphics and icons.
- A facility to merge two tournament file has been added to allow: a) to build large open tournaments registering players from different PCs and then merging data, b) to reuse old tournament file, c) to convert team to individual tournament. A checkbox field has been added to the 'open tournament' window at the right side. It is available if a tournament is in memory (read/written) and no rounds are yet generated (or they are completely removed). n.b. To export player's data it is suggested to use the 'report to FED' procedure.
- According to FIDE rules, delta performance for score 0% is now -800.
- Added flag 'F' for Elo FIDE

rev. 1.9 December 2010

019 topic: improvement; output

- Elo variation page has been extended taking into account games between Fide rated players (for both k=25 and 15). For the others the 'Fide tranche' is calculated if applicable. 

018 topic: improvement; engine

- improved messages on due colour and constrains (colour and floats)

rev. 1.8 November 2010

017 topic: improvement; engine

- completed development of the 'swiss Dutch' engine. It passes three test tournaments: NORDICWC, TU_29_11 and TU_31_13, can exchange S1<->S2 for any number of elements.

rev. 1.7 October 2010

016 topic: improvement; engine

- started development of the 'swiss Dutch' engine with the goal to be full compliant to FIDE official rules. At the moment it passes three test tournaments: NORDICWC, TU_29_11 and TU_31_13, but obviously could not pass other tests ... The current implementation can exchange S1<->S2 for one player at a time and the last two of S1 with the first two of S2.

rev. 1.6 July 2010

015 topic: improvement; general

- added report to German Federation
- for schema to import players from a variable length fields database it is allowed to specify the character '|' to join two fields and the character '-' to extract only the first part of a field (introduced to read German database format of the file SPIELER.TXT)
- added sorting capability of teams/players in the 'Registration' grid. Can be done by clicking on the column header or by using the right button of the mouse. It is allowed until the first round is built
- fixed a bug on the 'ritired' flag for individual tournaments
- cross tables present now the current round instead of '-BYE' if results are missing, by using the symbol '?'. For team tournaments the holders are temporarily paired. For tie break it is correct to consider a not yet played game as BYE. To have the ranking at the last really played round, go to the 'Rounds' page and decrease the round counter 
- fixed a bug on partial ranking by age
- improved the layout of the text obtained by the 'copy as plain text' in the 'Output' page

rev. 1.5 June 2010

014 topic: improvement; general
- completed the user manual
- added the 'implementation of the Swiss' document in the folder 'SWISS'
- added the German translation of the program and of the user manual
- added some tests
- improved the graphic user interface
- improved import schema handling 
- added option to load as a whole players from an own database
- fixed a bug in TPR calculation
- replaced the nonsense tie-break 'Not Played Games' with 'Score %', coherently calculated with the so called 'FIDE rule' ((sum of points in played games + draw for BYE)/number of rounds*100) even if it is not completely convincing. Then a win due to BYE or Forfeit scores half point less.
- added the Board number in the 'Results' page
- added a button to abort the engine

rev. 1.4 May 2010

WARNING: previously saved files need to be modified, inserting an empty row after "... // max Players per Team" and before of "...   // total Teams entered", by using a text editor. This procedure creates a row for tie-break criteria storage.

013 topic: improvement; engine
Problem:
Amalfi rating algorithm did not exclude absent players.
Solution:
Amalfi rating algorithm now correctly exclude absent players from pairing list 

012 topic: improvement; general
- the max number of Teams is now 500
- added icon 'Reset' in front of the 'Load player from database' field 
- in the same field are now handled <Enter> and <Del> keys
- in the list is now reported the status of the search if negative
- latecoming pairing has been improved
- tie-break criteria is selectable in the setup page
- non standard results are correctly handled
- IDs are reported on ranking
- output has been improved by adding page breaks. To evaluate it use the 'send to browser' button and do 'print preview' on that

rev 1.3 March 2010

011 topic: improvement; output
Problem: Missing FSI-Italia report.
Solution: Added FSI-Italia report and modified the program to allow easy national implementations.

010 topic: improvement; output
Problem: Missing alphabetic list of Teams/Players.
Solution: Added alphabetic list of Teams/Players after list by IDs. Also, ID assignment at first round has been improved.

009 topic: improvement; engine
Problem: Dubov implementation did not fulfill specifications about sorting players.
Solution: Now 1.st half (colour due White) is ordered by increasing ARO, increasing Rating and alphabetically; 2.nd half (colour due Black) is ordered by decreasing Rating, decreasing ARO and alphabetically (rule 6.2 first part).
 
rev. 1.2 January 2010

008 topic: improvement; engine
Problem: It happens in a swiss tournament, especially if there is a wide range of playing strengths, that two or more teams (players) may finish with perfect scores. This especially apply if the number of teams (players) is greater than 2 elevated the number of rounds to play.
Solution: Acceleration of the swiss tournament was introduced if the number of teams (players) top scored is grater then 2 elevated the number of rounds left to play, excluding the last round.
The aim is to pair the higher graded (ie "top half") players together, and then to use the top half non-winners to wipe out the 100% scores of "bottom half" players as quickly as possible. Acceleration ceases when this is achieved or when there is only one round left, whichever comes first. The system assumes that lower graded players will not repeatedly beat higher graded players. Like any probability-based system, it can be upset by a sequence of unlikely results. This does not alter the fact that in the long run it is the system most likely to avoid joint winners on 100%.
Acceleration is invoked automatically, there is no option in the setup. If you need to force acceleration, reduce the number of rounds to play. To exclude acceleration, increase the number of rounds to play.

007 topic: bug fix; engine
Problem: Dutch system revealed instable mainly if colour constrains are to be removed.
Solution: The algorithm for selection of a heterogeneous score group was fixed. Permutations due to color constrains were optimized.

006 topic: improvement; output
Problem: Scores were not shown on rounds.
Solution: Current round output now shows scores. 

005 topic: bug fix; other
Problem: Sometimes report to FED and WEB site were not generated.
Solution: A control on file extension was added.

rev. 1.1 November 2009

004 topic: improvement; output
Problem: Output is generated in HTML, that is great for printing and WEB publishing, but I need to have plain text too, i.e. to prepare a non standard report to FED.
Solution: A new button was added 'copy as plain text', that composes a plain text from HTML and copies it to the clipboard, so you may paste it in a text editor. The quality of conversion is quite good, better than you can do by yourself with standard copy/paste on each item! The procedure is optimized for screen resolution 1024x768.

003 topic: improvement; printing
Problem: In each board card it is not reported the round number. This may cause some confusion mixing board cards on the arbiter desk.
Solution: In each board card it is now clearly reported the tournament title and the current round number. 

002 topic: bug fix; printing
Problem: 'send output to printer' obtain unpleasant zoom factor. This is the main reason why a 'send output to browser' was added, so in their print preview some browsers offer a resize facility.
Solution: Thanks to a modification of Jan Michael Soan PrintMe class (www.jmsoan.com), the printing procedure was personalized setting the paper size to A4, margins at about one centimeter and a dynamic scale factor to correct the problem.

001 topic: improvement; engine
Problem: For Dubov and Dutch swiss systems, an extreme situation was found in a real tournament when the score group becomes too much heterogeneous (i.e. at the 6th round of a tournament with 10 teams, because of many pairs broken, a score group with 8 teams is generated, ranging from 8 to 0 points!!). In this situation the normal swiss 'S1 vs S2' may alter the tournament and secondary prizes win! This seems to be a problem related to swiss systems, not to the implementation. In fact, other programs fall too!
Solution: I inserted a suggestion to temporary switch to Amalfi rating system, because it proceed sequentially and can obtain more homogeneous pairs.

Associazione Sportiva Dilettantistica
Circolo Scacchi Bobby Fischer Cordenons

webmaster

Welcome to the JavaPairing home page!

Welcome to the
JavaPairing
home page!