When Tournament Directors Lose the Plot

 The last Ballarat Tournament had 32 players in 3 divisions:
Division A: 10 players with rating range 1255-1611
Division B: 12 Players with rating range 747-1211
Division C: 10 Players with rating range 500-861

 

Don’t ask why the lowest rating in division B is lower than the highest rating in division C. You would not get a satisfactory answer. It could be related to the weight, height or age of players.

 The serious matter is that the TD in her wisdom, or lack thereof, told the congregation that it was a round robin, where each group of more than 7 players would play 6 rounds. Pairing was done in advance using a secret recipe known only to the ingenious TD. What she missed was that 32 players was an ideal figure for a round robin of 7 games and 4 groups. She would have returned home in Mornington even faster than she aimed to achieve by her innovation of "partial round robins".  

 

  

Riddling the Riddle

We have two versions of the results: the vesion declared by Marj Miller (MM) on the Ozscrabble forum, which we may call version 1; and the version declared by Carol Johnsen (CJ), which we may call version 2. The latter version was finally accepted and recorded by Martin Waterworth. Martin can only go by the files forwarded to him. I assume that he had the two versions of the file. I also assume that he would rather not involve himself by comparing the two versions.

We also have two explanations of this anomaly, one given by Mohammed (MH) and another given by CJ, after being challenged by Harry Malcolm.

MH claims that if a single error, in recording the result of game 10 between him and Lina Camilleri (LC), is properly corrected by reversing the scores of 382/276, the outcome would be LC winning third place by a total margin of 378.

CJ claimed that another error cropped up later, which is a second error recording the result of game 6 between LC and Eileen Mills (EM). A win by LC in version 1 was corrected by reversing the scores. So instead of LC winning by 8 points, she lost to EM by that much.

Without any further documentation or much ado, the smoking gun becomes clear: If LC lost to EM by only 8 points, and won against MH by the huge margin of 106 points, how can you explain her total margin in version I being identical to her total margin in version 2? Also, the only correction applied to EM results is a reversal of her score against LC, which involved a margin of 8 points only. How could you explain the total margin of EM leaping from 79 in version 1 to 109 in version 2? There has to be a series of corrections applied to version 1 in order to produce version 2. You don’t have to be a mathematician to come to this conclusion.

Version 1 of results published on Ozscrabble:

       W     M                                   Old   Chg  New

  1    9   +670  Gillian Renwick     974  +56   1030

  2    8   +303  Elize Plaganyi        803  +67  870

  3    7   +320  Norma Engel          911  +24  935

  4    6   +362  Lina Camilleri        802  +20   822

  5    6   +95   Gayle Cameron      810   +10  820

  6    6   -11   Carole Eden           874   +12  886

  7    6   -92    Brian Rowe          (589)            (731)

  8    5  +105  Angie Winkler       565   +35    600

  9    5   +79   Eileen Mills          768    +8     776

 10   5   - 49  Mohammed Hegazi 943   -24    919

 11    5  -115 Ann Goodwin         792    -9      783

 12    5  -134  Gianna Devcic       906   -17     889

 13    4  +290 Irlande Alfred        971   -58     913

 14    4   +67  Marisa Nuccitelli 770   -27     743

 15  3.5  -257 Judith Westwood 762   -29     733

 16    3  -444 Mary Mackie         594   -12     582

 17  2.5 -402 Balada Catanchin   632   -31     601

 18    0  -787 *Trudy Kennedy                       ( 500)

 

Version 2 of results accepted and recorded as final

Rank	Player	Wins	Margin	Rating change
1	Gillian Renwick	9	670	974	+5	1030
2	Elize Plaganyi	8	303	803	+6	870
3	Norma Engel	7	320	911	+2	935
4	Lina Camilleri	6	362	802	+2	822
5	Eileen Mills	6	109	768	+2	796
6	Gayle Cameron	6	95	810	+1	820
7	Carole Eden	6	-11	874	+1	886
8	Brian Rowe	6	-92	589	)	731
9	Angie Winkler	5	105	565	+3	600
10	Ann Goodwin	5	-115	792	-	783
11	Gianna Devcic	5	-134	906	-1	889
12	Irlande Alfred	4	290	971	-5	913
13	Marisa Nuccitelli	4	67	770	-2	743
14	Mohammed Hegazi	4	-79	943	-4	899
15	Judith Westwood	3.5	-257	762	-2	733
16	Mary Mackie	3	-444	594	-1	583
17	Balada Catanchin	2.5	-402	632	-3	601
18	Trudy Kennedy	0	-787			500

 

Final accepted results of Lina Camilleri

Round	Opponent			Scores		Margin	Wins	Cum. Margin
1	594	Mary Mackie	[H][S]	394	264	130	1	130
2	810	Gayle Cameron	[H][S]	380	295	85	2	215
3	974	Gillian Renwick	[H][S]	326	385	-59	2	156
4	565	Angie Winkler	[H][S]	423	294	129	3	285
5	803	Elize Plaganyi	[H][S]	295	367	-72	3	213
6	768	Eileen Mills	[H][S]	316	324	-8	3	205
7	911	Norma Engel	[H][S]	270	320	-50	3	155
8	762	Judith Westwood	[H][S]	407	329	78	4	233
9	906	Gianna Devcic	[H][S]	367	344	23	5	256
10	943	Mohammed Hegazi	[H][S]	382	276	106	6	362

Final accepted results of Eileen Mills

Round	Opponent			Scores		Margin	Wins	Cum. Margin
1	906	Gianna Devcic	[H][S]	343	387	-44	0	-44
2	762	Judith Westwood	[H][S]	394	332	62	1	18
3	770	Marisa Nuccitelli	[H][S]	378	335	43	2	61
4	810	Gayle Cameron	[H][S]	405	298	107	3	168
5	974	Gillian Renwick	[H][S]	338	421	-83	3	85
6	802	Lina Camilleri	[H][S]	324	316	8	4	93
7	803	Elize Plaganyi	[H][S]	278	356	-78	4	15
8	874	Carole Eden	[H][S]	366	397	-31	4	-16
9	943	Mohammed Hegazi	[H][S]	336	321	15	5	-1
10		Trudy Kennedy	[H][S]	381	271	110	6	109

 

Final accepted results of Mohammed Hegazi

Round	Opponent			Scores		Margin	Wins	Cum. Margin
1	792	Ann Goodwin	[H][S]	410	387	23	1	23
2	770	Marisa Nuccitelli	[H][S]	340	296	44	2	67
3	803	Elize Plaganyi	[H][S]	317	381	-64	2	3
4	589	Brian Rowe	[H][S]	321	291	30	3	33
5	874	Carole Eden	[H][S]	289	330	-41	3	-8
6	565	Angie Winkler	[H][S]	417	293	124	4	116
7	974	Gillian Renwick	[H][S]	353	399	-46	4	70
8	911	Norma Engel	[H][S]	358	386	-28	4	42
9	768	Eileen Mills	[H][S]	321	336	-15	4	27
10	802	Lina Camilleri	[H][S]	276	382	-106	4	-79