English Chess Forum

In today's round eight at Mallorca, the top six-rated players are in order with white on the top six boards:
https://chess-results.com/tnr889534.asp ... ES&flag=30
I have no idea how to calculate the odds of that happening, but surely they must be 'astronomical'?

Pairings/Results
Round 8 on 2024/10/26 at 15:30h
Bo. No. White Gr Rtg Pts. Result Pts. Black Gr Rtg No.
1 1 IM Mascaro March, Pedro 2329 5 6 FM Voiteanu, Gabriel 2110 7
2 2 GM Bellon Lopez, Juan Manuel 2267 5½ 4½ Holliman, Bob 1930 17
3 3 FM Martin Rueda, Daniel 2251 4½ 4½ Spanton, Tim R 2013 11
4 4 IM Donchenko, Anatoly 2233 4½ 4½ Hamm, Rodrigue 1915 19
5 5 FM Hristodorescu, Daniel 2173 4½ 4 Yang, Sung-Jui 1882 24
6 6 WFM Trippold, Denise 2165 4 4 Bertz, Uwe 1892 21
7 9 Starley, Robert 2040 4 4 Hamm, Fabien 1972 15
8 28 Rossello Sanchez, Bartolome 1845 3½ 3½ Schenk Papke, Andres 2064 8
9 26 Stallinger, Vanessa 1871 3 3½ WFM Vilar Lopez, Monica 1973 14
10 31 Hops, Roberts 1794 3 3 Sayrol Clols, Josep Maria 2002 12
11 18 Fichtinger, Leopold 1927 3 3 Pujadas Cervantes, Gabriel Juan 1886 23
12 27 Cerdo Fernandez, Miguel 1849 3 3 Mach, Jan 1890 22
13 29 Abadia Pascual, Luis 1836 2½ 2½ Suarez Pozo, Juan Carlos 1876 25
14 20 Kamm, Guenter 1901 2½ 2 Donchenko, Irina 1990 13
15 30 Ramon Sanchez Campins, Pedro 1808 2 2½ Vasan, Navin 1677 33
16 34 Carranza Pons, Ricardo Nelson 1625 1½ bye
17 10 Kumar, Sanjit S 2037 0 0 not paired
18 16 Vives, Pepe 1938 3 0 not paired
19 32 Palou De Comasema Pou, Antonio 1742 2½ 0 not paired
20 35 Calzetta Torres, Raul 1527 0 0 not paired

I'd probably use a Monte Carlo method to simulate that sort of thing. It probably relies heavily on your assumptions and your pairing method (was this a tournament with accelerated pairings? It's probably more likely if it is).

Tim Spanton wrote: ↑

Sat Oct 26, 2024 1:28 pm

I have no idea how to calculate the odds of that happening, but surely they must be 'astronomical'?

I would doubt that because pairings aren't random. Instead they follow a set of rules which rank players in rating order. Also to the extent that ratings predict results, it's plausible that rankings towards the end of a tournament are the same as rankings at the end. There does however have to be only a limited number of circumstances where the alternating colour sequence forced at the start of a tournament becomes a same colour sequence at the end. Also towrds the end of a tournament it can be likely that the leading group have mostly played one another already thereby increasing the likelihood that they line up in sequence.

If you did try to calculate it, what's called a Monte Carlo method might be used. You take a tournament, use ratings and have an additional model to determine probabilities of win/draw/loss. You then use the pairing method and rules to generate pairings. Run that a thousand or tens of thousands of times and see how often the effect comes up. Bear in mind the outcome depends on the reliability of the ratings based model used to simulate results between individual players.

Hmm! Well, I can tell you the pairings were not accelerated.

When tournaments use accelerated pairings is it just for the first couple of rounds to kill off the chances of 'undesirable' players sneaking though later or do they continue to use them for the entire event?

Joey Stewart wrote: ↑

Sat Oct 26, 2024 4:17 pm

When tournaments use accelerated pairings is it just for the first couple of rounds to kill off the chances of 'undesirable' players sneaking though later or do they continue to use them for the entire event?

It tends to be four or five rounds in a nine/ten round event.

Joey Stewart wrote: ↑

Sat Oct 26, 2024 4:17 pm

When tournaments use accelerated pairings is it just for the first couple of rounds to kill off the chances of 'undesirable' players sneaking though later or do they continue to use them for the entire event?

If you do a Google search for Baku rules chess, you should get the official FIDE version of accelerated pairings. They never accelerate for the entire event as the top players eventually have to be given an opportunity to chop down "undesirables".

In recent years accelerated pairings have become less common, a possible reason being that they are messier than standard pairings when using computer software. There's a case for using them in five round weekend events when section numbers get to sixty and beyond, but it no longer happens.

Less of an oddity, perhaps: all six top boards were won by White