Russell Sherwood Saturday, June 17, 2023
In high end Correspondence chess draws are the most likely result. This can have a demoralizing effect on Correspondence chess players and make progression in multi-round events difficult to achieve or control.
Many solutions to this have been suggested, but almost all of them fail to gain support from players for three reasons:
- They change the fundamental nature of the game itself – so it is no longer chess but a sub-variant of the game.
- They tend to introduce randomized features, which reduce the relative skill component of the game.
- The rating system can become distorted through those randomized features.
The Predictor Tie break addresses all these concerns.
The game of chess is the same. The game is played in the same manner as currently, but with an important addition -Each game becomes a mini-match between two players, with the one who predicts the most moves of their opponent winning that mini-match. (2 for a win, 1 for a draw, 0 for a loss). These mini matches are then totalled for the event and used as a tie-break, after all, others (Wins, SB and so on) to determine placings. The prediction count is only used for
Looking at the issues above, (1) It's still chess (2) No Randomized features have been introduced (3) The rating system is not affected.
If adopted on a wider scale this could be programmed onto the ICCF server (similarly to conditional moves) and for test events an Umpire would be utilized, who the players would send a sealed file prior to making each move.
Other positive outcomes to this approach:
- Players who simply play Stockfishes 1st choice move will not do very well on this tiebreak.
- Players who rely too much on previously completed games or analysis will also struggle.
- To do well, players will need to go down new avenues of play, which can be risky but also a lot more fun!
It is worth mentioning that some players will believe that if they do not play the 1st choice of the engine, they will be on a path to defeat. This is true if they go down lines where there is only one “Good” move, however, if skill is applied, they will navigate to those positions where multiple playable moves exist (ideally for their own side!). This will be a fine line but means that skill will have a more say in the outcome of a tournament rather than just simple brute computing power.
An ICCF proposal from Slovenia seeks to add this officially to ICCF rules, as an option. Proposal (iccf.com) This is near identical to the outline above, with a few technical minor technical details.
The WCCF will be seeking to organise its own trial event in near future. It will be free and unrated to encourage both participation and risk taking.
ICCFProposalWCCF
