Sasku tournaments

At the moment, I would suggest: 15 hands are played in rounds, with as many different pairs as possible. This way, the tournament should last about 45 minutes. In each round, a new hand is generated and all pairs play this hand. The player receives a result of -6 to 6 from one round, depending on how the hand turned out. Now we need to calculate the percentage of the average result in this hand, so that the best one gets 100%, the worst 0%. The opposing pair then gets 100% minus the result of this pair. They get 50% for a free round. The question is how to calculate these percentages if the results of one hand are, for example, 4, 3, 3, 2, 1, -1. I will think of the exact formula another day. Who knows, can help. At the end of the tournament, we take the average of these percentages (analogous to bridge). The rating calculation would be similar to a doubles / sudoku tournament.

Interesting idea

Tournaments have to be organized later and it seems like I have a lot of tournaments left unfinished when, for example, friends come to visit, I don't play the tournaments to the end or I get a phone call saying I have to go to work now, for example, if I leave it unfinished then my partner will be offended too. How should my partner not suffer because of me thinking about it....

The minimum number of participants in the tournament will be 12 (to provide an adequate comparison with other tables)

I'll take the sasku tournaments next. I have some doubts about the format, though. In any case, the sasku tournament seems more exciting and concrete at the moment.

Tonight at 9:00 PM we will attempt to hold the world's first online sasku tournament. Don't miss out!

My proposal for tournament scoring. Let's assume that the same hand is played on a K table. Each table has 4 players: North-South-East-West. Each result is in the form "big points, small points" with respect to NS. For example "2, 62" or "-1, 58". 1. Each NS result is compared one by one with the remaining K-1 results. The results of these comparisons are summed. So, it remains to create a scoring system for comparing the results of the 2 tables. There are several ways to do this. 2. When comparing the two results, points could be divided on a 2:0 scale. If the division is tied, then the result is 1:1. 3. Only big points could be counted. If the big points are tied, then 1:1, otherwise 2:0. So in the case of 4, 3, 3, 2, 1, -1, each result would have 5 comparisons, 10 being the maximum. The points would be: 10, 7, 7, 4, 2, 0. This can be converted to percentages by dividing by the maximum.

Done! At first there was criticism about different solutions, but it seems that today's solution is also suitable - yesterday there were 5 tournaments, with 13-19 participants in each tournament. Now a little about the solution. Distribution into tables Currently, players are divided into tables on the principle that the strongest play each other, and the bottom of the table play each other. There are proposals to change this. One option would be to randomly pair players. However, the problem here is that the strong players do not want to pair up with a weaker player. Maybe today's solution is not the worst. Evaluation of the distribution Think of a solution where we evaluate the distribution's contribution to the average result. For example, if the results of the hand in the NS position are 4, 3, 2, 0, -1, -2, then the average result is 1. The largest deviation is 3. The results of the hand are then 100%, 83%, 61%, 39%, 17% and 0%, respectively. Players in the EW position get the opposite result, i.e. 0, 17, 39, 61, 83 and 100, respectively. For a free round, they get 50%. There have been suggestions to change this slightly (gauss' previous post in this topic). Maybe we'll take it. Comprehensibility of the tournament table In the tournament table, each user will have a row for each round result (for example) N:-2:60. This means that the user played in the NS position, EW won the game with a result of 2, but the NS result was still better than the average (60%). A simple method to evaluate your hand after the game is over: Search the table for more results in your position. If you see that someone got a better result (in this example N:0) or a worse result (N:-4), you will get an idea of whether your result is better or worse than average. Rating system We use a system familiar from dueling games for the rating system: At the end of the tournament, we take the table and for each user we add up the points that they would have lost in the duel to those they were behind in the tournament table and won from those they were ahead of. The system did not work with a large number of participants (at the Mind Sports Olympiad last year, a single sudoku tournament earned 2000 rating points). We set the limit at -420, which seemed like a lot, but at the moment it is 336 or something like that, still a lot. We will probably set the next one at 210. Since the tournament lasts about 1 hour, the question comes down to how much a particular user's rating could change in one hour. If you look at duel games, 210 would mean that grey won 5 games in a row against red (5*42 = 210), which is realistically possible in one hour. If black comes last or grey comes first in a card game tournament, then you have to accept either -210 or 210 rating points. PS! You can't watch the games in a Sasku tournament because all tables play with the same cards - other tables could then peek to see what cards are on the floor.

I also listed the tournament scoring system on the rules page: http://www.vint.ee/sasku-reeglid/