gjfeldman

The 299er and 0-20 were missing today. I did not see any announcement. Are they gone for good, or just taking a break?

A big improvement. It makes sense to have a standard system in these games in which you are playing with a partner you do not know and cannot communicate with.

A few weeks ago, the ACBL 299er daylong tournaments appeared. I wondered whether one would do better playing open daylongs or the new 299er, so I randomly selected 400 results for the C stratification from MP-1 and 299er daylongs. The results are shown below in a plot of masterpoint reward vs. percentage score. The blue dots are for MP-1 and the red dots are for 299er. In both cases, about 45% of the results get zero masterpoints. The first thing you notice is that there is a large range of rewards for each percentage score. The ACBL does this on purpose through sectioning and a hidden metasectioning so that the stronger players do not gobble up all the masterpoints leaving weaker players discouraged. This is generally accepted since it all averages out in the end. However, for a given percentage score, the 299er rewards are less than the MP-1 rewards. The dotted curve is a third order polynomial fit to the data, showing that to get a particular award it is necessary to score about 5 or 6 percentage points higher in a 299er tournament than an MP-1 tournament. But my experience is that, on the average, I do score about 5 or 6 percentage points better in 299er tournaments. I think that this expectation may vary for different individuals depending on their skill level and their stratification level in both types of tournaments.

I wrote a program to randomly deal a million hands to see what the statistics are for best hand tournaments. The results are shown in the two graphs below, one for the the best hand and one for the best hand partnership. Each shows the percentage of the time for getting HCP and total points. The total point count is basic with no corrections (e.g. honors in a short suit, partner's suit, etc.). Some conclusions: The most common HCP for the best hand is 14 and the average is 15.1; the most common total points is 16 and the average is 16.7. The same numbers for the best hand partnership are HCP: 23 and 23.4, and total points 26 and 26.6. If you take an opening hand to be 13+ total points, then you will get an opening hand 97.4% of the time. If you take game to be 26+ total points, then the your partnership will get game 60.4% of the time. Your opponents will have more HCP than you 11.2% of the time, the same number (20) 8.2% of the time, and fewer HCP as you 79.6% of the time.

The reason I posted this was that I recently played that tournament for the first time and I chatted to the table that I was playing GIB. I got a reply from an opponent that did not object to that, but implied that the standard was SAYC. I wanted to check if that was true.

In ACBL individual tournaments (non-robot, 4 rounds of 3 boards each with a different partner), is there a default set of conventions assumed?

In a two-table Howell movement tournament, when one table does not play a board due to time limitations, it correctly receives an average score, i.e., 1/2 MP for each pair. The same should happen to the other table, but it doesn't. The pair with the positive score gets 1 point and the other pair gets zero. The same thing happens in a 3-table tournament when 2 tables fail to play a board.

Thanks. This confirms my supposition. In large tournaments, the sectionization of overall awards is independent of sectionization into 15-player sections,

I understand about stratification. However in this case, both players are in the C strata. The player in 1st place in the C strata gets 0.63 MP and the player in the 2nd place gets 0.96 MP. It seems to me that there are only two possible explanations: either it is a program bug or there is some substratification going on, possibly involving the overall awards. I suspect the latter. Here is some more information. The player in 1st place has between 200 and 300 BBO points while the player in 2nd place (me) has between 20 and 50 BBO points. Both players are in the top 7% of the tournament.

These are ACBL points.

In the Feb 2 Daylong MP1 tournament, in section 4, the player who finished ABC = 211 got 0.63 MP and I, who finished ABC = 322, got 0.96 MP. How is this possible? I looked for other cases like this and only found one more. In section 44, the player who finished ABC = 211 got 0.63 MP and the player who finished ABC = 422 got 0.67 MP.

I cannot argue with that answer. However, it does seem to work well enough for every person to partner with every other. Since all boards are played simultaneously, it would not be hard to fix the algorithm so that each person would compete against every other person twice.

I set up a short 2-table individual full Howell movement tournament yesterday for one of my small clubs. 7 rounds, 2 boards a round. Every player played with a different partner each round, as they should. However, I played against 2 players 4 times, 2 players 2 times, 2 players 1 time, and 1 player not at all. This looks like a bug in the individual full Howell movement. The tournament number was 22822-1610460257.

There appears to be a change in the team match setup form. It used to stay active after the match started. It now disappears as soon as the invitations are sent. This leads to a waste of time, particularly when a player does not receive an invitation for some reason. Even without that, it is convenient to have it stay because usually the first two pages usually need minimal or no changes, and the third page often keeps some of the players in the same position. If others do want it to stay, they can simply close it. Can we get this change reversed, or at least give an option to keep it around?