Round By Round: How It All Works

Something New To (I Hope) Amuse You

I usually write a few sentences on how the leaders in each strat in two-session pair games got to where they ended up and you may have wondered how I derive this information. The initial version was written well before the pandemic but was always a bit suspect. I use a program that converts ACBLscore files into a massive dataset of everything in them, so theoretically I can derive whatever snippet I want from this data, and it only requires programming time with my meager skills to automate the process of making Bulletins. Until just after the 2023 Victoria Regional, I had each event produce a text file of all the data and I would peruse that. In addition, part of the Daily Bulletin program I wrote (3353 lines of Python and growing...) would track the matchpoints or VPs from each round and try to form a picture of everyone's rank after each round. At the end of the Victoria Regional, I had an unusually productive week, in which I corrected some nagging errors in this and added some charts to the pair games, then added the teams as well. These charts were added to the Victoria 2023 pages and from Penticton 2023 on, they appear with every daily page!

This is NOT at all the same as the 'one round to go' sheets that are posted near the end of every pairs session. My idea is different: if you get three scores in the first round that eventually (once played at all the other tables) turn out to be tops, you have 100% after round one. The posted sheets are calculated before the game ends, when the results in that last round are unknown; ACBLscore lists the standings based on the assumption that the final round was 'not played' and all of the boards are factored; in two-session events the leader sheets include results from the second session only. The charts I am now producing take into account only how many actual matchpoints (once play was complete and the boards were fully matchpointed) that you earned in every round. If you must complain that the last round leaders posted were different than the charts in the Bulletin, I will agree with you, because they were: they are calculated differently. If you insist that this means there is an error, I will direct you to this article rather than argue with you or agree with your suspicion that something is wrong and needs to be corrected, because frankly I don't have the time. This way is a better way to see where everyone really was. Nothing is "wrong" because two very different ways of calculating leaders going into the last round came up with different results. Just trust me on this.

For each game there are as many charts as there are strats. For each strat, there is a depth to which the overall awards pay out points, and the charts cover that many pairs plus about two-thirds more, so if there are 12 places, the charts will show the top 20 in that strat, in order of final finish. (If you finished below that, we figure you don't need your struggles publicized...) The names of the players are first, followed by a number of coloured boxes with their rank after each round to the end. In a game of more than one session, the first column after the names will be the carry-over total into the final session, the score you had in the first session for two-session pair games: this column does not appear for one-session games. Starting at, for example, '28' and moving to '5' is a good session in which you began in 28th place overall and ended fifth.

The background-colours of the boxes show your percentage score for the event so far in a colour-code:

Above 84%, which occasionally happens in the first round or in teams where you have 100% after round one for a blitz, the very dark green turns jet black. Since we are limiting the lists to overalls plus about two-thirds more, I haven't seen any pink or purple or ice blue yet, even in the lower strats.

If you are in a position that pays overall awards the rank will be white and in boldface, outside the overall depth your rank will be blue and normal. If you reach the top three your rank will be in gold, silver or bronze with a border around it of the appropriate colour. If you are viewing on a computer and you hover your mouse near one of these numbers (without clicking), you will see more details: what your actual score was on that round, your total score so far, and how many boards (or VPs in a team game) you are ahead of second place (if you are leading) or behind the leader in that strat. In a one session pair game, a full board is about 4% (100% divided by the number of boards played); in a two-session pair game, a full board is about 2%. On a phone or a tablet it is a bit more difficult to access these pop-up boxes, but sometimes holding your finger or stylus over the cell you want details on works.

(A few things to point out. The program calculates your score on percentages so far, so sometimes there are slight rounding errors that creep in and your final score is 0.01% away from the actual score. If there is a tie, the program has some random internal way of ordering the pairs and it takes the tied pairs in that order. Sometimes there are not enough pairs in a strat for overall awards and that strat is not listed. I probably will not write code for Board-A-Match teams, where the two pairs on a team usually play the boards at different times and trying to sort it all into some sequence of rounds would be impossible.

Time for a sample, from Saturday at Victoria:

This is the A-strat only from the Saturday Open Pairs in Victoria 2023. A bunch of things jump out right away from this graphic display that would not be possible from the text versions I was perusing before without a very close look. We immediately see that four of the top five pairs changed places in the final round, and hovering over the top five Round 8 scores shows that the fifth place pair was only 0.26 boards behind the leaders (in fact, the top two were tied) going into the last round. We can see that one pair began the evening in 23rd place overall at around average and was leading as the last round began! At the bottom of the overalls, the margin between the last overall and the next spot was only 0.02%! Even the bottom pair on the list, who got to the last listed spot from 32nd at the break, improved their event score from below 44.8% to 51.5% in the second session. The B and C charts have a lot more red in them!

All these charts add quite a bit of computer-generated data to the pages, and if you have a very slow connection you may have to wait a second or two for the page to fully load, for they are about nine times larger with all this stuff added. But we're going to see some very interesting races! Maybe we can put to bed this notion that many players seem to have, that a bad first round gives you no chance for eventual success! I hope it will be worth my limited-skill programming hours...