Trials and tribulations of a computer scorer

Brighton - Day 6

The pairs and multiple teams at Brighton are all scored using table-top scoring devices (Bridgemates). The devices communicate wirelessly with a server attached to the scoring program and the scores are read into the scoring program. All the scorer has to do is turn the system on at the beginning and print out the results at the end. The players are able to type in their (EBU) member number at the start of the session, which allows their names to be read in to the scoring program from the membership database. So the scorer does not even have to type the names in from the names slips.

Of course, it's not that simple! For almost all the events at the Brighton Congress, we do not know how many contestants there are until the event starts. So we can't select the movement for the event until after the event has started. Hopefully while the players are playing the first board: all the sections are set up on the computer, the events are combined, and details of the movements are sent out to the Bridgemates; in time for the players to enter their member numbers and the result of the first board.

Then the players make mistakes: they make mistakes with the scores that have to be corrected at the table or on the scoring computer, or they make mistakes against the laws that lead to rulings and adjusted scores. By the end of a large event, the scorer can end up with a sheaf of "score correction" forms: some a simple change of declarer, some with the back covered in three-way weighted score (something like 40% of 4H+1, 40% of 4H= and 20% of 3NT-2). When players see recap sheets at the end of the event, more mistakes come to light and a pair has to go off in search of their opponents on a given board, to get them to agree the new score (according to Law 79). The TDs inspect the travellers or frequencies and find more errors (e.g. scores obviously with the wrong declarer) and when those are corrected the final result can be displayed.

But the TDs can make mistakes as well and then the Bridgemate system has to be put right. In today's Mixed Pivot Teams, we had examples of many of the things that can go wrong with the Bridgemate system. We had set up for 60+ tables in two sections and had a range of mirrored web mitchell movements prepared in a spreadsheet preloaded in the scoring program. When the dust settled at "game time", we had two full 30-table sections and so we loaded the appropriated movements and fired up the Bridgemate system ...

... nothing happened. This computer had been used for Swiss Teams in the afternoon and the server had been disconnected. When everything had been plugged back in and the programs restarted, we found some Bridgemates would not talk to the server. These Bridgemates were mainly in a group in the far corner of the room. The problem was located: the server attached to the computer for the afternoon's open pairs was still active, and was working on the same channel we were using for the teams; so some Bridgemates were "talking" to the other server, in a world of their own. Then some of the Bridgemates had been programmed to the wrong channel and were talking to no one. Finally some Bridgemates were programmed with the wrong section/table, so they were submitting results that belonged to a different table. The Bridgemate at the proper table would not let the players input their results but mysteriously reported that the round was complete.

Bridgemates were reprogrammed, scores were deleted from the server, and new scores entered on the right Bridgemates from the player's score cards (thank heavens for pen-and-paper backup). From early in round two everything was running smoothly and round-by-round results could be produced as each move was called. Of course, there were some mistakes to be corrected but they were soon sorted out. The final result was displayed and then it was off to the bulletin office to get the results in tomorrow's bulletin and up on the EBU website; and finally off to the bar.