The match scheduling program gave several teams (including ours, 87) some interesting match pairings. We are paired with one team three times and another twice. At least five other teams had a similar “problem”, and six other teams that I asked did not have this “problem” at all.
Now, it might have something to do with the number of rounds and the number of teams. At VCU, there were eight rounds per team and 66 teams. At Philly, there are ten rounds per team and 45 teams.
However, it is definitely possible to produce match pairings that don’t have such lopsided results.
Is there a description in the rules that describes this process? I looked for such a description, but did not find one.
Other than that, this regional competition is very good. The quality of play has improved, and there are more ball handlers here than in Virginia.
hey RV person
i know what you mean.
we were paired with a robot 3 times who did not work AT ALL and was dead. we were lucky to pull victorys 2 out of the 3 times we were with them, its still a miricale how it worked
i would really like to know why the matchings where made this way.
we had the same exact problem as u gguys bu the team we were paired with three times fixed itself once and then broke down again and was unable to do anything except just sitting there
they did the same pairing thing at Johnson and Johnson Mid-Atlantic too. We were paired with 4 different robots two times each and even went against the same alliance twice.
i never went agasint the same aliance twice,
but the first time the robot broke down 30 seconds in
the second match they didnt move
the third match they moved and came back in and shut their robot off. it was a pain
I know someone who works closely with the computer side (mainly the webcast) of the VCU and Philly regional. Apparently it was the program that caused this (the guy I know said the person who wrote it has already been talked to). If you noticed, teams would play once, than 10 (+/- 1) matches later they’d play again. We noticed that every time we were right after team 21. So, basically the problem is the code did not do a completely random job, because they were trying to keep the matches evenly spaced out. There should be no problems at nationals (hopefully :))
The problem is not the code, it is the people that blindly follow the code. They could have shifted some of the matches around to avoid this mess. :mad:
*Originally posted by mrd_udhs *
**The problem is not the code, it is the people that blindly follow the code. They could have shifted some of the matches around to avoid this mess. :mad: **
Having run this year’s scoring system at 1 regional, and last year’s system at numerous events, shifting matches around is nowhere near as easy as it sounds for a number of reasons:
This year’s system is SQL Server-based, meaning that there is no easily accessable table to go into and edit the pairings. It would take a number of SQL statements to change even one match.
Even if the match table was accessable, internally, a team “index” based on the list of competing teams at an event is used rather than the actual team number. This index is based on the physical order of team records in the database, and may not be consistant from one import of a team list to the next.
At the majority of the regionals, the operation of the scoring system is placed on one person. After ensuring that all teams have arrived and passed inspection, this person tells the scoring system to generate matches sometime Thursday afternoon/night. If this person was to go over each match individually and “tweak” things as needed, you are looking at several hours of manual labor just tweaking the matches, not to mention the time entering this information into the match tables.
In a perfect world, there would be the time and the manpower to figure out each match by hand. However, due to the ever-increasing number of teams, and the small group of core staff at any given event, this is simply not realistic. While it may seem that we are “blindly following the code,” due to the logistics of the event, and the nature of the system, it is easier to follow the code, and perhaps tweak the code to get the desired result(which is being worked on, FYI) rather than eliminate the code entirely, and do the majority of the work by hand(and yes, I have both generated and “tweaked” matches by hand, and even with the system doing the initial generation, if hand modifications were done, that would be the majority of the work.)
