Quote:
Originally Posted by Ed Law
If FIRST wants to follow the sports model, tell me a sport where there is scoring that OPR is not applicable? The only recent game I know that does not work is 2007 Rack "N" Roll where the score goes exponentially up if you have a few tubes forming a chain.
|
OPR has biases in each and every year. In general, given the nature of FIRST scoring (finite number of game pieces and/or scoring locations), as the quality of teams at an event trends up, OPR for teams near the top and bottom flattens out - at some point, there are simply not enough tubes, or balls, or whatever for each team to score to their maximum ability. Compare the OPRs from the Michigan State Championship with those from a middle-tier Regional and tell me that they make sense.
Regardless, I think there is no need to balance scheduling with any consideration of robot performance or capability. Random is at least as "fair" as any other method. If you draw an easy schedule, enjoy your cakewalk but have fun getting bounced out of the playoffs if your bot isn't up to the task. Likewise, if you draw a tough schedule but the robot performs up to its abilities, you will be drafted and have a great chance to go deep into elims. IMO, the best way to be "fair" about it is to maximize the number of matches (random draws) that each team plays.