Observation about Selecting Alliances

[Alex Pelan]

At nationals, we scouted for all of Friday and Saturday, plus some of thursday. We had scouts assigned specific team numbers so that every team got scouted for, plus we had some general scouts. We met on Friday night, but didn't get a lot done. Finally, right before the eliminations, we compiled a list of the teams we wanted. I don't remember a lot about it except 782 was first and 365 was second (we wanted teams that could cap and hang to complement our small balls). However, at the last second, one of our drivers called up our human player who was making the selection and told her to pick 27 instead.

[RyanMcE]

This was team 492's first year picking at regionals. We had the following information:
1) We were a ball manipulating robot that could not hang.
2) In qualifications, we won every match in which we were paired with a hanging robot
3) Our highest scoring match (170) was with a hanging robot (1031)
4) Our last match, a very narrow win, was against a very tough hanging robot (1031)
5) Our excellent scout had kept track of every teams hanging attempts and successes, among many other things. At the top of the list in terms of percentage and total hangs was an excellent partner (1031).

Since we had first pick, we (obviously) chose 1031. The rest of our list was sorted, roughtly, based on how many times the team's robots got to hang in an actual match. As the number one seed, many teams talked to us and tried to convince us they would be good alliance partners. We entertained them all, but this really had minimal effect on our choices - since we had excellent scouting, we really had all the information we needed ourselves. We went to a quiet place and ranked the teams that we wanted for first choice, then all the possibilities in order for second choice. Admittedly, we got very lucky with 957 - we had some lemons on our list above them that got picked, because we ran out of time to flesh out the list.

Thats how we did it, at least.

[opnickc]

We figured out early on that our best alliance would be with a hanger. We could control the mobile goal, cap both goals, and sometimes hang (didn't always have time). We needed teams who could consistantly hang, and who agreed with us as far as strategy (robots who compliment each other well are worthless if your alliance can't agree on strategy). At Buckeye, we were the first pick by first seed (340) and then picked another great hanget (1126). Since we not only had great robots, but also a great strategy, we plowed through eliminations 6-0. So, it's not only important to have a partner that compliments your robot, but also an alliance you trust. If you end up with a partner who won't strategise with you or follow your strategy, you almost might as well be playing by yourself.

So, ranking really has nothing to do with it. Heck, there could've been a team with no robot that seeded by luck alone, while a great bot could've lost every match. It's about going with a robot that compliments yours, and a team you trust.