You should always think about the way the ranking system works early in the build season if you want to be picking on Saturday.
I don’t think approaching a match like this is very gracious or professional. Very rarely will you be “locked in” to rank 1, you will usually be competing with at least 2-3 other teams for that spot. Also, throwing a match will look really bad on your scouting, and if I approached a team and asked what went wrong in a given match, if they told me they were trying to prevent an alliance member from getting an RP, that is an instant red card to me. 0% chance I am picking that team for my alliance.
So I should be thinking about how my teammates might throw a match. How can I prevent that? How would I know that was their intent? My point is still the same.
I think you should be more worried about making sure your robot’s performance is good enough that edge case scenarios are meaningless.
Any advantage you’d gain from throwing a match is not worth the potential consequences.
Even if you didn’t get carded/thrown out officially I would personally never pick a team that made the decision to actively hurt the advancement of their teammate, and it is a serious enough offence that I would be taking it into consideration for the next few years.
This isn’t even a Noodle Agreement deal where both teams come out ahead, throwing a match is something that you should never do even if there was no rule against it.
I know that Woodie would come kill me in my sleep if I tried to actually approach a match like this, but I was only thinking of theoretical situations. I get that it’s a very scummy thing to do. I was just thinking theoretically, I could, Not that I should.
The probability of winning the event very likely does not strictly increase as your rank improves. There is very likely a “serpentine valley” of sorts where teams are more likely to win the event if they either seed better or worse. My rough guess is that this valley is centered at about rank 10 based on this graph:
Relevant post: paper: Miscellaneous Statistics Projects
If getting 4 RP will put you at about rank 10, it’s probably in your best interest to get 3 RP instead. Someday I’ll do a more thorough analysis as the above one is pretty crude, but I’m moderately sure event win probability does not strictly increase with increasing rank.
There has to be a fine line somewhere between being attractive enough to be a second pick, and being so attractive that you get picked second by a lower seeded alliance. Everyone wants to be picked by the alliance most likely to win the event!
Like Caleb, I’m interested in doing some more analysis, if I have time. I don’t think the Elo rank at the start of the event is the right metric to use here, though - the qualification ranking at alliance selection would be. Such an analysis has value beyond a “throwing the match” question as well.
Of course, intentionally throwing a match is horribly illegal, not GP, and not something I would ever recommend doing.
Besides being “a scummy thing to do”, I would refer you to these rules:
C3: Asking other teams to throw a MATCH – not cool. A team may not encourage an ALLIANCE, of which it is not a member, to play beneath its ability.
C4: Letting someone coerce you in to throwing a MATCH – also not cool. A team, as the result of encouragement by a team not on their ALLIANCE, may not play beneath its ability.
I’m not saying you should be thinking of how to get around this particular situation, but you should definitely consider ranking points and the way the ranking system works in your design process.
Sandbagging in any sport is frowned upon and I don’t think reflect the values of FIRST. There are even rules against it mentioned above but of course enforcement is tougher. When we win a match where the opposing alliance robot disconnects or something like that I always get a lot less excited. Sure we won, but why celebrate if you didn’t win against the alliance at their best?
There’s a stark difference between understanding how the ranking system works and thinking of underhanded ways to manipulate the rankings.
I have seen highly esteemed teams appear to do this, as recently as last year and as long ago as 2009, but that does not mean it is not very wrong. I don’t subscribe to win at all costs; I have to sleep at night.
You could, but why? If I were Team B, I would accept the first alliance’s pick every time, especially if I snuck into 8th place and would otherwise have to face them in quarters.
I think the author of the OP is trying to understand the purpose of the rule, not strategizing about how he can use it. It’s a “Why would anyone ever want to do that?” sort of question.
The answer is generally that it is extremely rare that such a strategy would be a good idea for a team, but just in case it happens to be one of those rare situations, that tactic is outlawed, because it would be a kind of scummy thing to do. Here’s another situation where it could be to a team’s advantage to throw a ranking point.
Team 9746 is in first place, with a three point lead, going into the last qualifying match. Team 9234 is in second place. Team 9087 is in third place, one point behind team 9234. Team 9746 and team 9087 (the 1st and 3rd place teams), are “friends”. i.e. they’ve worked together before. Their coaches know each other. They have complementary robots…whatever. They have met and agreed that if team 9746 has the opportunity to pick team 9087, they will do so.
In the last match, team 9234 and team 9807 are on the same alliance. A glance at the score sheet indicates that if team 9234 scores three points, they will be tied with the first place team and the tiebreakers wil fall to 9234, making them the number 1 seed, dropping 9746 into second place. At that point, team 9234 could pick the third place team and those teams, who had agreed to ally with each other if possible, could not be on an alliance together. (Because a team who declines an invitation cannot accept a later invitation. They can only form their own alliance.) So, it is in the best interest of team 9087 (the current third place team), to score two points by winning the match, but to not score any bonus ranking points.
Such a tactic would be considered unethical, so it is prohibited.
By the rules you are allowed to do whatever you want because you’re not letting a team convince you to lose. I agree it’s not professional but as I read the rules it is legal
Following up on this, here is a graph of each team’s win probability versus their end of quals rank:
Full post here.
Summary is that I don’t see good evidence of a “serpentine valley” using this methodology. Much deeper investigation would probably be needed to prove such a thing is real and large enough to be measurable.
The short answer is that often times qualification matches do a poor job at seeding. If the 2 best robots in the regional are seeded 2nd and 14th and those two robots face each other in the final match, then the 14th seeded alliance has a huge incentive to let the 2nd seed alliance win. In that case the 2nd alliance may become the first alliance, pick the 14th seeded team, and the two best teams go on to win the regional.
If the 14th alliance wins their last match maybe they move up to 12th, but there are no outcomes where they get paired with the other “best” robot in the regional. Now neither team is the favorite to win it all.
This is very much against the rules and the spirit of the game. If you truly want to be paired up with the other “best robot” the correct strategy is to design a robot that seeds high enough to do so.
All you need to do to get to pick your first alliance partner from a clear field is to have a strategy of getting the first seed? I wonder why only one team per event uses this strategy?
Seriously, if you throw the match on your own drive for strategic purposes that benefit your own team, it’s not against the rules. Doesn’t mean you won’t find yourself on some DNP lists for it.
Any game will create some incentive for some team to throw a match. Perhaps a team wants to fall farther in the serpentine draft by looking less reliable than they are. Perhaps they don’t want to give a loss to someone they want captaining an alliance with them. Maybe the RP they would earn would cause an alliance partner to seed higher than a team they would rather be picked by.
Intentionally sabotaging an alliance’s chances of winning is unethical. Everyone needs to work together to achieve the objectives of the alliance, which is usually to maximize seeding points in qualification matches. As quals go on and only a couple of teams are left in the hunt for the top seed, some alliance members may not care so much about the match results. But as long as somebody does, every team owes it to their alliance to try as hard as they can. Anything less is shortchanging your partners.
Not exactly what OP was about, but 2 RP > 1 RP. If you’re in a very close matchup and can score significantly faster on the cargo ship than on the rocket ship, by all means give up the rocket ship, and win the match instead. No need to get greedy and try and get the rocket RP along with the win.