Scouting and the Prisoner's Dilemma

[Leav]

anyone else see the similarity?
The lose-lose situation seems to be the most profitable for both sides.....


just my 0.02$...
-Leav

[Billfred]

Originally Posted by Leav

anyone else see the similarity?
The lose-lose situation seems to be the most profitable for both sides.....


just my 0.02$...
-Leav

I don't follow, but you have my interest. Could you explain it a little more?

[Swan217]

The Prisoner's Dillemma is a situation where there are two people with two choices: Collaborate or Betray. If both people collaborate, both players get something. If both betray, they both get nothing. If one betrays the other who wants to collaborate, he gets twice as much as if they would have collaborated, while the naive fellow loses that much.

If you are referring to penalties, or offense/defense, I disagree. It is best to both play by the rules and both be offensive teams. You get the highest QP's, and the game is determined by talent, rather than competition.

If both are defense, then one still wins, but doesn't get as high of a score, which affects the rankings a lot.

[Mr. Lim]

I don't see the similarity in scouting, but there is a tangent when it comes to alliance strategy.

There are a few criteria in the "Prisoner's Dilemma" which fail to apply to the FIRST competition.

1) No team can benefit from defection: An alliance is scored as a whole group, not as individual teams. A team who defects from their alliance only hurts themselves and their partners. The Prisoner's Dilemma states that the defecting team gains a considerable advantage over those who choose to co-operate. This is NOT the case in FIRST, and a defecting team just hurts everyone on the alliance, including themselves.

2) All teams co-operating is the most beneficial outcome for all teams: Whereas the Prisoner's Dilemma states that the most beneficial outcome for your team would be if you defect when the rest of your alliance chooses to co-operate. Again, this is not the case, as there is no advantage gained in defecting from your alliance. Three teams working together is always better than having one or two robots not on the same page.

2) The Prisoner's Dilemma also states that the case of everyone defecting is more advantageous than the cases of partial co-operation, where your team has chosen to co-operate. Again, this is NOT true. Partial co-operation will again be more beneficial than complete defection within an alliance.

I guess I'm not sure where you were really going with this, but that's about the only way I saw an application for the Prisoner's Dilemma in FIRST.

-SlimBoJones...

[Mark Pettit]

Let's not go down the collusion road again, as happened during Stack Attack. Woodie Flowers had the last word in that one in the form of a memo dropped off at our pits at Nationals. He said that the Gracious and Professional thing to do is to compete in the game the way it was meant to be played.
For Triple Play, it's 3 vs. 3, not 3 w/ 3.

[Andy Brockway]

Originally Posted by Mark Pettit

Let's not go down the collusion road again.......

I do not think collusion will be an issue. Lose/lose means you are on the bottom of the qualifying rank. With the new (in 2004) format for ranking, the key is to win every match. Gone are the days of three times the losers score where helping the other team was the best way to get the higest ranking score. BTW, Beatty in 2002 used a strategy that guaranteed a low ranking but allowed them to be the National Champion.

This game is sufficiently difficult so that each team needs to be very careful on how they use their partners. Those teams with too much pride will hinder their alliance. Know your strengths and weaknesses and be honest about them. It is OK to say you are working on improving part of your functions or strategy instead of saying it works well.