Winning a Match vs. Winning Strategically

[Anupam Goli]

Originally Posted by Chris is me

I am also saying that teams do that... I'm not sure where we disagree? All I am saying is that teams talk once they know their fate. 2791 did this at WPI once the top seed was secured. We didn't ask anyone to throw any matches, and never have.

I don't think we were disagreeing. I was just elaborating what actually happens at 10 AM isn't asking teams to throw away matches or back door deals. Sorry for the miscomm!

[Ivan Malik]

I highly suggest the you re-read his post. No where does it say "the top 8 teams" It says "the top teams." Which from 7 years of watching alliance selections and seeing this happen at almost every regional/district, participating in side conversations with other teams myself (not throwing matches or anything like that, but it could have looked like it, due to scheduling, to those not in on the conversation) I can say yeah this happens all the time. On top of that I know for a fact, from their own lips, that many well respected teams do some very VERY shading things during alliance selection that on no plane of thinking follow the tenets of GP or coopertition.

The post also doesn't explicitly refer to this as a "back door deal." That statement is followed by a "or whatever you would like to call them many alliances are prearranged." Hence foul play isn't necessarily what was meant.

As far as the OP: My personal opinion is that if you have to explain yourself to the community, then you are in the wrong. The community is the ultimate police of GP/coopertition. What strategies are acceptable are up to the group at-large. For instance I would say that "scorched earth" alliance selections are not in the spirit of coopertition, but they have been come to be accepted by the community.

[Anupam Goli]

Originally Posted by Ivan Malik

As far as the OP: My personal opinion is that if you have to explain yourself to the community, then you are in the wrong. The community is the ultimate police of GP/coopertition. What strategies are acceptable are up to the group at-large. For instance I would say that "scorched earth" alliance selections are not in the spirit of coopertition, but they have been come to be accepted by the community.

But how does coopertition have anything to do with alliance selections? Coopertition is about helping other teams compete in the sense that if team xxxx has a robot that is down, powerhouse team yy helps xxxx repair that robot so they can get on the field for the next match. If you wanted coopertition to happen in alliance selections, then you'd have to forbid teams declining eachother too!

[Ivan Malik]

Coopertition is also about not hurting your opponents to win and wanting to play against them at their best.

Scorched earth can make it so no one is at their best but you. Declines are in there so that you have the option to decide whether it is in your best interest to compete with a team and to decide whether you are fighting fit (as it were) for elims. Scorched earth games this and uses the decline rules to hurt the declining team. Its all about the intent. It isn't measurable or provable to 100%, but it is recognizable. This is why it is covered by something soft like coopertition and not something hard like a rule.

Edit: I recognize that this is my own personal view and is not shared by the vast majority of FIRSTers.

[TheMadCADer]

Originally Posted by Thunder910

It's not about playing to win just for your team. It's about your team, your sponsors, your pride, but most of all- it's about your alliance partners. If you play half-heartedly or completely throw a match, you're hurting your alliance partners who want to win to improve their position, seeding, and to prove their robot's abilities.
In your story, your alliance partner wanted you to play to win- not doing so would be a disservice to them.

If both your partners agreed that losing the match was the best course of action for all three teams- then I suppose this might not apply.

The thing is, you don't have to play half-heartedly or just sit there in order to lose the match. Just point your shooter at the other goal (especially if you can get some of their coloured discs and score them in their pyramid goal, that way they get 26 points each thanks to G17). That way you don't make yourselves or your sponsors look bad, it just makes it obvious that you want the other alliance to win. Get your alliance partners to help out, and they'll get an easy way to prove their abilities. You can probably put up a really high score together.

Obviously you should talk to your alliance partners about it, but even if they don't want you to, I don't see that as an issue. You shouldn't be obligated to win for some other team's sake, especially if it isn't in your own best interests. One thing I like to stress in the early stages of design is to never rely on your random alliance partners for anything. If they are only ranked high because of other teams helping them, then they probably shouldn't be ranked high. If they want to rank high, they should do it themselves. It gets completely different when there are two approximately equal teams on each alliance, and your behavior is basically choosing which will rank higher, but if it's just a couple teams that want to move up from 30th to 20th, I wouldn't hurt my own chances to help them, because it honestly wouldn't matter since they couldn't pick anyway. They should just focus on showing off.

I don't like to see teams making "pick deals" for losing matches, but rather I'd like to see this when there are two other teams that could seed first, and one is distinctly stronger than the other, and you are playing against that stronger team.

Imagine three teams, given numbers based on how many points they score. Team 200 scores 200 points each match, Team 150 scores 150 points, Team 80 scores 80, and every other team scores fewer than 80 points. You are on Team 150, and you already have a few losses with just one qualification match left, against Team 200, but you have better alliance partners and could win the match. Both Team 200 and Team 80 are undefeated, and Team 80 has already played all of their matches. If you win your last match, Team 80 is the #1 seed and picks Team 200. If Team 200 wins that match, they get the #1 seed and pick you, Team 150.

It should be obvious which situation gives you a better chance of winning, since no other robots can top Team 80, and you can't top Team 200. Also, this matches the expected outcome, where the best teams win the event. While we always want to see good competition, we also want to see the robots that are legitimately the best win the competition.

[Chris is me]

Originally Posted by TheMadCADer

Obviously you should talk to your alliance partners about it, but even if they don't want you to, I don't see that as an issue. You shouldn't be obligated to win for some other team's sake, especially if it isn't in your own best interests.

This is just wrong. Those other teams on the field want to win the tournament just as much as you do, and it is absolutely unethical to sabotage your allies for your own selfish gain.

If they want to rank high, they should do it themselves.

Excuse me? You're saying if the teams that are with you want to rank high, they should do it without your help? You're saying a team should expect to be able to overcome a *partner intentionally playing against them* in order to get into the top 8? How could you possibly justify this ethically?

Imagine three teams, given numbers based on how many points they score. Team 200 scores 200 points each match, Team 150 scores 150 points, Team 80 scores 80, and every other team scores fewer than 80 points. You are on Team 150, and you already have a few losses with just one qualification match left, against Team 200, but you have better alliance partners and could win the match. Both Team 200 and Team 80 are undefeated, and Team 80 has already played all of their matches. If you win your last match, Team 80 is the #1 seed and picks Team 200. If Team 200 wins that match, they get the #1 seed and pick you, Team 150.

It should be obvious which situation gives you a better chance of winning, since no other robots can top Team 80, and you can't top Team 200. Also, this matches the expected outcome, where the best teams win the event. While we always want to see good competition, we also want to see the robots that are legitimately the best win the competition.

I think most of us can understand situations when there is an incentive to not win. It's not hard to think of them. And if you were playing one on one, an argument could be made that you should take the opportunity presented to you. But as long as you have alliance partners, you have absolutely no ethical leg to stand on. Those teams have a right to have their partners do everything they can to help them to win a match. And you can't just ignore their legitimate concerns because you are selfishly glued to your own goals.

[DjScribbles]

Originally Posted by TheMadCADer

The thing is, you don't have to play half-heartedly or just sit there in order to lose the match. Just point your shooter at the other goal (especially if you can get some of their coloured discs and score them in their pyramid goal, that way they get 26 points each thanks to G17). That way you don't make yourselves or your sponsors look bad, it just makes it obvious that you want the other alliance to win. Get your alliance partners to help out, and they'll get an easy way to prove their abilities. You can probably put up a really high score together.

Obviously you should talk to your alliance partners about it, but even if they don't want you to, I don't see that as an issue. You shouldn't be obligated to win for some other team's sake, especially if it isn't in your own best interests. One thing I like to stress in the early stages of design is to never rely on your random alliance partners for anything. If they are only ranked high because of other teams helping them, then they probably shouldn't be ranked high. If they want to rank high, they should do it themselves. It gets completely different when there are two approximately equal teams on each alliance, and your behavior is basically choosing which will rank higher, but if it's just a couple teams that want to move up from 30th to 20th, I wouldn't hurt my own chances to help them, because it honestly wouldn't matter since they couldn't pick anyway. They should just focus on showing off.

I don't like to see teams making "pick deals" for losing matches, but rather I'd like to see this when there are two other teams that could seed first, and one is distinctly stronger than the other, and you are playing against that stronger team.

Imagine three teams, given numbers based on how many points they score. Team 200 scores 200 points each match, Team 150 scores 150 points, Team 80 scores 80, and every other team scores fewer than 80 points. You are on Team 150, and you already have a few losses with just one qualification match left, against Team 200, but you have better alliance partners and could win the match. Both Team 200 and Team 80 are undefeated, and Team 80 has already played all of their matches. If you win your last match, Team 80 is the #1 seed and picks Team 200. If Team 200 wins that match, they get the #1 seed and pick you, Team 150.

It should be obvious which situation gives you a better chance of winning, since no other robots can top Team 80, and you can't top Team 200. Also, this matches the expected outcome, where the best teams win the event. While we always want to see good competition, we also want to see the robots that are legitimately the best win the competition.

Let me lay out another situation. Our team was in a rather unique scenario at our St Joseph District (http://www.thebluealliance.com/event/2013misjo), we got to play king-makers in match 78 (the final quals match), where teams 2000 and 2959 were facing off as the undefeated 1st and 2nd seeded alliances (in the last match of the day no less). I wouldn't dare to figure out which of them was a better bot, they were both very good, and played a very similar game.

Our robot was a very attractive pick (3rd seed at the time) for either of the two amazing cycle runners, we had the only seven disk autonomous at the event, and considered ourselves to be likely for the first pick of either team.

I would be a liar if said we hadn't discussed the what-if's of throwing the match (although I wouldn't say we considered it seriously). If we chose to throw that match, we may have still won the event, but our friends on 2959 would have been the victim of something embarrassingly deceitful.

Even if we had been with another great robot in that final match instead of 2959, throwing the match would have had the same consequences for 2959, albeit less direct, and it still would have been wrong.

My personal opinion is, even when it IS worth it to throw the match, it isn't worth it. There is always the next event or the next year for winning, but you can't wipe away a seedy past as easily.

[pfreivald]

Originally Posted by Taylor

What about explaining to your sponsors that you aren't able to promote and advertise them at the championship event because your team was fixated on one match and not the event as a whole?

"We did our best" does not require explanation.

[EricLeifermann]

Originally Posted by pfreivald

"We did our best" does not require explanation.

Not condoning or vilifying purposely losing, but if you don't do everything in your power to win then you didn't "do your best".

[Anupam Goli]

Originally Posted by pfreivald

"We did our best" does not require explanation.

I don't support playing badly at your benefit and at the cost of your alliance partners, but in the case of 6v0:
Is it easier to show your sponsors a Blue banner showcasing a regional win that was the result of you seeding first due to a 6v0 match, or a finalist medal because you decided to not utilize the ranking system and play for the win in that one qualification match?

[dellagd]

Originally Posted by Chris is me

I'm sorry that I kinda jumped on this comment, but every year I hear people talk about alliance selection like its some huge political struggle when it's really all just tactics.

+1 to that. There was a lot of that at Bridgewater last weekend.

[Lil' Lavery]

This is a very nuanced issue that has been discussed at length several times in the past. And each new game has its own complexities in gameplay and rankings that make for an interesting twist on the discussion.

Here's another thread from after 2011:
http://www.chiefdelphi.com/forums/sh...ad.php?t=94916

While many will agree that throwing a match is undesirable, how to define "throwing a match" becomes unclear when actually examined. Should event strategy outrank match strategy? Does employing a sub-optimal match strategy count as throwing a match? Does avoiding strategic "dirty" play count as not employing the optimal match strategy?

[pfreivald]

Originally Posted by EricLeifermann

Not condoning or vilifying purposely losing, but if you don't do everything in your power to win then you didn't "do your best".

I knew someone was going to say that. My point was that no sponsor in the world is going to see a team that tried their hardest to win every game, and then ask for an explanation.

[Lil' Lavery]

Originally Posted by pfreivald

I knew someone was going to say that. My point was that no sponsor in the world is going to see a team that tried their hardest to win every game, and then ask for an explanation.

Why does trying your hardest to win every match outrank trying your hardest to win the event?

[Hoover]

Replying to DampRobot's first post of this thread. Which I have re-read.

I come away with a different logic on this now and hope not to over-Spock the issue. My level of strategy as a new coach is strictly match mode. But your thinking is a level above that which I will call tournament mode. This requires an order of magnitude higher level thinking, e.g. you must consider how many matches are left, who are all the members of your opponent alliance etc.

Is considering all these variables as a strategy and executing it precisely not a form of excellent gamesmanship? Better than just say... duh... win?

Now consider, what if a situation arose where the opposite alliance was trying to do the same thing. A sort of Nash equilibrium of losing if you will. It would be embarrassingly obvious if both teams didn't even try to score, but this is probably so rare it would never happen. Still please consider it.

Last question. Do you think this sort of thing ever happens in professional sports as a strategy? Are they bad for doing this or are they trying to do their best i.e. set up their seeds so they have a better shot at the ultimate win? Legal issues? Bookies?

If I get good answers and not just 'the rules say' or 'the spirit of', I promise to throw myself in the frying pan with you.