Intentionally Losing Matches

[Matthew2c4u]

Its a valid "tactics" not a strategy but its very Un-GP

Its valid because its a good idea, but theres lots of good ideas that break or bend the rules to work.

But its non-graciously professional, so you may not be solicited by that number 2 alliance, and you would have a hard time convincing kids to do it.

[JesseK]

Originally Posted by 1885.Blake
Somewhere along the line I got the impression that using brain power well and celebrating the skillful use of science and mathematics was part of that higher standard. I guess I was wrong about that.

You are right, we should stop keeping score during matches and play them only for fun and for the benefit of the audience. We should stop using success in the tournament as a way to recognize teams that have worked hard to understand fully how the tournament system works and how to use a well-designed and built robot to maximize their chances of being in the winning alliance.

See above - Also, since when did meticulous analysis of the mathematics and social interactions of a tournament become known as "finagling"? In my math, science and engineering studies it has been known as "optimizing".

Blake, you and I have been at odds about this before. Using your "brainpower" to do things correctly from the start will always trump trying to squeeze the system for what it's worth. "Optimising" also has the underlying assumption that you already have some measure of reliable success. Throwing a match hardly shows success, and I'd even venture to say that it'd backfire if the #1 seed recognises that it was your team that caused the alliance to lose. Mathematical analysis will always have the inherent flaw that it cannot predict human behavior.

I'd like to point out that Blake's posts here are NOT 1885's strategic, practical, or otherwise viewpoint(s). Our viewpoint has always been to do the best with the bot we have in its current state, regardless of where it would put us in rankings. I know and believe this because I've been the driver coach for the last 2 years.

If a hypothetical team has the opportunity to be in the top 8 yet has not done the legwork to scout other teams that have the chance to topple the #1 seed, then that hypothetical team deserves the fate it gets. Case Study: Newton 2007 Seed #8 -- they did their scouting and we all see where that got them.

[msd]

Originally Posted by 1885.Blake

since when did meticulous analysis of the mathematics and social interactions of a tournament become known as "finagling"? In my math, science and engineering studies it has been known as "optimizing".

Just because you /can/, doesn't mean you /should/. Optimize numbers, not people.

Also, let's not forget that the goal to be on the winning alliance is (or ought to be) a distant sub-bullet underneath the broader goal "to inspire young people to be science and technology leaders".

The way the tournament is structured is simply a mechanism, not the real goal.

The game is simply a mechanism, not the real goal.

The robot is simply a mechanism, not the real goal.

FIRST itself is simply a mechanism, not the real goal.

[Daniel_LaFleur]

This is a competition, and the goal of a competition is to win. however, it is what we do to accomplish that goal that shows the measure of the person.

While the strategy of throwing a match may be successful ... in the long run it may cost far more than you realize. Consider what you are teaching your students, that it is OK to hurt your allies if you can gain a little (possible) advantage for yourself. To me that sounds a bit too Machiavellian for me ... especially in a program that is supposed to be trying to change the culture.

[gblake]

Originally Posted by ebarker

... It is probably a pretty fair to assume that the game designers intend for everyone to compete in every match and make the best showing of their technological skills, not their superior application of game theory.

It's that simple.

Mathematics and in particular the ability to estimate the effect one's current actions have on ones progress in each match and in the tournament are obviously technical skills.

It is probably pretty fair to assume that the game designers intend for teams to compete in a tournmant conssisting of multiple matches and make the best showing of all their technological skills throughout the entire duration of the tournament.

I presume that they would frown upon only viewing each match in isolation, because that would mean that the teams were disregarding important information.

It's that simple.

[gblake]

Originally Posted by Brandon Holley

It appears however, we are not on the same page. It's been said forever, "winning isn't everythign", and thats because it isn't.

Why is it important to try to maximize your score in every match? I'm serious. Why?

And please point out the flaw in the scenario. Which of the statements I made to describe it is wrong/flawed?

Thanks,
Blake

[gblake]

Originally Posted by Daniel_LaFleur

This is a competition, and the goal of a competition is to win. however, it is what we do to accomplish that goal that shows the measure of the person.

While the strategy of throwing a match may be successful ... in the long run it may cost far more than you realize. Consider what you are teaching your students, that it is OK to hurt your allies if you can gain a little (possible) advantage for yourself. To me that sounds a bit too Machiavellian for me ... especially in a program that is supposed to be trying to change the culture.

Please read what I wrote - I specifically said consult your allies - explain your desire and ONLY if they support it, do you and your allies choose to score fewer points than your opponents.

[SL8]

slightly off topic, but I love this clean discussion.

[gblake]

Originally Posted by JesseK

... it'd backfire if the #1 seed recognises that it was your team that caused the alliance to lose.

Already covered in the scenario I described to create a solid foundation for the discussion.

I'd like to point out that Blake's posts here are NOT 1885's strategic, practical, or otherwise viewpoint(s). ...

Already made clear in a previous message I posted. I suppose it is wise to repeat that info again here: Remember, these are my personal opinions. I enjoy assisting several teams near my home and their members all have widely varying responses to this question. Plus, I think it is a very nice topic for a challenging debate. Intense, non-personal debate is healthy.

If a hypothetical team has the opportunity to be in the top 8 yet has not done the legwork to scout other teams that have the chance to topple the #1 seed, ...

Already covered in the scenario I described to create a solid foundation for the discussion.

[EricH]

See the 1vs3 thread for what else I have to say.

Remember, the #2 alliance wants you, BUT they can get you if you are in the top 8! The only way they can't get you is if #1 picks you, or they pick someone else as their first pick.

For Blake: The situation as originally described is that it's your last match and you want to throw it to get out of the top 8 (or rather not enter it). The question as given is whether or not this is GP. You are on a completely different topic.

With the described situation, 1) it doesn't matter unless you can't pick within the top 8 (which is not currently the case), and 2) throwing the match does nothing except make enemies. You don't advance in rank, true. However, because it is your last match (and therefore, presumably, most of the teams are either done or mostly done), you can't fall very far. Probably not even out of the top 15, ALL of whom can move up. So you might not be around.

The situation described is a case where it won't help you, period. It will probably hurt you.

One more point: numbers are NOT definitive. There were once two teams that a particular picking team rated approximately equally. When representatives from the team went around to one of the teams, they got nothing. Nobody talked to them. They went to the other pit and got the opposite response. That alliance went out and beat the other team in two matches.

The point? You are dealing with PEOPLE, not computers. People have long memories...feelings...attitudes...that sort of thing. By presenting the attitude that you don't care about your partners and how they feel (even if you talk to them, they may feel hurt), you probably will incite some feelings of annoyance. Those people who have those feelings will remember what happened, and may hold a grudge for some time. This hurts your team in the future.

Numbers are all well and good, but they only tell part of the story.

[David Noll]

Originally Posted by EricH

The point? You are dealing with PEOPLE, not computers. People have long memories...feelings...attitudes...that sort of thing. By presenting the attitude that you don't care about your partners and how they feel (even if you talk to them, they may feel hurt), you probably will incite some feelings of annoyance. Those people who have those feelings will remember what happened, and may hold a grudge for some time. This hurts your team in the future.

This is a very good point. Guys, while speculating about this is fun keep in mind that not every one gets that this concept is outrageous. Teams threw matches this year and then said that they were malfunctioning so no one else would pick them. It's a strategy that is despicable but some people care about winning more then gracious proffesionalism.

[gblake]

Originally Posted by EricH

See the 1vs3 thread for what else I have to say.

Remember, the #2 alliance wants you, BUT they can get you if you are in the top 8! The only way they can't get you is if #1 picks you, or they pick someone else as their first pick.

In the scenario I carefully described, I said that the #1 team intends to use their position to go down the line giving pseudo invitations to the other captains. This is not all that far-fetched. I think that this sort of thing happens reasonably often when highly seeded captains are worried about alliances between other captains. I postulated that the #1 captain would do exactly what you said (and if I didn't make it clear earlier, the presumption is that you and the #1 seed do not make a good alliance because of different approaches to playing the game or because of overlapping capabilities, or because your team colors clash, or...).

For Blake: The situation as originally described is that it's your last match and you want to throw it to get out of the top 8 (or rather not enter it). The question as given is whether or not this is GP. You are on a completely different topic.

I didn't say it is the team's last qual match. I said that it is the last Qual match of the tournament.

I don't think that I am far from a discussion of GP. If a strategy is sound and is supported by allies, then I have a hunch that the strategy is neither ungracious nor unprofessional. I also think that it satisfies the oft stated goal of competing like crazy.

With the described situation, 1) it doesn't matter unless you can't pick within the top 8 (which is not currently the case),

See above, the #1 captain plans to be a spoiler if this team is in the top 8.

and 2) throwing the match does nothing except make enemies. You don't advance in rank, true. However, because it is your last match (and therefore, presumably, most of the teams are either done or mostly done),

In my scenario I said the match is the last match of the Quals. There are no more qual matches for any team afterwards.

you can't fall very far. Probably not even out of the top 15, ALL of whom can move up. So you might not be around.

#2 gets to pick an ally (the team in question) before moving up would have an effect on the strategy.

The situation described is a case where it won't help you, period.

See above, I think we disagree about the statements preceding this one

It will probably hurt you.

One more point: numbers are NOT definitive. There were once two teams that a particular picking team rated approximately equally. When representatives from the team went around to one of the teams, they got nothing. Nobody talked to them. They went to the other pit and got the opposite response. That alliance went out and beat the other team in two matches.

The team I described is the sort of team that consults with its allies before a match and doesn't try an unusual strategy without support from those allies. I think that they are more like the second team you describe above.

The point? You are dealing with PEOPLE, not computers. People have long memories...feelings...attitudes...that sort of thing. By presenting the attitude that you don't care about your partners and how they feel (even if you talk to them, they may feel hurt), you probably will incite some feelings of annoyance.

I described ONLY using the strategy if the team received active support from their allies.

Those people who have those feelings will remember what happened, and may hold a grudge for some time. This hurts your team in the future.

I presume that the allies would only actively support the strategy if they agreed with it and that if they agreed with it they would have positive memories of the team that they supported.

Numbers are all well and good, but they only tell part of the story.

OK

Even though I disagree with Eric about some things above, - He is the most recent person to come the closest to giving the one actual reason I would agree with for not doing this. That reason is this:

"Other teams will not understand the logic behind the actions of the three allied teams and will (irrationally in my opinion - but that doesn't matter much - Logic has a hard time removing an idea that didn't spring from logic in the first place) look down upon the team(s) that purposefully score low in a match."

Even though my green dots have perversely gotten more numerous rather than shrinking because of this thread (go figure - I expected to see them disappear in a twinkling of the eye). I do see that many people would not give a team AND their allies the benefit of the doubt if that trio scored low on purpose. Apparently, without seeking out the reason for the team's actions, some folks (not necessarily anyone who posted in this thread, but some folks) might brand them with a scarlet "S" for being scalawags and scoundrels.

That might cause trouble in the future that might outweigh any improvement in the expected outcome of the tournament in which the incident occurred.

Blake
[EDIT]
PS: Kudos to Alan Anderson for hitting this paricular nail (that even if what the hypothetical team did was the 100% right thing to do, that might not equate to the team having post-tournament, long-term success because of how other people judge their actions) on the head in a PM he sent to me while I was typing the pre-edit version of this post. At least he and I agree about something
[/EDIT]

[gblake]

... Using your "brainpower" to do things correctly from the start will always trump trying to squeeze the system for what it's worth.

I'm pretty sure that what I am talking about is using your brainpower at every step of the way, so I guess we agree about that.

... if the #1 seed recognises that it was your team that caused the alliance to lose.

In the scenario I outlined, the team in question wishes to ally with the #2 seed and the #2 seed wishes to ally with them. Without any loss of generality one can assume that both of those teams have discussed the strategy and see merit in it. The goal of the strategy is to prevent the #1 seed from using FIRST's drafting rules (or some off-season events' rules) to deny #2 and the team in question the opportunity to ally.

Also, in the scenario I outlined, the one team does not cause the alliance to lose. The allied trio of teams all agree with and actively support using the strategy. If they don't then it doesn't get used. I tried to make that clear when I described the scenario.

Mathematical analysis will always have the inherent flaw that it cannot predict human behavior.

Then it is a good thing that employing this strategy does not depend on human behavior, except for the one possiblity that the #1 seed will pick the team in question, from the pool, with their first pick. Otherwise, the #2 and this team get to form the alliance they think will win the tournament.

Blake

[EricH]

Blake, in case you didn't notice, I did say the ORIGINAL situation. Hence the "completely different topic". I was not referring to your situation.

Your situation is a little far-fetched. Half the time, the #1 seed won't employ it at all; in fact, very few will. When it does occur intentionally, it is targeted at teams known to want to ally to create a formidable alliance. I'm not saying it won't happen; in fact, it might forseeably happen. The odds of it happening at full extent intentionally before the GDC figures up a stop for it are slim to none, but still, it may happen, at which point there will be massive complaints, and the GDC will say "It's legal by this year's rules" and possibly go figure up a solution.

I'm reviewing your original post. There are some things that don't sound right...

I assert that the notion that the only way to "do your best" during the field competition part of a FIRST FRC tournament, is to blindly employ strategies aimed at scoring as many points as possible in each match, is not mathematically sound.

Care to give a reason for the math? You know, not everyone plays that way. There are teams that make a living playing defense the right way. I don't care about the math behind this; I'm more of a tactician. "The best defense is a good offense", yes...but the defense is there to protect the offense's gains.

If I am right and if scoring as much as possible isn't a mathematically sound path to teams' desired end states, it would appear to me that we clever folks trying to inspire a true appreciation for science would be forced to examine alternative strategies and correctly employ the alternatives in appropriate situations.

You might be right...but the better strategy teams already look at alternatives.

I am using the term "strategy" in the sense of "a set of rules that are designed to maximize the likelihood of some desired outcome in a game, and that govern a players' actions in that game".

That's not strategy. That's tournament rules. There's a difference. Strategy is more of "What can we do, without breaking the rules, to win (or whatever it is the objective is)" at its highest level. Lower down, it's called tactics.

For the field competition portion of an FRC tournament, I think that most people have the same desired outcome: Being a member of the Winning Alliance. I am going to assume that outcome is nearly universally what the participants desire.

That's reasonable.

Game Theoretical analysis of many types of games shows that strategies which might at first seem counter-intuitive are actually the "best" strategies. Perhaps the situation we are discussing is one of those instances when a counter-intuitive move is the right move (akin to sacrificing a chess piece to obtain a better board position, and thereby improving your long-haul chances having the TOTAL game turn out successfully).

Interesting--but I see your point. Such situations are rare, however. And you MUST weigh ALL the risks and benefits of such a move beforehand. If you're doing it to get your opponent's rook and you sacrifice your queen, is it worth it? Opponent's queen for a bishop? Queen for queen? It's not always the right move--consider the options first.

In the situation Fred described (plus a couple of additions)

• You are about to play the very last Qual match of the tournament.
• The #1 and #2 seeds are already locked up.
• You are confident that the # 1 captain (expecting 7 "declines") is going to pseudo-invite all of the lower 7 captains in order to prevent them from allying with each other.
• You are confident that the #2 through #7 captains will decline the pseudo offers from the #1 captain.
• You wish to ally with the # 2 captain and they wish to pick you. You and they believe that paired together, you will be the foundation of the best alliance in the elimination rounds.
• If your Qual match alliance outscores the opposing alliance, you will become an alliance captain. If you don't, you won't be one of the original 8 alliance captains.

Odds of there ever being 7 declines: 0. Unless the team is SOO bad that they shouldn't be #1. That almost never happens. Also, you would need to ask all or almost all of those captains what their plans were. You won't get a straight answer for all of them, I guarantee it.

AND

• You ask your allies if they care whether they outscore your opponents or not. They do not care and are willing to help you avoid becoming a captain because they understand that you wish to avoid a pseudo-offer from the #1 captain. Additionally, perhaps they wish to avoid unnecessary wear and tear on their machines and don't care what the score of the match turns out to be. Instead, because the match is very largely irrelevant to them, and because it is very important to you, they want to graciously support their ally, i.e. you.

OK. But, if those teams are even thinking of elims and aren't in the top 8, they won't go along. You'd need fairly bad robots in the bottom of the field--or robots who all do the dominant strategy of the regional and not very well. I'm not saying it won't happen, but the odds are slim.

In order to "do your best" at attaining the outcome you, and just about everyone else, has been pursuing (on the field) (becoming a member of the winning alliance), please tell me why you would attempt to outscore your opponents, become the 8th seed, and thereby be unable to form what you believe is the strongest alliance possible for the elimination rounds.

Simple. Do the math, based on human relations, for the strategy. It won't happen. If it was sure to happen, I might use it.

If you offer an alternative strategy and back it up with testable propositions, please do so in the neutral language of science and math. I have to admit that I get just a little bit annoyed at pejorative terms like "throw the match". I get especially annoyed when, in the scenario being discussed, the alternative appears to be choosing to "throw the tournament".

Listen, science and math includes psychology, does it not? Psychology deals with human nature, does it not? Human nature says your situation won't happen for quite a while yet, if ever. You only have two alternatives--full bore playing or not even showing up (the only way to guarantee the match outcome is in your "favor"--see the thread about 0vs2 match and who won).

If outscoring your opponents in one particular match means you reduce your chances of winning the tournament; and doing the opposite increases your chances of winning the tournament; and if your allies are willing to support either option; then from a game theory perspective, the choice seems clear. Don't purposefully do badly in the total tournament

From a game theory perspective, yes, from a math perspective, yes, from human nature--nope.

PS: If it is wrong, as some seem to have suggested, for a team to aspire to using their analytical and mechanical skills to earn that piece of plastic, then I submit that all teams in the tournament should stop doing the "wrong" thing; and should instead make every match into a pro-wrestling style exhibition for the benefit of the audience.

It is not wrong. However, not all teams use them in this way. Some teams just go out there to win every match. And, pro-style wrestling is to real wrestling as Battlebots is to FIRST.

Please don't explicitly or implicitly denigrate the possibility that a team (supported by their allies) might actually take the time to think backwards from their desired end goal to their situation in a particular match, discover that a low score improves their odds of reaching their desired end-goal, and then act on that conclusion.

IF they do, then they better make sure that they actually reach it...and hope that #1 doesn't pick them.

Here is where your entire situation falls apart. Suppose that #1 picks all down the top 8 and they all decline. They look back at their list. They ask for YOUR team. Do you a) decline or b) accept, now knowing that you only have one possible chance at the eliminations? I will almost guarantee you--teams that have a working robot and no travel plans that interfere will ALWAYS accept if they are not in the top 9-10.

[gblake]

Originally Posted by EricH

Blake, in case you didn't notice, I did say the ORIGINAL situation. Hence the "completely different topic". I was not referring to your situation.

OK - Sorry

Your situation is a little far-fetched. Half the time, the #1 seed won't employ it at all; in fact, very few will. When it does occur intentionally, it is targeted at teams known to want to ally to create a formidable alliance.

OK - Let's say that they don't go down the entire list, and instead they only give a pseudo-offer to the team in question (in the #8 spot in this alternative scenario), knowing that the team in question will not accept (they would make a lousy alliance because their robots just don't go together well).

I'm not saying it won't happen; in fact, it might forseeably happen. The odds of it happening at full extent intentionally before the GDC figures up a stop for it are slim to none, but still, it may happen, at which point there will be massive complaints, and the GDC will say "It's legal by this year's rules" and possibly go figure up a solution.

I'm reviewing your original post. There are some things that don't sound right... Care to give a reason for the math? You know, not everyone plays that way. There are teams that make a living playing defense the right way. I don't care about the math behind this; I'm more of a tactician. "The best defense is a good offense", yes...but the defense is there to protect the offense's gains.

You are right - I was sloppy - How about replacing that "score the most points" statement with one about "maximizing the result of subtracting the opposing alliance's score from your alliance's score".

You might be right...but the better strategy teams already look at alternatives.

That's not strategy. That's tournament rules. There's a difference. Strategy is more of "What can we do, without breaking the rules, to win (or whatever it is the objective is)" at its highest level. Lower down, it's called tactics.

Not in the articles and texts I read, but that is a quibble.

That's reasonable.

Interesting--but I see your point. Such situations are rare, however. And you MUST weigh ALL the risks and benefits of such a move beforehand. If you're doing it to get your opponent's rook and you sacrifice your queen, is it worth it? Opponent's queen for a bishop? Queen for queen? It's not always the right move--consider the options first.

Odds of there ever being 7 declines: 0. Unless the team is SOO bad that they shouldn't be #1. That almost never happens. Also, you would need to ask all or almost all of those captains what their plans were. You won't get a straight answer for all of them, I guarantee it.OK. But, if those teams are even thinking of elims and aren't in the top 8, they won't go along. You'd need fairly bad robots in the bottom of the field--or robots who all do the dominant strategy of the regional and not very well. I'm not saying it won't happen, but the odds are slim.Simple. Do the math, based on human relations, for the strategy. It won't happen. If it was sure to happen, I might use it.

I didn't say that the #1 is bad. I said that the #2 and the team in question both believe that they are the best combination to create.

Listen, science and math includes psychology, does it not? Psychology deals with human nature, does it not? Human nature says your situation won't happen for quite a while yet, if ever. You only have two alternatives--full bore playing or not even showing up (the only way to guarantee the match outcome is in your "favor"--see the thread about 0vs2 match and who won).

Not showing up is in Fred's original hypothetical scenario; but it is never, ever an option in my hypothetical scenario.

From a game theory perspective, yes, from a math perspective, yes, from human nature--nope.


It is not wrong. However, not all teams use them in this way. Some teams just go out there to win every match. And, pro-style wrestling is to real wrestling as Battlebots is to FIRST.

I agree. That is why I wanted to remind everyone that you can not both tell teams not to worry about earning the Winning Alliance title, and at the same time surround it and those who win it with as much attention as they receive. Either it has non-zero value comensurate with its prominent, central place in FRC Regional and Championship Tournaments, or it doesn't. Once that debate is put to bed, one can move on to the ethics/GP question.

IF they do, then they better make sure that they actually reach it...and hope that #1 doesn't pick them.

Yep - In my reply to JesseK I acknowledged that being picked from the pool by the #1 in the first round is possible and would prevent an alliance between #2 and the team in question.

Here is where your entire situation falls apart. Suppose that #1 picks all down the top 8 and they all decline. They look back at their list. They ask for YOUR team. Do you a) decline or b) accept, now knowing that you only have one possible chance at the eliminations? I will almost guarantee you--teams that have a working robot and no travel plans that interfere will ALWAYS accept if they are not in the top 9-10.

See immediately above. We agree about that. The team in question would be briefly disappointed that they don't get to ally with #2. They will then turn their full attention to helping #1 beat #2 and the rest of the alliances.

Blake