View Full Version : 9.3.4 Match Seeding Points
theycallhimtom
09-01-2010, 12:26
9.3.4 Match Seeding Points
All teams on the winning ALLIANCE will receive a number of seeding points equal to the penalized score (the score with any assessed penalties) of the winning ALLIANCE.
All teams on the losing ALLIANCE will receive a number of seeding points equal to un-penalized score (the score without any assessed penalties) of the winning ALLIANCE.
In the case of a tie, all participating teams will receive a number of ranking points equal to their ALLIANCE score (with any assessed penalties).
Am I missing something or does the losing and winning team get the same number of points? This basically makes the game cooperative as both teams should only score for one alliance in an attempt to maximize their score.
Another significant rule:
9.3.9 Qualification Seeding
All TEAMS in attendance will be seeded during the qualification matches. If the number of TEAMS in attendance is 'n', they will be seeded '1' through 'n', with '1' being the highest seeded team and 'n' being the lowest seeded TEAM.
The Field Management System will use the following seeding method:
• TEAMS will be seeded in decreasing order by seeding score.
• Any TEAMS having identical seeding scores will then be seeded in decreasing order by their highest coopertition bonus.
• Any TEAMS having identical seeding scores and highest coopertition bonus will then be seeded in decreasing order by cumulative ELEVATED/SUSPENDED points earned by their ALLIACES throughout the Qualification Matches.
• Any TEAMS also having identical highest coopertition bonuses will then be seeded based on a random sorting by the Field Management System.
Matthew2c4u
09-01-2010, 12:49
Fixed for rules.
Read my post below
Monty Python
09-01-2010, 13:07
9.3.9, teams will be seeded based on highest seeding score. It's still a competitive game though, because cooperition points will be the king maker (two times the losing alliance score, if manipulated correctly, that's going to be a lot. It's definitely the G22, I think that's the number anyway, of last year). Remember that according to 9.3.5, only teams on the winning alliance earn cooperition points. Winning in qualifications is still very important.
Oh and just remember that the cooperition score is twice the losing teams score, so even if it were cooperative (which it isn't!) it wouldn't make sense to only score for one team anyway.
ScottOliveira
09-01-2010, 13:29
The way the rules read, the two teams get the same amount of points if there are no penalties by the winning team. If the winning team gets a penalty, the losing team actually receives more points than the winning team.
And while Matthew2c4u would normally be correct about the ranking points, the manual doesn't mention them in this game (that I have seen yet). It appears so far that the game is based solely on the seeding points.
theycallhimtom
09-01-2010, 13:37
I was missing:
9.3.5 CoopertitionTM Bonus
All teams on the winning ALLIANCE will receive a coopertition bonus: a number of seeding points equal to twice the un-penalized score (the score without any assessed penalties) of the losing ALLIANCE.
In the case of a tie, all participating teams will receive a coopertition bonus of a number of seeding points equal to twice their ALLIANCE score (with any assessed penalties).
So if red team scores X and blue team scores Y and red wins then
Red team gets X + 2Y
Blue team gets X
if blue wins
Red team gets Y
Blue team gets 2X + Y
ScottOliveira
09-01-2010, 13:48
I originally read that as the coopertition bonus being separate from seeding points, used to break ties, but I believe you're right. That makes the system a bit more sensible.
I caught that too.
There are two scenarios here; which one is the actual one can be resolved quickly around the time Q&A opens.
W is winner; L is loser. Points refers to the pre-penalty score.
1) Coopertition bonus is not added to seeding.
W gets Wscore (which is Wpoints-Wpenalties)
L gets Wpoints
L comes out ahead, if not even. Ouch.
2) Coopertition bonus is added to seeding, and tracked separately as well.
W gets Wscore+2*Lpoints
L gets Wpoints
W comes out ahead by quite a bit. Better.
Effectively, ranking points have been eliminated from the equation (despite a reference to them), as has W-L-T. This is quite an interesting method, as it is now theoretically possible for a team with a 0-X-0 record to seed above a team with a 1-(X-1)-0 record. I'm not sure if I like it or not, yet, but it'll still be interesting. (Hopefully, W-L-T will still be tracked.)
XaulZan11
09-01-2010, 14:13
At first glance, I don't think I like the change. Mainly because it causes it difference in Qualifcation and Elimination matches. As Eric pointed out a team (Team X) that has lost every match could seed higher than a team (Y) that went undefeated. But, when elimination matches come around, team Y will be more likely to take home the gold than team X. I know this is the extreme case, but I don't like the fact that what it takes to be sucessful can be drastically different from qualification to elimination matches.
To look at it another way, a team could have a fantastic stratedgy that just wins matches, although ugly and not resulting in many seeding points. While this team will likely do very well in elimination matches, it is unlikely they will seed very high.
(I think it is similar to college football's overtime rules, which completely remove special teams play, other than field goal kicking. A team could be dominating field position and returned a kick or two for a touchdown, but when overtime comes around their advantage or how they tied the game is now completely obsolete.)
At first glance, I don't think I like the change. Mainly because it causes it difference in Qualifcation and Elimination matches. As Eric pointed out a team (Team X) that has lost every match could seed higher than a team (Y) that went undefeated. But, when elimination matches come around, team Y will be more likely to take home the gold than team X. I know this is the extreme case, but I don't like the fact that what it takes to be sucessful can be drastically different from qualification to elimination matches.
To look at it another way, a team could have a fantastic stratedgy that just wins matches, although ugly and not resulting in many seeding points. While this team will likely do very well in elimination matches, it is unlikely they will seed very high.
Your first point is sometimes true but it is also possible that a team went undefeated with the old season not due to them being the best robot there, but due to luck of the draw giving them the good alliances (or easy opponents). The new system isn't perfect but neither was the old one. At least this way takes into account the competition you were up against a little better. In the old system a win counted the same (essentially) even if the other alliance was a no show. Of course, this system could give a no show alliance more points than the winning alliance (if i'm understanding it) depending on how its being implemented so i guess that raises a whole new set of problems.
I do agree with your second point though. I guess it discourages defense.
David Brinza
09-01-2010, 14:35
Basically if you win you get 3 points for your main ranking score, if you lose you get 0 if you draw you get 1. If theres a tie then the secondary score kicks in.
Those are secondary ranking scores so that a team who plays a good game against a great alliance gets credit for it.
In what section of the game manual do you find this?
Section 9: The Tournament
What I see as the intent of the rule is to discourage cut-throat competition in the qualification rounds. It is perfectly to your advantage to keep the score close (e.g. win by say, 1 point). The idea is that the scenario that is the most beneficial to both teams is a high-scoring, penalty-free tie. It's like the previous system (I think on that one it was called either ranking points or qualification points. I don't remember which was which) in which one portion of seeding was based off of the score of the opposing alliance. However, it is now included in the primary seeding value (your total seeding points) as well as being the first tiebreaker.
The scenario that would be the most beneficial to a single team is to have the opposing alliance lose to a ridiculous number of penalties while still obtaining a high score (and not have any penalties).
If the losing team's score is 0, then they get the same seeding points as the winner, discounting penalties.
If the winning team gets a penalty, and the losing team scores 0, then the losing team wins for seeding points.
The optimum strategy is thus to convince your alliance partners to score 0, and help the other team score as much as possible on you to maximize *your* seeding points. You must also defend against all attempts at a counter-strategy to score points in *your* goal to get you points, because their seeding points increase by two every time they score for you.
You can block both of your goals by moving 2 of your robots blocking your own goals, and having one of your robots pushing balls into their zone.
This has a few problems. Remember that you are being compared to other teams. Unless everyone does this, you would be much better off either tying or winning (but not losing). Because you get 0 points for the "coopertition bonus" you will be far behind those who win in matches where this doesn't happen, but only possibly ahead of those who lose in said matches.
This can only get you ahead of one team. Look at the big picture and it will become apparent that winning by a close margin in a high scoring (and penalty-free) match is much better for your team (as well as your opponents).
The reason for this is that while yes, the points from your proposition will be equal/better than your opponent depending soley on penalties. Even tying would be a better option because even if the score of each team is half of what happens in your proposition, if you have no penalties, you will recieve 50% more seeding points, putting you in a better position of those who you didn't play against, and may still come out ahead depending on penalties on your opponent.
Besides, your proposition in the end seems to simply be a malicious strategy where the primary aim is to hurt the opposing alliance rather than to help your own. This isn't exactly a zero sum game. Ties are possible, and what really matters is the amount of seeding points you get with respect to other teams.
Unless everyone does this, you would be much better off either tying or winning (but not losing)
Everyone being your alliance, or everyone being people in the competition? In the former, yes. In the latter, no; The cooperition bonus is only deterministic of the seeding rank with people whom you tie seeding points. If you don't score any points, the maximum points the other team can get (will get) is increased (even doubled or more) because you aren't going after any balls.
Actually, if you had read the thread, there are two points of view. You have one of them.
The other is that the coopertition bonus is added to your seeding points (hence the "bonus" part), to produce a much higher seeding point total. Instead of the winner getting their own score as seeding points, the winner gets their own score plus their coopertition bonus as seeding points.
Also note: 9.3.7 Seeding Score
The total number of seeding points earned by a TEAM throughout their qualification matches will be their seeding score.
Now, see this: 9.3.5 Coopertition™ Bonus
All teams on the winning ALLIANCE will receive a coopertition bonus: a number of seeding points equal to twice the un-penalized score (the score without any assessed penalties) of the losing ALLIANCE.
In the case of a tie, all participating teams will receive a coopertition bonus of a number of seeding points equal to twice their ALLIANCE score (with any assessed penalties).
In other words, the Coopertition Bonus is in fact seeding points, which in fact are added to the seeding score, which is your ranking.
Winners have a distinct advantage here, especially in close games.
Everyone being your alliance, or everyone being people in the competition? In the former, yes. In the latter, no; The cooperition bonus is only deterministic of the seeding rank with people whom you tie seeding points. If you don't score any points, the maximum points the other team can get (will get) is increased (even doubled or more) because you aren't going after any balls.
Everyone in this case == everyone (and I repeat... everyone), so it would be the latter
Actually, if you had read the thread, there are two points of view. You have one of them.
The other is that the coopertition bonus is added to your seeding points (hence the "bonus" part), to produce a much higher seeding point total. Instead of the winner getting their own score as seeding points, the winner gets their own score plus their coopertition bonus as seeding points.
Also note:
9.3.7
Now, see this:
9.3.5
In other words, the Coopertition Bonus is in fact seeding points, which in fact are added to the seeding score, which is your ranking.
Winners have a distinct advantage here, especially in close games.
I agree. How the manual is worded at this specific point in time, it is abundantly clear that the Coopertition Bonus is a part of your total seeding score as well as the first tiebreaker. It is a bonus not a separate score, otherwise it wouldn't say that it was a bonus of additional seeding points.
Joe Johnson
09-01-2010, 16:54
The Seeding of teams is a tricky thing, even when a you have many many comparisons between teams (think of MLB with a 160ish game season). It is even more difficult when you have 7 or 8 matches for each team and you have multiple teams compared each match.
FIRST has the added problem of trying to encourage GP while still having a game worth playing.
The proposed system for 2010 is not that much different from some of the systems that we have had in the past. I think there are going to be anomalies but there always are anomalies.
There are always going to be teams in the top 8 slots that are "undeserving" in someone's eyes.
The drafting process is the great equalizer... ...which is a topic for another thread...
Joe J.
Richard Wallace
09-01-2010, 17:06
This seeding system appears to be an embodiment of the invention protected under US Patent 7507169 (http://www.freepatentsonline.com/7507169.pdf).
The inventor? Dean Kamen. :)
mathking
09-01-2010, 17:08
This a fascinating exercise in applied game theory. I actually think collusion (that is the game theoretic term, not meant to imply anything negative) between the two alliances, where one scores 0 points and they all try to score as many as possible for the winning side so that all six teams get the same number of seeding points is not a bad strategy. There is definitely a problem in that there is an incentive for the "winning" side to cheat at the end of a match. But the fact that elimination competition is based solely on wins and losses means that if you always use this strategy you will likely not be as prepared for elimination rounds as a team that does not use this strategy. Nonetheless, there are a lot of chances for such collusion/coopertition.
The seeding system has always encouraged the winning alliance to ensure that its opponents had a somewhat high score, but to a lesser degree. Previously the opponent's score was used as the first tiebreaker. Now it's used as a part of the initial round as well as the first tiebreaker.
This a fascinating exercise in applied game theory. I actually think collusion (that is the game theoretic term, not meant to imply anything negative) between the two alliances, where one scores 0 points and they all try to score as many as possible for the winning side so that all six teams get the same number of seeding points is not a bad strategy. There is definitely a problem in that there is an incentive for the "winning" side to cheat at the end of a match. But the fact that elimination competition is based solely on wins and losses means that if you always use this strategy you will likely not be as prepared for elimination rounds as a team that does not use this strategy. Nonetheless, there are a lot of chances for such collusion/coopertition.
If you're working with the other alliance, it's much better for both alliances to rig up a tie.
AllenGregoryIV
09-01-2010, 17:12
This seems like an interesting system to me and I am not sure if I like it yet.
In a situation where the match is close it will always be better to score for your own team to try and win the match thus getting your penalized score and double the losing teams score.
However once the game gets lopsided say 10-0. The losing team, if they believe they cannot win, should attempt to score on the winning team’s goal because those goals will be given to their seeding score. The winning team should also try to score for their opponents because those scores will be worth double in their seeding score.
In the past it has been advantageous for the team that is winning to score for the losing team but never, to my knowledge, has it been better for a losing team to try to make the score differential higher. In past systems the losing team would get their score as their seeding score and would always continue to score for themselves.
ScottOliveira
09-01-2010, 17:17
If you're working with the other alliance, it's much better for both alliances to rig up a tie.
It's actually much better for both alliances to say this, and then one to cheat at the end (the cheating team will receive more points). However if a team cheats, it's less likely to be able to use this strategy on later matches, as other teams won't trust them. However, in the last match an alliance plays it is more likely to cheat, because the trust matters little after that point. Using game theory again, specifically backwards induction, it's possible to show that everyone will cheat every single game (game theory doesn't account for honor, only an effort to maximize points).
It's actually much better for both alliances to say this, and then one to cheat at the end (the cheating team will receive more points). However if a team cheats, it's less likely to be able to use this strategy on later matches, as other teams won't trust them. However, in the last match an alliance plays it is more likely to cheat, because the trust matters little after that point. Using game theory again, specifically backwards induction, it's possible to show that everyone will cheat every single game (game theory doesn't account for honor, only an effort to maximize points).
Maybe I should rephrase that. It would be most beneficial to both alliances to do so, but that is an ideal scenario which will likely not happen due to competition. It is the situation in which the gain of one team is not necessarily at the expense of the other per say (or at least not as large of an expense both in overall seeding and the tiebreaker).
Matthew2c4u
09-01-2010, 17:24
Actually, if you had read the thread, there are two points of view. You have one of them.
The other is that the coopertition bonus is added to your seeding points (hence the "bonus" part), to produce a much higher seeding point total. Instead of the winner getting their own score as seeding points, the winner gets their own score plus their coopertition bonus as seeding points.
Also note:
Now, see this:
In other words, the Coopertition Bonus is in fact seeding points, which in fact are added to the seeding score, which is your ranking.
Winners have a distinct advantage here, especially in close games.
Basically If you win you get coopertiton plus seeding, Lose = seeding
Secondary ranking factor = Highest coopertiton from any one match.
3rd is a point value based on hanging robot points for your alliance from every match that you played in combined.
mathking
09-01-2010, 17:27
If you're working with the other alliance, it's much better for both alliances to rig up a tie.
There is an incentive to work for one side getting all the points so that both alliances get the same (hopefully high) score, but there is an incentive for the side with all the points to try to score on the side with none at the end because then they will get more points than, and move up in the rankings relative to, the teams on the losing alliance. On the other hand, you can only "cheat" like this once, because other teams will no longer trust you.
If everyone were to try for the n to 0 score strategy, it could end up being quite arbitrary as to who ended up as the top 8 teams for alliance selection. And it might well not be clear who those top 8 teams should select as partners. Still, I think we will see some examples of coopertitive collusion.
Matthew2c4u
09-01-2010, 17:31
Coopertition bonus is always worth the win unless you have a very low chance then collusion is valued.
For any fixed total number of points (S) that are scored in a match, if you score x points, then your seeding score (if the game is penalty free) is:
S - x (if x < S/2)
2S - x (if x >= S/2)
Within each case, your seeding score decreases with the number of points you scored. The result is that the "optimal" options are to get shutout, or to win by the smallest possible margin (or even tie).
If the game becomes lopsided to the point where one alliance is guaranteed to win, then:
The winning alliance scoring 1 pt gives both the winning alliance and the losing alliance 1 seeding pt.
The losing alliance scoring 1 pt gives the winning alliance 2 seeding pts and the losing alliance 0 seeding pts.
This makes for a balance of competition when the game is close enough to be "competitive" and cooperation (scoring for the other team) when the game is decided.
Also, defensive robots are punished more in this game. A team that gets shut out 0 - 20 in every game beats a team that wins every game 19 - 0.
IMO, this is worse than last year's <G14>, but at least this doesn't apply for elimination matches.
Lil' Lavery
09-01-2010, 18:00
This a fascinating exercise in applied game theory. I actually think collusion (that is the game theoretic term, not meant to imply anything negative) between the two alliances, where one scores 0 points and they all try to score as many as possible for the winning side so that all six teams get the same number of seeding points is not a bad strategy. There is definitely a problem in that there is an incentive for the "winning" side to cheat at the end of a match. But the fact that elimination competition is based solely on wins and losses means that if you always use this strategy you will likely not be as prepared for elimination rounds as a team that does not use this strategy. Nonetheless, there are a lot of chances for such collusion/coopertition.
This is essentially what I was going to post. 6 robots working together to shuttle the balls from the midfield and score them on the same goals is has a high probability of achieving a higher score than a 3v3 match with defense, ball possession fights, traffic jams, etc.
Obviously the strategy won't be applicable to all robot designs (defensive, and even potentially end-game specialists). It may well also run into issues dealing with <G29>, depending on the implementation and robot design. And obviously not every alliance will be willing to do it. And the tie-based scenario is much higher-reward (though also higher risk, both in terms of cheating and just honest mistakes and miscues).
However, I can guarantee you we will see it attempted at least once per regional event.
Matthew2c4u
09-01-2010, 18:04
So if your going up a set of excellent teams and your alliance partners aren't @ the same level. You simply dont score @ all and protect your goals so they cant score and your seeding points come out to
SS(your seeding points)
SSO(your opponents seeding points)
WP(Winning teams points)
LP (Losing teams points)
PP(penalty points)
SS=WP
SSO= WP+ (2xLP) - PP
So basically if you help your opponent score and hold down any scoring on your goals to 1 or 2 you will get more seeding score than your opponent if they get double as many penaltys as you let goals by, Which is totally feasible if they arent prepared for your strategy.
Also if you help them score you will definatly get more points than you would have if you were normally facing them. So it helps your alliance and it slightly hurts they're alliance but they still get more points for winning. Its like acting symbiotically with a slight drawback.
My guess is it will be team updated to be fixed quite fast.
- my 2 cents.
This is all theorectial i doubt it will every be employed.
I agree with MathKing and Lil' Lavery that a 6-team cooperative effort could be very fruitful. Let's try some numbers, ignoring the tower bonuses for the moment.
1) Suppose that in a "normal" round, the Red Alliance can score 8 points against the Blue defense, while the Blue Alliance can only score 3. Red gets 8 + 2*3 = 14 seeding points; Blue gets 8 seeding points.
2) In a second scenario, Red scores 8 times, 6 for Red, and twice for Blue. Blue scores 3, making the final score 6 to 5. Red gets 6 + 2*5 = 16; Blue gets 6.
3) Now suppose the alliances cooperate - this makes both sides far more productive. I'll suggest that Red can now score 12 balls, while Blue scores 4 (in Red's goal). Red gets 16 + 2*0 = 16 seeding points; Blue also gets 16 points.
I think playing this way is highly advantageous for both sides. Seeding-points wise, option 2 is the best win for Red. However, option 3 doesn't come with the risk that Blue could make a sudden comeback and win with the help of Red's two goals. This is where the tower points might really make a difference. If Blue had a stronger defense, the balance tips further in favor of option 3. Even if you can beat an alliance handily, having their help might still put you ahead.
I agree with MathKing and Lil' Lavery that a 6-team cooperative effort could be very fruitful. Let's try some numbers, ignoring the tower bonuses for the moment.
1) Suppose that in a "normal" round, the Red Alliance can score 8 points against the Blue defense, while the Blue Alliance can only score 3. Red gets 8 + 2*3 = 14 seeding points; Blue gets 8 seeding points.
2) In a second scenario, Red scores 8 times, 6 for Red, and twice for Blue. Blue scores 3, making the final score 6 to 5. Red gets 6 + 2*5 = 16; Blue gets 6.
3) Now suppose the alliances cooperate - this makes both sides far more productive. I'll suggest that Red can now score 12 balls, while Blue scores 4 (in Red's goal). Red gets 16 + 2*0 = 16 seeding points; Blue also gets 16 points.
Here's another proposition with respect to your third one.
Say that that the score is 8-8
Each alliance gets 8+8*2=24, which is 50% more. In every scenario in which there is an even numbered total score, it is always more beneficial to both teams to split the score evenly.
StephLee
09-01-2010, 19:32
Here's another proposition with respect to your third one.
Say that that the score is 8-8
Each alliance gets 8+8*2=24, which is 50% more. In every scenario in which there is an even numbered total score, it is always more beneficial to both teams to split the score evenly.
I agree with this, with one corollary: I think most people are assuming that two teams, each playing defense on the other, will always score fewer points than two teams scoring in coopertition with no defense being played. I agree with this assumption for the most part.
SteveGPage
09-01-2010, 19:34
However, I can guarantee you we will see it attempted at least once per regional event.
I think you will see it once or twice per regional, but probably only that. If there are 4 different basic scoring outcomes for an alliance {zero points, score points, score on both alliances and win, and tie} then there would be basically 5 different combinations {zero points v score points, zero points v score on both, score points v score points, score points v score on both, and tie v tie}. Let's say for the sake of argument, the Blue alliance is on the losing side of the match (except for a tie, of course). In that case, the Blue alliance will get the same number of seeding points as the winning alliance (not including penalties) - so therefore derives no advantage over the Red alliance in terms of moving up the ranking - in 2 cases. This also includes the tie situation. Therefore, it is in the best interest of the Red alliance to not only win, but insure that the Blue alliance also scores. This is true in 3 of the 5 combinations. So by a 3:2 ratio - it is in the best interest of an alliance to win and to have the losing alliance score as many points as possible. The best case is to win by as close a margin as possible, even if that means scoring for the opposing alliance.
Therefore, if the Blue Alliance recognizes that the Red Alliance's best strategy is to have a Med - High score for both alliances, and will score on the opposing alliances goal if necessary, and visa versa. You have the Nash Equilibrium as far as the game strategy is concerned. If you attempt to have collusion, and have a zero point v score high points scenario, and a team defects, by breaking this agreement (scoring on the opposing alliance to maximize their score, or by accident), then you will now have "Tit-for-Tat" and you could never have that agreement again, since everyone would assume you would cheat again, and would then defect from any agreement you would attempt to make – even if they initially agreed to it. (You might get forgiveness for the first time you “cheated”, but not twice!)
Best regards,
Steve
I agree with this, with one corollary: I think most people are assuming that two teams, each playing defense on the other, will always score fewer points than two teams scoring in coopertition with no defense being played. I agree with this assumption for the most part.
Since right now we're purely theoretical. This setup could potentially generate higher scores (with respect to total points) than the shutout collusion.
Reasoning:
Given that the shutout collusion ideally has six robots scoring on one side, the bottleneck will likely be the return of balls to the field. If an arranged tie takes place, you can score in both sides and will therefore be able to generate higher output due to approximately double capacity to return balls to the field.
StephLee
09-01-2010, 19:53
Since right now we're purely theoretical. This setup could potentially generate higher scores (with respect to total points) than the shutout collusion.
Reasoning:
Given that the shutout collusion ideally has six robots scoring on one side, the bottleneck will likely be the return of balls to the field. If an arranged tie takes place, you can score in both sides and will therefore be able to generate higher output due to approximately double capacity to return balls to the field.
I definitely see your point. I do, however, see a higher incentive to "cheat" in this situation than in the collusion shutout, and a wider gap between the winner and loser if cheating does occur. Thus, I as a strategist would be less willing to try this strategy if I know or suspect that my opponent has the capability to outscore me.
Matthew2c4u
09-01-2010, 19:59
So basically, if you can trust your opponent enough and you have enough skill to go for the tie its the most benefical for both teams
If have very little chance against a team its better to help them score
and if your ahead by more than 15 points you should be helping other team score.
Sounds like a plan.(Sarcasm)
This system is definatly unique, but its feasibilty is in question in my mind.
Stephen Kowski
09-01-2010, 20:01
This seeding system appears to be an embodiment of the invention protected under US Patent 7507169 (http://www.freepatentsonline.com/7507169.pdf).
The inventor? Dean Kamen. :)
I believe this patent is from 2003 where there were significant complaints of 'collusion' between opposing teams during seeding at many regionals. It will be interesting to see if this makes a comeback.
SteveGPage
09-01-2010, 20:03
I definitely see your point. I do, however, see a higher incentive to "cheat" in this situation than in the collusion shutout, and a wider gap between the winner and loser if cheating does occur. Thus, I as a strategist would be less willing to try this strategy if I know or suspect that my opponent has the capability to outscore me.
Exactly. Zero Point v High Points OR Tie score collusions are extremely suspect in that the "cheater" has a strong motive to cheat and defect from the agreement. The upside is that they move up in the ranking on the 3 opposing teams, the downside is that they won't be trusted in one of these agreements again. Also - you would have to get all three members of the alliance to agree and not cheat individually, when all three of them may be thinking, those other three teams could be the difference between my team being in the top 8 or not.
Richard Wallace
09-01-2010, 20:04
I believe this patent is from 2003 where there were significant complaints of 'collusion' between opposing teams during seeding at many regionals. It will be interesting to see if this makes a comeback.The game described in the patent is indeed Stack Attack (FRC 2003). Dean filed the patent application in 2004 and it was issued in 2009.
Does anyone else (who's been in FIRST long enough to remember) have a very uneasy presence of the ghost of 2003?
Do a search on the word "collusion" to see the very heated arguments over what people did back in the day to seed better. I can't believe FIRST would want to invite that kind of ugliness back again
I definitely see your point. I do, however, see a higher incentive to "cheat" in this situation than in the collusion shutout, and a wider gap between the winner and loser if cheating does occur. Thus, I as a strategist would be less willing to try this strategy if I know or suspect that my opponent has the capability to outscore me.
That is very true. But then from an audience standpoint, a collusion shutout won't be very popular. It would be boring, and for those who don't know how seeding works (e.g. a spectator just walking in like somebody's parents), it will really confuse them. With respect to the audience, an intentional tie will be more entertaining, because you might not know if somebody made a mistake, cheated, or got a penalty, which would change the outcome of the match.
From a scouting perspective, that pretty much tells me almost nothing about the robost. Sure, I can see that it's scoring, but I can't see how well the driver reacts under the pressure of a defender. I can't see how well the bot in the middle of the field is at tracking down balls and avoiding the other or causing the other to lose a ball. I can't see if the defensive bot is good at its job. The only thing it really tells me is whether the human player with the trident is fast or not.
Sure, the intentional tie doesn't really either, but I think it'd be a lot less boring to watch. Not to mention then I can at least see both alliances' human players in action.
StephLee
09-01-2010, 20:11
Does anyone else (who's been in FIRST long enough to remember) have a very uneasy presence of the ghost of 2003?
Do a search on the word "collusion" to see the very heated arguments over what people did back in the day to seed better. I can't believe FIRST would want to invite that kind of ugliness back again
I wasn't around until two years after that, but I remember it being a topic of discussion even when I joined. I don't think I'd like to see it first hand.
That is very true. But then from an audience standpoint, a collusion shutout won't be very popular. It would be boring, and for those who don't know how seeding works (e.g. a spectator just walking in like somebody's parents), it will really confuse them. With respect to the audience, an intentional tie will be more entertaining, because you might not know if somebody made a mistake, cheated, or got a penalty, which would change the outcome of the match.
From a scouting perspective, that pretty much tells me almost nothing about the robost. Sure, I can see that it's scoring, but I can't see how well the driver reacts under the pressure of a defender. I can't see how well the bot in the middle of the field is at tracking down balls and avoiding the other or causing the other to lose a ball. I can't see if the defensive bot is good at its job. The only thing it really tells me is whether the human player with the trident is fast or not.
Sure, the intentional tie doesn't really either, but I think it'd be a lot less boring to watch. Not to mention then I can at least see both alliances' human players in action.
I am hesitant to agree that an intentional tie would be likely to entertain the audience either, especially if the audience is expecting competition and realizes halfway through a match that the teams are simply scoring in their own goals with no real intention to actually compete against one another.
Lil' Lavery
09-01-2010, 20:12
In that case, the Blue alliance will get the same number of seeding points as the winning alliance (not including penalties) - so therefore derives no advantage over the Red alliance in terms of moving up the ranking - in 2 cases.
I agree with some of the concepts in your post, but you're missing an important point. In this case it does indeed not result in any notable advantage for the red alliance to gain points in the standings over the blue alliance, but if they end up getting more seeding points than a "standard match," they get a net bonus when compared to the other teams in the field (not playing on either alliance).
Say team X is on the red alliance and team Y isn't playing. Team Y is done playing matches and has 100 seeding points. Team X has one match remaining and has 80 seeding points. In a competitive match, the red alliance might win 6-5 and get 16 seeding points (not enough for Team X to catch Team Y). In a colluded match, the red alliance might win 21-0, and Team X could pass Team Y. Or they could attempt to tie 7-7 and earn 21 seeding points for both alliances, in which case Team X passes Team Y as well.
SteveGPage
09-01-2010, 20:15
Sure, the intentional tie doesn't really either, but I think it'd be a lot less boring to watch. Not to mention then I can at least see both alliances' human players in action.
The big risk however to an "intentional" tie agreement (and I still see that there is a lot of risk that someone would also cheat) is that due to the shear magnitude of penalties that are possible, you would have to have an intentional tie, with absolutely no penalties. How many times have we seen matches that appeared to go one way, to only after the penalties were assessed, the match went to the other alliance. Many of the penalties - case in point "Overdrive" - were not obvious to the team/alliance at the time because they couldn't tell that had even violated a rule due to limited vision what the robot was doing on the far side of the track.
I am hesitant to agree that an intentional tie would be likely to entertain the audience either, especially if the audience is expecting competition and realizes halfway through a match that the teams are simply scoring in their own goals with no real intention to actually compete against one another.
Note that I said more entertaining than a colluded shutout, which is a valid point. The reason for this is because you still don't know if one team will cheat at the end or of somebody will make a mistake or get a penalty be accident.
The big risk however to an "intentional" tie agreement (and I still see that there is a lot of risk that someone would also cheat) is that due to the shear magnitude of penalties that are possible, you would have to have an intentional tie, with absolutely no penalties. How many times have we seen matches that appeared to go one way, to only after the penalties were assessed, the match went to the other alliance. Many of the penalties - case in point "Overdrive" - were not obvious to the team/alliance at the time because they couldn't tell that had even violated a rule due to limited vision what the robot was doing on the far side of the track.
I'm well aware of this. However, I don't see many of the penalties this year being like <G22> and maybe <G35> from 2008. I'm fairly certain that the only one that might be like this not crossing the center line in autonomous.
SteveGPage
09-01-2010, 20:24
Say team X is on the red alliance and team Y isn't playing. Team Y is done playing matches and has 100 seeding points. Team X has one match remaining and has 80 seeding points. In a competitive match, the red alliance might win 6-5 and get 16 seeding points (not enough for Team X to catch Team Y). In a colluded match, the red alliance might win 21-0, and Team X could pass Team Y. Or they could attempt to tie 7-7 and earn 21 seeding points for both alliances, in which case Team X passes Team Y as well.
True, but I think with X, Y, and Z playing against A, B, C the motives of the individual teams gets a little messy! The hardest part of this concept is keeping all 6 teams in the match honest with the "agreement" made, when each may have a very pressing reason to defect from the agreement.
mathking
09-01-2010, 20:41
The colluded shutout agreement runs some significant risk for the "winning" alliance in that if they get any penalties, they will actually come out behind the "losing" alliance in seeding points for that round. You could argue that if they two alliances cooperate they might still score more points than they would have scored if the match had been played with both alliances trying to get the most points. But I suspect that a lot of teams will be unwilling to risk what they might view as an unfair result.
One thing I think is clear is that in most unbalanced matches without any collusion, there will be a point at which the sides start trying to score for each other. If there are 30 seconds left and you are up 12-1, you have more incentive to put in balls for the opponents than for yourself. Furthermore, unless you have a reasonable chance to overtake the other side, you have no in-game incentive to try to hang from the bar if you are the losing side. All you will be doing is giving twice as many points as you earn to the winning alliance, while giving none to yourself.
Again I will say that this is a REALLY interesting game theory problem. I wish I were still teaching discrete math, because I would love to have a test question asking students to devise and justify a good strategy for the game.
SteveGPage
09-01-2010, 20:45
Again I will say that this is a REALLY interesting game theory problem. I wish I were still teaching discrete math, because I would love to have a test question asking students to devise and justify a good strategy for the game.
So true! I think an alum at my very, very, very small college in Southwestern Virginia (John Forbes Nash, Jr) would love this game, too!
StephLee
09-01-2010, 21:17
Has anyone pointed out that perhaps you COULD benefit from giving your opponent more seeding points by scoring for yourself, even in a no-win situation at the end of the game?
I know in the past, we have ended up with excellent alliance partners based on our performance against them in qualification matches. Giving a likely top-8 team a run for their money is, in my experience, a good way to get them to notice you, and such a game played without nastiness is a good way to get them to notice you favorably. If the primary overall goal during the qualification matches is to seed highly yourself, the secondary should be to put on a good showing for the teams that do seed highly enough to impact your fate for the elimination rounds.
Not to mention, if you have the choice between helping your opponent at no cost to yourself or not helping your opponent also at no cost to yourself (as in, you wouldn't win either way)...well, it'd be within my personal definition of GP to help my opponent.
SteveGPage
09-01-2010, 21:25
I know in the past, we have ended up with excellent alliance partners based on our performance against them in qualification matches. Giving a likely top-8 team a run for their money is, in my experience, a good way to get them to notice you, ...
I remember that well ... but, trust me, our scouting had all ready noticed you that year! :)
This year we get to see each other's teams at both the Chesapeake and Buckeye Regionals ... maybe this year, we will get you to notice us! :)
Tom Line
09-01-2010, 22:09
The colluded shutout agreement runs some significant risk for the "winning" alliance in that if they get any penalties, they will actually come out behind the "losing" alliance in seeding points for that round. You could argue that if they two alliances cooperate they might still score more points than they would have scored if the match had been played with both alliances trying to get the most points. But I suspect that a lot of teams will be unwilling to risk what they might view as an unfair result.
One thing I think is clear is that in most unbalanced matches without any collusion, there will be a point at which the sides start trying to score for each other. If there are 30 seconds left and you are up 12-1, you have more incentive to put in balls for the opponents than for yourself. Furthermore, unless you have a reasonable chance to overtake the other side, you have no in-game incentive to try to hang from the bar if you are the losing side. All you will be doing is giving twice as many points as you earn to the winning alliance, while giving none to yourself.
Again I will say that this is a REALLY interesting game theory problem. I wish I were still teaching discrete math, because I would love to have a test question asking students to devise and justify a good strategy for the game.
You may call it interesting. I call it insanity. I have to wonder that since entertainment and audience inclusion was a big part of kickoff, what part of this system makes it:
#1 Easier for the audience to understand
#2 More entertaining for the audience
I'm pretty sure that if half way through the game the teams start scoring on themselves, the audience is going to go "what the <insert word>". If then someone explains to them that the teams on the floor are artificially manipulating the scores to obtain a result rather than simply competing against one another, I'm pretty sure most of the "audience" are going to wonder exactly what type of competition their watching.
Simple it isn't. Entertaining.... I don't think so. I certainly hope they fix this quickly. Though at this point it doesn't really matter - we're all still going to design robots to score as quickly as possible.
Everyone seems to be missing two key factors:
1. If the two alliances are not colluding, then the best strategy for the alliance planning to score 0 is to block both goals with its robots and push the balls into the other alliance's zone.
2. Competition should lower the maximum amount of points a given alliance can get, given competency on the part of both alliances. If we both get 6 of the balls, then that's a 6 point margin that's being taken away. If one alliance plans to get 0 points, then the other alliance gets 12 points plus however many they can put back onto the field.
Some say (http://www.chiefdelphi.com/forums/showpost.php?p=895009&postcount=31) that the cooperition bonus "is for this purpose", while shying away from the stronger statement that the cooperition bonus would compel both teams to compete. This simply isn't true; they fail to factor in the heightened maximum points you can gain from not scoring any goals.
For instance, take the numbers that have been thrown around. Say the score is 8-8. That means that 4 balls have been re-entered into the field. A total of 16 balls. The winning team in kirtar's scenario is 8-8, or 8-7, or something close to a tie. The winning alliance gets 12 seeding points minus penalties, while the losing team gets 8 points.
Compare this to the strategy of scoring 0: using the same number of active balls, one alliance scores 16 and the other 0. The final seeding points distributed is 16 points to each team. 4 points higher for the winning alliance, and double the losing alliance's potential score. It is simply not the case that it is always beneficial to tie a game.
In fact, the original scenario assumes that 8-8 is the score, assuming that the number of active balls is 16. The two teams fight for the 16 balls, making sure that they are not efficiently delivering goals and re-entering balls into the play field. If they did, by teaming up to score the most points, they could increase the number of balls higher, thus making their score even higher.
Why does everyone think this strategy is counter to the spirit of FIRST? It's not counter to the spirit of the rules, since Dean Kamen himself foresaw this use of the rules in his patent (http://www.chiefdelphi.com/forums/showpost.php?p=895009&postcount=31). Second, it effects an environment in which both teams work together in order to maximize their benefits. Is this not more close to gracious professionalism than competing to make sure you beat the other team just hard enough to get the win, but just soft enough to get the most cooperition bonus? Surely once you have established the win, it's almost manipulative to behave this way. If you are at 1 minute left with your alliance at 8 points and the other alliance at 4, a team looking for seeding points (and who isn't?) would wait for the other alliance to score up to 3 points to gain that extra 6 seeding points. The losing team knows this. They know it, and they know that they're going to lose and any attempts to win would just increase the other alliance's seeding points at the end of the match. What hopelessness. Scoring for yourself actually drags you lower relative to the mean seeding points.
I'm not advocating this as a main strategy; when I presented it to my team our mentors repeatedly said that it would probably be changed, it might be entirely eliminated, and that we shouldn't use it as our main strategy. The strategy we could use, if this isn't changed, is determine whether we're going to lose or not beforehand. If we are, then convince our alliance partners to not score. If they won't be convinced, score like mad. If they will be convinced, don't score and block the goals, helping the other team while preventing them from lowering our score relative to them.
Again, most teams will likely examine this strategy and find it too risky. What if the other team scores on us - then we actually lose points. I'm just pointing out that this is a legal, effective strategy, and worthy of discussion, even as a matter of pure game theory. Questions of fairness or fidelity to our foundations are tangential at best.
Enigma's puzzle
10-01-2010, 10:07
I believe that these seeding rules were actually implemented to encourage offensive teams that tried to accomplish scoring instead of just going for the defensive approach. And actually i think that this will do a more accurate job of calculating the most potent scoring robots, and actually in an effort to raise the score of the match many teams will not put someone on the "defensive" end to encourage scoring and quick ball recycling. This strategy would benefit both sides because hopefully the score by both teams will increase.
One thing I think is clear is that in most unbalanced matches without any collusion, there will be a point at which the sides start trying to score for each other.
Yes, this seemed like the "missing" conclusion to me too.
At some point of imballance BOTH teams are better off scoring on each other.
Of course, at that point, the whole "This 1020 game is easier for the crowd to understand." concept goes out the door.... Will it make sense to anyone not on this forum that the red is now scoring for blue and that blue is now scoring for red?
It will be interesting if the term "Tipping Point" starte getting applied to matches.....
Annoucer..... "Looks like we've hit the tipping point. Teams are starting to score in each other's goals."
You read it here first :)
Phil.
Daniel_LaFleur
10-01-2010, 12:32
So true! I think an alum at my very, very, very small college in Southwestern Virginia (John Forbes Nash, Jr) would love this game, too!
You're right. :D
This is a great example of 'Nash equilibrium', which is why I don't believe we'll see as much collusion (but we will see some) as is being posted here.
Each team will work to their own benefit ... which will mean that they will be very motivated to break their 'agreements' when the crucial time comes.
You're right. :D
This is a great example of 'Nash equilibrium', which is why I don't believe we'll see as much collusion (but we will see some) as is being posted here.
Each team will work to their own benefit ... which will mean that they will be very motivated to break their 'agreements' when the crucial time comes.
The nash equilibrium refers to a strategy which, when all players are aware of it, is the optimum strategy while accounting for any changes other teams may make in their strategy. No player has any reason to change, since the strategy they have is optimal, and any change would lead in a decrease in benefits.
The nash equilibrium of this game is not every team working to break agreements; An alliance has much more to gain if they score 0 than if they work hard, given that the other alliance is at least somewhat similar in scoring ability.
I made a long post that never appeared; expect this post to be edited ~6:30pm EST.
mathking
10-01-2010, 13:06
This game does not really have a Nash equilibrium. With the "n to 0" strategy the side with the n points has an incentive to unilaterally change strategies to score for the opponent. This incentive increases if they have received (or think they received) any penalties. If the sides are playing for an "n to n" tie they both have an incentive to score n+1 and end up with 3n +1 points instead of 2n points. I would guess that the game designers looked at all this when they designed the game.
FIRST really wanted to make sure that we're all winners this year.
waialua359
11-01-2010, 03:54
The way I see it, seeding points is just like W-L, except in the way you can position yourself higher/lower than another team.
In the W-L method, the only way you can overcome a team with a better record, is if they start losing matches, and often times is out of your control.
In the seeding system, you have more control to a certain extent, of positioning yourself with higher seeding points vs. other teams.
Penalties in the past sometimes was the difference to a W-L. The same situation this year would be a loss of coopertition bonus points. That's pretty substantial in a high scoring close match, where the penalties were a difference maker.
Daniel_LaFleur
11-01-2010, 08:38
The nash equilibrium refers to a strategy which, when all players are aware of it, is the optimum strategy while accounting for any changes other teams may make in their strategy. No player has any reason to change, since the strategy they have is optimal, and any change would lead in a decrease in benefits.
The nash equilibrium of this game is not every team working to break agreements; An alliance has much more to gain if they score 0 than if they work hard, given that the other alliance is at least somewhat similar in scoring ability.
I made a long post that never appeared; expect this post to be edited ~6:30pm EST.
I believe you are incorrect here.
Both alliances benefit better in a high scoring tie (the reason for this, over your 'x to 0' strategy is that coopertition points are the tiebreaker ... therefore teams would wish to maximize them), so early on in the competitions they are well motivated to do just that (high scoring ties).
But as the competition goes along and rounds will score differently, teams will become more motivated to change that strategy so that they will move up in the seeding.
Applying the Nash Equilibrium to game strategy will mean that teams will change their strategy towards scoring based upon their seeding position as the regional goes on, and therefore the Nash Equilibrium must be re-evaluated each match.
... it's going to be a fun year :D
hipsterjr
11-01-2010, 09:26
Lets look at a hypothetical match:
Say winning team "W" scores 6 points and gets a 3 point penalty. Losing team "L" has a final score of 1. This would mean that W gets 5 seeding points (3W + 2x1L). L would get 6 seeding points (W's points without penalies).
This means that a team will have to be very aware of the score during the match. I find it a litte odd that this kind of scenario could happen.
My $0.02
Jonathan Norris
11-01-2010, 12:42
Like alot of people I am still very confused with this seeding system, to me it looks like 'collusion' will be rampent and the winning team will be determined by a coin flip at the beginning of the match between the two alliances. Here's my logic:
Teams will be ordered first by seeding score. then the cooperation bonus is the tie braker. Therefore teams want the highest seeding score, making it far more important then the cooperation bonus. I would much rather loose a match help the other alliance and get double the seeding score then compete and get the cooperation bonus that only counts as a tie braker. here are the applicable rules:
9.3.9 Qualification Seeding
All TEAMS in attendance will be seeded during the qualification matches. If the number of TEAMS in attendance is 'n', they will be seeded '1' through 'n', with '1' being the highest seeded team and 'n' being the lowest seeded TEAM.
The Field Management System will use the following seeding method:
• TEAMS will be seeded in decreasing order by seeding score.
• Any TEAMS having identical seeding scores will then be seeded in decreasing order by their highest coopertition bonus.
9.3.7 Seeding Score
The total number of seeding points earned by a TEAM throughout their qualification matches will be their seeding score.
where seeding points are defined as:
9.3.4 Match Seeding Points
All teams on the winning ALLIANCE will receive a number of seeding points equal to the penalized score (the score with any assessed penalties) of the winning ALLIANCE.
All teams on the losing ALLIANCE will receive a number of seeding points equal to un-penalized score (the score without any assessed penalties) of the winning ALLIANCE.
In the case of a tie, all participating teams will receive a number of ranking points equal to their ALLIANCE score (with any assessed penalties).
9.3.5 CoopertitionTM Bonus
All teams on the winning ALLIANCE will receive a coopertition bonus: a number of seeding points equal to twice the un-penalized score (the score without any assessed penalties) of the losing ALLIANCE.
In the case of a tie, all participating teams will receive a coopertition bonus of a number of seeding points equal to twice their ALLIANCE score (with any assessed penalties).
so lets take two sernerios one where all the teams score for RED and one where they compete.
6 teams scoring for RED:
RED 20 points
BLUE 0 points
therefore the seeding points are:
RED 20 seeding points
BLUE 20 seeding points
Coopertition bonus would be:
RED 2*0 = 0
BLUE 0
In my interpretation the Coopertition bonus is totally separate from the seeding points, it says nothing about adding the Coopertition bonus to seeding score in the rules. This is where I think I may be wrong, someone please correct me if so.
the two alliances compete:
RED 10
BLUE 7
Seeding points:
RED 10
BLUE 10
Coopertition bonus:
RED 2*7 = 14
BLUE 0
To me with my interpretation I dont see a situation where teams would actually want to compete against one another... I'm not even going to get into the situation where the loosing teams actually receives more seeding points due to penalties on the winning team.
I am confused and don't see how this seeding system makes any logical sense, what am I missing? :confused:
Alan Anderson
11-01-2010, 12:49
In my interpretation the Coopertition bonus is totally separate from the seeding points, it says nothing about adding the Coopertition bonus to seeding score in the rules. This is where I think I may be wrong, someone please correct me if so.
You quoted it yourself:
All teams on the winning ALLIANCE will receive a coopertition bonus: a number of seeding points equal to twice the un-penalized score (the score without any assessed penalties) of the losing ALLIANCE.
In the case of a tie, all participating teams will receive a coopertition bonus of a number of seeding points equal to twice their ALLIANCE score (with any assessed penalties).
The bonus does add to seeding points.
Jonathan Norris
11-01-2010, 12:59
The bonus does add to seeding points.
Ok that was the part that I thought was not worded well, to me it was not clear if the Cooperation bonus was a totally separate tally of points being used as a tie breaker or added to the seeding points.
9.3.7 should really read something like this to make it more clear:
"The total number of seeding points and CoopertitionTM Bonus points earned by a TEAM throughout their qualification matches will be their seeding score."
I got confused when they didn't reference Coopertition Bonus points when describing the seeding score.
GaryVoshol
11-01-2010, 13:34
The Coopertition(tm) Bonus has units - seeding points. The points earned by Coopertition are added into your total. Coopertition bonus points are also used - separately - as a tiebreaker.
Penalties are a double whammy this year, so don't do things to get penalized. You lose seeding points, and then you could also lose the match meaning you don't get any bonus.
This game does not really have a Nash equilibrium. With the "n to 0" strategy the side with the n points has an incentive to unilaterally change strategies to score for the opponent. This incentive increases if they have received (or think they received) any penalties. If the sides are playing for an "n to n" tie they both have an incentive to score n+1 and end up with 3n +1 points instead of 2n points. I would guess that the game designers looked at all this when they designed the game.
If the score is 0 to n, they have 1 seeding points to gain for every point they score on themselves, whereas if they score on the opponent, they have 2 seeding points to score for themselves. This is true.
However, the 0-alliance is allowed to have 2 robots blocking their own goals. They can also forcibly score on the other alliance's goals, which the winning alliance has no incentive to block.
I believe you are incorrect here.
Both alliances benefit better in a high scoring tie (the reason for this, over your 'x to 0' strategy is that coopertition points are the tiebreaker ... therefore teams would wish to maximize them), so early on in the competitions they are well motivated to do just that (high scoring ties).
But as the competition goes along and rounds will score differently, teams will become more motivated to change that strategy so that they will move up in the seeding.
Applying the Nash Equilibrium to game strategy will mean that teams will change their strategy towards scoring based upon their seeding position as the regional goes on, and therefore the Nash Equilibrium must be re-evaluated each match.
... it's going to be a fun year :D
Both alliances don't benefit in a high-scoring tie. Only one does. If an alliance score is n-u with p penalties on the team with n, with u<n, the n-alliance gains 2u + (n - p) seeding points, larger than n seeding points that the losing team gets.
If u is 0, then the winning alliance gets n-p seeding points, and the losing team gets n seeding points. I don't see how the losing team has any benefit to scoring higher in a tie; only the winning alliance does. You could say it increases their chance of winning, but that's not what we're talking about.
In my interpretation the Coopertition bonus is totally separate from the seeding points, it says nothing about adding the Coopertition bonus to seeding score in the rules. This is where I think I may be wrong, someone please correct me if so.
It's not; the coopertition bonus is added to their running total seeding score. See 9.3.5:
9.3.5 CoopertitionTM Bonus
All teams on the winning ALLIANCE will receive a coopertition bonus: a number of seeding points equal to twice the un-penalized score (the score without any assessed penalties) of the losing ALLIANCE.
In the case of a tie, all participating teams will receive a coopertition bonus of a number of seeding points equal to twice their ALLIANCE score (with any assessed penalties).
To me with my interpretation I dont see a situation where teams would actually want to compete against one another... I'm not even going to get into the situation where the loosing teams actually receives more seeding points due to penalties on the winning team.
They want to compete to win for the Coopertition bonus. If they win, they get it. We can't go assuming what the score will be, then postdict the team's behavior.
I never got to get my post back; I don't know why - it said the post was moderated and it never showed up. Here's the gist of what I was saying:
Some would say (http://www.chiefdelphi.com/forums/showpost.php?p=895072&postcount=41) that this will encourage a lot of colluding. But who do they think will be colluding - two alliances? Hardly. The only collusion that will happen if this strategy is employed is between the teams on one alliance planning to score 0. What is the incentive for the other alliance (the alliance who stands to win the Coopertition bonus) to collude with the losing alliance? If they don't, then they get the Coopertition bonus. Intuitively, though I may think otherwise, it seems that most players view this strategy as counter to GP or the spirit of the rules / FIRST. This would be an additional disincentive to use this strategy.
Here's another proposition with respect to your third one.
Say that that the score is 8-8
Each alliance gets 8+8*2=24, which is 50% more. In every scenario in which there is an even numbered total score, it is always more beneficial to both teams to split the score evenly.
This is actually false. Even with the example you gave, it's false. If alliance W gets 8 points, and alliance L gets 8 points, then it's a tie. As per 9.3.4, they get seeding points (it says ranking points, but probably means seeding points) equal to their own alliance's score. In this case, each alliance gets 8 points, plus twice their alliance score (as per 9.3.5). This results in a net gain of 24 points.
Only in a tie, though. That's an edge case. And it still doesn't apply retroactively to this strategy.
If alliance L gets 0 points, then alliance W gets 8 points for their seeding score. Alliance L also gets 8 points. Primia facie, this seems to be a loss relative to other teams that are playing other matches. However, the previous case is only in the case of a tie. If the previous scenario had a much more likely difference in score, then the points become unbalanced:
Say alliance L gets 6 points, and alliance W gets 8 points. W wins, netting 11 seeding points. L gains 8 seeding points.
This is much more likely than 8-8; consider last year's game. FIRST data shows that at an average regional (picked one at random: New York City Regional 2009 (http://www2.usfirst.org/2009comp/events/NY/matchresults.html)) there was only 1 tie. That's hardly enough to offset the median or mean ranking score with this new system. Even when counting for the fact that the range and standard deviation were both higher, the number of ties will still be significantly low enough.
Another important fact is being glossed over: the number of balls is doubled when one team is playing to score 0. If they are constantly fighting for each ball, then of course they'd only get 8. That means that 4 balls total were recycled. This number would be quite higher - perhaps 12 or more (speculative) - increasing the total ball count. This would, I think, overcome the incentive to work for a coopertition bonus.
In case anyone missed it, here (http://www.chiefdelphi.com/forums/showthread.php?t=79800) is the new algorithm run against last year's championship divisions.
Ricky Q.
12-01-2010, 15:45
This section has been clarified in Team Update #1:
9.3.4 Match Seeding Points
All teams on the winning ALLIANCE will receive a number of seeding points equal to the penalized score (the score with any assessed penalties) of the winning ALLIANCE.
All teams on the losing ALLIANCE will receive a number of seeding points equal to un- penalized score (the score without any assessed penalties) of the winning ALLIANCE.
In the case of a tie, all participating teams will receive a number of ranking seeding points equal to their ALLIANCE score (with any assessed penalties).
9.3.7 Seeding Score
The total number of seeding points (Match Seeding Points plus Coopertition Bonuses) earned by a TEAM throughout their qualification matches will be their seeding score.
This game does not really have a Nash equilibrium. With the "n to 0" strategy the side with the n points has an incentive to unilaterally change strategies to score for the opponent. This incentive increases if they have received (or think they received) any penalties. If the sides are playing for an "n to n" tie they both have an incentive to score n+1 and end up with 3n +1 points instead of 2n points. I would guess that the game designers looked at all this when they designed the game.
If the score is 0 to n, they have 1 seeding points to gain for every point they score on themselves, whereas if they score on the opponent, they have 2 seeding points to score for themselves. This is true.
However, the 0-alliance is allowed to have 2 robots blocking their own goals. They can also forcibly score on the other alliance's goals, which the winning alliance has no incentive to block.
I believe you are incorrect here.
Both alliances benefit better in a high scoring tie (the reason for this, over your 'x to 0' strategy is that coopertition points are the tiebreaker ... therefore teams would wish to maximize them), so early on in the competitions they are well motivated to do just that (high scoring ties).
But as the competition goes along and rounds will score differently, teams will become more motivated to change that strategy so that they will move up in the seeding.
Applying the Nash Equilibrium to game strategy will mean that teams will change their strategy towards scoring based upon their seeding position as the regional goes on, and therefore the Nash Equilibrium must be re-evaluated each match.
... it's going to be a fun year :D
Both alliances don't benefit in a high-scoring tie. Only one does. If an alliance score is n-u with p penalties on the team with n, with u<n, the n-alliance gains 2u + (n - p) seeding points, larger than n seeding points that the losing team gets.
If u is 0, then the winning alliance gets n-p seeding points, and the losing team gets n seeding points. I don't see how the losing team has any benefit to scoring higher in a tie; only the winning alliance does. You could say it increases their chance of winning, but that's not what we're talking about.
In my interpretation the Coopertition bonus is totally separate from the seeding points, it says nothing about adding the Coopertition bonus to seeding score in the rules. This is where I think I may be wrong, someone please correct me if so.
It's not; the coopertition bonus is added to their running total seeding score. See 9.3.5:
9.3.5 CoopertitionTM Bonus
All teams on the winning ALLIANCE will receive a coopertition bonus: a number of seeding points equal to twice the un-penalized score (the score without any assessed penalties) of the losing ALLIANCE.
In the case of a tie, all participating teams will receive a coopertition bonus of a number of seeding points equal to twice their ALLIANCE score (with any assessed penalties).
To me with my interpretation I dont see a situation where teams would actually want to compete against one another... I'm not even going to get into the situation where the loosing teams actually receives more seeding points due to penalties on the winning team.
They want to compete to win for the Coopertition bonus. If they win, they get it. We can't go assuming what the score will be, then postdict the team's behavior.
I never got to get my post back; I don't know why - it said the post was moderated and it never showed up. Here's the gist of what I was saying:
Some would say (http://www.chiefdelphi.com/forums/showpost.php?p=895072&postcount=41) that this will encourage a lot of colluding. But who do they think will be colluding - two alliances? Hardly. The only collusion that will happen if this strategy is employed is between the teams on one alliance planning to score 0. What is the incentive for the other alliance (the alliance who stands to win the Coopertition bonus) to collude with the losing alliance? If they don't, then they get the Coopertition bonus. Intuitively, though I may think otherwise, it seems that most players view this strategy as counter to GP or the spirit of the rules / FIRST. This would be an additional disincentive to use this strategy.
Here's another proposition with respect to your third one.
Say that that the score is 8-8
Each alliance gets 8+8*2=24, which is 50% more. In every scenario in which there is an even numbered total score, it is always more beneficial to both teams to split the score evenly.
This is actually false. Even with the example you gave, it's false. If alliance W gets 8 points, and alliance L gets 8 points, then it's a tie. As per 9.3.4, they get seeding points (it says ranking points, but probably means seeding points) equal to their own alliance's score. In this case, each alliance gets 8 points, plus twice their alliance score (as per 9.3.5). This results in a net gain of 24 points.
Only in a tie, though. That's an edge case. And it still doesn't apply retroactively to this strategy.
If alliance L gets 0 points, then alliance W gets 8 points for their seeding score. Alliance L also gets 8 points. Primia facie, this seems to be a loss relative to other teams that are playing other matches. However, the previous case is only in the case of a tie. If the previous scenario had a much more likely difference in score, then the points become unbalanced:
Say alliance L gets 6 points, and alliance W gets 8 points. W wins, netting 11 seeding points. L gains 8 seeding points.
This is much more likely than 8-8; consider last year's game. FIRST data shows that at an average regional (picked one at random: New York City Regional 2009 (http://www2.usfirst.org/2009comp/events/NY/matchresults.html)) there was only 1 tie. That's hardly enough to offset the median or mean ranking score with this new system. Even when counting for the fact that the range and standard deviation were both higher, the number of ties will still be significantly low enough.
Another important fact is being glossed over: the number of balls is doubled when one team is playing to score 0. If they are constantly fighting for each ball, then of course they'd only get 8. That means that 4 balls total were recycled. This number would be quite higher - perhaps 12 or more (speculative) - increasing the total ball count. This would, I think, overcome the incentive to work for a coopertition bonus.
So maybe I'm wrong, but is this how rank is determined?
1) w/l/t
2) seeding points
3) coopertition points
Am I missing some steps?
GaryVoshol
12-01-2010, 20:27
So maybe I'm wrong, but is this how rank is determined?
1) w/l/t
2) seeding points
3) coopertition points
Am I missing some steps?Nope. You're just adding one - there is no w/l/t ranking this year. And coopertition points are included in seeding points. The tiebreaker is max coopertition points, if seeding points are equal.
So maybe I'm wrong, but is this how rank is determined?
1) w/l/t
2) seeding points
3) coopertition points
Am I missing some steps?
You're both missing and adding steps.
9.3.9 Qualification Seeding
All TEAMS in attendance will be seeded during the qualification matches. If the number of TEAMS in attendance is 'n', they will be seeded '1' through 'n', with '1' being the highest seeded team and 'n' being the lowest seeded TEAM.
The Field Management System will use the following seeding method:
• TEAMS will be seeded in decreasing order by seeding score.
• Any TEAMS having identical seeding scores will then be seeded in decreasing order by their highest coopertition bonus.
• Any TEAMS having identical seeding scores and highest coopertition bonus will then be seeded in decreasing order by cumulative ELEVATED/SUSPENDED points earned by their ALLIANCES throughout the Qualification Matches.
• Any TEAMS also having identical highest coopertition bonuses will then be seeded based on a random sorting by the Field Management System.
Note the lack of W/L/T , and note that your Coop bonuses are added into your seeding score per Update #1. Your highest Coop bonus is then used as the first tiebreaker.
Anne_droid
13-01-2010, 12:44
I must admit, I'm a little leery of this seeding system, due to the many strange scenarios that immediately come to mind, or have been described in this thread.
What I'm trying to do, though, is stay focused on my team's goals. Besides learning, having fun, sense of accomplishment, etc, when we go to a competition, we want to win. It would be strange if we didn't. But would we really feel good about winning, if it was because we came up with the best strategy to manipulate the scoring system? That would be a strange accomplishment. I think I would advise my students to just play the game as best we can, keeping in mind the seeding system in that offense is more productive than defense, and you shouldn't blank the other alliance, and leave it at that.
Maybe at the competitions, the top 8 alliance captains will be made up of those who are best at "working" the seeding system, and then their alliance partners will be the ones who are best at playing the game!
Daniel_LaFleur
13-01-2010, 13:24
Both alliances don't benefit in a high-scoring tie. Only one does. If an alliance score is n-u with p penalties on the team with n, with u<n, the n-alliance gains 2u + (n - p) seeding points, larger than n seeding points that the losing team gets.
If u is 0, then the winning alliance gets n-p seeding points, and the losing team gets n seeding points. I don't see how the losing team has any benefit to scoring higher in a tie; only the winning alliance does. You could say it increases their chance of winning, but that's not what we're talking about.
Here, again, I beg to differ.
2 alliances decide to work together to attain higher seeding. Both are capable of 10 scores without defense:
In a '0 to x' game scenario -- each will score 10 points for a total of 20 to 0. each will get 20 {winners score (20) + 2x coopertition score(0)} seeding points and 0 coopertition points.
In the 'tie game' scenario -- each will score 10 points for a 10 to 10 tie. Each will get 30 {their own score (10) + 2x coopertition score (20)} seeding points and 10 coopertition points for the tiebreaker.
I see this happening early in the regionals as teams jocky for seeding points ann I see it breaking down as teams get closer to the elimination rounds and need to 'remove' potential opponents from the elimination rounds (at least as team captains).
Again, because of the changing structure of seeding, teams strategy will shift forcing each team to re-evaluate each strategy (and thus the Nash Equilibrium) before each match.
Both alliances don't benefit in a high-scoring tie. Only one does. If an alliance score is n-u with p penalties on the team with n, with u<n, the n-alliance gains 2u + (n - p) seeding points, larger than n seeding points that the losing team gets.
If u is 0, then the winning alliance gets n-p seeding points, and the losing team gets n seeding points. I don't see how the losing team has any benefit to scoring higher in a tie; only the winning alliance does. You could say it increases their chance of winning, but that's not what we're talking about.
Please, please, please work out your math. No matter what, there is no configuration in which with the same total score that a win will give you more seeding points than a tie unless the losing team had penalties. Regardless of whether you win or lose, you automatically get points based on your own score after penalties, and then twice the opponent's score prior to penalties. Say for an example that n points are scored in a match (after penalties). In a tie, each team has score n/2 after any penalties. Then each team gets n/2+2n/2 in seeding points. If a team shuts out the other, it will simply get n-p. If the match has a winner (winning margin of even number x), you get (n/2)+x/2+2((n/2))-x/2+o) for the winning team, yielding 3n/2+o-x/2. Unless the losing team has penalties, given a certain total score you will always get more seeding points by tying. As a note, this function is actually applicable to many scenarios (including a shutout) Oh by the way, the losing team does benefit from the tie since it actually gets a coopertition bonus instead of not getting one.
This is actually false. Even with the example you gave, it's false. If alliance W gets 8 points, and alliance L gets 8 points, then it's a tie. As per 9.3.4, they get seeding points (it says ranking points, but probably means seeding points) equal to their own alliance's score. In this case, each alliance gets 8 points, plus twice their alliance score (as per 9.3.5). This results in a net gain of 24 points.
Only in a tie, though. That's an edge case. And it still doesn't apply retroactively to this strategy.
If alliance L gets 0 points, then alliance W gets 8 points for their seeding score. Alliance L also gets 8 points. Primia facie, this seems to be a loss relative to other teams that are playing other matches. However, the previous case is only in the case of a tie. If the previous scenario had a much more likely difference in score, then the points become unbalanced:
Say alliance L gets 6 points, and alliance W gets 8 points. W wins, netting 11 seeding points. L gains 8 seeding points.
This is much more likely than 8-8; consider last year's game. FIRST data shows that at an average regional (picked one at random: New York City Regional 2009 (http://www2.usfirst.org/2009comp/events/NY/matchresults.html)) there was only 1 tie. That's hardly enough to offset the median or mean ranking score with this new system. Even when counting for the fact that the range and standard deviation were both higher, the number of ties will still be significantly low enough.
Another important fact is being glossed over: the number of balls is doubled when one team is playing to score 0. If they are constantly fighting for each ball, then of course they'd only get 8. That means that 4 balls total were recycled. This number would be quite higher - perhaps 12 or more (speculative) - increasing the total ball count. This would, I think, overcome the incentive to work for a coopertition bonus.
Your entire argument falls apart given that both scenarios imply cooperation (in the context of this thread). If both alliances are aiming for a tie, there will still be an increased total score. My point is that given a certain total score that no matter what, both teams receive maximum points in a tie situation (barring more than 2 points in penalties as shown above).
If you even read the entire topic, you would see that I am arguing that if you are organizing with the other alliance to maximize your seeding points, that a tie requires far fewer points to be scored than a shutout. In a tie, you receive 50% more seeding points than if you were score all of them in one goal. In addition, 2/3 of these seeding points are in the coopertition bonus, which is the first tiebreaker (and the last before random as well).
I'm not even sure what you're even saying in your first paragraph because it is unclear. However, I am able to tell that you completely misinterpreted my entire statement. My statement is that in pure theory, if the alliances worked to score a tie (in which case scoring output would still be doubled) you will always receive 50% more than a shutout. I included a proof above to show why this is the case. As for why I put down 8-8, it is because it was an arbitrary value. I could have easily said that a score could be 12-12 or even 10000-10000 because it does not change the fundamental relationship
This actually means that the losing alliance can get higher seeding points than the winning alliance. As a simplified case, consider that the winning alliance scores 10 points and is penalized 8, but the losing alliance scores zero. The winning alliance gets two points, but the losing alliance gets 10!
This actually makes it unprofitable to play defense if you are winning. In fact, if you are winning by a lot, it makes it profitable to SCORE ON YOURSELF to drive up the losing alliance's score!
I don't like this change. It penalizes defensive bots, when defense is an integral part of the game itself.
Quite a bit of interesting reading here guys. However, in and among all your talk of collusion, scoring for the other alliance, etc, etc, you need to consider what each of those behaviors will mean when it comes time to choose alliance partners. It is assumed that your team would prefer to be one of the top eight, right? Is there anything about the seeding situation that changes that? Now, if you don't make it to the elite 8, what about your behavior during the elimination rounds with respect to seeding maximization will make you stand out as a desirable partner for the elimination rounds?
SteveGPage
16-01-2010, 16:54
Quite a bit of interesting reading here guys. However, in and among all your talk of collusion, scoring for the other alliance, etc, etc, you need to consider what each of those behaviors will mean when it comes time to choose alliance partners. It is assumed that your team would prefer to be one of the top eight, right? Is there anything about the seeding situation that changes that? Now, if you don't make it to the elite 8, what about your behavior during the elimination rounds with respect to seeding maximization will make you stand out as a desirable partner for the elimination rounds?
I think we will see two outcomes from this:
1. Teams who don't do any scouting, and end up in the top 8, will select alliance partners based on final seeding number (as often happens).
2. Teams who have an effective scouting team, will recognize what you can do, regardless of seeding points. I plan on scouting shots on goal vs number of goals, regardless of which goal it is. If we are an alliance captain, we will look for teams that have that kind of firepower.
... Now, here is the problem though, and I think the comment reflects this ... the strategy during the elimination rounds is different than the qualification rounds. In the quals, seeding points are the focus. In the elims, wins vs loses are the focus. So, things like defense, ability to score when being defended, etc... will be hard to judge if everyone is just scoring. Quals will be primarily offensive games, the elims will be defensive games.
Steve
Chris is me
16-01-2010, 17:09
Here, again, I beg to differ.
2 alliances decide to work together to attain higher seeding. Both are capable of 10 scores without defense:
In a '0 to x' game scenario -- each will score 10 points for a total of 20 to 0. each will get 20 {winners score (20) + 2x coopertition score(0)} seeding points and 0 coopertition points.
In the 'tie game' scenario -- each will score 10 points for a 10 to 10 tie. Each will get 30 {their own score (10) + 2x coopertition score (20)} seeding points and 10 coopertition points for the tiebreaker.
I see this happening early in the regionals as teams jocky for seeding points ann I see it breaking down as teams get closer to the elimination rounds and need to 'remove' potential opponents from the elimination rounds (at least as team captains).
Again, because of the changing structure of seeding, teams strategy will shift forcing each team to re-evaluate each strategy (and thus the Nash Equilibrium) before each match.
Any team that tries to collude for ties won't meet success this year.
First, the kind of teams that game Ranking Points by intentionally going "Hey, let's make a match 10-10, then BOTH STOP" aren't the most honest teams around. I'm unfortunately willing to bet at least one team will set something like that up, then backstab the alliance they colluded with.
Secondly, both alliances need to keep track of every penalty and counteract them. These penalties then need to not be overturned after the match. Both the number of penalties and the chance for review make forcing ties unreliable.
These two together makes me think predetermined matches won't be a feasible strategy this year.
Yes, you have a point about "making friends". But that's not the point of the discussion.
FIRST wants to promote coopertition. I get that. I even support it. But this scoring system is not well thought out, and will actually DETRACT from coopertition. Consider a few cases:
1) The score is a blowout; let's say the score is 30 to 8, with 45 seconds to go. I'm on the alliance which is going to win. What is my best strategy? To turn around and score goals in the other team's goal (what is called in soccer an "own goal". Why? Because goals scored for me give me one point for seedings; goals scored for the other team give me two points in coopertition bonus.
2) Same scenario, but I'm on the losing alliance. What is my best strategy? Again, to turn around and score goals for the other team. Because goals scored for me are worth nothing to me and my teammates, whereas goals scored for the other alliance give us one point.
3) I'm on a losing alliance, and I have balls coming into the corral with about 30 seconds to go. What is my best strategy? I HOLD THE BALLS and take intentional penalties. I'm going to lose anyway, and this improves my proportional score vs. the opposite team.
4) Whether I'm on the winning or losing alliance, do I go after the bonus points? Absolutely not. No matter which alliance I am on, those points benefit my opponents more than they do me.
Yes, they are trying to promote coopertition. And yes, that is a composite containing "cooperation". But it's also a composite containing "competition". This scoring system is just not well thought out.
And no, I don't think that pursuing these strategies would hurt us unduly in the selection process. Everyone knows what the point structure is. Everyone also knows that seeding points no longer matter in the elimination rounds, so the strategy changes then.
Alan Anderson
16-01-2010, 23:55
FIRST wants to promote coopertition. I get that. I even support it. But this scoring system is not well thought out,...
This scoring system is an embodiment of the Coopertition patent granted to FIRST last year. That patent has been in the works for years. I think it's been adequately thought out.
All your points make good sense, except where you say the strategy changes between qualification and elimination rounds. That assumes that the strategy in qualification rounds is merely to get more seeding points than your opponents. I believe that the primary strategy is to get more seeding points than all the other teams, and the way to do that is to win matches (preferably close ones). The secondary strategy is to make yourself likely to be chosen as an alliance partner by high-seeding teams, and the way to do that is also to win matches.
I disagree that this system promotes coopertition. Yes, that is the GOAL -- but it encourages the teams to look for what their own long-term best interest is. Whether they pursue that or not, for the rules to ENCOURAGE it is just wrong.
Nor is it right or true to the spirit of coopertition for the two alliances (all six teams) to pursue something that is beneficial to all six. That would be cooperation. Coopertition includes competition, and a fair and honest competition is part of what we should be encouraging. These rules detract from that.
Michael Corsetto
17-01-2010, 01:24
If I am about to play a match vs. an alliance I know I have little or no chance of beating in head to head competition, I see no reason to try and score for yourself. I have every reason to have the two weakest robots on my alliance blocking my goals, with the 3rd bot scoring for the opponents. The opponent now has no choice but to boost their score as much as possible, giving more points to us, the "losers" of the match, as well. "Collusion" isn't the issue, it's the fact that not colluding can still result in a boring, onesided match, where the weaker alliance can strong-arm the stronger one into scoring for them. In another game (09 for instance) this seeding system might have worked. But not for a game where an alliance can easily lock up their score at 0 or close to it. Please change if you can, I still can't believe FIRST actually thought this was a good idea. Great game GDC, just go back to 09's competition section of the manual please.
3) I'm on a losing alliance, and I have balls coming into the corral with about 30 seconds to go. What is my best strategy? I HOLD THE BALLS and take intentional penalties. I'm going to lose anyway, and this improves my proportional score vs. the opposite team.
Breakaway Rules Section 9.3.5:
All teams on the winning ALLIANCE will receive a coopertition bonus: a number of seeding points equal to twice the un-penalized score (the score without any assessed penalties) of the losing ALLIANCE.
So earning penalties would not affect the winning alliance's seeding score for the game.
If you are losing, then scoring a goal for the other alliance gives all six teams on both alliances 1 more seeding point, which improves their seeding scores over the teams that are not in that match. So, this strategy of hoarding balls in your alliance station is not necessarily a good idea. (Unless you think that there is no chance for a goal to be scored for the other alliance.)
Daniel_LaFleur
18-01-2010, 17:34
Any team that tries to collude for ties won't meet success this year.
First, the kind of teams that game Ranking Points by intentionally going "Hey, let's make a match 10-10, then BOTH STOP" aren't the most honest teams around. I'm unfortunately willing to bet at least one team will set something like that up, then backstab the alliance they colluded with.
Secondly, both alliances need to keep track of every penalty and counteract them. These penalties then need to not be overturned after the match. Both the number of penalties and the chance for review make forcing ties unreliable.
These two together makes me think predetermined matches won't be a feasible strategy this year.
First off, where did I say STOP? I said shoot for a tie.
Next, Even if a tie doesn't happen, high scoreing close matches will yield the highest number of seeding points ... and thus help each team accordingly.
If you read the last line of my post, the Nash Equilibrium will force a team that is 'on the bubble' to change it's strategy and want to win for the bonus coopertition seeding points.
Seems to me like the closer we get to the elimination rounds, the less we'll see any collusion ... Hmmm, didn't I say that before :P
ugh. FIRST wasn't kidding when they said that this year our brains will hurt. I'm So confused at the moment. So W/L/T's aren't going to matter at all? Could a Powerful team end up lower than a weaker team because they have more points?
Daniel_LaFleur
18-01-2010, 22:33
ugh. FIRST wasn't kidding when they said that this year our brains will hurt. I'm So confused at the moment.So W/L/T's aren't going to matter at all?
Correct
Could a Powerful team end up lower than a weaker team because they have more points?
This has happened in the past and will happen again.
ugh. ... Could a Powerful team end up lower than a weaker team because they have more points?
Depends on your definition of a powerful team. Do you have another scale for power other than the seeding ranking scale as defined in this year's rules?
@Refresh:
Yes. Here is an example of what you might be talking about:
Team A wins all 10 of their matches with a score of 5-0.
Team A has 50 seeding points.
Team B's robot breaks down in the first match and they lose 0-51. They go back to the pits in an attempt to fix their robot, and don't show up in any of their remaining 9 matches.
Team B has 51 seeding points.
So Team B is seeded higher than Team A.
Obviously this is an extreme example, but it is possible.
Please, please, please work out your math. No matter what, there is no configuration in which with the same total score that a win will give you more seeding points than a tie unless the losing team had penalties.
I don't understand. Who do you think is getting the seeding points? Only the winning alliance gets the coopertition bonus. How about one alliance gets 12 and the other gets 11. Losing alliance gets 12 seeding points, winning alliance gets 34. No matter what the losing alliance does, if they lose, then it doesn't matter how much they lose by, except for a tie. In fact, they have an incentive to score, if they know they are going to lose, as little as possible in order to minimize the effects of increasing the opposing alliance's coopertition bonus.
Regardless of whether you win or lose, you automatically get points based on your own score after penalties, and then twice the opponent's score prior to penalties.
No, read 9.3.4 and 9.3.5 again, noting the bolded phrases. Losing alliance does not get points based on the losing alliance's score:
9.3.4 [...] All teams on the losing ALLIANCE will receive a number of seeding points equal to un- penalized score (the score without any assessed penalties) of the winning ALLIANCE.
.3.5 CoopertitionTM Bonus
All teams on the winning ALLIANCE will receive a coopertition bonus: a number of seeding points equal to twice the un-penalized score (the score without any assessed penalties) of the losing ALLIANCE.
In the case of a tie, all participating teams will receive a coopertition bonus of a number of seeding points equal to twice their ALLIANCE score (with any assessed penalties).
The scenario you mention only occurs in the case of a tie. How many ties were there in a random regional? I found one in that regional I linked; ties are infrequent and shouldn't be expected.
Oh by the way, the losing team does benefit from the tie since it actually gets a coopertition bonus instead of not getting one.
No, it doesn't. See above. Coopertition bonus is only awarded to the winning alliance.
I'm not even sure what you're even saying in your first paragraph because it is unclear. However, I am able to tell that you completely misinterpreted my entire statement. My statement is that in pure theory, if the alliances worked to score a tie (in which case scoring output would still be doubled) you will always receive 50% more than a shutout.
Again, please read the rules. The losing alliance does not get the Coopertition bonus.
You also failed to address my point that in a shutout, the number of potential maximum points is increased, possibly doubling or more.
martin417
24-01-2010, 21:29
Wow, someone just pointed out this post (http://www.chiefdelphi.com/forums/showthread.php?p=851425#post851425) I made almost a year ago.
At every regional I attended, including championships, The seedings did not reflect the ability of the robots or the teams. If FIRST wants this to be a "sport", and be as popular as football, they need to come up with a better way to rank teams prior to alliance pairings....... I do have a problem with some of the best teams not even being in the top 8. Throughout the season, I saw teams that could not score at all ranked as the number one seed, while top scoring bots were not even in the top 10. Perhaps the seeding should be done based on the scoring of the bots rather than a win-loss record.
(emphasis added) Prophetic isn't it?
I don't understand. Who do you think is getting the seeding points? Only the winning alliance gets the coopertition bonus. How about one alliance gets 12 and the other gets 11. Losing alliance gets 12 seeding points, winning alliance gets 34. No matter what the losing alliance does, if they lose, then it doesn't matter how much they lose by, except for a tie. In fact, they have an incentive to score, if they know they are going to lose, as little as possible in order to minimize the effects of increasing the opposing alliance's coopertition bonus.
No, read 9.3.4 and 9.3.5 again, noting the bolded phrases. Losing alliance does not get points based on the losing alliance's score:
The scenario you mention only occurs in the case of a tie. How many ties were there in a random regional? I found one in that regional I linked; ties are infrequent and shouldn't be expected.
No, it doesn't. See above. Coopertition bonus is only awarded to the winning alliance.
Again, please read the rules. The losing alliance does not get the Coopertition bonus.
You also failed to address my point that in a shutout, the number of potential maximum points is increased, possibly doubling or more.
Maybe I should explain this again from the beginning since you obviously don't even reread the posts that I'm responding to. First of all, you yourself said that:
I don't see how the losing team has any benefit to scoring higher in a tie; only the winning alliance does. You could say it increases their chance of winning, but that's not what we're talking about.
In a tie there is no "winner" or "loser". In every single case that you quoted, I was using the case of a tie, so go back and read the rules. Besides, the first part that you quoted from me, while not correct is not actually how the rest of my post was worked out. In every case that I mention the "losing team" it means a team that may have lost by say 1 point in a different situation. My proof remains valid since I was under the assumption that the entire match was fixed as much discussion beforehand had been about. Learn to read not only the rules but at least the thread that you're replying to. While I did make a few errors in what I said, my calculations were according to the rules. No matter what, given total score 2n, the most beneficial to each alliance with respect to maximizing seeding points and precluding a certain number of penalties, the best score is n-n.
Say for an example that the total in one case is 30 (no penalties). If the score is 17-13, then the winning team gets 43 points for seeding and the losing gets 17. Given the same total, if the score is 15-15, both teams get 45 (15+2*15=45) and 45>43. If it is a shutout, both teams get 30 exactly, which is 33% lower than 45. If you read the second half of 9.3.5, you'll see why the above is the case.
As for your last point. The doubled part is that if both teams are cooperating for a tie, the scoring capability is still theoretically doubled, thus addressing your point, but you have not addressed mine. The rules have nothing to do with this point. Coopertition bonus is as above. If it is a tie both alliances get the coopertition bonus.
If you had actually worked out the math, you would have realized this. In the case of a forced shutout (defend against yourself), your scoring output is only increased by a theoretical value of 33%. Even otherwise, you still will probably bottleneck on the return. On a organized tie, you have a doubled rate of return since there is scoring in both. In either case, scoring ability is increased, but for a given total score, a tie will yield 50% more seeding points for both teams.
engunneer
25-01-2010, 16:18
Wow, someone just pointed out this post (http://www.chiefdelphi.com/forums/showthread.php?p=851425#post851425) I made almost a year ago.
This was certainly true. Our team makes a good example on that for 2009.
In Portland, we were the stongest scoring robot and won all of our quals, so 1st seed made sense.
In Seattle, we lost a few matches, but were still a strong robot, but ended up on the #8 alliance
On Galileo, we were nowhere near the strongest robot, since we had Wildstang, HOT, Exploding Bacon, and many other top teams in our division. We did have a very favorable match schedule, and OPR analysis said we'd be in the top 3. We ended up seeding 2nd. We were 3rd most of the time, with 111 and 67 both having the same W/L/T record, but the Qual points were much higher than ours. The only reason we hit 2nd was because HOT lost a match at the very end.
We had a slight chance at Finals on Galileo (except for sudden technical difficulties and a match stuck on another robot's bumper), but 111's alliance would have thoroughly thrashed us if we had made it that far.
Basically what I'm saying is that I agree with you. The new (to most people) system rewards the strongest scorers that win by the slightest margin possible.
That being said, I would have preferred that the coopertition bonus not be twice the loser's score (just 1x), and the loser's seeding points should have been their own score plus half (or some other fraction) of the winner's score. Then it would never be /more/ beneficial to score for your opponent, but only equally beneficial at best. Then a ball in either goal benefits the winner, but the losing team only wants to score for themselves.
For example (under the 'engunneer rules' just described), if the match is currently R20-B15, Blue wants to score a point the normal way (20-16), because it adds 1 to both seeding scores (R36-B26), whereas if they scored against themselves (21-15), Blue would only get 0.5 of a seeding point, while Red still gets the whole point. (R36-B25.5)
Red /could/ score in either goal, since both give them a seeding point, but they would prefer to score (21-15) because of two reasons: It blocks Blue from 0.5 seeding points, and increases the Spread that Blue needs to score to catch up by 1.
Unfortunately, that removes the incentive for the winner to win by as close a margin as possible. To correct that would probably make the rules pretty complicated (while still making each team want to score the normal way instead of against themselves). Without that incentive, the matches will be blow-outs, which is somehow seems like it would be less exciting to watch.
engunneer rules also include 1 point for a robot entirely within the tunnel at the end of the match :)
One other scoring thing in the official rules that I agree with - hanging scores just enough to break close games in an exciting fashion, but not so much that you want a robot that only hangs (as opposed to ramp robots as in 2007, which often decided the match no matter what was hanging on the rack.) There are also no crazy multipliers in the game, which always confuses the general public. There is no analogous sport that they can take that experience from. At least now we can describe the hanging as being like an extra point, and suspending as a two point conversion.
That rambled more than expected, sorry.
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.