What are your thoughts on the scoring system for 2004?
As mentioned in another thread, here are mine.
I would like to see the qualifying points be the difference in scores between the winner’s score and the loser’s. If you win 50 to 40, you get 10 points. If you win 20 to 10, you get 10 points.
The idea would be to encourage defense as well as offence, and get rid of any reason for opponents to make agreements.
True, that would give the loser a negative score but if that was a problem, each team could start out with 10000.
In the elimination rounds, I would like to go back to best 2 out of 3.
However the above scoring system would work for a 2 match system also. Basically the alliances’ total scores for both matches would be compared and the higher score would move on. There would therefore be no reason to reduce your own score in a match.
I’m proponent of FIRST moving to a much simpler scoring system. I would like to see team be seeded on the basis of their won-loss record, using some form of qualification points as a tiebreaker. Under this system, a highly defensive robot would not be punished.
Too often we have had scoring systems where defensive robots could seed high. Last year three of the most dominant robots, teams 60, 71 and 469 all had difficulties obtaining QPs. Under my proposed system a defensive robot who wins all of their matches in a low scoring manner would still seed very high.
With the first tiebreaker being QPs (say three times the losers score), teams would still have incentive to avoid blowing their opponents out. Remember with 50+ teams playing 7-10 matches at a regional, it’s easy to see that many teams would have the same record.
Also if FIRST moved back to best 2 out of 3 for the eliminations, we would have consistency between qualifying and eliminations. In both rounds teams would be rewarded for winning.
This proposed system also makes scoring fairly easy for the inexperienced viewer. I struggled very much in explaining this game to my friends. For most, the scoring system was too complicated for them to want to try and figure out. As a result most were disinterested in the game.
This system may encourage more defensive strategies than FIRST would want, but I think enough teams would still play offensively and keep things interesting.
The reason FIRST went to a 3x the losers score and 2x loser + winner score for QPs was to dissuade teams from blowing out less experienced opponents in the qualifying rounds.
Using a winner-loser QP score would do the exact opposite. It would motivate teams to blow out their opponents in order to get high QPs.
That doesn’t sound very graciously professional. Does it?
Karthik’s idea is brilliant, I think this is one of the best solutions to seeding and scoring problems I have heard. It happened that there were some people winning a lot and not being seeded at all. And I think there is definitely a majority that wants to go back to 2 out of 3 for elims. I saw some second matches in elims that quite frankly, pissed me off, because there was no competition, just everyone trying to get no points.
*Originally posted by Warren Boudreau *
**The reason FIRST went to a 3x the losers score and 2x loser + winner score for QPs was to dissuade teams from blowing out less experienced opponents in the qualifying rounds.
Using a winner-loser QP score would do the exact opposite. It would motivate teams to blow out their opponents in order to get high QPs.
That doesn’t sound very graciously professional. Does it? **
I see what you are saying. I think one of the reasons for the Qualifying Points system was to ensure that dominant teams would keep playing hard, as opposed to getting a lead and then coasting through a match. It would not be audience friendly if one team achieved a dominant position in the first 20 seconds and then didn’t have to try for the rest of the game.
We have seen the problems that can occur with this year’s system. I think tying qualifying points to the gap in scores would encourage teams to do their best and would also prevent any tendency towards colllusion. The problem then is how to prevent new teams from being “blown out”. Well, what happens in other sports?
I think one way to prevent “blow out” is to provide the rookie teams with reliable basic drive train components with some choice of gearing. FIRST took a step in that direction this year. Another way to help balance the game is to incorporate elements in the game that favor new teams. This year’s midfield bar is an example: basic low robots could go under it, where some of the more elaborate robots couldn’t.
Personally I don’t like the “let’s help the new guys” method of scoring. I find that to be artificial and patronizing. I would rather score what our team deserves to score. Then we can work on improving. I say, let’s find other ways to balance the game, starting with reliable basic drive train components. We might then find newer teams beating out some of the veterans who are experimenting with advanced drive systems.
Bottom line: I don’t know any other sport where they decide to boost the loser’s score so they won’t feel bad. And I don’t think it really helps. It just creates artificial scores and makes it hard to tell which designs and strategies are the most effective, and that is part of what we want our students to learn.
The truth is, I think being blown out by an opponent is not nearly as bad as tipping early in the game and lying there with your wheels spinning in the air. That happened to a lot of robots this year, including ours on a couple of occasions.
So let’s balance the game as above, and have real scores. If our robot needs work to be competitive, I want to know about it.
I just didn’t like the scoring system in eliminations.
In the first match, both teams are trying to rack up the points, so what you get is a huge point total.
In the second match, the team that won the first match is just trying to stall. Hence, low point total.
So the team that wins the first match almost always wins.
Why can’t they just make it 2 out of 3 like it used to be?
The best scoring system is for 2 minute robotics games is the “Winning Bonus” system. We have been using this system in the Michigan OCCRA Robotics League for the past 3 years and it is much better than any scoring system FIRST has ever used.
In the winning bonus system, you get a fixed point bonus plus your score if you win…you just get your score if you lose. We scale the bonus value to be approximately what we expect the winning score to be in an average match. This gives you about a 2x score if you win. We like low scoring games (easier for the audience), our games usually have a max possible score of 30-50 points. Last year we gave a 10 point winning bonus. The beauty of this system is it encourages both offense and defense. If you win 1-0 you still get a half decent score. If you lose 30-35, you get a good score even though you lost. Basically, the better you are, the less winning matters because you can “outscore the system”.
FIRSTs systems put too much emphasis on scoring, but make defense too easy. Our way is better.
*Originally posted by Jim Zondag *
**The best scoring system is for 2 minute robotics games is the “Winning Bonus” system. We have been using this system in the Michigan OCCRA Robotics League for the past 3 years and it is much better than any scoring system FIRST has ever used.
In the winning bonus system, you get a fixed point bonus plus your score if you win…you just get your score if you lose. We scale the bonus value to be approximately what we expect the winning score to be in an average match. This gives you about a 2x score if you win. We like low scoring games (easier for the audience), our games usually have a max possible score of 30-50 points. Last year we gave a 10 point winning bonus. The beauty of this system is it encourages both offense and defense. If you win 1-0 you still get a half decent score. If you lose 30-35, you get a good score even though you lost. Basically, the better you are, the less winning matters because you can “outscore the system”.
FIRSTs systems put too much emphasis on scoring, but make defense too easy. Our way is better. **
That system definitely sounds better than the scoring system we had this year. However it seems like it would still encourage “opponent agreements” to leave stacks up, let all the robots on the ramp, etc. since there is an advantage to both alliances to have high scoring games.
I would like a system which doesn’t reward “collusion”, does reward a team for being better than their opponent, and keeps both teams competing to the end of the match. I think just making the qualifying points equal to the gap in the scores would accomplish that. The losing team gets negative qualifying points equal to the winners positive qualifying points. The only complaint I have heard so far is that this system would encourage “blowing out” your opponent, but that doesn’t worry me. I personally would much rather be “blown out” than tipped over and immobilised (as happened to us this year on several occasions).
I am strongly against the new elimination scoring because of what happened at nationals. In the semifinals of the Newton division, our alliance went up against the hot + rage alliance. It was a tough, high scoring match due to hot’s ability to stack and since most of the match was spent with our robots fighting each other rather than descoring the opponent’s zone. It was an exciting match, with hot tipping over in autonomous, only to be righted seconds later, gunn getting stuck on a box, us saving gunn only to get stuck on rage (we had run up their robot so far our wheels were off the ground and our robot was at a 60 degree angle). We eventually used our arms to get off rage and make it to the ramp after descoring a few bins. In the final 10 seconds, it looked like we were going to win with our 2 robots on the ramp versus their 1. Hot was still off the ramp building / protecting their stack. In the final seconds, hot was able to push one of our robots off the bridge and get on themselves.
The final score, them: 94, us: 47. EPs, them: 188, us: 47, a deficit of 141. If we had gotten the 2 ramp robots, it would have been 72 for us and 69 for them, a defecit of 141 in the other direction. There is no way to overcome a deficit like that. We had 2 great alliances facing off, and the entire thing was decided in 5 seconds of one match. The second match was basically not even worth playing. We tried our best to stop them from descoring the zones, but I bet we still would have lost even if they had never touched their controls.
Maybe it was the insane amount of points for getting on the ramp that screwed up the scoring, but I doubt it. I think they should just go with the 2 out of 3 method, but I am concerned that they might stick with the current method because it reduced the number of matches teams had to play in the elimination rounds.
By the way, sorry to hot for any damage we may have caused in the second round. We were trying to get our wedge, designed for removing KoH robots, under you to try to get your wheels off the ground and keep you from descoring. It was hard to see though and I think we ended up doing more damage than good with that thing. I hope we weren’t the reason your robot shut down in the finals.
I would like to see the scoring system change a bit next year. During the qualifying rounds I would still like 4 robots on the field but they would be competing for themselves no alliances. Qualifying points would be 1st 4 times low score, 2nd 3 times low score, 2nd 2 times low score and 4th your score. After the qualifying rounds the top 8 teams would select partners (alliances) for the finals. The finals would be 2 vs 2 with best 2 out of 3.
*Originally posted by Jay Lundy *
**I am strongly against the new elimination scoring because of what happened at nationals. In the semifinals of the Newton division, our alliance went up against the hot + rage alliance. It was a tough, high scoring match due to hot’s ability to stack and since most of the match was spent with our robots fighting each other rather than descoring the opponent’s zone. In the final 10 seconds, it looked like we were going to win with our 2 robots on the ramp versus their 1. Hot was still off the ramp building / protecting their stack. In the final seconds, hot was able to push one of our robots off the bridge and get on themselves.
The final score, them: 94, us: 47. EPs, them: 188, us: 47, a deficit of 141. If we had gotten the 2 ramp robots, it would have been 72 for us and 69 for them, a defecit of 141 in the other direction. There is no way to overcome a deficit like that. We had 2 great alliances facing off, and the entire thing was decided in 5 seconds of one match. The second match was basically not even worth playing. We tried our best to stop them from descoring the zones, but I bet we still would have lost even if they had never touched their controls.
Maybe it was the insane amount of points for getting on the ramp that screwed up the scoring, but I doubt it. I think they should just go with the 2 out of 3 method, but I am concerned that they might stick with the current method because it reduced the number of matches teams had to play in the elimination rounds. **
I have observed a number of elimination rounds, some from www.soap108.com, where similar things occurred. It particularly bugs me when I see a team lose their first match by 50 to 52 and then win their second match 80 to 10 but not advance. True, the other team was probably descoring, but it sure isn’t clear which is the best team. Some luck in the game is fine, but this year’s game had too much luck.
Anyway, you are definitely not alone in wanting to see the end of this year’s scoring system in the elimination matches. I also vote for 2 out of 3 in the elimination rounds. If we have to stick to highest score over 2 matches, then each team should just get the total of their scores. Under that system, your team would have had a shot to come back. Also no one would be descoring, which must seem really weird to spectators.
I would like to go further and arrange the scoring so that we don’t see “opponent agreements” again. That also tended to throw things off the rails.
*Originally posted by DougHogg *
**Also no one would be descoring, which must seem really weird to spectators. **
Even when you know the scoring in and out, it still seems weird, and is really annoying to watch.