Seeding System

I was just curious as to what people’s thoughts were on the new scoring system, as opposed to previous years. Is the number of wins a preferable method of determination of ranking over the accumulation of your opponents scores?

Two years ago, we had a match where we won, but we accidently pulled the last goal out of our opponents zone and ended up getting no QPs. It was very upseting to see our robot work perfectly and then still lose because of a driver mistake. Plus, the other team got no points either. I LOVE this system.

I love the new system and I’ve been asking for it for years. I remember the great thread back in 2000 when Joe Johnson made the analogy that we were “having a bunting contest to determine who plays in the World Series” (with the old system). I think it worked out well at VCU. I think it should continue to work well, but time will tell.

I really liked the system this year. Wins were important for seeding, but you still had to think about QPs and letting your opponents score and even helping them score points.

I believe that the new system is a much needed improvement to the previous system they had and it worked well at the NJ regional were tribe never lost a match but I don’t believe they had the highest average but they were still the 1st seed which they really deserved. So this system works well

I like the win/loss thing much better than previous years’ systems, but I’d rather have scoring within the tiers be based on the team’s score rather than the opponents’ score.

the concept of basing your qp’s on the oppents score times two was initiated with gracious professionalism in mind. It’s FIRST guys :wink:

While GP is certainly important to the whole FIRST idea, I think basing the seedings off of the opponents’ scores strays a bit from professionalism. Professionals play fair and respect their opponents, but they still play to win. Scoring points for your opponent to increase your own score seems selfish and patronizing–you are telling the other team their best efforts aren’t good enough. Further, because of the 2x rule last year, 254 in at least one final played to lose–we just cleared out all the boxes from both sides to prevent the other team from catching up in QPs. We played to win the final, but not the match, and it feels like we cheated another team out of a good round because of it–not terribly GP.

I believe the new system will work much better, though I agree that seedings should be based on a teams own score. Let gracious professionalism pervade our lives…but let us still be professionals out on the field.

Or maybe I don’t get GP at all =), feel free to yell at me.

Cheers,

Love the new system. It’s more team friendly AND much more understandable for the audience (parents, sponsors, media).

I like the new system BUT what about this for the future.

#1 Seed: Best win/loss record
#2 Seed: Highest QPs
#3 Seed: Second best win/loss record
#4 Seed: Second highest QPs
#5 Seed: Third best win/loss record
#6 Seed: Third highest QPs
#7 Seed: Fourth best win/loss record
#8 Seed: Fourth highest QPs

Obviously there would need to be tie-breakers … for win/loss, it would be QPs and for QP, it would be win/loss.

This would certainly change strategies and some teams might change their strategy midway through a tournament if they come out of the gate with a few losses. Admittingly, this would be confusing for spectators but most people just look at a scoreboard to get standings.

Just some “crazy” thoughts,
Lucien

That would certainly mix things up…but I don’t really see the advantage behind it. Rather, it might penalize defensive robots when they lose and artificially raise up high scoring teams (those that lose). Neither strategy should be favored directly by FIRST over another…which is why the direct win/loss is so beneficial.

It’s funny that Paul wrote this… I was having the exact same thought. I couldn’t figure out why this was… and though I have no official contact with FIRST, I think I figured it out.

Let’s say that the #1 seed has an 8-0 record with an average opponent score of 80, and the #2 seed has an 8-0 record with an average opponent score of 20.

If you make the assumption that in order to win, you need to play better offense than defense this year, (which I believe to be true, though perhaps other will dsisagree), then you know that the average score of the #1 seed MUST be higher than 80, AND they’ve been going against tougher opponents.

In a lot of ways, the average score of your opponents is sort of like a “strength of schedule” factor that you see in professional sports. I think this is why they have it in place.

Just my two cents,

Matt

How about we implement the 'Bot Championship Seeding" (BCS) system:

The coaches vote and rank the teams

The media votes and rank the teams

You get debits for strength of schedule (based on who you beat in the top ten)

You add 1 point for each loss

You have 4 computer systems that noone ever heard of come up with a ranking based on some bizarre scientific reasoning.

The 2 teams with the lowest total score from the above play each other.

Then, regardless of who wins the game, you still vote for the number 1 team.

Nah, no sport would ever consider something as silly as that.

That’s just too much.

Nice Picture.

How about we implement the 'Bot Championship Seeding" (BCS) system:

NOOOOOOOOOOOOO!!!

anyway, I personally prefer the new seeding system. It make the game more viewer friedly and attractive to “outsiders”. It adds drama to the game, where there are more last second scrambles to win through defense and offense rather than just trying to score points. Teams are rewarded for both defense and offense in the rankings as well as scouting reports. And that is important for some teams, as many teams only talk to the top 15-20 or so teams when choosing allies, especially if they dont have very good scouting systems in place. Also it allows defensive teams the chance to choose, rather than be chosen, thier alliance partners for the eliminations. The tie-breakers once again reward offensive teams, but they do show SoS(strength of schedule) somewhat. My team was amognst the top 2-5 teams in Annapolis(we were 3-3 in richmond and made the semi-finals) with just under 30(29.7) tie-breakers. That shows we had a tougher schedule than many teams(we faced the #3 ranked team[1083] 2 matches in a row). I needs a little bit of work, but I like much better than the old system.

I LOVE the new seeding system! When people ask what the game is about, they still regret it because of my long detailed explanation, but it’s cut down the time of explaining by 72.99887 %! Lol, besides that I just like the idea that to get ahead you have to win… Gee who ever came up with a crazy idea like that… Lol.

Agreed. The new system is much better.

Using losing team’s score as the “tiebreaker” within tiers actually rewards teams that have close, exciting matches. I don’t mind this at all, I just wish we could’ve had to face some harder bots during the prelims. Improve the “random” pairings?

Why not just have a tie breaker based on stength of schedule: total opponent record and record vs. common opponents. I think that the NFL tie breaking system is pretty consistent and fair…

I know we were invited to our alliance as a defensive robot, and filled our role well even though we seeded poorly…1002 had a very organized scouting system and found 2 solid partners @ peachtree…

I agree that strength of schedule should play into the formula somewhere. It seems like it is implying that the average opponents score is something that the winning team conscientiously somehow contributed to. I just don’t see that occuring. If the teams were truely randomly selected as opponents - the capabilities of those robots and human players have more to do with how they performed than what the opposing alliance did for them. When winning is the more important factor than score differential AND the game dynamics don’t really allow the field participants to know the score accurately (penalty points and hanging or not hanging cannot be comprehended until the match is over) very few if any teams are gonna try to barely win or allow their opponent to score very many points. This is very true when a robot hanging might just beat 2 robots that can’t double or hang. Seeding methodology MUST match the game dynamics or the tie breaker rule that attempts to dissuade blowouts is meaningless.