New vs old schedule advantages

[fuzzy1718]

Every year I hear some one say "this team had a bad showing at a regional because they had a really tough schedule" or "they slipped through the cracks because their schedule was easy, and landed in the top 8." This year I havn't really heard any of this talk. So does the new ranking system fix the "easy schedule" debockle? or have I just been under a rock this season?

[Chris is me]

After week 1, yes I think so, but we need to get to Championships and other deep fields to find out for sure.

I'm growing to love this schedule. Teams with bad Fridays can sometimes make big jumps into the Top 8 with a lot of effort, and you can never count yourself out of the competition.

[Alex Cormier]

No.

[BJC]

The new schedual benifits (gives a higher seeding score point value to):

-teams with good robots playing matches against good robots.

-teams with bad robots playing matches against good robots.

-teams with bad robots playing with good robots.

It does not benifit (generally gives a lower seeding score point value to):

-teams with good robots playing matches against bad robots.

-teams with bad robots playing matches against bad robots.

-teams with good robots playing with bad robot

Robots can still get lucky in their matches, but in order to be really sucessful in this seeding system you simply have to have a good robot.

In Win/Loss/Tie

Teams benifit the same amount no matter what the score is. It becomes much easier for bad robots to sneek into the top 8.

Therefore, I like this new seeding system because I think it provides for a better top 8 then the Win/Loss/Tie method did. I also like the way the standings can rapidly change which in my opinion makes qualifications much more exciting. The only thing that needs to change is that scoring on yourself can sometimes be more productive than scoring on the opposition. This needs to change next year, but hey, its the first time its ever been tried out and FIRST just needs some time to work out the bugs .

[FIRSTgirl675]

This is actually not the first year rankings have been done based off of Coopertition, it has just been a long time. I was looking up stuff on FIRST today and found out that in the 2000 game Co-Operition FIRST, the winning alliance got 3x of the losing alliances score.

[EricH]

It doesn't fix it, but it improves it to the point where it isn't an issue. Discussion on the L.A. webcast amounted to the fact that the top 6 or so teams at the event did in fact have the top 6 or so spots in the rankings. That's a good thing. One or two teams have slipped out, but that may be because they've been playing not-so-good teams.

For reference, 330 is the top seed at Los Angeles. Here is their schedule for today, along with such approximate rankings as I can remember (noted only the first time a team appears):

3027, 589, 233 (5th) vs 867, 330, 1515 (10th)
330, 696, 190 (3rd) vs 3328, 691, 2085
3271, 233, 330 vs 3309, 1438, 1644
1452, 1160 (8th), 1759 vs 3120 (6th), 330, 1702
1388 (7th), 1836, 1717 (2nd) vs 330, 2272, 1160
2404, 2659, 1692 vs 3021, 330, 2339
606, 597, 190 vs 330, 1671, 1836

Result: 7 wins, 1st seed. Note that they've been playing both with and against good teams all day, and had a pretty tough schedule. (Facing 233, 190, 1717 and 1388 in the same match, with 190. Tomorrow, they don't face anyone ranked, but pair with 294 (#4).)

At one point, 330 was #1 and 190 was #2. 190 played a match and took #1. 330 played the next match and sent 190 back to #2.

Overall, I really like the ranking system, especially after TU#16. It's fair to say that it sifts the top robots out from the lower robots in the shortest amount of time--far faster than W-L-T does.

[Lil' Lavery]

Originally Posted by fuzzy1718

Every year I hear some one say "this team had a bad showing at a regional because they had a really tough schedule" or "they slipped through the cracks because their schedule was easy, and landed in the top 8." This year I havn't really heard any of this talk. So does the new ranking system fix the "easy schedule" debockle? or have I just been under a rock this season?

Yes and no.

It does eliminate the factor of having one or two really tough matches all but eliminate your chances of being a top 3-4 seed (which are often occupied by undefeated or one-loss teams, especially at events which only run 7-10 matches). You can lose a match or two and still have a shot at the top seed.
On the other hand, it can weight a single match more heavily than it should be (a really big seeding score in a single match can often result in a massive rankings jump). A "fluke" match with good partners (and/or opponents) can result in huge rankings boosts for an average or sub-par team.
And beyond eliminating the detriment of a small set of matches being "tough," if an entire schedule is difficult, it still harms the teams' standings. The 5 bonus seeding points for winning and the coopertition bonus are large seeding factors. And especially with upper caliber robots that can actually lower their opponents' scores by "fighting" for balls, it can be difficult to seed well with several losses.

More than anything, I think the current ranking system (its' not a scheduling system) just redefines what a "bad" schedule is. It doesn't fix it.

If there is any perceived improvement to the standings (which is a debate, not a fact), it may also have to do with the larger quantity of matches being ran at many events (as mentioned with the promise to strive for 10+ matches in Bill's Blog). Most events are running at least 9 matches/team, and a very large amount are surpassing even the 10.

[Alex Cormier]

Perfect example, why the answer is no.

http://www2.usfirst.org/2010comp/eve...edulequal.html

Check out 1114's schedule. 4 times they had basically back to back matches.

Time......Match #...................Teams
10:05 AM 5 1114 3190 2505 3161 2702 1334
10:31 AM 9 1565 1334 2185 1114 781 2702
10:43 AM 11 296 1503 1334 1000 1114 772
11:25 AM 17 771 3190 1535 1114 296 854
1:34 PM 25 3396 3161 1246 1310 2185 1114
1:50 PM 27 1565 1219 1114 843 1310 2625
2:46 PM 34 781 3396 2505 1075 1114 1535
3:02 PM 36 2361 1114 1246 772 771 2505
9:54 AM 44 610 2056 854 3161 1114 1503
10:10 AM 46 1334 1114 3396 1503 2056 2361
10:50 AM 51 771 1114 2056 1075 2609 843

[Daniel_LaFleur]

Originally Posted by Alex Cormier

Perfect example, why the answer is no.

http://www2.usfirst.org/2010comp/eve...edulequal.html

Check out 1114's schedule. 4 times they had basically back to back matches.

Time......Match #...................Teams
10:05 AM 5 1114 3190 2505 3161 2702 1334
10:31 AM 9 1565 1334 2185 1114 781 2702
10:43 AM 11 296 1503 1334 1000 1114 772
11:25 AM 17 771 3190 1535 1114 296 854
1:34 PM 25 3396 3161 1246 1310 2185 1114
1:50 PM 27 1565 1219 1114 843 1310 2625
2:46 PM 34 781 3396 2505 1075 1114 1535
3:02 PM 36 2361 1114 1246 772 771 2505
9:54 AM 44 610 2056 854 3161 1114 1503
10:10 AM 46 1334 1114 3396 1503 2056 2361
10:50 AM 51 771 1114 2056 1075 2609 843

With only 30 teams at Waterloo, and 6 teams on the field, and not wanting to play with and against the same teams every match, you will get schedules like this ... especially if you want 10+ matches in qual rounds.

Also, I don't think it hurt them too much as they seeded 2nd. Teams like 1114 usually take this as 'just another challange'

[Alex Cormier]

Originally Posted by Daniel_LaFleur

Also, I don't think it hurt them too much as they seeded 2nd. Teams like 1114 usually take this as 'just another challange'

And what about the other teams, or the rookies?

[Daniel_LaFleur]

Originally Posted by Alex Cormier

And what about the other teams, or the rookies?

2 answers:
1> Veteran assistance
2> Learning to deal with adversity and pressure.

People complained last year because teams only got 6-7 matches during their regional ... now your complaining because you get 11 matches and they are closer together (and will be unless more money can be raised to allow for 'extra' days at the competition).

Do the math. 6 teams on the field and 30 teams total. Average time between matches is 5 matches ... and this would pair you with the same teams every match (Ugh!).

So let me challange you. Come up with a better algorithm using only 30 teams and 6 teams on the field at the same time, with not playing the same teams all the time.

[Alex Cormier]

Originally Posted by Daniel_LaFleur

2 answers:
1> Veteran assistance
2> Learning to deal with adversity and pressure.

People complained last year because teams only got 6-7 matches during their regional ... now your complaining because you get 11 matches and they are closer together (and will be unless more money can be raised to allow for 'extra' days at the competition).

Do the math. 6 teams on the field and 30 teams total. Average time between matches is 5 matches ... and this would pair you with the same teams every match (Ugh!).

So let me challange you. Come up with a better algorithm using only 30 teams and 6 teams on the field at the same time, with not playing the same teams all the time.

I do not have the knowledge or get paid to, so no.

[Chris is me]

Originally Posted by Alex Cormier

And what about the other teams, or the rookies?

This happens at most small events. You just deal with it and keep playing.

[EricH]

Originally Posted by Alex Cormier

I do not have the knowledge or get paid to, so no.

Then you have three options: Learn the knowledge and find a place that will pay you to do so, talk to the people that developed the algorithm, or be quiet.

It might have been better for Waterloo to go to 10 matches with more time, but I don't think they had much of a choice.

[Nawaid Ladak]

Originally Posted by Alex Cormier

9:54 AM 44 610 2056 854 3161 1114 1503
10:10 AM 46 1334 1114 3396 1503 2056 2361

2 great matches, the back to back made it so much better