Incredible Rank Boost

[PowerfulKitty]

I had an amazing experience at the NorthStar regional this weekend, and I thought others might find this interesting. In an incredible match towards the end of qualifications team 2826 Wave (ranked #1) and team 3130 (ranked #2) Error were put together with the rank #55 robot. After their incredibly high scoring match of 268 points, the rank 55 team jumped up to rank #15!

[Ginger Power]

Originally Posted by PowerfulKitty

I had an amazing experience at the NorthStar regional this weekend, and I thought others might find this interesting. In an incredible match towards the end of qualifications team 2826 Wave (ranked #1) and team 3130 (ranked #2) Error were put together with the rank #55 robot. After their incredibly high scoring match of 268 points, the rank 55 team jumped up to rank #15!

Corrected Wave's number. That match was the most impressive thing I've seen live this year.

[cglrcng]

Originally Posted by PowerfulKitty

I had an amazing experience at the NorthStar regional this weekend, and I thought others might find this interesting. In an incredible match towards the end of qualifications team 2825 Wave (ranked #1) and team 3130 (ranked #2) Error were put together with the rank #55 robot. After their incredibly high scoring match of 268 points, the rank 55 team jumped up to rank #15!

__________________________

That is called the Recycle rush "Great Bot QPA field & bot INFLATION factor." **Unless you see it w/ your own eyes** Scouting this year...OPR and many of the other stats means....Ummm, not very much.

[Alex Webber]

This happened at Greater Kansas City too for 1810.

We were ranked 46th going into the regional on Saturday, came out at 16.

Thanks to 1730 and 1288 we scored 126 points
Thanks to 1737 and 1939 we scored 114 points

We finished 16th place!

http://www.thebluealliance.com/team/1810

[Isaac501]

Some teams got VERY luck this year.

There was a team that fell over or got stuck on the scoring platform 38.5% of the time at one of our district events that was ranked in the top 40%. They scored a total of two totes the entire first day of qualifications.

This team got picked.. Most likely because they were the highest ranked among the remaining unpicked teams.


The Qualification Ranking System is nearly meaningless - Scout well or be prepared to handle your poor pick.

[Nemo]

Originally Posted by Isaac501

The Qualification Ranking System is nearly meaningless - Scout well or be prepared to handle your poor pick.

I'd say the ranking system is pretty good at finding the top teams, but it doesn't do a lot to help you differentiate the middle and lower tier teams.

[shhrz]

The problem with this year's system is that it is unusally suspicable to luck in qualifications. If there are two average teams in a regional with, say, ~6 good teams, and one of these average teams is drawn to play with 4 of these - compared to the other team who is drawn to play with 1 of them, the gap in QA will be substantial.

This is where the average team with the worse QA, who might in fact be obviously better than the other team in terms of points it can score, needs to hope that scouting is done properly by teams.

Still, like Dan (Nemo) said above - the top teams will usually come out on top since they consistently score well. The middle and lower tier teams rank according to luck of the draw (not ONLY, but it is a HUGE factor).

My main issue with this game is that it is REALLY hard to recover from either a mechanical problem or just some bad luck - your robot fell, another robot knocked some stacks, etc.:

1. In one of our regionals, a really good robot fell over at the start of the match. He was ranked in the top-3 at the time, and were he to be playing with middle/lower tier teams - he would have fallen far out of the top 5, in all probability. Luckily, he was playing with another good team, so the team's average was almost unaffected. But again, luck of the draw. In other years, a loss would hurt, but not be as subsantial.
2. In the playoffs, these issues simply 'kill' your chances. Especially in the QF, where you have only 2 matches, there is no room for error. Unlike other years, where the best-of-3 system allowed you to lose a match and still have a shot.

I find this ranking system to be ... just not good enough. Less so in qualifications, but it is horrible for the playoffs.

[Alex Webber]

The current scoring system really points out the top 3 or so teams, but for everyone else: not so much.

At 1810, we created a custom scouting program this year, and tested it at the Greater Kansas City Regional. We scouted 75% of the matches, and our top 3 teams were our top 3. We scored by Average Scores by each Robot, not alliance.

The results were well, shocking.
1730 averaged 90+ points individually per match.
The last few teams averaged 5 or 10 points per match.

We think this really comes to show how much the alliance based scores impacts the alliance selection.

Cheers to good scouting

[Rachel Lim]

Last year, the top few teams were mixed together based on luck, the middle teams were mixed together based on luck, and the bottom ones were mixed together based on luck.

This year, the top teams are separated fairly well, the middle teams are mixed together based on luck, and the bottom ones are mixed together based on luck.

The alliances you play with in qualification rounds will always have an effect on your ranking. We definitely benefited from this at SVR this year. However, in general, I feel like QA has done a better job of separating out the top few teams, and the absolute top team in specific, than WLT has done last year. It's sort of like OPR: it does a good job for the top few teams, but isn't that helpful for everyone else. And everyone knows who those top few teams are, so it's not particularly useful in that aspect either.

Ranking alone is rarely an accurate measure of a team's ability. Good scouting data is always necessary.