Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Regional Competitions (http://www.chiefdelphi.com/forums/forumdisplay.php?f=10)
-   -   toughest regional in 2011? (http://www.chiefdelphi.com/forums/showthread.php?t=90565)

Racer26 02-02-2011 15:22

Re: toughest regional in 2011?
 
Quote:

Originally Posted by xSAWxBLADEx (Post 1014234)
i think the rookie factor kills everything like 3357 and the two from Canada so lets wait til we are sitting in the stands watching the compitition to see what the "toughest" regional is :) what you guys think i think its a good idea


I can only assume that one of the "two from Canada" is 2056, known around here as the "Rookie Sensation". Nowadays, they're no rookie, but are poised this year to become the sole holders of the record they currently share with the Cheesy Poofs (254). They have won their first 8 consecutive Regionals. They have never competed in a Regional event that they didn't win. 2007 was their rookie season, and they swept WAT and GTR with 1114's help. 2056 has been regional champions at Waterloo and Greater Toronto every year since 2007. 7 out of the 8 times they've won, 1114 was at the top of the podium with them (exception being 2009 WAT). The two teams have excellent programs, and are leaders in their communities.

You're right. You can't account for the rookie factor. We can only assess regional difficulty in a numerical form by assessing the past performance of the teams in attendance by one metric or another, and no matter what metric you use, you're bound to be surprised at some point.

xSAWxBLADEx 02-02-2011 15:22

Re: toughest regional in 2011?
 
Quote:

Originally Posted by Andrew Schreiber (Post 1014242)
We try to find the hardest regional purely for fun. Realistically the numerical rankings mean next to nothing to anyone.

As a side note, I'm really struggling to understand what you are saying with this post. I would strongly suggest utilizing the English language when trying to communicate on these boards. Capitalization and punctuation would also help.

Im just saying that with as much time as most people spend on trying to predit the toughest regional, I think they should be trying to make the toughest robot

pandamonium 02-02-2011 15:28

Re: toughest regional in 2011?
 
Perhaps look at highest seeded teams as I feel it is more impressive to win qualifications and lose the regional than get lucky as a 3rd pick team by the 2nd alliance. Although luck also plays a factor in this but if 7 teams with an average qualification ranking over the past 4 years of 2 are all attending the same regional you better bet money that it will be competitive!

Tetraman Totally agree with your post as well. FlR is different and the wining and losing scores would be a good indicator!

IKE 02-02-2011 15:33

Re: toughest regional in 2011?
 
Quote:

Originally Posted by Andrew Schreiber (Post 1014226)
Describe the algorithm and a good way of harvesting data for it and I'll do my best. (After my hw of course)

http://www.chiefdelphi.com/media/papers/2383

Do this for every team at an event, and then do some stats for the event. Highest top 2 deviation from number 3 should be a good indicator of how tough it would be to win the event (assuming you are not 1, 2, or lucky number 24).
Averaging the values would give a pretty good idea of how tough it would be to compete at an event.
Looking at the average of the top 24 will give you a pretty good idea of how competitive elims will be.
Yes, this will give a ranking advantage to teams that compete at more events, but they are generally more competitive teams anyway.
Rookies would be wild-cards.

After the season is done, it is really good to look at things like OPR (if OPR is a reasonable metric for that year). Going into the championship, it was pretty reasonable to predict that the highest seeding score match would come from Archimededs.

Alpha Beta 02-02-2011 15:40

Re: toughest regional in 2011?
 
If a regional doesn't have alot of teams traveling to it, or if those teams do not visit other regionals to any large extent, then success in that event does not mean much in comparison to the rest of the world. Someone has to win regardless of how great or poor the competition is. Could we instead see what percentage of the teams in each regional get picked up for elims at Champs? Maybe weight 1 point for being picked up in elims, 2 points for semis, 3 points for division finals, 4 points for making it to Einstein and 5 points for a world championship? Divide all of that by the number of years of experience so that a successful rookie last year doesn't penalize the regional for not having years of data.

indubitably 02-02-2011 15:43

Re: toughest regional in 2011?
 
I think the only way to do a decent regional ranking would be to utilize, in some way, every bit of statistical data that is widely available for every team.

A regional win could be weighted high, a regional finalist would be considered but not weited as high. Seeding from previous regionals should also be looked at.

I just feel that basing it off of just blue banners or seeding ranks is not accurate enough. In many cases, the last picked team can win a regional. Also, in 2009, we seeded like 50th in the Midwest regionl only because we were consistently paired with bots that either didn't show up, or didn't move. We ended up making it to finals that year, which just shows that seeding can be completely off; although getting that unlucky doesn't occur very often.

Chris is me 02-02-2011 15:47

Re: toughest regional in 2011?
 
Ultimately, the reason BBQ is effective is because, for discussing the hardest regional to win, it doesn't matter how well you do unless you won.

Don't construe the above statement as "winning is all that matters" - but when we're talking about difficulty of winning, all that is in the discussion is winning - catch my drift?

That's why a regional with 2 powerhouses and otherwise completely terrible robots will be just as hard to win as a very deep event. Strength of the top is what will ultimately stop you from winning overall - unless you can assemble a quality alliance of 3 second tier robots. This alliance may beat 2 top tier robots and a box on wheels.

pfreivald 02-02-2011 15:49

Re: toughest regional in 2011?
 
Quote:

Originally Posted by Chris is me (Post 1014281)
Ultimately, the reason BBQ is effective is because, for discussing the hardest regional to win, it doesn't matter how well you do unless you won.

Is it effective? That is to say, how well does BBQ correlate with future wins?

How well does 2007 BBQ rankings relate with 2008 wins? 2008 with 2009? 2009 with 2010?

Does anyone know the answer to that question? Anyone care to do the work to find out?

klmx30302 02-02-2011 16:22

Re: toughest regional in 2011?
 
I'm not sure why anyone hasn't mentioned granite state regional yet. It is tough for 2 reasons: 1. It is a week 1 competition so unless a team has attended a week 0 competition no one has any had any practice with other bots on a field (unless they have 2 practice bots). 2. GSR has some of the oldest teams in FIRST, this year there are about 10 teams attending with a rookie season of 1996 or earlier, just to name a few: 20, 126, 151, 166, 131, 61. I know that just because they are some of the original teams does not make them the best but the competition up here can be really intense (and exciting to watch).

Joe Ross 02-02-2011 16:33

Re: toughest regional in 2011?
 
If 2056 never existed, how much would Waterloo or GTR's BBQ or SAUCE change? Probably not at all.

Every regional gives out the same number of blue banners. If 2056 didn't exist, other Canadian teams would have won 8 of those 9 blue banners (excluding the championship division winner from 2010), and the Canadian regionals BBQ and SAUCE would hardly change. However, the Candian regionals would be easier regionals, since 2056 wasn't there.

The same does not hold true for 1114. 1114 has blue banners from many events (Waterloo, GTR, Midwest, Pittsburgh, Long Island, and the Championship). If they never existing, American teams would have won Midwest, Pitsburgh, and Long Island, and there would be a net loss in BBQ and SAUCE for the Canadian regionals, and both regionals would be easier.

Thus BBQ and SAUCE reward regionals that can attract successful teams from outside their area. That's why Las Vegas has done well in this category in the past, because they get teams from all over the country.

Last year, 33 did not win any regionals, while 330 won 2. However, had we swapped positions, 33 would have easily won the Arizona and Los Angeles regionals, while 330 would not have won the Kettering and Troy Districts, because 33 was better then us.

While BBQ and SAUCE are convenient, I wouldn't use it for anything other then a talking point. Any method that only uses regional results will fall victim to this.

To evaluate how strong regionals are, I would probably compare teams performance at a regional compared to that same teams performance at the championship, against teams from all regionals. However, this is much harder to do, both for data collection, and computation.

Nawaid Ladak 02-02-2011 16:45

Re: toughest regional in 2011?
 
Quote:

Originally Posted by dodar (Post 1013928)
Florida BBQ/Sauce
BBQ: 89/61=1.460
Sauce:22/61=.361
...
Rookies:18

If my memory serves me correct. You receive a Blue Banner for winning a division. 179 won Archimedes with 233 and 71 at the 2007 Championship Event.

Quote:

Originally Posted by Andrew Schreiber (Post 1014096)
Got a better way of numerically ranking events? I'm listening and will do my best to implement it into my scripts.

A regressive formula that calculates winning percentages from prior years? ie: a win in a match from 2008 counts less towards your total score than a win last year.

Also, the toughest regional is usually the one your playing in.
P.S. Michigan State Championships will be the toughest "regional". if you don't count that. then i'd prbobaly say Midwest or Finger Lakes.

Joe Ross 02-02-2011 16:49

Re: toughest regional in 2011?
 
Quote:

Originally Posted by pfreivald (Post 1014030)
2009 saw the worst-performing robot that 1551 has ever put together. Neat design, just didn't play the game for squat.

2010 saw us picking 217 and 174 to join us in the top-seeded alliance.

So now we have a blue banner, but I don't think we're a fundamentally different team than we were in the past. We nailed the strategy for last year's game (though not as well as 469), and built a robot that excelled at enacting that strategy (at least on the first-week regional level... it kinda took a beating and decided to fall apart repeatedly at Championship). Yet there were other teams there that consistently do very well that last year were not all that impressive on the field.

My point being that the BBQ might not be all it's SAUCEd up to be -- sometimes teams are very surprising in both good and bad ways. Some teams are consistently awesome every year, teams to look up to and to aspire to be. Some teams are consistently middle-of-the-road in terms of robot performance, but can break out and do great things or break down and do poorly (on the field) in one particular game.

A further reason that BBQ might be skewed is that a single robot winning multiple regionals in one year can net a team multiple Blue Banners, while a team that has a robot as good or better that only wins one regional gets only one Blue Banner. Being tournament champions four times in one year, methinks, means a lot less than being tournament champions once for four separate years.

It's an interesting metric, but I'm not sure it's a good one for determining regional difficulty... And that's ignoring the fact that the number of data points are so small that if you tried to do any meaningful statistics on them, you'd get GIGO.


A previous look at year to year seeding showed there was very little correlation between year to year performance. This is obviously not the case for some small percentage of teams (1114, 67, etc) but your experience is the norm.http://www.chiefdelphi.com/forums/sh...ad.php?t=56631

pfreivald 02-02-2011 17:02

Re: toughest regional in 2011?
 
Quote:

Originally Posted by Joe Ross (Post 1014323)
A previous look at year to year seeding showed there was very little correlation between year to year performance.

I thought so.

XaulZan11 02-02-2011 17:22

Re: toughest regional in 2011?
 
Quote:

Originally Posted by Joe Ross (Post 1014323)
A previous look at year to year seeding showed there was very little correlation between year to year performance.

Thats cool, but I have yet to see a regional where the top 8 seeds were actually the top 8 teams. I think the lack of a correlation is more of an indication on how poorly teams are ranked with only a couple of qualification matches. I'd be interested in seeing correlations between selection order, but then you get into teams being biased and picking the famous teams.

XaulZan11 12-02-2011 21:33

Re: toughest regional in 2011?
 
I'm not saying this is the hardest regional, but Wisconsin is probably one of the most underrated regionals. Overal, I think Wisconsin was comparable to Midwest last year, even though Midwest gets the attention and no one really talks about Wisconsin. Midwest had the edge at the top (16, 71, 111, 1625, 1732) but it wasn't that far off of Wisconsin's (111, 706, 1714, 1732, 2481). But Wisconsin had the huge advantage after the top tier as 537, 1018, 1306, 2062, 2194, 2202, 2338, 2826, 3362 were all really solid scorers and would have been 1st round picks at Midwest.

Wisconsin has all those teams returning (besides 2481 and 1018), but gets 81, 135, 234, 279, 868, 1625, and 2041 to attend this year. I always expect comeback seasons from 93, 1652, 1736, 2039 and 2506, after a disapointing last year. There is no doubt in my mind that Wisconsin will be very deep, leading to some potentially wild elimination rounds (in my opinion, the best alliance as only won 1 time in the last 4 years, while the #1 seed as won only once in the history of the Wisconsin regional).


All times are GMT -5. The time now is 17:18.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi