Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Chit-Chat (http://www.chiefdelphi.com/forums/forumdisplay.php?f=14)
-   -   Top 25: Week 1 (http://www.chiefdelphi.com/forums/showthread.php?t=75518)

Lowfategg 05-03-2009 17:32

Re: Top 25: Week 1
 
Hmmm, I find is laughable how 25 ended up a place ahead of us even though we beat them in NJ. :P

I like to see a ranking list like this based off the OPR data.

wilsonmw04 05-03-2009 17:42

Re: Top 25: Week 1
 
Quote:

Originally Posted by Lowfategg (Post 832251)
Hmmm, I find is laughable how 25 ended up a place ahead of us even though we beat them in NJ. :P

I like to see a ranking list like this based off the OPR data.

Numbers don't tell the full story. All you have to look as is the BCS ranking systems to know the truth in that.

Kyle Love 05-03-2009 18:02

Re: Top 25: Week 1
 
These results make no sense. I do not agree with allowing a team who hasn't played yet in the top 25. Like someone said earlier, there are plenty of teams to fill the top 25 that played in week one. Also, I don't see how two robots that got beat in the quaterfinals are still in the top 10. I guess they have the namesake of a team like Ohio State or Texas in the AP football polls :rolleyes:. Eventually, namesakes will fall, and true contenders (maybe the "namesakes" will be a contender) will rise to the top.

Jonathan Norris 05-03-2009 18:17

Re: Top 25: Week 1
 
Quote:

Originally Posted by joshsmithers (Post 832204)
Question to everyone in general: Can anyone think of better methods of doing this? I'm open to options.

Your biggest issue is your sample size, set up a website where anyone can vote on their top 25 teams every week and use that as your data. Compare it against real scouting data (i.e. average points scored per match) and you should be able to get a more accurate list. If you have 40+ people voting on their top 25 (and ignore the ones that are obviously biased) you will see better data. There are major flaws in that list that just show me that its not a good representation of the top 25 at all.

XaulZan11 05-03-2009 18:23

Re: Top 25: Week 1
 
Quote:

Originally Posted by Jonathan Norris (Post 832267)
Your biggest issue is your sample size, set up a website where anyone can vote on their top 25 teams every week and use that as your data. Compare it against real scouting data (i.e. average points scored per match) and you should be able to get a more accurate list. If you have 40+ people voting on their top 25 (and ignore the ones that are obviously biased) you will see better data. There are major flaws in that list that just show me that its not a good representation of the top 25 at all.

I don't think just sample size is the problem. You could have 1,000 people who didn't watch any regionals but read who one and you probably wouldn't get much better results.

Instead, if we really wanted to make this something worthwhile and cool, we could nominate say 25 people to be the voters. Each person could nominate 3 people and the top 25 people who recieved the most votes (and would be willing to vote each week), would be designated the voters. Everyone would know who they were and they would be credible voters.

PS: how is 1625 not in the top 3? I only saw Midwest, but they were clearly the best team there.

Jonathan Norris 05-03-2009 18:28

Re: Top 25: Week 1
 
Quote:

Originally Posted by XaulZan11 (Post 832269)
I don't think just sample size is the problem. You could have 1,000 people who didn't watch any regionals but read who one and you probably wouldn't get much better results.

Instead, if we really wanted to make this something worthwhile and cool, we could nominate say 25 people to be the voters. Each person could nominate 3 people and the top 25 people who recieved the most votes (and would be willing to vote each week), would be designated the voters. Everyone would know who they were and they would be credible voters.

PS: how is 1625 not in the top 3? I only saw Midwest, but they were clearly the best team there.

That sounds like a good idea, basically 7 'credible' people is not a large enough sample size you need 20+. The simple fact that 1625 and 121 are not in the top 3 while 111 is, means that who ever voted this week simply got it wrong. It was pretty clear in week 1 that 1625, 45, and 121 were the best robots last week.

Tottanka 05-03-2009 18:52

Re: Top 25: Week 1
 
Everyone,
This is not the "top 25 Lunacy robots" thread, this is the "top 25 in FRC" thread, pleace try rememering that, and thus said, with all due respect, it's hard for me to see why 2753 is in that list, no rookie should be there, maybe only one who wins the Chamiponship all star.

People keep thinking that this list is supposed to somehow "predict" who wins, or rank teams by how well they perform in Lunacy - well, it's not the purpose of this list, and i can say that i had "performance", as a less important consideration than others.

Ryan Dognaux 05-03-2009 19:06

Re: Top 25: Week 1
 
Quote:

Originally Posted by Tottanka (Post 832275)
Everyone,
This is not the "top 25 Lunacy robots" thread, this is the "top 25 in FRC" thread, pleace try rememering that, and thus said, with all due respect, it's hard for me to see why 2753 is in that list, no rookie should be there, maybe only one who wins the Chamiponship all star.

Pretty sure this IS the Top 25 Lunacy robots thread. At least that's what I thought when I read it. Doesn't make much sense to arbitrarily rank FRC teams without regards to this year's game.

That being said, this list is very flawed. Don't put teams on it that haven't competed yet.

smurfgirl 05-03-2009 19:17

Re: Top 25: Week 1
 
Quote:

Originally Posted by Jonathan Norris (Post 832271)
That sounds like a good idea, basically 7 'credible' people is not a large enough sample size you need 20+. The simple fact that 1625 and 121 are not in the top 3 while 111 is, means that who ever voted this week simply got it wrong. It was pretty clear in week 1 that 1625, 45, and 121 were the best robots last week.

111 did win a regional week 1, and they do have a long history behind them. I can see why people might have voted for them. I can also see why people might expect to see 1625 or 121 at the top. We can disagree about what criteria makes a team "the best". This list is by no means a scientific determination of who are "the best" 25 teams. I think we can all agree on that. It's just for fun... don't take it so seriously.

Edit: Just curious, how did you determine what "extreme bias" was?
Quote:

Originally Posted by joshsmithers (Post 832180)
Seven voters participated this time, but one list was thrown out due to extreme bias.


Alexa Stott 05-03-2009 21:02

Re: Top 25: Week 1
 
Quote:

Originally Posted by Tottanka (Post 832275)
Everyone,
This is not the "top 25 Lunacy robots" thread, this is the "top 25 in FRC" thread, pleace try rememering that, and thus said, with all due respect, it's hard for me to see why 2753 is in that list, no rookie should be there, maybe only one who wins the Chamiponship all star.

People keep thinking that this list is supposed to somehow "predict" who wins, or rank teams by how well they perform in Lunacy - well, it's not the purpose of this list, and i can say that i had "performance", as a less important consideration than others.

I, like many others, take issue with this thread.

You say that the current game has no bearing on the rankings, yet this list is apparently based off week 1, hence the title of this thread.

Also, if this list has nothing to do with performance, then what criteria is being used? The Championship Chairman's Award is given to the "best" team, but many of those teams are not frequent competitors on Einstein or in regional finals. Which definition of "best" are you using? The definition applying only to robot performance, or the one applying to the team's impact on FIRST as a whole?

Assuming that you are using the former definition, are the people submitting these lists from various regions? People tend to always think that their region is better, due to both familiarity with the teams in question, as well as general regional pride (a quick example--when watching college sports, I always root for Big East in nonconference games). Since I'm from NJ and was at that regional, I'd probably tend to rank teams like 2753 higher than other teams from other week 1 regionals because I watched them perform firsthand.

I know you said that this list was "just for fun," but you can't expect people not to question the integrity of the list. We don't expect anybody to be all-knowing, but we'd at least like to see some fact-based evidence for the placement of the teams on the list.

Also, a side note, you may want to reconsider your thoughts on 2753. They are a rookie to FRC, but not to FIRST. They won the FTC championships last year. There is nothing in your explanation of this list indicating that rankings are solely based on FRC involvement.

joshsmithers 05-03-2009 21:05

Re: Top 25: Week 1
 
Everyone,
These replies are all AMAZING! The website idea sounds like an awesome way to get a better polling base. I'm still undecided about ranking teams that haven't played yet, but it sounds like most people think it's a bad idea. I'm all for developing a way to factor OPR into these rankings. I'll sleep on thes ideas tonight. Thanks again. (keep 'em comin'!)

Quote:

Originally Posted by smurfgirl (Post 832282)
Edit: Just curious, how did you determine what "extreme bias" was?

The voter unwarrantedly put their team in the #1 spot and failed to list teams that should've been on their list, like 71, 1625, 121, 111, 234, 45, 2753, etc. I figured it was best to scrap the list completely.

DonRotolo 05-03-2009 21:13

Re: Top 25: Week 1
 
Quote:

Originally Posted by Elgin Clock (Post 832207)
45 & 2753 should be tied for 1st place after week 1.
End of story.

I have seen 2753, and they are like 1114 was last year. Just incredible to see, it seems they can do no wrong. Robot, drivers, strategy - all top notch.

Don

Aren_Hill 05-03-2009 22:34

Re: Top 25: Week 1
 
from the looks of most of these replies, almost containing votes in some way or other, your top 5 should look like this 1625(woot), 45, 121, 2753, 234, in no particular order.

thanks for all the support for 1625

EricH 05-03-2009 22:36

Re: Top 25: Week 1
 
Quote:

Originally Posted by Jonathan Norris (Post 832271)
That sounds like a good idea, basically 7 'credible' people is not a large enough sample size you need 20+. The simple fact that 1625 and 121 are not in the top 3 while 111 is, means that who ever voted this week simply got it wrong. It was pretty clear in week 1 that 1625, 45, and 121 were the best robots last week.

Jonathan, if that's the way you feel, contact Josh and submit a list!

That goes for everyone who doesn't like this ranking. SUBMIT A LIST if you don't like the ranking!

Every time a list like this comes out, you guys say the same thing. "So and so should be higher." "This list is messed up." Et cetera. Et cetera. Nobody submits a list! Nobody takes action to change the lists! All you guys do is complain, or that's what it seems like. So submit a list.

Jonathan Norris 05-03-2009 23:28

Re: Top 25: Week 1
 
Quote:

Originally Posted by EricH (Post 832365)
Jonathan, if that's the way you feel, contact Josh and submit a list!

That goes for everyone who doesn't like this ranking. SUBMIT A LIST if you don't like the ranking!

Tell me when and where to submit a list and I will. As far as I know there was no open thread asking for people to submit their lists. If someone makes an open invitation for people to submit lists more people will.


All times are GMT -5. The time now is 16:43.

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