Unofficial Rankings C

Locked
Ionizer
Member
Member
Posts: 32
Joined: March 20th, 2014, 6:03 pm
Division: Grad
State: PA
Has thanked: 0
Been thanked: 0

Re: Unofficial Rankings C

Post by Ionizer »

Ashernoel wrote:its kinda crazy that last year East's formula predicted New treir to get 11th. they got 13ths by getting DQd in wright stuff and were <40 points off of 11th :0 and had mira loma winning, Stevenson 6th, so I trust it XD
I just hope we do better than it predicts lol
Correct me if I'm wrong, but I believe New Trier got 14th last year. They were less than 40 points from 12th, but they were more than 60 points from 11th.
Class of 2016 Bayard Rustin High School Alumnus
State Event Supervisor
My Page
User avatar
Ashernoel
Member
Member
Posts: 345
Joined: January 27th, 2017, 1:31 pm
Division: Grad
State: IL
Has thanked: 0
Been thanked: 2 times

Re: Unofficial Rankings C

Post by Ashernoel »

Ionizer wrote:
Ashernoel wrote:its kinda crazy that last year East's formula predicted New treir to get 11th. they got 13ths by getting DQd in wright stuff and were <40 points off of 11th :0 and had mira loma winning, Stevenson 6th, so I trust it XD
I just hope we do better than it predicts lol
Correct me if I'm wrong, but I believe New Trier got 14th last year. They were less than 40 points from 12th, but they were more than 60 points from 11th.
I'm sorry yeah :( In my defense I wasnt on the team, but i should have been able to read the table and nt confuse 14th with 13th

but still, East's predictions were pretty great.
NT '19
Harvard '23
User avatar
EastStroudsburg13
Admin Emeritus
Admin Emeritus
Posts: 3203
Joined: January 17th, 2009, 7:32 am
Division: Grad
State: PA
Pronouns: He/Him/His
Has thanked: 48 times
Been thanked: 204 times
Contact:

Re: Unofficial Rankings C

Post by EastStroudsburg13 »

Ashernoel wrote:but still, East's predictions were pretty great.
They have... moments. Mira Loma 1st was good. Stevenson 6th and Northville 8th were good.

Seven Lakes 11th, Charter School of Wilmington 17th,and Chattahoochee 31st were less good.

It's a work in progress. :lol:
East Stroudsburg South Class of 2012, Alumnus of JT Lambert, Drexel University Class of 2017

Helpful Links
Wiki
Wiki Pages that Need Work
FAQ and SciOly FAQ Wiki
Chat (See IRC Wiki for more info)
BBCode Wiki


So long, and thanks for all the Future Dictator titles!
User avatar
Alex-RCHS
Member
Member
Posts: 539
Joined: November 4th, 2016, 3:46 pm
Division: Grad
State: NC
Has thanked: 3 times
Been thanked: 1 time

Re: Unofficial Rankings C

Post by Alex-RCHS »

These are my predictions. I've never done national predictions before so excuse any glaring misplacements.
1. Harriton
2. Troy
3. Mira Loma
4. Solon
5. LASA
Gap
6. Mentor
7.  Adlai E. Stevenson
8. Clements
9. Northville
10. Munster 
11. Fayetteville-Manlius
12. Mounds View
13. New Trier 
14. Acton-Boxborough
Gap
15. Bayard Rustin
16. Columbia
17. WWP South
18. Ladue
19. Iolani
20. Fossil Ridge
21. Boca Raton
22. Chattahoochee
23. International Academy Central
24. William G. Enloe
25. NC School of Science and Mathematics
I put "gap" in places where I think there will be a big score difference between teams.

Also, I predict that the top score will be about 200.
About me!
Raleigh Charter HS (NC) 2018
UNC-Chapel Hill 2022
Private Wang Fire
Member
Member
Posts: 100
Joined: June 1st, 2015, 3:43 pm
Division: Grad
State: OH
Has thanked: 0
Been thanked: 0

Re: Unofficial Rankings C

Post by Private Wang Fire »

I made these like 2 weeks ago, but I forgot to post them until now, literally 5 minutes before team awards are about to be announced, and these predictions high-key fell apart after seeing all these event placings, but I just wanted to post them anyways for the record. Clements is really popping off today as well! Might throw a lot of top 3 or 5 or maybe even 1 picks off (including mine).
1. Mira Loma 
2. Harriton
3. Solon
4. Troy
5. LASA
6. Northville
7. Clements
8. Stevenson
9. Mentor
10. Mounds View
11. New Trier
12. Acton-Boxborough
13. Munster
14. Fayeteville-Manlius
15. Chattahoochee
16. WWP-South
17. Columbia
18. Boca Raton
19. Ladue
20. Bayard Rustin
Edit: my top 5 is actually really close, I'm surprised... they were all safe picks tho hmm...
MASON HIGH SCHOOL '18
User avatar
windu34
Staff Emeritus
Staff Emeritus
Posts: 1383
Joined: April 19th, 2015, 6:37 pm
Division: Grad
State: FL
Has thanked: 2 times
Been thanked: 40 times

Re: Unofficial Rankings C

Post by windu34 »

Unome wrote:
kenniky wrote: 13. Boca Raton - stepping it up, builds are op, did well at MIT
-Boca Raton: doesn't seem any better than last year; same goes for Chattahoochee. I'd expect no higher than 15th for either.
Boca Raton Community High School Alumni
University of Florida Science Olympiad Co-Founder
Florida Science Olympiad Board of Directors
[email protected] || windu34's Userpage
maxxxxx
Member
Member
Posts: 284
Joined: November 30th, 2015, 8:11 am
Division: Grad
State: PA
Has thanked: 0
Been thanked: 1 time

Re: Unofficial Rankings C

Post by maxxxxx »

maxxxxx wrote:I think part of the fun is making a ranking before you even know half of the teams

1. Troy
2. Harriton
3. Seven Lakes
4. Solon
5. Northville
6. Adlai E. Stevenson
7. New Trier
8. Mason
9. Mounds View
10. Mountain View
11. F-M
12. Chattahoochee
13. WWP South
14. Acton-Boxborough
15. Carmel close enough...
16. Clements
17. Columbia
18. Boca Raton
19. 'Iolani
20. Troy(MI) close enough...
Bold are within 2 places of actual, italics are actual.

This was pretty good for March 15th.
Lower Merion Class Of 2017
User avatar
EastStroudsburg13
Admin Emeritus
Admin Emeritus
Posts: 3203
Joined: January 17th, 2009, 7:32 am
Division: Grad
State: PA
Pronouns: He/Him/His
Has thanked: 48 times
Been thanked: 204 times
Contact:

Re: Unofficial Rankings C

Post by EastStroudsburg13 »

My computer rankings did alright! Overall, the average error on any one team was 4.2, which represents an improvement over last year's 5.0. Hopefully I'll be able to get it under 4 for next year. Some notes about key teams (not all teams are listed here):

The good (Predicted ranking, actual ranking):
  • Harriton (2nd, 2nd): Getting a team exactly right is good. The team ahead of them was less well-predicted (more later).
  • Mounds View (8th, 7th): There was a prevailing opinion that MV may fall out of the top 10, but the formula persisted and placed them 8th. As such, being off by one and underestimating them slightly is pretty good.
  • DuPont Manual (31st, 31st): Again, hitting on a team exactly right is good. Hitting on a team exactly right where the team hasn't qualified for nationals in several years and places almost exactly in the middle of the pack is even better. Thanks DuPont for making me look kind of smart, you're a pal.
  • Fairfax (34th, 33rd): I was off by one, but again, when you're a new-ish team and you're in the middle of the pack, being off by one is what I consider good.
  • University (43rd, 42nd): See above notes for Fairfax.
  • Cathedral, Kelly Walsh, Huntingdon: Just like the top of the rankings, the bottom of the rankings are also important for judging a predictive measure. Getting 57th, 59th, and 60th right indicates that the formula is working pretty decently. Now, if Maine had actually released scores, then perhaps I would have ended up with Cathedral 58th, but that's a separate issue.
The alright:
  • Solon (4th, 3rd): Use of the geographic factor really helped here, bringing Solon's prediction from 7th to 4th.
  • Northville (6th, 4th): 2 places is quite a bit for this high in the rankings. However, Northville placed higher than most expected so 6th wasn't that bad of a shout.
  • IA Central (21st, 18th): Did slightly better than the formula predicted, but for a new team with no prior data, that's not too bad.
  • Mat-Su Career and Technical (45th, 48th): One weakness of the algorithm that I'm seeing is that it is not very good at dealing with big shifts from year to year. Catching Mat-Su's fall from 41st to 48th to some extent, thus,
    is relatively fine.
  • Maple Mountain (46th, 44th): A slightly different case, where the big shift happened last year, and the algorithm had to predict if the team would have a bounceback or continue falling. Luckily for me, it chose right.
  • Most of the returning teams that didn't jump or fall: Many returning teams that held steady found themselves to be pretty accurately predicted, which is expected.
The bad:
  • Mira Loma (1st, 5th): In the grand scheme of things, 4 places isn't bad, but when you're trying to predict 1st, you want to be close. Mira Loma's low state score failed to indicate their relatively low ceiling compared to last year, and the indicators from invitationals were not strong enough.
  • Mentor (9th, 14th): Just as the geographic factor helped the Solon prediction, it hurt this one. The difficulty in assigning historical values to teams that did not make nationals is also difficult in the strongest states, and that affected this one.
  • Clements (12th, 6th): This was just hard. With a team with as volatile of a nats history as Clements, it was hard to say if historical results or recent results would rule. And since the algorithm takes strong account into historical results, it really underestimated Clements. It could have been worse, but also could have been a lot better.
  • Ladue Horton Watkins (19th, 11th): Ladue, as rapid a riser as any in recent years, is a classic sort of team the algorithm has trouble with. It did predict an increase from last year, but really underestimated how much. Yet another indication that the algorithm is not sensitive enough to volatility.
  • Ed W. Clark (36th, 28th): The same sort of this as with Ladue applies here, with the added bonus that Clark also placed 2nd and 3rd at the state finals, which the algorithm couldn't sense. This may have to be applied somehow next year.
  • Auburn (37th, 27th): 10 places isn't as bad in the middle as other places, but it's still bad. Their small state competition relative to overall state size meant that the algorithm was less impressed with their state score than it should have been. Dealing with states that have a smaller state competition than other states of similar sizes is one of the big issues that the algorithm will need to fix.
  • Merrimack (44th, 53rd): If New Hampshire had actually released scores, maybe this would have been more accurate, but they didn't and so I had to estimate. If I estimated poorly, that could have resulted in how far off this prediction was. If I had given them 30 more points than I initially estimated, they would have been ranked 51st, which would have looked much nicer (and brought the overall error down to 4.0). Unfortunately, without the scores, there's not much I can do.
  • Waterville (58th, 50th): Same idea applies here, except it underestimated Waterville. Without released results, I had to assume a score, and unfortunately the assumed score pushed them really far down. A score 5 points better than what I assumed would have predicted them at 56th, and 15 points better would have put them at 53rd. In small states, score is volatile, and so missing scores can have a big effect. Not that I'm saying that scores are the only factor, but they really help for the lower parts of the ranking.
The ugly:
  • Lincoln Southwest (22nd, 35th): Perhaps it's because I didn't have their score last year, but their state score this year really didn't translate. Even with the cap I instituted for % state score (which I use to normalize score for competition size), Lincoln SW still ended up ridiculously overranked. And even if I had tripled their state score, they still would have been overranked by 5 places. This may have been a case of other factors at work that I just couldn't account for.
  • Medford (28th, 40th): Oof. Their state score was decent, but it definitely didn't hold up to past Wisconsin teams. The new-team effect may have been stronger here than in other cases.
  • Camas (38th, 25th): The algorithm definitely didn't catch Camas's bounceback. This could have been another case of big state, small state finals. Washington was underranked in both divisions this year, so there's probably some underlying factor that was missing.
  • White Station (54th, 43rd): After Farragut's jump last year and then loss to White Station, it was difficult to see where they'd finish. Unfortunately, the big state small finals effect was strong here as well, as the algorithm was thoroughly unimpressed by their score, and placed them accordingly low. Not the finest moment.
  • St. Martin's Episcopal (55th, 45th): Somewhat surprisingly, St Martin's didn't fall to the new-team effect, managing to beat Baton Rouge's 46th from last year. The state score also had a negative effect here. Thus, the surprising jump went unpredicted.
East Stroudsburg South Class of 2012, Alumnus of JT Lambert, Drexel University Class of 2017

Helpful Links
Wiki
Wiki Pages that Need Work
FAQ and SciOly FAQ Wiki
Chat (See IRC Wiki for more info)
BBCode Wiki


So long, and thanks for all the Future Dictator titles!
kenniky
Member
Member
Posts: 283
Joined: January 21st, 2016, 6:16 pm
Division: Grad
State: MA
Has thanked: 0
Been thanked: 0

Re: Unofficial Rankings C

Post by kenniky »

kenniky wrote:1. Troy - duh Got this one right
2. Mira Loma - great performance at Nats last year actually 5th, unfortunate
3. Harriton - 2nd at Nats last year, 3rd at MIT. I think they sandbagged Penn and Princeton? got 2nd by a tiebreaker, so pretty close
4. Solon - always a strong contender for a top spot this one too
5. LASA - obligatory Texas is good got 8th??? must have been a bad day for LASA
6. Northville - strong Wright State on top of being a generally decent team stronger than expected showing I think, solid work
7. Stevenson - idk much about these guys but this seems ok got 10th, must have been an off day
8. Mentor - obligatory Ohio is good. they died at Wright State though... ? got 14th... oops
9. Fayetteville-Manlius - decent past performances at Nats, last year seems like it was a fluke. They didn't do great at MIT though got 16th..... missing Nationals that one year must have killed their momentum
10. Acton-Boxborough - pls did not actually expect us to do as well as we did after recent events
11. New Trier - strong MIT + Wright State but NT doesn't seem to believe in itself lol they did even better lol
12. Mounds View - falling but probably still around this level? I could see them anywhere from 9th to 14th didn't expect them to do so well tbh
13. Munster - did bad at Wright State and also didn't go to Nats last year, but they did really well at Nats the last time they went and they did well at States. could see them here up to 9th welp
14. Clements - Clements is good but they don't do so well at Nationals usually Clements saw my predictions and were like "nah bro"
15. West Windsor-Plainsboro South - being new to Nationals since the last few years probably means they'll underperform, I could see them going up to 10th though underestimated their underperformance
16. Chattahoochee - good MIT performance, comparable to last year, seems they have momentum welp
17. Boca Raton - comparable to Chattahoochee it seems evidently not, did really well
18. the other Pennsylvania team - I literally can't tell where these guys are at but this seems like a safe bet 21st ehhhh??
19. Columbia - they usually do about this well. I'd put them higher but it seems kind of unrealistic actually 15th
20. Fossil Ridge - sleeper pick, they did decently last year got a couple of medals but 24th overall
21. Iolani - how do they get their builds here without breaking them welp 29th, at least still top 50%?
Automated Event Assigner!
UMich 2018: Chem Lab, Fermi

[url=http://tinyurl.com/kenniky-so-test]Rate my tests![/url]
[url]https://scioly.org/wiki/index.php/User:Kenniky[/url]

[url=https://scioly.org/forums/viewtopic.php?t=10008&start=34]2017 Nats = rip[/url]
[url=https://youtu.be/MCo8IAovjfw]ABRHS 2016[/url]
User avatar
Ashernoel
Member
Member
Posts: 345
Joined: January 27th, 2017, 1:31 pm
Division: Grad
State: IL
Has thanked: 0
Been thanked: 2 times

Re: Unofficial Rankings C

Post by Ashernoel »

Ashernoel wrote:1. Troy - they are op. They were op
2. Solon - they have been super strong every time i have seen them, and their history is great... nice
3. William P. Clements - Beat Seven Lakes (which was 2nd at mit) and have gotten super good super fast. Who knows where they will stop.. woot I knew they were awesome. :D 2 off 5th
4. Lasa - narrowly beat Clements at state but clements rate of improvement is greater bc of last rip off day
5. Harriton - lower than texas teams bc of MIT but still strong. rip I posted this before PA states. :{ new they were in contention for #1 after that!
6. Mira Loma - They didn't do that well at golden gate but won nats last year so... got 5th so close
7. Northville - did super well at wright state they killed it !!! Woooooot!~
8. Stevenson - Lost a lot of ppl so it will be hard but they will be strong prob. rough year but they will only get better
9. Acton Boxborough - idk seems like a good place to plop them. idk what happened to them
10. Mentor - Ohio teams are ruthless and had a good state performance motivation dies I guess
11. Chattahoochee - faith in the Unome to get his team where he wants to go. unome won spirit award so kinda counts?
12. WWP - idk much about these but they seem good. rip
13. New Trier - similar place we got last year. Only two people at practice today, so team is ded. ayyy we only bombed two events :p
14. Fayettville Manlius - They have good build events (towers?) i think which will help them get points and do well. MANLIUS!
15. Munster - poor wright state, no nats last year, but indiana W and strong. close? ikd
16. PA2 = good? meh
17. Iolani - did well at golden gate for the events they did and cute name still cute hame
.
NT '19
Harvard '23
Locked

Return to “2017 Nationals”

Who is online

Users browsing this forum: No registered users and 3 guests