I think you mean an alliance composed of a Michigan, a Florida, and a Canadian team 
Yay!
Here are my predictions for two members of each winning alliance, and one of each finalist:
Archimedes: 1114/2056 over 624
Curie: 217/254 over 79
Galileo: 67/71 over 45
Newton: 234/1625 over 148
Woo, time to make a very distant and probably completely inaccurate prediction! Einstein alliances!
Archimedes: 1114 & 2056
Curie: 217 & 68
Galileo: 71 & 67
Newton: 1625 & 121
Galileo vs. Newton determine the champ (leaning towards Galileo)
I would say 67/71 and BACOONNN!! 1902
I counted 40 wins (!) for Curie.
Would that be called āHOT Hammond Baconā??
Sign me up for that!!!
or Beastly Hot Baconā¦lol
Whoever gets out of Galileo isnāt going to have an easy time with it. I have a feeling there will be a lot of ābreaking upā of good alliances to level the playing field in the division, but it will ultimately hurt the alliance when they get to Einstein.
Curie 2007 was a prime example of this. 330 and 1114 could have gotten together and left Einstein in ruins. However, a team ranked above 330 and kept that alliance from forming. 330 ultimately won the division, but were defeated on Einstein in two matches.
In any case, there are enough very strong teams where they can still defeat the Newton alliance champions on Einstein. Newton isnāt nearly as deep as Galileo, but they still carry a lot of weight with 234, 1625, and 148, to name a few. If the right teams rank in the right spots, Galileo may have their second championship in two years.
Losing only one match during the entire Michigan State Championship is incredibly impressive. 217 will probably take the Curie title, and go on to play Galileo in the finals. Who wins from there is all up to whether the right teams rank in the right spots in their respective divisions. Iām going to say Galileo has the favor.
Archimedes has a chance in the semi-finals if things donāt go as planned in Curie. Yeah, they have 1114 and 2056. When together, these teams have won every regional.
However, this game is prone to upsets, so who knows exactly what weāll see?
Anyone else think that ALL of the divisions are stacked this year? Atlanta is going to be pretty crazy, I think. Weāll be watching the webcasts intensely this year.
how were the divisions chosen this year ?
it seemed so random
I completely agree! Iām sad I canāt go, but you bet Iāll spend three entire days glued to the webcasts with my friends, while other people wonder what is wrong with us. 
I think so, whatever method they used for sorting teams looks pretty random and pretty level.
You know your division is stacked when every time you look at it, you recognize another powerhouse team you didnāt notice before. Yeesh, Galileo here we come :]
http://spreadsheets.google.com/pub?key=pThjnO_ReGYCtCaOgz4x5Yw&gid=1 is a graph of Teams v Division (Divisions correspond to Archimedes = 1 and proceed alphabetically) I can give anyone access to the google doc used to create this graph or you can copy it and create your own in about the same amount of time anyway.
I cannot see any pattern that jumps out. The possibility of doing a completely random distribution would seem foolish because it could mean that all the low teams end up in one division and the higher teams end up in another. Does this mean FIRST wouldnt do it? Heavens no.
I will continue to look at some numbers, I have a few hunches but I am betting they are fruitless.
Kudos to anyone who can break the formula before Championship
The order doesnāt seem to have anything obvious to do with team number, location, team name, or order of registration (yes, I checked by looking at the divisions of the last few teams that were added to the list this morning and yesterday 
).
So itās either random, or generated by some suitably complex but deterministic function from whose obfuscated nature Dave is undoubtedly gleaning pleasure. :]
Tough divisions. A lot of good teams, tons of strategiesā¦
Should be an amazing competition. Donāt count any team out thoughā¦we all know how easy it is to get into trouble with this game.
It must be a game hint!
It almost seems as if they just randomized the list and went ran the Serpentine algorithm on the random list, or something of that nature. I donāt see any readily noticable pattern emerging from this.