Re: States sending two teams
Posted: March 19th, 2012, 7:49 pm
Sweet! So Indiana is sending two teams!Luo wrote:Well, this was way earlier than expected.
http://www.soinc.org/sites/default/file ... umbers.pdf
Sweet! So Indiana is sending two teams!Luo wrote:Well, this was way earlier than expected.
http://www.soinc.org/sites/default/file ... umbers.pdf
Awww.. If we had kept all teams from last year we probably would've gotten a second team!Luo wrote:They posted the complete breakdown of B/C registration numbers today at http://www.scienceolympiad2012.com/tour ... nationals/.
Yikes. They had to break a tie between Indiana and Colorado for a second team in B Division, and take a look at that pileup in C division with 6 states between 65 and 67 teams, right beneath the two-team cutoff, to boot.
They could be broken by the date the registration numbers were sent to the national office...Luo wrote:I wonder how they break ties. The two methods I think seem most fair would be to either use last year's registration numbers as a tiebreaker or else use the state's most recent placing at nationals as a tiebreaker. Either of these tiebreakers would have resulted in Indiana getting the second Division B spot over Colorado.
Everywhere else on the list, they ranked tied states in alphabetical order. However, the only place it actually mattered (the Indiana/Colorado Division B tie) was the only place that the alphabetically latter state was listed ahead of the alphabetically former state.
Although it's probably an extremely remote possibility, it makes me wonder how they would break the tie at 66 if two states dropped out in Division C. If they broke the tie based on last year's registration numbers, Minnesota would get the spot. If they broke the tie based on last year's national placings, New Jersey would get the spot. If they broke the tie based on the current ordering, which appears to be simply alphabetical, Kansas would get the spot.
Or is there another possible tiebreaker they could be using that I'm missing? It just seems weird that they ordered tied states alphabetically in all cases except in the one place it actually mattered.
Yeah, I thought of that, but it seems so grossly unfair I didn't think it was a viable possibility. On second consideration I suppose it's reasonable.rfscoach wrote:They could be broken by the date the registration numbers were sent to the national office...Luo wrote:I wonder how they break ties. The two methods I think seem most fair would be to either use last year's registration numbers as a tiebreaker or else use the state's most recent placing at nationals as a tiebreaker. Either of these tiebreakers would have resulted in Indiana getting the second Division B spot over Colorado.
Everywhere else on the list, they ranked tied states in alphabetical order. However, the only place it actually mattered (the Indiana/Colorado Division B tie) was the only place that the alphabetically latter state was listed ahead of the alphabetically former state.
Although it's probably an extremely remote possibility, it makes me wonder how they would break the tie at 66 if two states dropped out in Division C. If they broke the tie based on last year's registration numbers, Minnesota would get the spot. If they broke the tie based on last year's national placings, New Jersey would get the spot. If they broke the tie based on the current ordering, which appears to be simply alphabetical, Kansas would get the spot.
Or is there another possible tiebreaker they could be using that I'm missing? It just seems weird that they ordered tied states alphabetically in all cases except in the one place it actually mattered.
I would guess the alphabetical ordering is to avoid anyone reading too far into what state would be ahead if there was a scuffle over getting a second team - i.e. NSO only breaks ties when they need to. I don't think alphabetical order would (or should) be a tiebreaker, since that's just arbitrary. I don't think last year's national placings would be a tiebreaker either - too volatile, especially in Div B, and it seems too dependent on team ability instead of interest. Last year's registration numbers seems to be the most reasonable tiebreaker (since number of teams determines the rankings anyways), although perhaps it might be growth in teams instead of just straight numbers?Luo wrote:I wonder how they break ties. The two methods I think seem most fair would be to either use last year's registration numbers as a tiebreaker or else use the state's most recent placing at nationals as a tiebreaker. Either of these tiebreakers would have resulted in Indiana getting the second Division B spot over Colorado.
Everywhere else on the list, they ranked tied states in alphabetical order. However, the only place it actually mattered (the Indiana/Colorado Division B tie) was the only place that the alphabetically latter state was listed ahead of the alphabetically former state.
Although it's probably an extremely remote possibility, it makes me wonder how they would break the tie at 66 if two states dropped out in Division C. If they broke the tie based on last year's registration numbers, Minnesota would get the spot. If they broke the tie based on last year's national placings, New Jersey would get the spot. If they broke the tie based on the current ordering, which appears to be simply alphabetical, Kansas would get the spot.
Or is there another possible tiebreaker they could be using that I'm missing? It just seems weird that they ordered tied states alphabetically in all cases except in the one place it actually mattered.
Also, an FYI, our state grew more than Indiana, they gained 1 team, we gained around 6AlphaTauri wrote: I would guess the alphabetical ordering is to avoid anyone reading too far into what state would be ahead if there was a scuffle over getting a second team - i.e. NSO only breaks ties when they need to. I don't think alphabetical order would (or should) be a tiebreaker, since that's just arbitrary. I don't think last year's national placings would be a tiebreaker either - too volatile, especially in Div B, and it seems too dependent on team ability instead of interest. Last year's registration numbers seems to be the most reasonable tiebreaker (since number of teams determines the rankings anyways), although perhaps it might be growth in teams instead of just straight numbers?
Like, really or sarcastically? I didnt think either of them was a good poststycman wrote:thank you for good post)