Event and Team Placing Correlations

[Infinity Flat]

So, after tournaments, I like to find which events are the best indicators of team success. In order to do this, I found the average absolute difference between a team's placing in an event (e.g 27th Disease Detctives)) and their overall placing (e.g. 14th) The following work I've done so far is posted below.

EDIT: If it wasn't clear, this is meant to give some numerical context for when people complain about a test being too easy/difficult, or an event being poorly run. For example, although the disease detectives test was extremely long, it seemed to be the closest to having the perfect difficulty. That is, the best teams had the best scores, and the worst teams had the worst scores.

EDIT2: Division B added. Working on analysis for top 6 teams.

DIVISION C:

All teams (Click to Show Hidden Text)

Organized from smallest average difference to highest

Overall: 10.88

[color=#008000]Life, Personal, & Social Science – 9.11[/color]
[color=#FF0000]Earth & Space Science – 9.72
[/color][color=#0000FF]Physical Science & Chemistry – 10.89[/color]
[color=#FF8000]Technology & Engineering – 11.95[/color]
Inquiry & Nature of Science – 12.258

[color=#008000]Disease Detectives – 7.87[/color]
[color=#008000]Ecology – 8.03[/color]
[color=#FF0000]Astronomy – 8.23
[/color]
[color=#FF0000]Dynamic Planet – 8.87[/color]
[color=#0000FF]Forensics – 8.95[/color]
[color=#008000]Microbe Mission – 9.2[/color]

[color=#008000]Anatomy – 10.03
Ornithology – 10.4[/color]
[color=#FF8000]Towers – 10.6[/color]

[color=#0000FF]Optics – 10.77[/color]
[color=#FF0000]Fossils – 10.77[/color]
[color=#0000FF]Protein – 10.93[/color]

[color=#FF0000]Remote Sensing - 11[/color]
[color=#0000FF]Wind Power – 11.07[/color]
Technical Problem Solving – 11.13

[color=#0000FF]Sounds of Music – 11.4[/color]
[color=#FF8000]Helicopters – 11.73[/color]
[color=#0000FF]Chem Lab – 12.23[/color]

[color=#FF8000]Mousetrap Vehicle – 12.27[/color]
Experimental Design – 12.37
[color=#FF8000]Mission Possible – 13.2[/color]

Sumo Bots – 14.47
Write it Do it – 14.63

Top 6 Teams (Click to Show Hidden Text)

Overall - 7.83

[color=#FF8000]Technology & Engineering – 4.96[/color]
[color=#008000]Life, Personal, & Social Science – 5.77[/color]
[color=#0000FF]Physical Science & Chemistry – 7.00[/color]
[color=#FF0000]Earth & Space Science – 10.08[/color]
Inquiry & Nature of Science – 12.25

[color=#FF8000]Helicopters – 2.33[/color]
[color=#008000]Disease Detectives – 2.5[/color]
[color=#0000FF]Wind Power – 3[/color]

[color=#008000]Ecology – 3.83[/color]
[color=#FF8000]Towers – 4[/color]
[color=#0000FF]Forensics – 5[/color]

[color=#FF8000]Mission Possible – 5[/color]
[color=#0000FF]Chem Lab – 5.83[/color]
[color=#008000]Anatomy – 6.5[/color]

[color=#008000]Ornithology – 6.67[/color]
[color=#FF0000]Astronomy – 7[/color]
[color=#FF8000]Mousetrap – 8.5[/color]

[color=#0000FF]Optics – 8.67[/color]
[color=#FF0000]Remote Sensing – 8.67[/color]
Write it Do it – 9

[color=#008000]Microbe mission – 9.33[/color]
[color=#0000FF]Sounds of Music – 9.5[/color]
[color=#FF0000]Dynamic Planet – 9.5[/color]

[color=#0000FF]Protein Modeling – 10[/color]
Sumo Bots – 11.5
Experimental Design – 13.3

[color=#FF0000]Fossils – 15.17[/color]
Technical Problem Solving – 15.17

Division B:

All teams (Click to Show Hidden Text)

Overall: 11.17

[color=#008000]Life, Personal, & Social Science – 9.77[/color]
[color=#FF0000]Earth & Space Science – 10.54
[/color][color=#0000FF]Physical Science & Chemistry – 10.95[/color]
[color=#FF8000]Technology & Engineering – 12.39[/color]
Inquiry & Nature of Science – 12.78

[color=#008000]Disease Detectives - 8.10[/color]
[color=#0000FF]Optics - 8.33[/color]
[color=#008000]Ornithology -8.85[/color]

[color=#008000]Anatomy - 9.2[/color]
[color=#FF0000]Dynamic Planet - 9.57[/color]
[color=#FF0000]Fossils - 9.88[/color]

Awesome Aquifers - 9.97
[color=#0000FF]Science Crime Busters - 10.18[/color]
[color=#008000]Microbe Mission - 10.33[/color]

Compute This - 10.45
[color=#0000FF]Storm the Castle - 10.55[/color]
[color=#FF0000]Meteorology - 10.58[/color]

[color=#FF8000]Bottle Rocket - 11.17[/color]
[color=#FF0000]Road Scholar - 11.22[/color]
[color=#FF0000]Solar System - 11.45[/color]

Experimental Design - 11.70
[color=#FF8000]Towers - 11.73[/color]
[color=#FF8000]Battery Buggy - 12.00[/color]

[color=#0000FF]Shock Value - 12.15[/color]
[color=#008000]Ecology - 12.40[/color]
[color=#0000FF]Can't Judge a Powder - 13.52[/color]

[color=#FF8000]Junkyard Challenge - 14.65[/color]
Write it Do it - 18.98

[kentuckyfan1001]

So, after tournaments, I like to find which events are the best indicators of team success. In order to do this, I found the average absolute difference between a team's placing in an event (e.g 27th Disease Detctives)) and their overall placing (e.g. 14th) The following work I've done so far is posted below.

EDIT: If it wasn't clear, this is meant to give some numerical context for when people complain about a test being too easy/difficult, or an event being poorly run. For example, although the disease detectives test was extremely long, it seemed to be the closest to having the perfect difficulty. That is, the best teams had the best scores, and the worst teams had the worst scores.

Great post, infinity. What's clear is this: the events that have been continually been in Div. C are the best indicators for team success. For example, Disease, Astronomy, and (to some degree) Ecology have been an event for 5+ years in a row. The "bluebloods" of Science Olympiad, notably Solon, Centerville, Troy, Penncrest, Grand Haven, and Harriton obviously have tons of resources to keep developing these events to a point where these teams are constantly doing well. I can tell you that at least at my school, we have years of resources to go from. The reverse is true; with new events, such as Proteins, Sumo Bots, and Sounds of Music, these events tend to give more opportunity for new teams to get even or an edge over these bluebloods. Finally, events that are less "systematic," such as WIDI and XPD (which are pretty wide open events), are not as strong for the good teams. Most importantly, however, is the correlation between mechanicals and team placing. From a quick glance, one can see all of the mechanicals, such as Sumo, Mission, Mousetrap and Helicopters have lower correlations. This is likely because of the ability to "bomb" more easily. If you look at the teams, notably Solon and Penncrest are good at these; however, LASA had some trouble with mechanicals, and so did Centerville. Perhaps the bombs by good teams have skewed the results downward.

Either way, this was a great analysis of the events. Low correlations don't necessarily mean they were worse-run events than other events; it just means that they're more wide open in terms of competition.

[paleonaps]

That's really interesting.

[Infinity Flat]

So, after tournaments, I like to find which events are the best indicators of team success. In order to do this, I found the average absolute difference between a team's placing in an event (e.g 27th Disease Detctives)) and their overall placing (e.g. 14th) The following work I've done so far is posted below.

EDIT: If it wasn't clear, this is meant to give some numerical context for when people complain about a test being too easy/difficult, or an event being poorly run. For example, although the disease detectives test was extremely long, it seemed to be the closest to having the perfect difficulty. That is, the best teams had the best scores, and the worst teams had the worst scores.

Great post, infinity. What's clear is this: the events that have been continually been in Div. C are the best indicators for team success...
Low correlations don't necessarily mean they were worse-run events than other events; it just means that they're more wide open in terms of competition.

That's a very interesting interpretation, and one that I honestly hadn't really considered.

Division B information is now posted.

Working on info for Top 6 teams only.

Also, does anyone know how to make spoiler tags?

[paleonaps]

Hit the Hide button on the posting toolbar.

[Infinity Flat]

Hit the Hide button on the posting toolbar.

Ah, thank you

[quizbowl]

Here's a question - is there a correlation between number of years as an event and event placement as well?

[Infinity Flat]

Here's a question - is there a correlation between number of years as an event and event placement as well?

I'll take a look at that as soon as I'm finished with the top 6 teams for both divisions.

The "event history chart" from the soinc.org page only goes up to 2007, so I'd need some info on 08. 09, and 2010 first.

[Luo]

The "event history chart" from the soinc.org page only goes up to 2007, so I'd need some info on 08. 09, and 2010 first.

For that, you could look at the Nationals results spreadsheets from those years. Remember to count name-changed events (i.e. Reach for the Stars / Astronomy) as the same event.

[Kokonilly]

This is a really interesting set of data, Infinity. Even if you used Excel, it must have been taken a while just to put in all of the data for every event for every team.