The Gonzalez System is a computer ranking model similar to Elo and is a rating exchange system based on research performed by World Rugby. It has been adapted by the NCDA to the demands of College Dodgeball, but can be tuned and customized endlessly to incorporate accurate data. It has been used to help determine seeds for the Nationals bracket since Nationals 2014, and was used exclusively for the Nationals 2017, 2018, and 2019 bracket.
Technical Upset Spotlight
A technical upset in the Gonzalez System is when a lower rated team defeats a higher rated team. The overall success rate of the system is currently 80.62% based on 354 technical upsets in 1847 ranked matches played since 2010-09-25.
There were 10 technical upsets since our last standings update.
Technical Upsets in September
ZM: Think that we had a lot of technical upsets this past month? You’re right. 11 technical upsets in the first 31 ranked matches is about ~14% difference than the past six seasons.
In 2011, 2012 and 2013 we had only a handful of matches in September. We’ll exclude those for the purposes of comparison of later seasons in which teams started to play earlier into their school year.
2014 – 3 technical upsets in 16 ranked matches (81.25% success rate)
2015 – 4 in 32 (87.50%)
2016 – 5 in 17 (70.59%)
2017 – 10 in 34 (70.59%)
2018 – 8 in 35 (77.14%)
2019 – 6 in 33 (81.81%)
2020 – 11 in 31 (64.52%)
While a “significant upset” can often be a floating definition for the Content Team, an important measure of the system is whether a technical upset is statistically significant. By design, the system should be able to ensure these statistically significant upsets are few.
We generally settle that upsets beyond the second deviation (95.45%) would include technical upsets with an initial rating gap greater than -6.685 (1.668 exchange), spanning all the way up to -9.848 (1.985 exchange). There are 16 statistically significant technical upsets in the second percentile, of 354 technical upsets. These are the current numbers, updated to include the weekend’s two significant upsets, presented as the first two matches below.
There are five significant technical upsets that occurred in the first matches of September:
2020: JMU def Towson 4-3 OT at this weekend’s Double Header
2016: MSU def GVSU 1-0 at Impact at IM West
2018: BGSU def MSU 4-2 at Early Access
2020: OSU def BGSU 6-2 at this weekend’s Buckeye Invitational
2017: Ohio def Pitt 3-0 at Party on Fifth Ave
-zm
JMU def Towson 4-3 (OT)
46.833 def 54.552 (+1), exchanging 0.936 points
In the first match of the JMU-Towson Double Header, the Dukes beat Towson in overtime to pull off the upset. Overall, the upset ranks 4th all time out of 64 overtime upsets in NCDA history. The match ranks 6th of all technical upsets, and classifies as a significant upset in the 98th percentile.
OSU def BGSU 6-2
37.736 (+1) def 46.151, exchanging 1.741 points
At their home event, the Buckeyes pulled off a significant upset over in-state rivals BGSU. It was one of two technical upsets for the team in the same weekend. Overall, the upset ranks 11th all time and falls in the 97th percentile of all technical upsets, and classifies as a significant upset.
UC def Kent 3-2
35.349 def 38.811, exchanging 1.346 points
Cincy matched their win total from last year during their first tournament of the year by beating Kent State. Overall, the upset ranks 111th all time and falls in the 69th percentile of all technical upsets.
CSU def Kent 3-1
37.223 def 40.529, exchanging 1.331 points
Cleveland State pulled off their first of two technical upsets at the Buckeye Invitational in their victory over Kent State. Overall, the upset ranks 120th all time and falls in the 67th percentile of all technical upsets.
CSU def Akron 4-1
38.554 def 41.537, exchanging 1.298 points
The Vikings’ second upset came against Akron. Overall, the upset ranks 134th all time and falls in the 63rd percentile of all technical upsets.
Ohio def Miami 4-2
42.752 def 45.714, exchanging 1.296 points
In a matchup of likely top two teams in Ohio, the Bobcats came out on top thanks in part to solid teamwork. Overall, the upset ranks 135th all time and falls in the 63rd percentile of all technical upsets.
Ohio def BGSU 5-0
41.456 def 45.714, exchanging 1.295 points
Ohio pulled off the largest win of the Buckeye Invitational by point spread, beating BGSU in a technical upset. Overall, the upset ranks 136th all time and falls in the 63rd percentile of all technical upsets.
WVU def UVA 5-4 (OT)
35.917 def 38.492, exchanging 0.629 points
West Virginia upset Virginia in the highest scoring and potentially best game of the weekend. Overall, the upset ranks 30th out of 64 overtime upsets all time.
OSU def Akron 2-1
39.477 (+1) def 42.766, exchanging 1.229 points
The Buckeyes also pulled off a technical upset against Akron during their home event, defeating the Zips in a tightly contested match. Overall, the upset ranks 173rd all time and falls in the 52nd percentile of all technical upsets.
VCU def UMD 4-2
44.543 def 44.662 (+1), exchanging 1.112 points
In the last game of the weekend, the Rams pulled off the technical upset over East Coast foe Maryland. Overall, the upset ranks 260th all time and falls in the 28th percentile.
Net Rating Changes
Rating Changes | Pre | Post | Change |
OSU | 37.736 | 41.205 | 3.469 |
Ohio | 41.456 | 44.628 | 3.172 |
CSU | 37.223 | 39.272 | 2.048 |
VCU | 43.812 | 45.655 | 1.843 |
JMU | 46.833 | 47.454 | 0.621 |
UC | 35.672 | 36.197 | 0.524 |
WVU | 36.186 | 36.545 | 0.360 |
Miami | 45.327 | 45.287 | -0.039 |
Towson | 54.552 | 53.931 | -0.621 |
UMD | 44.211 | 43.550 | -0.661 |
UVA | 39.405 | 37.863 | -1.542 |
Akron | 42.443 | 40.239 | -2.204 |
Kent | 40.529 | 37.465 | -3.064 |
BGSU | 46.151 | 42.244 | -3.906 |
The Buckeyes took home the most points this weekend and rightfully so. With three wins, two of which were upsets, it should come at no surprise they moved up as much as they did. Ohio also had a great weekend, finishing 3-0 and getting a leg up on Miami in their first matchup of the season. Cleveland State rounded out the biggest gainers from the Buckeye Invitational thanks to two upset wins over Akron and Kent.
Where there are big winners there are also big losers, however. BGSU, Kent, and Akron all took significant damage in their Gonzalez ratings after lackluster performances. After a pivotal win over Michigan State in their first tournament of the year, BGSU felt the effects of missing some of their top players such as Owen Sill. Kent and Akron both saw significant losses this offseason by graduating talented Seniors. This year was expected to be a bit of a down year for both programs. With more experience for their young players, expect both teams to hit their stride a bit more by November.
Ratings, sorted.
Mov. | Rank | Rating | Team |
---|---|---|---|
— | 1 | 53.931 | Towson |
— | 2 | 53.600 | GVSU |
— | 3 | 49.138 | CMU |
↑ from 5 | 4 | 47.454 | JMU |
↓ from 4 | 5 | 47.185 | MSU |
↑ from 10 | 6 | 45.655 | VCU |
— | 7 | 45.629 | SVSU |
— | 8 | 45.287 | Miami |
↑ from 14 | 9 | 44.628 | Ohio |
↓ from 9 | 10 | 43.550 | UMD |
↓ from 6 | 11 | 42.244 | BGSU |
— | 12 | 42.091 | PSU |
— | 13 | 41.502 | UWP |
↑ from 24 | 14 | 41.205 | OSU |
— | 15 | 40.916 | UK |
↑ from 17 | 16 | 40.424 | ZAG |
↑ from 18 | 17 | 40.401 | OS |
↑ from 19 | 18 | 40.346 | UNG |
↓ from 11 | 19 | 40.239 | Akron |
— | 20 | 40.007 | WMU |
↑ from 25 | 21 | 39.272 | CSU |
— | 22 | 39.111 | SIUE |
— | 23 | 38.126 | UNL |
↓ from 21 | 24 | 37.863 | UVA |
↓ from 16 | 25 | 37.465 | Kent |
↑ from 27 | 26 | 36.545 | WVU |
↓ from 26 | 27 | 36.228 | MC |
— | 28 | 36.197 | UC |
— | 29 | 35.607 | DePaul |
— | 30 | 34.795 | GSU |
— | 31 | 34.539 | BSU |
— | 32 | 33.990 | GVSU-B |
— | 33 | 33.713 | NSU |
— | 34 | 32.793 | MSU-B |
— | 35 | 32.000 | CMU-B |
See the Resource Center for more documentation.
Records, Master Spreadsheet: 2005-Present
Records, Individual Docs: 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020
Systems: Gonzalez Current
Spec Document: Gonzalez System Spec Doc
Prediction Calculation: Gonzalez Predictor
Spec Document: Gonzalez System Spec Doc
Prediction Calculation: Gonzalez Predictor