College Dodgeball closed out the opening month of the 2018 Season with an action packed Saturday. Fifteen teams played across three venues spread over 1215 travel miles, where 21 ranked matches were played, and in which 118 points were scored.
Think about all those lazy Shot Clock reset throws.
Of these Saturday games, five were technical upsets according to the Gonzalez system. None were extreme upsets but interesting regardless. We have 244 technical upsets in 1285 ranked matches.
BGSU def Kent 5-4 OT
Looking at the historical chance of Overtime in the records, we would have expected about 1.7 of the matches to be decided in OT. The only Overtime match of the day claims a new record for being the highest scoring OT in history. There’s been five Overtime matches that needed 7 points to determine the winner, and this is the first to require 9 points.
With the largest initial rating gap of any of these upsets (4.783), this match wasn’t the most influential exchange of the day and falls in the middle of the road. Overtime exchanges are halved in the Gonzalez System, with this match exchanging 0.739. It is 10th largest upset decided in Overtime, out of 40 technical OT upsets, of a further 104 Overtimes on record.
UVA def UMD 4-1
In a developing rivalry utilizing Dr Pepper gifs, Virginia claims the most influential upset of the day by grabbing 1.463 off Maryland. This upset is ranked #53 of 244 technical upsets, in the 77th percentile. They last met at Nationals 2017 with UMD nabbing the win.
Akron def CSU 5-1
We might often see technical upsets when established teams play new schools, but that is not always the case. The historical data doesn’t determine any reliable connection. Cleveland State has played enough games to develop a rating, but their recent performance is causing an adjustment to their rating. The system is adaptive. This technical upset exchanged 1.290, ranking #93 of 244 in the 61st percentile. That’s within range of most upsets. This is an adjustment match concerning the ratings of both Akron and Cleveland State. It’s early in the Season, and we should see plenty of performance by both these teams in the coming months.
JMU def Towson 4-2
Apparently there has been some doubt that JMU actually has talent. I will point you at their performance versus eventual National Champions GVSU in the later rounds. And because one datum is never enough to build a case from, I’ll point you to their 2017 season history, where they have consistently ranked as a top 10 team a posting well above a 43.000 rating. Generally, if a team is 43+ that team is going to rank in the top tens and will be a caliber above teams posting in the low 40s. It’s not that JMU has dropped, it’s that Towson has grown to challenge JMU’s rating by playing and winning more games than James Madison.
This match exchanged a tiny 1.106, ranking #179 of 244 and in the 26th percentile. This was an even match between second tier competitors, and it was Towson’s match to lose. Both these teams have talent, and that second ranking tier really is quite a bit ahead of the tiers of the lower 80%.
We need to see these teams play more, play more against each other, against their neighbors, and further cross region competitors. For now, the lever has teetered in JMU’s direction.
Ohio def CSU 8-0
While there was a heavy differential in the box score, the system doesn’t notice that. This was the second tiniest technical upsets on record, with both teams going into the match with just 0.007 difference in ratings. That gives a 1.001 exchange. This is a good demonstration that our system uses a floating decimal point for arbitrary precision, while we only display ratings to three decimals. A gap of about .00665 gap will give an exchange toward 1.000665. Rounding will display this as 0.007 and 1.001. This ranks #243 of 244 technical upsets and can be considered an adjustment match in the short term. We would need to see more matches from both teams in order to attribute a retroactive evaluation of this result.
|↑ from 8||4||47.240||JMU|
|↓ from 4||6||45.774||Towson|
|↓ from 6||8||44.839||Kent|
|↑ from 10||9||43.626||UWP|
|↑ from 11||10||43.171||BGSU|
|↓ from 6||11||42.524||OSU|
|↑ from 14||12||42.386||VCU|
|↓ from 12||13||41.615||PSU|
|↑ from 15||14||40.934||DePaul|
|↑ from 16||15||40.502||UNT|
|↑ from 17||16||40.139||WKU|
|↑ from 22||17||39.642||UVA|
|↑ from 26||18||39.021||Ohio|
|↓ from 13||19||38.825||UMD|
|↓ from 18||20||38.586||UWW|
|↓ from 19||21||38.557||NIU|
|↑ from 23||22||37.885||MC|
|↑ from 24||23||37.722||Pitt|
|↑ from 25||24||37.489||GSU|
|↑ from 27||25||37.119||UNG|
|↓ from 21||26||36.702||UNL|
|↑ from 28||27||36.381||BW|
|↑ from 29||28||36.169||NSU|
|↑ from 31||29||35.966||Akron|
|↓ from 20||30||35.834||CSU|
|↓ from 30||31||34.659||SU|
This weekend touched nearly all slots in the rankings.
The First Tier: Notably, Kent’s loss to BGSU caused CMU to miss out on #1 by a tiny .114. GVSU is close to losing their top of the hill position. CMU need only defeat a competitor ranked higher than the current #9 to knock into first.
The Second Tier: The jostling continues in a competitive 3-8, though these ratings are starting to widen. SVSU had solid victories and point gains in previous weeks and remains at the top. With competitive, high ranked neighbors, they have the ability Kent has slipped a little but is still a top eight team.
The Enterprising Third: UWP is leading the pack and knocking on the second tier. Their program is doing well and their ability to play and travel is only helping their cause. Despite the tough schedule this weekend, BGSU is not far behind after posting wins. The rest of this cluster is hard to judge objectively. Should we lump #8 Kent in and halt this at #12 VCU? Those ratings would be about 2.5 from high to low. Penn State has a tournament coming up, and they could gain a few wins at home to jump into this tier. Every competitor of this section needs to be looking for a win for a number of unique reasons. Is OSU a return to form or are they slated to slide in a rebuilding year? PSU stated they wanted to play more, and you can’t win if you don’t play. Just a few thoughts… this third tier is one of the most active and entertaining groups to watch adjust over the years. Teams come and go and these represent good matches.
Remember the Fourth (and fifth): This roughly spans from #13 PSU to #21 NIU, and #22 MC to #32 Miami. In many ways, the latter half of the standings is a lot more competitive than the top is. We would be better off splitting this section into a lot more micro sections to represent better. Games between these competitors can go either way for any number of reasons, but a hallmark is that games will have a much smaller rating gap. This section is volatile but not inefficiently so; and that is healthy. It shows the system is doing the job. The goal of teams belonging in this section should be to play more dodgeball. I would look toward the capitals of this tier for inspiration. DePaul and Penn State make a fine bunch of role models.
Ohio claims the largest rating gain this weekend by winning two even strength matches against Akron and CSU, being able to jump up a competitive section of the board as a result.
Please keep in mind that there is a Exec Board vote scheduled on the NCDA Ranking Algorithm; so the Gonzalez system as it appears before you today may or may not be the single system we use for Nationals 2018 at VCU. We’ll release information on the vote as soon as it comes up.