The state existing before the war
In the same way I explored last year, I will preview the War happening this weekend at Akron. This year Akron plays host again, with some new teams brought into the mix.
|36.364 (#33)||Akron (H)||4-11|
I’ve marked this as opinion, since a predictive pre-event preview article deals with a lot of predictive analysis. It is different than the quantitative approach of the Standings. When we work with predictions, we can only be so accurate until the opinion floats in.
First, some housekeeping.
- A technical upset is when a lower rated team defeats a higher rated team.
- West Virginia (Welcome WVA!), as a new team, enters the Gonzalez system with 40.000, the same for all new teams. This functions as the rough average rating for the system, though depending on League expansion or home court advantage factoring in, the real average is slightly higher (currently 40.825).
- The predictive nature of the Gonzalez System hovers around correctly picking four victors of any five matches (currently 80.65%). The wider the rating gap between the two teams, the greater the chance that the higher rating will win.
In the early days of the League, this amount of games would constitute a whole season.
|High Seed||Low Seed||Rating||Rating||Predicted Exchange||Upset Exchange|
The following chart focuses on technical upsets. The Gonzalez system lets us predict which team might win in any particular matchup. This is incredibly important for working out a fair Nationals bracket, for example. But it is possible to see where a potential technical upset might fall among the records, and how significant it might be in terms overall.
|High Seed||Low Seed||Upset Rank / Percentile|
|JMU||Miami||#1 / 100%|
|PSU||WVU||#166 / 39%|
|VCU||Akron (H)||#35 / 87%|
|BGSU||PSU||#13 / 95%|
|WVU||Akron (H)||#95 / 65%|
|JMU||SVSU||#112 / 59%|
|Kent||Towson||#166 / 39%|
|PSU||Ohio||#207 / 23%|
|Kent||JMU||#238 / 12%|
|WVU||Miami||#163 / 40%|
|Miami||Akron (H)||#137 / 49%|
|BGSU||VCU||#16 / 94%|
|Kent||SVSU||#26 / 90%|
|VCU||Ohio||#124 / 54%|
|BGSU||Towson||#113 / 58%|
|Towson||SVSU||#200 / 26%|
|JMU||Ohio||#1 / 100%|
|SVSU||CSU||#1 / 100%|
|Towson||Ohio||#6 / 98%|
|Towson||Akron (H)||#1 / 100%|
|CSU||Akron (H)||#221 / 18%|
Of the 1395 ranked matches, 270 of those have been technical upsets. Most have been between teams of similar rating, and thus relative strength. Upset rank is thus out of 270. The percentile presented is in terms of other upsets only, where that 124th ranked upset falls within the 270 total upsets. Ohio def VCU would rank in the 54th percentile, providing no technical upsets occurred in prior games.
A key component in the predictability of the Gonzalez system is to decrease the number of massive upsets between opponents with a huge rating gap, but also allowing upsets between shallower rating gaps. Ohio (40.227) def JMU (48.483) should rightly rock the Akron Rec Center, given a hearty rating gap of some eight rating points.
But Ohio’s match versus Penn State is more of a proving ground for these two ratings. The predictability between similar ratings is less effective by design. The system will allow Ohio’s rating to adjust if they defeat PSU in a technical upset, but either outcome won’t sway the rest of the standings unnecessarily. There’s a good balance inherent via the system.
And now your game time previews
For the fans of the old AJP section breaks
9:30 – JMU v Miami (10.265 Gap)
Yes, they have been back a while now, but it really is a privilege to see Miami back in the game, and what better event to attend than a fellow Ohio school? Miami has played 10(!) games this season, posting a 6-4 record before they meet JMU’s 7-2 record. The first match of day is markedly in JMU’s favor given such a high rating gap. The predicted exchange is also the minimum value, 0.010 so won’t affect either team’s rating too much. JMU first played Miami way back at the first BEAST on 2012-02-25. Since then, the historical matchup record is 6-0 in JMU’s favor. Those games were a long time ago.
9:30 – PSU v West Virginia (1.525 Gap)
Hot dog! An induction match! West Virginia is making the trip to Akron to compete in a massive tournament, good on them. As is the case like other teams, WVU enters the Gonzalez system with a 40.000 rating. PSU is above average at 41.525, so the exchange favors Penn State slightly. The League will be watching to see how this one shakes out. Everybody loves to see new teams competing!
10:45 – VCU v Akron (H) (5.268 Gap)
The home court advantage (+1) falls to Akron as host, and their first match of the day is Nationals 2018 host VCU. Virginia Commonwealth holds the favor here, with nearly half a point at stake and the potential for an even .500 win percentage. A technical upset would fall in the 87th percentile, so isn’t out of the question for an Akron home team, but it would be significant news. They both narrowly missed playing each other at Akron’s Pink Out in the fall and at The War last season. VCU is 3-0 over Akron in their matchup history.
10:45 – BGSU v PSU (6.835 Gap)
Bowling Green is favored in their match versus Penn State by a relatively comfortable margin. A potential upset would be bordering on a statistically significant second deviation, ranking 13th in upsets. That’s not a great reason alone to sleep on PSU. The matchup history between these two competitors is BGSU 3-1 PSU. It was only last season (2017-03-26) that PSU def BGSU 2-1 in The War. It’s a rematch!
12:00 – WVU v Akron (H) (3.262 Gap)
West Virginia’s second match is against home team Akron. On the surface, WVU is slightly statistically favored here because they are a new team. Statistically new teams tend not to win games until they put a couple matches under their belt. So the new team rating is a small issue within the system. Admittedly, it’s not a huge hit to the predictabiltiy of the Gonzalez system, but an issue to be considered fully. Established teams that face a new team playing their first few matches, before the new team stabilizes their rating into a more appropriate value, will get a slightly larger boost. In Akron’s case, they stand to make a top 100 upset (out of 270 technical upsets), but still within our “allowable” technical upset range.
12:00 – JMU v SVSU (2.691 Gap)
A doozy for both teams, this match will be a good test for either traveling team. SV is looking for a reachable upset here. It’s good practice playing any Michigan team, and JMU is looking to keep their Top 4 ranking while making that opponent SV. Earlier this Season, James Madison came out with a win over Saginaw after having lost to SV at Nationals 2017. It’s a lower risk, higher reward match for SV. Should they win, they won’t jump any Standing places (Towson has 1.438 points on SV) but it will work as a sum of the whole. A solid upset win against their second toughest opponent on their schedule will favor them when the Tuesday standings release comes around. JMU is 4-2 over SVSU in their matchup history.
1:15 – Kent v Towson (1.575 Gap)
Another big test for Kent and Towson. Given their geographical distance, these two teams have played oodles over their shared historical matchups: Kent is 6-1 over Towson. The only win of this matchup for Towson was late last season on Towson’s home court in the Battle at Burdick. We’ve seen comparable wins over opponents for both Kent and Towson’s schedules, but Kent might have the edge given their matches this year. The Gonzalez system predicts Kent by a slight edge, but don’t count Towson out. Unless you like hidden home court advantage, given Kent’s potential ease to field a better roster closer to home.
1:15 – PSU v Ohio (0.982 Gap)
This is a close one. Penn State is 2-0 over Ohio historically, having only first played Ohio last season. The first match for Ohio will be the third and final match for Penn State, so on that hidden factor alone we could see a potential technical upset. PSU’s games earlier in the day could also alter the starting exchanges for this game, as the Gonzalez system is a rolling system. The exchanges are updated after each match, and a loss to WVU or win over BGSU or even an overtime contest could put either PSU or Ohio into favor for this close matchup. It’s also the second closest rating gap of the day, beaten only by our next match:
2:30 – Kent v JMU (0.433 Gap)
A late game between two opponents with such a close rating and even closer shared history. Kent was the team that welcomed JMU and the other East Coast teams into the League. Kent has a healthy 7-2 over their rivals JMU. I’d consider it a rivalry. Maybe not as bloody, but these two have met up at least once a Season except 2016. This is a real close match that could go either way and becomes especially difficult to predict should anything occur in the earlier matches. But it’s results will be a solid marker for which team deserves to stay in the top 4.
2:30 – WVU v Miami (1.618 Gap)
WVU’s third match of the day will qualify them for Nationals! But they’ll ideally need another three before we can consider their rating out of the provisional stage. Six matches is a good marker that allows most teams to to float into a solid rating. A team at the near bottom of the chart can jump to the near top with 10-12 wins, changing their stars. Miami might get a soft nod for this match, but a technical upset would rank low overall, around the 40th percentile.
3:45 – Miami v Akron (H) (2.154 Gap)
Miami may be coming off a win or a loss in the previous round, but will likely be favored or closely nodded at when the 3:45pm round comes knocking. Akron is 4-0 over Miami including a win earlier this fall. But Miami has posted a positive win percentage. They could be a different team, and with a win over the home team they are likely to get a solid valued exchange. It’s the third and final game for both competitors. This is a bit of a toss up, but should be a good indicator for the crowded lower half of the teams in this college League of ours.
3:45 – BGSU v VCU (6.572 Gap)
Bowling Green claims 3-0 in their historical matchup over VCU, and this could be one of the larger rating gaps of the day. It will be a good test for VCU’s traveling participants, getting solid experience in which VCU could use to make a splash at Nationals 2018 on their home court.
5:00 – Kent v SVSU (5.543 Gap)
In The War, Both Kent and SVSU had an impressive schedule against their similarly ranked brethren. SVSU has a narrow lead over Kent 5-3 in their overall matchup history, but this season Kent’s rating gives them a fairly strong favor in terms of the Gonzalez System. It will be interesting to see, Kent will be entering their third game, but SV only their second. SVSU would be posting a fairly significant upset if they could pull it off. Another lower risk, high reward match for Saginaw.
5:00- VCU v Ohio (2.437 Gap)
Tangentally similar to their record vs Penn State, Ohio has historically lost two matches to VCU, both last season. The end of day matches are always super interested to see. More than once I’ve considered trying to do some data crunching to see if there was a measurable effect we could apply to teams deep along their schedule, but it would be complicated. As part of the rolling schedule, VCU has the edge here but only slightly. While not as close as other matches, the potential for an technical upset isn’t out of order.
6:15 – BGSU v Towson (2.622 Gap)
The last match of the day is truly a test for Nationals and the endurance that comes with that element. I hope this match is fun for both teams, because if you fall into the feeling that this match is it I believe you are thinking in the wrong terms. That’s an old DePaulian giving unwanted advice, I’m sure. In any case, this match does have the potential to be valuable for ranking, but depends a lot on other close matches earlier in the day. I’m hesitant to give a nod to either team.
Day 2 Exchanges
Sunday’s exchanges rely heavily on Saturday’s results, so I plan to write a little article covering it (Status Quo Intra Bellum?, the state during the war)
Cleveland State returns to The War on Sunday, with a mini schedule featuring Towson, JMU, Ohio, SVSU, and Akron. Towson has a grueling three games scheduled on Sunday after playing two games on Saturday. I can only imagine Ohio and Akron taking advantage of that!