Policy Proposal by Hunter Ford
Reset all Gonzalez rankings at the beginning of the year to the baseline average of 40.000
Obviously the biggest event of the season is Nationals, however that does not take away from the importance of the regular season as a whole. One of the things I have found an issue with this year is the ranking system which can determine seeding for Nationals, and more specifically the Gonzalez ranking system. In general I do find that it is an effective power rating system except on one front, the fact that it takes historical data into account.
As an example, my team, VCU, has historically not been a very strong team. This year we are by no means a powerhouse, but we are definitely better this year than what our history would suggest. So to start the season, our Gonzalez rating was around 32 points, carried over from the previous season, and we have clawed our way up to where we are now at 38.097. Another team like Maryland started the season at 41 because they have historically been better than average. However, to end the year their Gonzalez rating is 39.006 because they did not have a strong season.
For the year, VCU’s record is 5-9, and Maryland’s record is 5-10-2, giving each team a winning percentage of 0.357 and 0.333 respectively. In that time span, VCU has won the overall series between the two schools 3-2.
So in this scenario, the team that has 1) played the better season and 2) won more games head to head is ranked lower. This should not be the case. This is the only league where a team’s record/ranking from previous seasons has an effect on the ranking of the current team, and it can adversely affect teams or make teams ranked higher than they should be.
As an alternative, each team should start the year at the same Gonzalez rating, 40.000, and then move up or down relative to how that team plays throughout the year. This gives a much better estimate for how strong a team is during the current year. Additionally, if a new team is introduced to the league, they will start at the same rating, 40.000. This will actually help newer teams for when they play against weaker teams, because those weaker teams will not have incredibly low ratings which swing more points overall.
To illustrate this point, I bring up the example from UVA, a brand new team this year whose inaugural match this year was against us, VCU. UVA comes in with a 40.000 rating because they are a new team, but our rating going into that game was about a 34. Even though UVA is the statistical favorite, the reality is that brand new teams struggle against relatively experienced teams. So by losing to us, they lost more Gonzalez points than they should have, around 1.600, and we gained more Gonzalez points than we should have. If we started the year at a 40.000 rating, and lost/gained points progressively, this huge swing would not have occurred because we would not have been at such a low rating.
This would give much more accurate rankings going into Nationals, and make setting up schedules and overall rankings much easier. If not just for the points factor, consider this. College athletics, and especially dodgeball, have very high turnover rates in terms of players who graduate or leave the team. One team that exists one year, almost certainly will not have the same team 4 years from now, so the fact that a team of current players can be affected by a team 4 years ago is a little bit absurd, and doesn’t accurately measure the quality of the current team.
This is something which I believe needs to go into consideration, and can make major impacts on our standings for years to come.