![]() |
![]() |
![]() |
![]() |
![]() |
This is the New Zealand Paintball Team Ranking Page. The idea is to rank all New Zealand teams form all regions across the country against every other team.
The original Idea was to place teams into one of 3 grades based on the number of National level Competitions that they attend in a season and then rank them against other teams in the same grade.
This system worked in as much as it "Made Sense" when you looked at the final list, the teams you expected to be higher were higher, and the teams you expected to be lower were. However it ignored a large (and increasing every year) number of tournaments, some of which had a greater claim on being a "National event" than some counted in the formula. In other words it Sacrificed "Scope" for "Sense". So with that in mind, I have revised the method for the third and hopefully final time.
Definitions
Seasons: are simply New Zealand’s playing year i.e. February to November.
Events: The event is a Tournament "Product" that Tournament providers offer on an annual basis. The tournament on the day may change structure; format, scale or any number of other ways but the "Brand Name" remains basically the same. The list of events in the database is shown below:
Tournaments: The Tournament is an event on a specific date. A lot of things in a Tournament tend to be specific for that date of play rather than to the Event or "Brand Name" to which that tournament belongs. The list of events in the database is shown below:
Rounds:Each tournament has one or more rounds in which all or some of the attending teams participate and accumulate points. Each round has a maximum possible number of points available. A round is further classified in one of two ways:
Teams: are scored by there name and sub name, some teams are able to field more than one team at an event these teams are usually given a sub team name i.e. Saints Gold or Katipo
Each sub team is treated as a separate team for scoring purposes, so Saints Gold and Saints Black do not have their points lumped together from the Northland Open and added to the Saints points from the South Pac’s. Ah! I hear you say, what happens if:
In the 2000 season the Saints field two sub teams, Saints Gold and Saints Black at the Northland Open. Then at the South Pac’s field one team called Saints. Then at the Nationals they field two sub teams Saints Gold and Saints Green.
Method:
With such a possible diversity in the way tournaments are structured the following method has been developed to produce a Team Ranking for a given year.
Every score entered into Valhalla Database for every round within every Tournament is evaluated to see if it is the first round, then assessed to see if it is a final or not and the appropriate formula applied. All of the resulting scores for that year are then averaged for that team (i.e. a team attending one single round tournament and one elimination/elimination/final Tournament will be averaged across 4 scores if they made it to the final).
Finals are weighted to give a score above the 75th percentile because Tournament finals are often won with 50% of the Maximum Available score. If you were to use the raw score Saints would end up in the middle of the pack, simply because they participate in more finals than everyone else and that drags down the average.
Conclusion:
There are some sense issues with this method, for example 'Damage Inc' finishing ahead of 'Flashpoint' when in a toe-to-toe match up I know where my money would be. This has happened because of the widening of the scope of events measured. Flashpoint attended several events with much stronger opposition so took lower points total in those Tournaments. Where as Damage Inc attended one 'Regional' event where the only serious opposition were umpiring for the day. This will happen as soon as you introduce all of the smaller competitions. 100% in the Kapiti Coast 3 mans will not mean you are the best Team in the country and when and if Valhalla tells me that "Team Who?" is the best team around because of that I'll look at revising it again. So in a way Valhalla 2.0 has sacrificed "Sense" for "Scope".
At this point I am still missing some results from the four seasons being ranked (1997, 1998, 1999 and 2000) and so rankings will change without notification.
Please feel free to email any submissions, corrections or updates to me and I will attempt to verify and correct the information within the week.