As used for the Warhammer Fantasy & 40K Rankings for 2003 – 2006
This tournament rankings system was initially put together in by Geoff Tewierik in 2003 for only Warhammer Fantasy (WFB) tournaments. The system proved so popular that the players of Warhammer 40,000 (40K) tournaments wanted one as well. So Geoff compiled a retrospective 2003 rankings for 40K. Geoff maintained the ranking for both WFB and 40K through to the end of 2006. The Warhammer Fantasy Battle and Warhammer 40,000 rankings for the 2003 to 2006 seasons can be found on Geoff's web site, http://figmentia.com.
At the end of 2006 Geoff announced that his desire to compare players across the Australasia region was no longer there. I would like to thank Geoff for the work he did over those 4 years, and his help in getting me started with the stewardship of the rankings in 2007.
The Australasian Tournament Rankings uses the term "GT Score" to refer to scores calculated using the same mechanism that has been used in previous years by Geoff Tewierik and posted on the http://figmentia.com web site. Geoff's system was originally inspired by a system put together by Andrew Galea, editor of Irresistible Force, in 2001. Geoff acknowledges input or assistance from Andrew Bishop, David Capon, Pete Dunn and Alan Borthwick.
The details of both 40K and WFB rankings for 2003 - 2006 were published on http://figmentia.com.
The following information is Geoff's description of how it all works.
Raw scores. They are a necessary evil for this system to work. The more we have the better.
Each player earns a particular score in a criteria of a tournament, i.e. 80/100 for Battle, 25/30 for Sports and so on. The players score is compared to other scores achieved by players in the same event and the whole field is spread over a range of 1 - 100, where 1 is awarded to the lowest score achieved and 100 for the highest score achieved.
Once that calculation has taken place the new scores that players get is multiplied by the average number of players at all events in the Rankings. This helps spread the field out a little and awards the players that do well at very large events. Players that do well at a big event will get a minor bonus of a 1.25 multiplier added to their score, and those that go to very small events will have their multiplier limited at 0.50.
So that determines your score for that event. We then only consider the top 3 scores, so all your results are compiled, the top 3 determined, added together to produce that final score. And for those of you that have followed along closely you will have worked out the maximum ranking score a player can get is 375. The number is pretty much meaningless except for determining ranking, and helping to show that "degree of separation" between players at a ranking level. If you've followed us by now you can probably tell it's damned near impossible to reverse engineer these scores to find out your actual scores for events that haven't published them. That's if we even got them.
All of the above is (thankfully) managed by a complex spreadsheet (12 sheets just to compile results!) so as events are added the rankings will change due to the average changing. You can change ranking position without having attended any further events just by the addition of a new event. It's possibly you could even go up in rankings as more players are added.
Well... maybe... it's one system of many that could be used, but this one revolves around averages. Tournament weighting is completely independent of what anyone "thinks" an event should be, and purely on where it sits in the statistical distribution. It may even change as new events get added in.
Well, mostly because, as it's for fun, everyone (who reads the list) is interested in where they sit. If they don't even rate a mention because they haven't attended 3 events, then it'll get brushed off and never looked at again. On the other hand, if you find you rank, to your own surprise, 1000th out of 1350 players, then you may be encouraged to attend a second event to see how that improves your ranking.
For example in 2007 the number of people who attend 3 Warhammer events stops at around 179 (of 1344), two events peters out at 390. So, if it were only to include 3 event people, we'd only be picking up 13% of the people who *might* be interested. The average number of events attended is 1.6