Due to demand from some other TO's I've put together some guidelines to tournament sanctioning for our ELO tracking and list database. I thought I'd outline that here.
The below criteria are required to be sanctioned by our gaming group. What does sanctioning do for you? Well for starters I'll promote your event right here. After the event I'll put up all the results complete with any data you provide for me. I'll publish all army lists and add the winners to our growing database. I'll also input all participants into our ELO database and track them in our rankings list. High ELO scores can translate into a slot at our invitational. Any sanctioned event will be honored.
1) Swiss Pairings
Swiss pairings must be used in all sanctioned tournaments. You can use rankings to generate opposite seeds pairings if you wish, but winners must always play winners in subsequent rounds where possible. I strongly recommend using a pairing software like Swiss Perfect to achieve this.
2) Number of Participants
All sanctioned tournaments require a minimum of 8 players per event. You must hold an appropriate number of rounds based on the number of players. Tournaments with an uneven number of participants must issue byes. "Ringers" are not allowed.
8 Players = 3 Rounds
9-16 Players = 4 Rounds
17-32 Players = 5 Rounds
32-64 Players = 6 Rounds
64-128 Players = 7 Rounds
3) Round length
In addition to having enough rounds to accommodate your players, you must provide a minimum amount of time per round based on the point total. I'll honor point totals between 1500 and 2000 so long as they have the minimum amount of time per round outlined here:
1500 points - 90 Minutes
1750 points - 105 Minutes
1850 points - 120 Minutes
2000 points - 135 Minutes
4) W/L/D or W/L - No Battle Points
I recommend w/l/d but if you wish you use game tiebreakers like secondary objectives or victory points we'll honor that. No battle points that carry over from round to round. I've explained why I don't like them here.
5) No soft scores applied to generalship scores or pairing
You're welcome to use painting, sportsmanship or comp but it can't have any impact on generalship scores or player pairings. We are only interested in generalship results for the purposes of the data collection.
6) Approved Missions
The easiest way to get approval is to simply use book missions and deployments. That's what we use and what I'll strongly recommend for other TO's. I'll consider custom missions if we deem them balanced and in the spirit of 5th edition deployments and missions so long as you can vouch for their fairness and so long as they are play-tested. This means that even in a custom mission environment I expect to see dawn of war deployments, kill point missions, capture and control and the like. Tournaments featuring "wacky*" missions won't be sanctioned.
* I reserve the complete right to define any mission not in the 5th edition rulebook as "wacky." But this does not mean that all custom missions will fit that description.
7) Approved TO's.
Someone in our group, or friends of our group needs to vouch for you and your event. No fake event reporting with fake results. Also, pictures or it probably didn't happen!
If you are interested in sanctioning please let me know about your event prior to running it and then send me the results following that. Full results will allow me to plug your players into our ELO database. Full army lists will get posted on this blog and top entries will be added to our list "hall of fame" which serves as a great data point for our readers.
Note that just because we don't sanction an event doesn't mean I don't support it or think it's unworthy of people's time. There are plenty of fantastic competitive events out there that use different formats aside from our own. There are also plenty of great events out there that use soft scores. Events that don't qualify for sanctioning do not imply a lack of endorsement or approval, sanctioning is for the sole purpose of ELO tracking and growing our list database. Any endorsement beyond that varies by each individual event.
I applaud you for trying to get a larger ELO database going. However, all you really need is a tournament with balanced missions. ELO scores are calculated around 2 players and how they did compared to how they were expected to do. Even if I have 30 players and only 3 rounds, the ELO scores are still valid based on those rounds.
ReplyDeleteWe've spoken before about what K value your using. I'm using 20 to get my players' scores spread apart. I did do a little digging and found out that FIDE uses a K factor that depends on how many completed evens and the rating. For new players (under 30 games) the K factor is 25. I believe this is there to speed up their rating from the base to where they might belong. From there it drops to 15 until they reach a high enough score. After that it is permanently 10.
All the other requirements you list above are only required to meet your sanctioning standards.
Agreed with your assessment Ryan. ELO is basically localized. However the other purpose of sanctioning is to determine valid list design. That's why I am looking for a standard for sanctioning. I want the list database to mean something across the formats.
ReplyDeleteFor the 2010 season I use a k value of 10. In hindsight this was a mistake for reasons you cited: people's scores didn't move enough. For the 2011 season RTT's we sanction will have a k value of 20. You can see the results already from that with the SLO RTT I just posted, scores are much more spread out. For our Grand Prix this weekend, being an even larger event the k value will be 40. For hypothetical mega-GT events it will be even higher.
Magic: The Gathering and the DCI apply K value based on the size and scope of their events. Small store tourneys are weighted much lighter than qualifiers or the Pro Tour. I plan on following that model.
Valid list design? What's that? Making sure they're using the FOC?
ReplyDeleteI mean "valid" for tournaments using our format. If I publish lists from tournaments using different formats (tournaments lacking missions that use KP's for example) then it will skew the results of the lists in favor of armies that excel in formats without KP's. For better or worse, that dataset will be very different than the data-sets I've already created using only book missions.
ReplyDelete