I already posted about the way in which I intend to seed lower rated players in the upcoming Masters of Arena tournament:
What about the upper skill bracket though? Do what was done in pretty much all major tournaments in the past?
Guesstimate, don't use a fixed formula. Pick and chose information, try to be as intelligent as possible.
In theory this should lead to a better result than simply using hard numbers. You can take in context, most recent form, more specific information. You will however add quite a bit of subjectivity to the end result, recent discussion and uproar over seeding have shown this.
If you used a fixed formula there's none of that, the result is as objective as the information that was feed into it. Combine current and highest 1v1 rank, add extra points for non-invite Top-10 1v1 RM tournament results in the last 12 month. Everybody knows exactly what he needs to do to improve his seeding. But you lose context, information. You need to ask how objective 1v1 ratings are in the first place.
Option 1 can lead to more nuanced results, option 2 is more objective.
I lean towards the fixed formula. Everyone will be on the same page and will have the same chance of tweaking his seeding position in the next few days, weeks. What do you prefer?
edit
The end result will influence the decision heavily but won't be the ultimate decider.
If you are a lower rated player (<1900) and/or don't have any noticeable tournament experience to your name expect your seeding to be highly dependent on a mix of your current and highest RM 1v1 rating. I don't have time for extensive research these days and won't go out of my way to investigate the lower skill bracket.
If you want to increase your chances I suggest that you try and bring up both. 50-100 rating points can make a huge difference.
What about the upper skill bracket though? Do what was done in pretty much all major tournaments in the past?
Guesstimate, don't use a fixed formula. Pick and chose information, try to be as intelligent as possible.
In theory this should lead to a better result than simply using hard numbers. You can take in context, most recent form, more specific information. You will however add quite a bit of subjectivity to the end result, recent discussion and uproar over seeding have shown this.
If you used a fixed formula there's none of that, the result is as objective as the information that was feed into it. Combine current and highest 1v1 rank, add extra points for non-invite Top-10 1v1 RM tournament results in the last 12 month. Everybody knows exactly what he needs to do to improve his seeding. But you lose context, information. You need to ask how objective 1v1 ratings are in the first place.
Option 1 can lead to more nuanced results, option 2 is more objective.
I lean towards the fixed formula. Everyone will be on the same page and will have the same chance of tweaking his seeding position in the next few days, weeks. What do you prefer?
edit
The end result will influence the decision heavily but won't be the ultimate decider.