Seeding in tournaments

Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you:

No adverts like this in the forums anymore.
Times and dates in your local timezone.
Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance.
Email notifications for threads you want to watch closely.
Being a part of the oldest wargaming community on the net.

If you are already a member then feel free to login now.

**olympia**

I'm going to a tournament (King Kon in Naas, Ireland http://www.naaswargamesfraternity.com/kingkon ) next month that is going to seed players in the first round based on RHQ rankings. Let's talk about the upside of this system first. The primary advantage, which far outweighs any downside, is that it reduces the so-called bunny run. For example, WAAC player 1 plays babseal in round 1 and wins. Meanwhile babyseal 2 plays babyseal 3 in round 1 and wins. So it can be the case that WAAC player 1 plays babyseal 2 in round 2 and wins. So WAAC player 1 does not get a serious test until game 3 of a 5 game tournament. Meanwhile WAAC players 2 and 3 meet in game 1 which makes things even easier for WAAC player 1. Seeding means that potentially all the WAAC players get a babyseal round 1 (the poor things).

**Mannahnin**

Sounds cool to me.

**mikhaila**

Would work better if a good chunk of tournaments turned in results. I've tossed a few results to them, but it's not something I really think about doing much. I've heard that he has some good software for running tournaments, but the fees he wants to charge are way too much for a termporary rental of software.

**Timmah**

Basically, and I think its already been kinda stated. This plan is predicated on RHQ correctly ranking all players.

Someone who is not ranked is not necessarily a baby seal. And likewise, someone who is high ranked could potentially have just got lucky at a few events. (based on the way RHQ ranks players)

It mainly means higher ranked players are just going to have a tougher road to the top than lower ranked or unranked players because they will be playing more mentally exhausting games each and every round.

Not saying that this is a bad system, just something to consider.

**nosferatu1001**

SOlent used the same system, and it works very well

Yes RHQ dont rank everyone, however it reduces the bunny run chances and means you are more likely to get a series of challenging games from the get go.

**Horst**

I always thought the best system for seeding of players would be to have a panel of judges rank the lists on nastyness according to some rubric, and assign each list a score.

That score would be added to their total battle points, and players with the same score + battlepoints would play each other.

At the end of the tournament, subtract that judge score from everyone's total score.

So, all the hard lists who win round 1 fight against other hard lists who won, while the fluffy bunny lists who win fight other fluffy bunnies who win, OR hard lists who have lost.

In any case, I think its a better way to ensure matching than hard comp scores, because in the end it only effects who you play against, not what you can bring.... and a guy with a fluffy bunny list, in the end, will match against hard lists. Best of both worlds.

**olympia**

@mikhaila: there is some good tournament software that is free. German company makes it. If I can dig up the link I'll PM it to you.
Ireland is lucky because it is a small enough country so RHQ can encompass most of the player base. For the last few years tournament organizers have done a great job working together to implement basic standards for what counts as a ranked event. There is some talk of switching to a Glicko ranking system but for now RHQ works well enough.

**Green is Best!**

Actually, a properly seeded system is designed to keep your two best players not meeting until the finals. While this sounds great in theory, there are problems with it:

1. A seeded system requires knowledgable ranking system. (Think March Madness. You have an entire season of college basketball designed purely to generate a ranking going into the college tournament).

2. Most tournaments do not have enough matches to adequately field a tournament of this style. If you are going off a bracketing system, you need 4 games to crown a winner if you have 16 teams.

While I love the concept, I don't think it really works. Well, it won't work until we establish the WH40K league with feeder conferences submitting their numbers.

**olympia**

Green is best,
I would never contend that using RHQ to seed is the best system. It is, however, better than simply throwing names into a hat for round 1 pairings.

**biccat**

Green is Best! wrote:Actually, a properly seeded system is designed to keep your two best players not meeting until the finals. While this sounds great in theory, there are problems with it:

You missed one of the most significant issues: it isolates top players from serious competition until later in the tournament.

Under a random draw scenario everyone has an equal chance of going up against either a WAAC or a less experienced player.

Hypothetically, if the seedings are accurate (#1 seeds are actually better than #2 seeds, etc.), then a random draw system for the NCAA tournament should deliver the same results for each group of 16 teams (#1 wins). The idea of the NCAA tournament, however, isn't necessarily to sort out the best teams, but to provide more exciting games (closer matchups) as the tournament progresses.

**nosferatu1001**

From what I've seen this is still swiss pairing - just the initial pairings arent random but based on rank. So #1 plays #2, etc.

Tennis seeding isnt widely used in 40k tournaments, certainly not in europe.

**Green is Best!**

biccat wrote:
Green is Best! wrote:Actually, a properly seeded system is designed to keep your two best players not meeting until the finals. While this sounds great in theory, there are problems with it:

You missed one of the most significant issues: it isolates top players from serious competition until later in the tournament.

How did I miss it? My whole point was that seeding systems were designed to keep your top 2 teams / players from meeting until the finals? In the semis, 1 plays 4, 3 plays 2. This keeps your "exciting" matches late in the tournament, just as you pointed out.

In a seeded system, 1 is getting the weakest opponent in the first round. How is this any different than what the OP spoke about originally?

The bottom line is that your typical tournament only has 3 games. This is simply not enough rounds to eliminate everyone completely to properly crown a champion. Personally, I like the system (swiss pairing?) where they assign points based on outcomes of the games. At the end of the round, 1 plays 2, 3 plays 4, etc. Rinse. Repeat after 2nd round. But that is juts me. My friend HATES that system and thinks it should just be random the entire time.

**ArtfcllyFlvrd**

It’s a cool idea but I just don’t think it matters that much. Seeding only really matters in elimination tournaments where places other than first mean something. Very few tournaments fit that bill. For most RTTs there in only 1 undefeated player, and they are small enough that the best two players will meet eventually. It just doesn’t matter if it’s in the first or last round (although the last round is more dramatic).

I large tournaments where there is a second day winners bracket it would be nice, really nice. But the competitive community is so spread out and uncoordinated I just don’t see how it could be done right.

**Jimsolo**

Olympia, you say you aren't trying to claim RHQ seeding is the best system. I believe you, and not to be contentious, but what in your opinion IS the best system?

**biccat**

Green is Best! wrote:
biccat wrote:
Green is Best! wrote:Actually, a properly seeded system is designed to keep your two best players not meeting until the finals. While this sounds great in theory, there are problems with it:

You missed one of the most significant issues: it isolates top players from serious competition until later in the tournament.

How did I miss it? My whole point was that seeding systems were designed to keep your top 2 teams / players from meeting until the finals? In the semis, 1 plays 4, 3 plays 2. This keeps your "exciting" matches late in the tournament, just as you pointed out.

Because unlike the NCAA, wargaming tournaments aren't spectator sports and defeated teams are not eliminated from play. The intent of the tournament is for everyone to have a good time, declaring a winner is a secondary consideration.

Note that in "high-stakes 40k" where declaring a winner matters, there will be more games, so seeding would be unnecessary.

Green is Best! wrote:The bottom line is that your typical tournament only has 3 games. This is simply not enough rounds to eliminate everyone completely to properly crown a champion. Personally, I like the system (swiss pairing?) where they assign points based on outcomes of the games. At the end of the round, 1 plays 2, 3 plays 4, etc. Rinse. Repeat after 2nd round. But that is juts me. My friend HATES that system and thinks it should just be random the entire time.

Personally, I prefer the idea of random games as well because it gives everyone an equal opportunity to win at the start of the tournament. I've been in environments where there were 10-12 players (3 rounds) and someone feels cheated because they went undefeated and still didn't win a prize.

Ideally, the first round should be random pairings. The second round should be randomized by groups - all 1-0 players play one another and all 0-1 players play one another. The third round again randomizes by groups - 2-0, 1-1, 0-2.

If you group people based on points earned (3 for a major victory, 2 for a minor victory, 1 for a tie, 0 for a loss) you end up in a situation where a player benefits from playing "worse" players. Which is the problem that the OP is trying to solve.

**Phazael**

Rank the lists, but not the players. Too many potential arguments then. Plus, lets be honest here. Seeding players by ability means either the seals advance farther and some stronger players get rubbed out early (if you seed strong with strong) or it gives every shark in the room an easy first gamer or two while all the seals get clubbed simultaneously.

If someone brings a dickpunch list, they should play others of their kind in the first two rounds. After that, they face what their record dictates. This way, the fluff bunnies get a couple of games without worrying if they randomly draw tripple manticore guard guy/18 ML Razorwolves/30 Purifier 6 Psyfleman GK in the first round or so. Our club has done things this way for a while and it keeps things more competitive and seperates the truely good from the flowchart netlister players.

**olympia**

The tournament in question is no-comp, standard missions, no fething soft scores. Thus, ranking lists would not be a sound way to proceed. There is no best way in a world made of gak, only better ways. Seeding the players is a better alternative than not seeding. As I said in the OP, which evidently some of you glossed over, the key goal is too avoid a WAAC player getting a bunny run to game three.

**skyth**

Am I the only one disturbed by people playing powerful lists or being good at the game being refered to as WAAC players?

**olympia**

Winners And Accommodating Citizens...what's offensive?

**Phazael**

Seed the armies and no one gets a bunny run, like I said. As long as the judged comp assesments only affect pairings and in no way directly alter scoring, people should be amiable to that.

**olympia**

Juding army comp is problematic for obvious reasons. In addition, as was noted, plenty of gak players use copy-and-paste lists that are hard as nails. A much better system is to seed players. The King Kon tournament (link to in OP) uses the RHQ ranking to seed playesr. As I previously mentioned, Ireland is an interesting country in that there exists an agreed upon set of standards for what counts as a ranked tournament. It's not a perfect system but a significant improvement.

**blood angel**

I have spent more months at the top of that list in the US than anyone else and I would be fine going into a tournament with this idea with a couple of suggestions...

Soft scores would have to be used in a way that don't overly skew the results at the end.

Good players don't show up to tournaments to be penalized before they even get their foot in the door. Some will consider this sort of thing a penalty. I would suggest offering some sort of benefit.

Only a small number of players in the top 20 are WAAC players. I can count 3, maybe 4.. please stop generalizing

Even in a 3 round tournament I rarely end up facing a 'baby seal' in round 2. Let's not even talk about that the game is rapidly descending into a coin flip where even the freshest face newbie can win a game

**vhwolf**

olympia wrote:@mikhaila: there is some good tournament software that is free. German company makes it. If I can dig up the link I'll PM it to you.
Ireland is lucky because it is a small enough country so RHQ can encompass most of the player base. For the last few years tournament organizers have done a great job working together to implement basic standards for what counts as a ranked event. There is some talk of switching to a Glicko ranking system but for now RHQ works well enough.

If you find that link I would love to see it too. RHQ's tournament software is just too expensive to use.

**olympia**

blood angel wrote:

Good players don't show up to tournaments to be penalized before they even get their foot in the door. Some will consider this sort of thing a penalty. I would suggest offering some sort of benefit.

Huh

How is playing a low-ranked player in round #1 a penalty?

**redkeyboard**

If it is an elimination tournemant don't bother as it would mean loads of players are going home straight away and won't have much fun against WAAC players. WAAC vs WAAC and seal vs seal. But I don't think it really matters if it is a regular tournemant.

**nosferatu1001**

Use swiss pairings, but instead of random for round 1, use Ranking to determine pairing

So 2 highest ranked players in attendance play each other, then next two, then next two. NOT seeding.

Every player i've met thinks this is a good idea. Yes, you may end up with great players who arent ranked - but thats only for the first tournament they attend that uses this system.

**Phazael**

Ranking players is a dubious endeavor, at best. You can always look at army lists and rate those, however. A lot of good players play in my area and I get a lot of experience against them, but I hardly ever play in GTs because fantasy if my game of choice. In your system, I could show up with the ugliest list imaginable and face Necrons Like a Boss! in the first round or two because I would have no record as a player. Meanwhile Reece shows up with his fun Nid list and gets paired off against the hardest guys in the room, because he has more activity on RHQ. Reece is certainly a better player than me (though my lifetime record against him is somethign like 3-1-1), but he would tell you that I certainly do not belong in the Seal Clubbing end of the room. How exactly is that even remotely fair to everyone involved?

**nosferatu1001**

And random is more fair?

The first time you show up you would have no record. Second time onwards you would.

**ryan3740**

nosferatu1001 wrote:Use swiss pairings, but instead of random for round 1, use Ranking to determine pairing

So 2 highest ranked players in attendance play each other, then next two, then next two.

Nosferatu, you are confused. You are not describing swiss pairings, but GW's battle point pairings. There is a difference. From wikipedia - The basic rule is that players with the same score [# of wins] are ranked according to rating. Then the top half is paired with the bottom half. For instance, if there are eight players in a score group, number 1 is paired with number 5, number 2 is paired with number 6 and so on.

if a high ranked player always beats a low ranked player..
Round 1 - high ranked play low ranked players
Round 2 - high ranked play high ranked (winners), low ranked play low ranked (losers)
Round 3 - high ranked vs high (2-0), high vs low ranked (1-1) and low vs low (0-2)

It keeps your best players apart until the end and the middle ranked players get sorted out by the end of the tournament. The biggest advantage to this is that if you have 16 players but only have time for 3 rounds, then theoretically your two undefeated will be #1 and #2 seed.

This is different from battle points, where your top two players are the undefeated, and the person who drew the first round and then went on to massacre his opponents. Just look at the recent ard boyz results with me adding in the W - D - L record

mikhaila wrote:And the final results:

1 Rick Puig 19 23 23 = 65__Imperial Guard --- 3 - 0 - 0
2 Mark Billings 12 24 22 = 58__Grey Knights --- 2 - 1 - 0
4 Bill Hennessey 22 9 22 = 53__Dark Angels --- 2 - 0 - 1
3 Charles Peters 11 23 16 = 50__Grey Knights --- 2 - 1 - 0
5 Mattew Bennett 15 16 19 = 50__Grey Knights --- 3 - 0 - 0
3 Aaron Aleong 21 14 9 = 44__Grey Knights --- 2 - 0 - 1
11 Brad College 12 9 23 = 44__SpaceMarines --- 1 - 1 - 1
7 Mike Twitchell 19 0 23 = 42__Space Wolves --- 2 - 0 - 1
1 Dustin Brinson 19 20 1 = 40__Imperial Guard --- 2 - 0 - 1
6 Chris Johnson 5 20 15 = 40__Imperial Guard --- 2 - 0 - 1
2 Dustin Stevenson 12 23 2 = 37__Imperial Guard --- 1 - 1 - 1
6 Nicolas Aad 5 22 9 = 36__Dark Eldar --- 1 - 0 - 2
8 Jayze Eckerman 11 3 22 = 36__Eldar --- 1 - 1 - 1
4 Al Edel 11 23 1 = 35__Imperial Guard --- 1 - 1 - 1
9 Leigh Brady 4 8 23 = 35__Dark Angels --- 1 - 0 - 2
5 Max Rodriguez 21 6 5= 32__Imperial Guard --- 1 - 0 - 2
10 Douglas Hepwert 9 1 11= 21__Imperial Guard --- 0 - 1 - 2
7 Matt Schultz 2 15 0= 17__Dark Eldar --- 1 - 0 - 2
10 Chris Bradly 4 0 11= 15__Orks --- 0 - 1 - 2
8 Jesse Newton 11 1 1= 13__Imperial Guard --- 0 - 1 - 2
9 Edward Ennent 5 6 1= 12__Blood Angels --- 0 - 0 - 3
11 Mike Mirobelli 10 0 0= 10__Space Wolves --- 0 - 1 - 2

**nosferatu1001**

Which is normally described as modified swiss pairings