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.

OK, so I decided to go with a southwest cup format to include Arizona and Nevada as most tournaments are all fairly close to each other in that geographic area.

This is the scoring system I am thinking about going with. Please poke holes in it, find mathematical loop holes, etc so that I don't look like a jack ass if someone exploits the system in the inaugural year! haha.

I want to have this done this month so that it will be ready to go for the 2011 season.

I want to be as inclusive as possible so that as many gamers are motivated to participate as we can get. The more the merrier! I am also a huge proponent of K.I.S.S. (keep it simple, stupid) so that people can quickly and easily understand how the system works. Lastly, this is for competitive gaming as it is meant to show who the best player is. I love painting and hobbying and the fluff, too, but in the context of this particular event, it's all about winning and losing. So with that philosophy in mind, here's the nuts and bolts:

We take the first, second and third place Generals (battle points, win/loss only) of each event in this geographic location as reported by TO's and keep a running score of their points on a website which I will create and on which players can track their scores. Points are awarded according to a simple formula:

Only GT events (2 days and 5 to 7 games) count.

1st: 30 points
2nd: 20 points
3rd: 10 points

Attendance: This is the part I am most wary of as it can really heavily skew the system. Do you all think a bracketed system would be more equitable? Like, 10 points for less than 30 people, 20 for 30 to 50 and 30 for 51+? Or would it be better to give a point per person in attendance? A monster event with 100 people could really throw this out of wack, but as of now we don't have any that big in in this region and the nice thing about this is that it clearly stratifies the field for the purposes of seeding. This of course would heavily favor players who play in areas with lots of big events. I am open to any ideas on a compromise or alternate system here.

How many events count? We would take the top three scores for each player and at the end of the year, invite the top 16 to an invitational event (most likely in LA for ease of travel) and hold a two day event.

Tournament Structure: Now, I am thinking a seeded, single elimination, 2,000 point, book missions (4th round would be a random draw for mission/deployment), 4 round system with KP's as tie breakers in objective games and VP's as tie breakers in KP games would work best given time constraints, but that means half the field will be out of the event after game 1. That also means two thirds of the field would be done after the first day which for people who have to travel can be crappy. So, my solution to this was a losers bracket but then you can end up with weird situations where a guy can go 3 and 1, having only lost his first game, and end up placing below someone who went 1 and 3 having only won his first game. If you go off of straight W/L records then you will have a whole boatload of ties. Ideally I would like to have a system wherein we could rank the players 1 through 16 and wherein everyone got to play 4 games. Does anyone know of a system of tie breakers that would work well in this scenario? Perhaps something based off of seedings? I am open to ideas here.

The event itself would have sponsorships, prize support, a trophy and a banner for your sig stating your ranking in the Southwest.

So please, give me your two cents here. I am open to ideas and would love to hear them in regards to the items I listed above.

Thanks!

You're not going to get what you want by doing anything other than a swiss style tournament. I don't think there's any benefit to doing single elimination unless you want the players standing around watching the semifinals and finals (which may be a good thing). I don't think there is any way to guarantee that 4th game for the people who lose in the first round while also sticking with the single-elimination idea. I think your best bet is to do something like use 4 games Swiss style as an eliminator and then for game 5 just have two games - a first place game and a third place game and everyone else watching. Similar to how they appear to do the finals of the Nova Open, though I think they make it a semi-finals and then one more finals, which also works if you can sqeeze in a 6th game for those two people.

I think swiss style is important also because really, this is a game that lends itself to ties and most tiebreakers are pretty arbitrary. Give the players a chance to actually shake things out, or come back from a couple of close losses with a big win. While in reality most of the time you're basically eliminated if you lose the first game anyway, you can call on personal experience about how that isn't always the case in the right scoring format.

Yeah, initially I wanted to do double elimination but with a loser's bracket it takes too long. I guess just making that first round all important is where it's at. I will play around with some bracketing systems to see what would work best.

Thanks for the input.

Yeah, double elimination with 16 players you end up with the losers ultimately playing 8 games - that is definitely too many.

With a normal single elim and a losers bracket you can end up with everyone getting a placement and the people who lose round one playing three games (first round, then a round to determine 15th-16th and a 14th-13th and then a game to decide who is who). Everyone else gets 4 games or more. This seems fine to me if people know what they're getting into and doubly fine if there are prizes or something that go down that low to make every game mean something.

Reecius wrote:Ideally I would like to have a system wherein we could rank the players 1 through 16 and wherein everyone got to play 4 games. Does anyone know of a system of tie breakers that would work well in this scenario? Perhaps something based off of seedings? I am open to ideas here.

Swiss style tournament. Sort first round based on your tournament point system. Then pair 1 vs 9, 2 vs 10,... 8 vs 16. This will average out the gap between players. Second round rank on # wins, then tournament points. 1 vs 5, 2 vs 6... Repeat for two more rounds. At the end sort on # of wins, head to head, sum of opponents' wins and then tournament points. Who should get ranked #2? One of the 3-1 players that only lost to #1, not another 3-1 or a 2-2 player. Strength of schedule helps this. As an alternate for sorting in rounds 2 - 4 you can use battle points instead of your tournament system points. Keep in mind that battle points are a tie breaker, not a sign of skill.

I used note cards in my last 16 player tournament to keep track of things. That way you can easily see who they played, what their battle points are and what table they played on. It also allows you to qucikly add up sum of opponnent's wins at the end of the tournament.

No system will give you 1 - 16 accurately in four rounds. You really need a 15 round-robin style to have a true standing.

As for your tournament scoring system, you could use an ELO type scoring with varying K values based on # of entrants. This may be very time consuming to calculate and keep track of though. Go ask a local chess group how they handle ELO scoring for their events. Is it based on the number of entrants?

Or you can give points to each player at an event, not just the top three. This would be similar to NASCAR regular season scoring.

Or just give out a golden ticket at all the events. Give out a number based on the # of people that enter.

If you are looking for a system for determining invites, here is what we were looking at when we were considering something similar for Fantasy:

Any 1st Place in a 2 day event gets an automatic invite.
Any Best General in a 2 day event gets an automatic invite.
Any combination of five total Best Generals/Best Overalls in RTTs consisting of 12 or more people or top 10% finishes in two day events gets an invite.

We were also going to invite people involved in running events under the same sort of criteria (counting running events as a "win" essentially), but thats neither here nor there.

Not sure if that sort of thing helps you establish a seeding system, or if one is even needed. If we are talking a 5 round event, do random pairings for the first two rounds and use W/L or BP (depending on what you are going with) for the remaining three rounds.

@Lambadomy
Yeah, 8 games is too long for two days with travel time and what not.

@Ryan3740
Thanks for the tips there are some really good ideas in there. I think a ranking based off of seedings or battle points as tie breakers would work best to stratify the field. That way we can rank players. It may not be ultra accurate but it is better than simply declaring a winner, IMO. The round robin would be awesome, but obviously with time restraints it isn't realistic.


Automatically Appended Next Post:
@Phazael
That's not a bad idea as for one, there are not that many large events to draw from. Two, not everyone who gets an invite will be able to make it. Plus, some larger RTTs like thos up at Great Escape Games in Sac draw 80+ players on the reg.

Maybe we could do that as it would be easier.

Best general: Automatic invite.
2nd Best General: Automatic invite.

If either player already has an invite, it goes to the next in line.

Then we could use random pairings for round one of the actual tournament and W/L after that with true Swiss, single elimination. That is super simple which I like.

Reecius wrote:@Lambadomy
Yeah, 8 games is too long for two days with travel time and what not.

@Ryan3740
Thanks for the tips there are some really good ideas in there. I think a ranking based off of seedings or battle points as tie breakers would work best to stratify the field. That way we can rank players. It may not be ultra accurate but it is better than simply declaring a winner, IMO. The round robin would be awesome, but obviously with time restraints it isn't realistic.


Automatically Appended Next Post:
@Phazael
That's not a bad idea as for one, there are not that many large events to draw from. Two, not everyone who gets an invite will be able to make it. Plus, some larger RTTs like thos up at Great Escape Games in Sac draw 80+ players on the reg.

Maybe we could do that as it would be easier.

Best general: Automatic invite.
2nd Best General: Automatic invite.

If either player already has an invite, it goes to the next in line.

Then we could use random pairings for round one of the actual tournament and W/L after that with true Swiss, single elimination. That is super simple which I like.

FYI Reece Swiss and Single Elimination are mutually exclusive. You can't do both. You'll need to decide on what format you want.

Benefits of single elim is you can bracket the tournament after ranking people. Benefits of Swiss is everyone gets to play all 4 rounds, you can have ties at the game level, and you have a foolproof tiebreaker system at the tourney level.

I would defin give an auto invite to the best general of the tournament. Then use a point system to determine the "at-large" invites and then invite those people accordingly. What you'll end up with is a bunch of tourney winners and then runners up from the larger tournaments and I think that is fine.

Good point, that allows for a broad range of players to come in case people can't make it.

Also, thanks for the clarification on Swiss/single elimination. I had a misconception of what Swiss pairings were and that is good to know in advance! Haha.

A lot depends on the format style you go with. Personal bias aside, I suggest Swiss for a number of reasons. First, its much more flexible with an uncertain number of players. Second, it can handle ties without breaking down. Third, it can be worked out in 5 rounds with 40 people. Finally, no one is going to want to drive down, have a bad break in round one, and then stand around with their d!ck in their hand for the entire rest of the weekend.

If you are trying to have as many people as possible involved than I don't think only allowing two day events to count is a good idea. There are a lot of good players out there that for what ever reason just don't go to many two day events. Aard Boyz is a one day event up until 2011 the tournaments at Adepticon were one day events (they had a lot of different ones)and some of the better players go to them.
I would say the criteria should be based on the number of players at an event not the number of days.
One system I have been working on is to award players points based on wins at a tournament so that the better players will generally have more points at the end of the season. For the local level I am including all events, but for a Tri-state area I would suggest only using events with 20+ players as a qualifer.

If you're really so into the idea of single elimination you could do it world up style:

4 groups of 4 players who all play each other (3 games each)

top two players from each group move on to a single elim 8 player bracket. Bottom players have their own bracket if you really still feel the need to separate them out. This means that the "finals" would be game 6 but would only be two players.

Another way to do it is to do 4 swiss games and then take the top 4 players and do two rounds of playoffs for them - 6 total games for two players, 5 games for 2 more and 4 for everyone else.

As for actually rating everyone - I think people are a little too hung up on the number of players. If you're playing 5 games, the best you're really doing is picking the best player out of 32 people. If you have more than 32 people, you're ending up with people who are roughly equally successful who have not played each other. Of course, I'm coming from the perspective that a game of warhammer should have a winner and a loser, not a massacre/minor win dichotomy. That stuff only exists because of the fact that 5 games isn't actually enough to differentiate 32+ players. I don't think it's actually meaningful in terms of skill, mostly it just influences army choices because it forces people to try to table their opponents because just winning isn't enough.

OK, I feel that a Swiss system with tie breakers for identical records using the methods you guys outline in the Chaos Wins blog is going to work out great. We won't have any tie games, we will use VPs and KPs as tie breakers. That makes the pairings and match-ups easy and determines placings with a fairly high degree of accuracy. Everyone gets a game and no one is eliminated. I like that.

Thanks for the input on that guys, it makes things super easy and straight forward.

As for invitations, we will go with top general from every event of a certain size and then rank everyone else based on a score determined by placing and size of the event. Since we won't be needing seedings we can go with a straight points system. I was thinking:

20/40/60 for 3rd, 2nd, 1st and 1 point for each player in attendance.

You get 10 points per the number of games, so in a 3 game RTT, you get 30 additional points, in a 4 game event like the Ironman you get 40 and a 5 game GT 50, etc.

So, that way a big RTT can carry close to as much weight as a mid sized GT.

I feel that this will keep things fairly even. A huge RTT like those in Sacramento will carry a lot of weight where a small GT like the LA Conquest event will carry less as it should be.

For a typical RTT with 20 players the winner will get a point range of 70 points for third place (30 for 3 games, 20 for 20 players and 20 for third place) up to 110 for first place.

A 30 person, 5 game GT will have a spread of 100 for third up to 140 for first.

An 80 person RTT will have a spread of 130 for third up to 170 for first.

And a 70 person GT will have a range of 140 to 180.

Does this seem like an equitable spread? I feel it gives a fair shake to events based on the numbers that indicate difficulty, which are primarily number of games and players.

This also includes as many people as possible in the event. It means for those unable to travel to a lot of GTs, they can still get involved if they have some good sized RTTs to go to. If they win 3 mid sized RTTs they can have a good shot at getting into the top 16.

What do you guys think? I am open to adjusting those points levels.

Also, for a minimum qualifying size for an event to get an invitation based on coming in 1st, I would say a starting point of 80 is fair. That means at least 3 games and 50 players or 5 games and 30 players. Anything smaller and coming in first counts for points only, not an automatic invite. Thoughts?


Automatically Appended Next Post:
@Vhwolf
I agree, I think only allowing 2 day events is too restricting for a lot of people.

@lambadomy
If we do a straight Swiss system, we will have a single player having gone 4-0 after four games, 4 players at 3-1, 6 at 2-2, 4 at 1-3 and 1 at 0-4. We will then use tie breakers to stratify these guys from 1 to 16. That seems easy and fair to me, what do you think?

Yes, the straight swiss system with 4 games for 16 players and no ties will get you one winner, exactly how you laid it out.

I think you need to figure out generally what events are going to be qualifying before you worry about a qualifying system. If it's win-and-in then you can only have 16 qualifying events, for example. If you want to have more than 16 events to qualify for the 16 slots then you have to do some kind of ELO rating system, and that system needs to be meaningful across events (so the events kind of need to have the same rules). If you have less than 16 events then you're fine, I think you just pick the winners plus 2nd place from the largest X events to fill out your 16.

I still feel that winning something with 100 people when you have only 5 games is more about dumb luck and weird tiebreakers and bonus points and less about who actually deserves it, but that's just me. The top 4-8 players probably won't even have played each other and will have small differences in scores.

You are correct, with tournaments that size and that few of an amount of games, it gets really tough to accurately determine who is best. A small tournament with higher caliber players is a tougher challenge, however quantifying that without an extensive ranking system is very difficult. ELO would be sweet but on this scale would take a lot of bookkeeping. The system we are proposing is very simple and I think for the inaugural year at least, will suffice.

You are right about 16 events total for auto qualification. When you include RTT's with varying attendance, that makes things very unpredictable. A straight scoring system will be the most flexible system. And that also eliminates situations where someone wins a single event and then doesn't go to any other tournaments.