There is no seeding when it's done that way. As such, your top 16 are unlikely to actually be your top 16 since somewhere along the way they will get eliminated from other folks.

In large events, you'll do ~4 rounds of swiss to "figure out seeding" and then cut to top 16 or whatever and then do single elimination. But when you're running <16, a cut to the top 4 doesn't really mean much.

I will say that I have issues with swiss style pairing in addition to FFG's tie breaking system. All too often it seems like the top two players meet at the final table (in the case of 16 players and 4 rounds... going into the final round there's only 2 people 3-0... winner takes 1st. But common sense to me seems like the loser should take 2nd. But all too often, one of the folks that was 2-1 ends up 3-1 and either via SoS or modified wins ends up taking 2nd place. And whomever lost at the "final table" comes in 3rd or 4th or something like that.

So I haven't competed in a tournament yet but will be in may. My question is why don't they just do it single eliminatation? Get all the players, draw the matches and set the byes, then run the tourney that way? Seems like it would clear alot of headaches and be the simplest way of running it. Granted it sucks losing in the first round but that's just competition.

But second, and more importantly, single-elimination matches are very bad at finding the best player. Given random seeding (which is the most feasible approach in X-wing), you're very likely to eliminate some of your top players early. No one would really be interested in watching a "championship" NFL game, for instance, where the final score was an embarrassing 43-8 blowout -- clearly that wouldn't be a game that put the best two teams against one another.

So you need a tournament structure that accounts for the fact that even very good players can lose a match, and over the course of the tournament allows the best to rise to the top. Swiss accomplishes that, and does so in a more reasonable amount of time than, say, double-elimination.

Yup, that happened to me. Though I ended up dropping to fifth.

Yup, that happened to me. Though I ended up dropping to fifth.

Well, you stated they did their tiebreakers wrong anyways. But even if they did the TB correctly, it is possible. There will be 1 person at 4-0, and 4 at 3-1 (and 6 at 2-2, 4 at 1-3, and 1 at 0-4) assuming a binary win/loss combination. So if you managed to get the one and only modified win between the 4 people at 3-1, even though you were at the final table, you will drop out of the top 4 and into 5th. SoS tie breaker will have a hard time doing that to you, but it is indeed possible. For example, in the rare occasion that the 4-0 defeated all 4 of the 3-1 folks, the final table loser doesn't have the "well, at least my SoS is boosted by the 4-0" 'advantage' that he normally has.

I don't consider people "top people" if they get eliminated early. If they have a bad game or make a minor mistake and the opponent is able to capitalize on that and win then they deserve the win. That's the pressure of being the best, you got a target on your back and can't make any mistakes or have chinks in the armor.

All tiebreakers are terrible. There is no one that doesn't screw someone.

Not all tiebreakers are terrible; for instance, in a round-robin tournament you can resolve any tie by the result of a head-to-head match. Overall, though, I'll agree with you that most of them can be pretty bad.

But strength-of-schedule in a Swiss tournament with random initial pairings is uniquely terrible, in that it's so close to random selection. In Swiss, you have the most information about the skill levels of the players at the ends of the distribution (e.g., 5-0 players and 0-5 players), and you lose information as you go; you have the least information about the people in the middle of the distribution (for instance, 3-2). Each player's record is an estimate of their skill, and those estimates have error associated with them; the error associated with of the sum of multiple estimates is the sum of the individual errors.

Basically we use strength-of-schedule as an indicator of the reliability of a player's record--but we do so without acknowledging that the error in strength-of-schedule swamps the value of the actual estimate. And if you add in the effect of players dropping out of the tournament, the picture gets even worse.

SOS is really, really terrible, and it needs to go away from every competitive event that can't claim to have reasonably accurate rankings for each participant and use those rankings to establish initial seeding.

So if the #1 and #2 people got paired together in the first game, the #2 person really isn't "top tier" because he lost to the world champ?

So if the #1 and #2 people got paired together in the first game, the #2 person really isn't "top tier" because he lost to the world champ?

You beat me to it. Randomly seeded single-elimination is a really bad tournament format for this reason.

So if the #1 and #2 people got paired together in the first game, the #2 person really isn't "top tier" because he lost to the world champ?

that's an unfortunate drawback to single eliminaton, if they are the 1&2 guys then they'd go against each other anyway. No matter the system the best person will win, a scrub won't make it to the finals and win. I think there is an inherent bias when you use seeds. In my opinion everyone should start at equal footing at the tournament regardless of past results.

So if the #1 and #2 people got paired together in the first game, the #2 person really isn't "top tier" because he lost to the world champ?

that's an unfortunate drawback to single eliminaton, if they are the 1&2 guys then they'd go against each other anyway. No matter the system the best person will win, a scrub won't make it to the finals and win. I think there is an inherent bias when you use seeds. In my opinion everyone should start at equal footing at the tournament regardless of past results.

I don't think ranking is particularly feasible in X-wing anyway, so it's a moot point. But if we can't seed players by their rank--or simply choose not to--then that constrains the tournament structures we can use. And, as I said, randomly seeded single-elimination is a huge problem in that case: you don't want strong players eliminated early by other strong players, because it makes the tournament as a whole weaker.

In what way does it make it weaker? Really at the end of the day it's about who wins. Again in my opinion it should be the guy who didn't lose a single match. 2nd, 3rd, 4th, and so on don't really matter at the end of the day. I know that last sentence kind of sounds conceited and jerkish but I don't mean it as such. "Stronger" people getting eliminated early doesn't make it weaker in my opinion. They got eliminated because the person across from them was either a better player, or made less mistakes.

Instead of a a single elimination you could go to a true (or modified) double (or even triple) knock out system, that way if a good player does lose a game they still have a chance to make it to the finals.

Edit: Spelling

2nd, 3rd, 4th, and so on don't really matter at the end of the day.

I think this is where some/many people would disagree with you. Examining your ranking can give a sense of your progress, and if there are prizes involved and you care about that then it would matter as well

If you think tournaments are all about "which one single person is the best at all times regardless of any other factors" then sure, single elimination might be fine, but to anyone else it is not the most desirable solution

2nd, 3rd, 4th, and so on don't really matter at the end of the day.

So your store championship gave all 4 rulers and both card boxes to the top player?

I just like that everyone gets to play a set amount of games, no matter what. Tournaments are also about having fun and finding more competition to improve your game play. You won't learn much if you lose immediately. This will mean new players won't be able to get as good as the veterans as quickly, creating larger skill disparities. Just a lot of draw backs. It's not all about finding first place, it's also about community growth and participation in a very fun game. While swiss isn't necessarily the best solution, at least everyone gets to play some games!

Not to mention, people who have to travel large distances will be discouraged from driving if they might only get 1 game. This lowers the player pool quite a bit.

I haven't done a tournament yet, my first one will be the ohio regional one. I'm coming from a place of ignorance when it comes to the swiss format and no doubt will have a much better understanding and appreciation for it after the regional. When I say they don't matter I was referring to the statement about the tournament being stronger or weaker. I dont see how who coming in 2nd and so on would have any bearing on that. Obviously if you have gone from not placing to taking 2nd then you have made serious improvements. I hope I'm not coming across as douchey, I'm a very competitive person and when I dedicate my time to something competitive it's obviously to win, I'm not that hyper competitive ****** who pouts and gets angry at a loss and acts like a fool but I definitely want to take 1st lol.

When I say they don't matter I was referring to the statement about the tournament being stronger or weaker. I dont see how who coming in 2nd and so on would have any bearing on that.

I think the exception people are taking to this statement is that a tournament which determines only 1st place and does nothing to rank other players is inherently a weak system, on several different levels.

Most tournament structures have a few goals... they want to rank competitors, they want to make an interesting event, and they may want to try and maximize the enjoyment of those involved (depending on the event).

Swiss does this in a lot of ways. Matching best-to-best ranks people similarly to elimination. It ensures games are interesting by matching people with comparable scores, and theoretically therefore comparable skills, which also reduces the number of blowout games that are rarely fun for anyone involved. It also structures so that people get to play a set number of games regardless of how they do, which is a major factor in enjoyment. And, finally, at the end of the day it ranks ALL players, from top to bottom.

Comparatively, a randomly seeded elimination event fails on most of these. It has no control over the comparable skill of each player, which means games can frequently end up lopsided, and half the people who show up are going to play one game and go home, and 75% of them are done after two games. When you're done, you know first place, but the system tells you nothing at all other than that. That's far less interesting to most people, especially those who aren't going to be in those top two spots, and gets to be very problematic if you've got more than two prizes to give away.

In short, elimination may be fast and efficient at finding the #1 player, but it pretty much sucks completely for everything else.

I was wondering are asteroids mandatory?

I did my first tournament last weekend, had to drive over 2 hours, and after registering and then signing up with my squad the TO said that we were not going to use asteroids which I wasnt to happy about.

This gives tie swarms huge advantages and had I known I probably would have used a tie swarm.

Also according to the faq, I'm sure that I read in the that the player who had initiative, uses their asteroids and dmg deck, where as everyone had used their own dmg deck which I also thought wasnt quite right, as there are 7 crit hit cards, if 2 players are using one deck then I would think chances of hitting those cards or any other cards is a little more balanced is instead of just one deck per person. You don't burn through the cards as quickly

So when playing tournament do you have to use asteroids?

I really was disappointed, as that makes a huge difference

Definitely very valid and good points! I'm excited about my first tournament!

Ive only been to a couple of tournaments so far and it amazes me how little some of the TO know about the rules.

I mean one store told me that they werent going to play the standard dogfight, they were going to play scenarios from the books with random number of asteroids, some tables with none, others with 10. Thats the official FFG store championship as well.

The tournament rules are only a few A4 pages, not a hard slog really to get it right.

All of us to varying degrees, wants to take 1st. Some of us temper our expectations so we're not disappointed, but its bred into our DNA to be the best, or at least Darwin thinks it is =P. That said, being at the final table and losing is a lot more fun than being "#2" and losing in the first round. In large tourneys, swiss does seem to work better. For example, I believe there were 84 people at worlds last year. Lets go ahead and say that this year there's 128 people, and after 5 rounds of swiss, they'll cut to the top 16.

4 people will be undefeated and 20 will be 4-1. And the way swiss pairs people, it's unlikely you're in the top 16 if you aren't 4-1. You could potentially lose to one of those that's undefeated, but that would drop you to x-1. And you'll be matched up against other x-1 people. Potentially you lose to one of those (say 5th place) x-1 people, putting you down to a 3-2 record. It is plausible for this to happen. But even if we said you were the 16th ranked person, you would have to be matched against one of the top 4 and one of the 5-15 placed people. Let's look at odds of that happening.

So, you have to lose to the 5-0 in the first 4 rounds. If it's the 1st round, you had a 4/127 chance of getting matched up. Second round would be 4/63, 3rd would be 4/31, and 4th would be 4/15. So, playing against the top 4 players has a 49% chance of happening in this case... thats rather reasonable to happen. So now you've dropped one game.

Now you have to be placed against 5-15 in another game to drop you to 3-2. This could only happen in rounds 2-5. And actually, you could not possibly play against the 1-4 and 5-8 since they're both undefeated until they match up against each other. As such, we're looking at the chance that you play 9-15. I'm going to over estimate this by making the assumption that they have lost one game whenever you need them to have 1 loss (aka, they lost in the first round if you were matched against them in the 2nd)... so then its 7/63 in the second, 7/63 in the third, 7/47 in the 4th, and 7/39 in the 5th... adding these together, the chance of you playing against 9-15 under these circumstances is 55%.

So, the overall probability that as #16 you get matched up against two players better than you, kicking you out of the top 16 is 27%. This is not a completely unreasonable thing to happen... And will probably happen to a few people in a large tourney. But, if 12 of the top 16 truly belong in the top 16, I'd say the Swiss rounds did their job right.

And while I'd be a little miffed that I missed the top 16 because I played against the top players, at least I can (hopefully) draw some happiness from the fact that I held my own decently against the top players.

I could see them updating the scoring, they did it for SWLCG once they saw how skewed the results were becoming. I've actually warmed up to the idea of:

4pt win

3pt mod win

2pt draw

1pt mod loss

0pt loss

That way every game has 4 points up for grabs, and you're playing for how those 4 points are distributed.

They also need to do something about their tie breakers. Not only does their tiebreaker not go deep enough, it isn't based on player performance. I can't help it if in the first round I'm matched up against the worst player at the tourney. I had a swift victory and got my 4 points (according to my system)... but then ended up tied with someone who got a decent player at the beginning, and had a stronger SoS. I have no control over how that works. Plus, it severely handicaps the player that loses to someone undefeated in the first round and then won the following 3 games, vs someone who won 3 games and then lost to the same undefeated person in the 4th round. I've done the math on the SoS, and ignoring modified wins, the person that lost in the 4th round has a variance of 6 games on his SoS... the guy that lost in the 1st has 10 variance, and is down by 3 on SoS to begin with.

So if you split the variance, the guy that loses in the first round loses SoS by 1 game. If we give each game a 50% win ratio, that means that the person that lost in the final round to the same person that lost in the first will win the tie breaker only 22.7% of the time. Meanwhile the guy that lost in the final round will win the tie breaker 59.8% of the time. They will tie 17.4%... And that 17% of the time, there is no further tiebreaker to determine who wins.

I like the idea of a point for a modified loss. It gives the person who is losing in a close games something to fight for still and does not necessarily kill your SoS as much when your opponent only gets 3 points for the modified win. But I still think 5 points for an outright/unmodified win should still be used - make a full win worth more than a combined modified-win/modified-loss. That is following the idea of soccer that uses 3 points for a win but 1 point each for a tie - just modified to fit the Star Wars tourney point structure.

And be design in a Swiss tournament, you have to accept that if you lose in the first round you probably are not going to play nearly as tough of opponents over the course of the tourney as you are if you lose in the last round - which will affect your SoS. Hopefully the person you lost to in round 1 has a good tournament and finishes near the top, but that doesn't always happen. I have lost to a person in the first round and won the rest of my games and the person that beat me I was his only win. In that case my SoS was not the best. It happens and you just have to deal with it.

Is FFGs tie-break system the best - no. Chess uses Swiss format and has different tie breakers the TO can choose from that seem to work fine. Some of the tiebreaks used have decently complex math so I am not going to go into them here - just stating a fact. When I ran tournaments, I let the computer handle that. But chess does have the rating system that eliminates a lot of the first round pairing randomness that Star Wars still needs to deal with though.

I would like to see improvements in the FFG tournament structure, but they are not going to be easily implemented. Until then, we have to use the rules they have provided us. A good place to start would be a standardized software program for running tournaments that does all the pairings and SoS calculations for you automatically. I have looked and couldn't find anything out there that worked well - pairings were fine, but SoS calculations were something else. A program directly from FFG for TOs to use would be nice - or at least if they could work with a software developer to come up with something. Most local tournaments are small enough to run by hand but getting into Regionals/Nationals/Worlds - it is so much smoother if the tournaments can be automated as much as possible.

Lee

Not all TO's are created equal. You just deal with them a like a bad ref call in a sports game.

So if the #1 and #2 people got paired together in the first game, the #2 person really isn't "top tier" because he lost to the world champ?

Perhaps this is arrogant of me but this is what my Farmageddon experience was like. Draw KineticOperator round 1. I'd put myself as a favorite for the final table at that tournament with a few others.

Here's a question. Does Fantasy Flight provide official 3x3 play mats for Store Championships? What about for Regionals?

I've played in 2 Store Championships and mats were not provided. They had to rely on people bringing their own. I have a feeling that we got lucky that there were enough.

I kind of assumed that FF would provide the stores with "official" mats for Regionals at least because it seems really bad if the store didn't have enough to cover the event and not enough players brought them.