|
Hey there Starcraft II community, I have a question for you.
In a perfect tournament world, the number of entrants is always a power of two (2, 4, 8, 16, 64, 128, 256, 512, ...). This requires everyone to play the same amount of matches, essentially leveling the playing field. However, after a given number of entrants, reaching that magical number becomes harder and harder. The number needed to get to the next benchmark doubles every time. If 128 players are registered, only 128 more are necessary to hit the next magical number. If 256 are registered, 256 more are needed.
So what happens if a tournament has 500 registrants? Do we give 12 players a first-round bye and have 500 participants (and hundreds of more exciting matches to watch)? Or do we kick out 244 hopeful registrants in order to make the tournament fair? Like I said, hitting the magical number means everyone has a fair chance at winning. Presenting only some players with a bye gives them a better chance at advancing (who's to say those who received byes would have made it out of the first round?).
You guys can probably infer what my position is on this, but I want your opinions. It's important that I ask because it's possible that I have to make a similar decision. So what do you say? Bye or no bye?
One last word:
Tournaments are a great way not only to showcase your skills, but also test them and push them to their limits. That's why we love playing in them, that's why we love viewing them. Thus, it only makes sense to allow as many people as possible to participate in tournaments - specifically SC2 tournaments. It only leads to further development of the game we love. Think of tournaments as a way to inch closer to a game's competitive potential.
Love and Peace!
-MeistR
|
Um... obviously you just have 12 byes instead of kicking 200+ people out of the event ... lol.
|
Hehe, that's the response I was anticipating (and the one I want). But there might be players out there who are against it? If they have an argument, I'm willing to hear it.
|
If the tournament is seeded properly, the 12 who get byes would've won their first round easily and giving them a bye just means they can sit back for another 15minutes until they have an opponent for the 2nd round. It wouldn't make much difference whether they played a first round or not if you know they're going to win. For a 512 man bracket, odds are it'd be putting someone like Idra, Silver, or HuK against some random silver ladder player if the tournament was full, which is a guaranteed win for those guys
|
Even if there were an argument for byes being unfair, isn't a bit more unfair to the people who get kicked out because there weren't enough entrants? I mean, the people getting byes didn't HAVE to play to get to the next round, but the people being kicked out didn't even GET to play at all.
Also, G2Wolf has the right idea, the people getting byes should be the ones who cakewalk through the first round anyway.
|
What about making groups of 3-4 players? Like the King of the Beta had.
|
I agree with giving the high seeded players a bye in round 1. This is the reason i think most tournaments put a player cap on them so this sort of thing doesn't happen.
|
If it's only 12 people I think it's a no brainer. The problem comes up when there are like 400 registered people. Then you'd have to give 112 people a bye. Maybe you should have a limit where you say like until 382 ppl you make it 256 and above 382 you give out bye's. Maybe you'd just favour ppl who are known to be good and who'd advance anyway.
I know, it's a hard decicion^^
|
On August 04 2010 16:08 Lighioana wrote:
What about making groups of 3-4 players? Like the King of the Beta had.
You've never organized any sort of thing like this have you? Making squares of four instead of pairs of two increases the work involved of organizing and tracking everything by a power of three (thanks, math!).
Edit: For my two cents, a bye should always be granted regardless of the number as long as the maximum number of matches are filled.
If, in your 128 to 256 example you had 255 people show then 127 people should obtain a bye. This is because kicking them out could possibly change the outcome of the tournament. Also, despite the fact that nearly half the tournament would be receiving a bye, you are still having a marvelously large tournament granting such a huge number of people a chance at testing themselves. Some may be resentful but many, many more will be grateful.
I have seen a great number of tournaments occur where I work (a game store). Often employees are asked to step out, on their day off, to even the brackets. I would like to emphasize, we work here. That is why having people step out works. If you ask 127 people to "kindly go home" you will have a very disgruntled mass of people.
|
You don't need to have a power-of-2 tournament, that's doing things too simply. Rank all players in random order for the first round, pair them off.
After the first round, place the people with one win in group one, no wins in group two, order alphabetically (or however) within groups and pair the top player in each group with the bottom, and so on.
Rinse and repeat. Each round sees people ranked again, and then groups are divided up again. The finals are the top 16 players after rankings.
|
im totally voting for byes, even if its 513 players id give 511 byes and let one game be played.
Seriously, especially first round results mean nothing to the tournament in large fields especially, and that way 255 more ppl will have had a nice tournament.
Ow and btw: Even though most pros tend to reg earlier in the tournaments since they know when they are and stuff, this will also make the tournaments alot tougher later on, since of course skill also comes through numbers (i.e. some very good players might be thrown out without byes).
|
Using my suggestion (swiss power pairing), you need never have more than one bye per 1v1 tournament. It even copes with people who pull out mid-comp. 
|
I always try to maximise participation, and say if there are 96 players signed up I give 32 byes to 32 'seeds'.
|
Okay, so it's been established that giving a bye is the way to go, regardless of the number of entrants. But who do you give the byes to?
Seeding a 512 player tournament is no easy task. On what basis would you seed them? I could understand seeding players based on their performance in a previous competition of the same tournament (e.g. seeding them for cup #23 based on their cup #22 performance). Unfortunately, I'm working with a first-time tournament.
I understand the likelihood that a pro would demolish some random silver ladder player. However, tournaments aren't about presumptions. They aren't about what we think will happen. They're about proving competence in the present. Like I said, who's to say any participant will win for sure? Each tournament - each match, for that matter - is an independent event. Why do the pros deserve a bye any more than the next participant? I want to make this thing as fair as possible.
The only way I can see byes happening is if they are given randomly.
|
On August 04 2010 16:16 daveydweeb wrote:
You don't need to have a power-of-2 tournament, that's doing things too simply. Rank all players in random order for the first round, pair them off.
After the first round, place the people with one win in group one, no wins in group two, order alphabetically (or however) within groups and pair the top player in each group with the bottom, and so on.
Rinse and repeat. Each round sees people ranked again, and then groups are divided up again. The finals are the top 16 players after rankings.
Haha, well that sounds nice! But a couple of questions:
The initial rankings don't really mean anything, right? They are simply for the purpose of pairing up participants. After the first round of matches, #1 from group one is no better than #2 from group one because they both won one match a piece.
|
Correct, initial rankings are arbitrary by design. You can seed people high if you want using any method you want, but it makes no difference unless you want it to.
The nice thing about the method is that it can be modified in pretty much any way you want, and I have done for inter-varsity debating tournaments in the past. If you demand that some group of high seeds make it to the second round, you just add them to the top of the seed table for the start of that round. If you demand that amateurs be given a handicap, just assign them additional points whenever you want. And so on.
|
I see. I'll be organizing the PN Open #1 this Saturday. Obviously our first tournament. I want to get the brackets up as fast as possible in between rounds, and this seems like a pretty time-consuming method. Do you know of any computer programs (other than a TI-89, lol) that quickly and randomly assigns numbers?
Honestly, it would probably be easier for us to copy and paste the registered players list in a bracket creator.
|
Sorry for the delay between posts, I was using my phone as a modem until both it and my laptop were caught in my broken locker at work. I won't get either of them back 'til tomorrow. 
To fill in the details of the Swiss system, you can find the Wikipedia article (for a better description than the one I provided) here, and this looks like a reasonable enough application to run the tournament. To be honest, it really isn't much more time-consuming than elimination or round-robin tournaments on a per-game basis because it only requires one more piece of information (the score). It is more time-consuming in that it demands that everybody get to play a minimum number of games, but that's something that you can choose to pitch as a feature of your tournament if you think it's appropriate to do so - that's entirely up to you, of course. Depending on the resources you have available, you might find that people are more willing to register if they can expect to play in more than just the first elimination round. 
I've used a more complicated version of the Swiss tournament system for British Parliamentary debating tourneys, in which the two sides of the debate are represented by two discreet teams of two speakers (for a total of four teams per "game", and eight individual participants) - so I was recording eight times as much information per game as you would for a Starcraft 2 tournament. I was able to keep the turnaround between rounds to under 15 minutes with ~30 teams in the competition using only a spreadsheet, but a simplified system and a little practice can massively improve upon that. You can also get incredibly nerdy after the debate and release your results, so the participants can see their overall rankings, the rankings of the people they played, etc. You can data-mine a surprising amount of information out of that. ^_^
|
Its a bye in the FIRST ROUND who cares. if someone is good enough to win they will have to go through much better players in the later rounds anyways.
|
As stated above multiple times, debate tournaments have run swiss-style with power matchups forever. It's very easy to do. It also allows everyone to play every round until the playoffs.
Debate tournaments use what is called a "tab program" -- if you Google it, I am sure one exists for free and you can just use it for sc2
|
|
|
|
|
|
EPT
