[Jul 26, 2014] The Salt Shaker - Circuit Championship 300$$ In Pot Bonuses (Indianapolis, Indiana)

[Wiisnake]

MM for first date Wiisnake?

Possibly, I just enjoyed playing a MK again.

 

[Doctor X]

That's not how pools work. You don't decide "who deserves to be in the bracket".

You do, however, consider the estimated skill levels of known players, and try to avoid putting too many potential, say, top-8 players in the same pool. This is considered good practice, because without it you wind up with death pools and a lopsided bracket. These TO's were aware of this-- that's why XeroXen wasn't put in with SneakyTako, and why you weren't put in with Village Mascot.

I was not part of this consideration because again, my name was mistakenly not copied down by the staff. I was in none of the pools. Neither was Drephen, for the same reason, though it seems like he noticed much later than I did. The following attempts to solve this problem simply created more problems, but as I said before? Every pool was already underway. There was no easy solution.

Perhaps it would have been bigger of me to simply take a refund and sit out. I would happily have done so, had I known how badly the switch would ruin Wiisnake's day. Wiisnake's a great sport, though, and didn't want my charity. He did everything he could to beat me-- and almost did-- and I admire him for that. What sucks is, though, our match didn't matter since he dropped a game to OS, and I didn't. Just another reason why top-2 was just a little too brutal for our tastes.

For real though, @ Wiisnake , if that offer to team still stands next time I see you? We're totally doing it. Little boys will wreck the world with fire and ice. It is known.

@ Xiivi :

No worries here, man. Thanks for the tournament, and for being so understanding about our delay. We'll leave an hour earlier than usual next time just in case.

 

[Overswarm]

Things Xivii did right for Indiana tournament:

• Multitude of CRTs available
• Separating PM and Melee in each room and separating when each game was being played with only slight overlap, encouraging enrollment in all events and leaving friendly setups available for those who weren't. This also made it easier to find people when you are forced into two rooms.
• Had change on hand for the tournament, so that it was not required for people to make their own change
• Had multiple TOs helping him, so he didn't have to do everything himself
• Had a stream setup with an overlay and multiple commentators
• Had a pot bonus to encourage outside registration (successfully), involving MD/VA/KY/OH/IN at minimum
• Had pool captains
• Had a dedicated TO area
• Had payouts pre-counted for distribution immediately after completion of a tournament
• Did his best to make up for mistakes that occurred ex post facto
• Stuck to a ruling that, while poorly designed, was a ruling that needed to be stuck to so as not to punish players who had participated with it in mind
• Had clear beginning times for the events, so it was clear which event was being played and why
• Had fans located strategically throughout the building to accommodate players as best as possible
• Had multiple bracket sheets and pens ready and available at the TO table
• Consistently went back to the TO table to do his work, so he was easily found
• Created brackets while another event was occurring via staggered events, so there was no delay in gameplay
• Finished a tournament that started a bit late due to an accident with a very large attendance for the venue at a decently reasonable time (over limit, but I left at 11:30 which isn't too bad for a 2 hour drive)
• Had actual honest-to-God pre-registration for the tournament that was used and posted prior to the tournament for those looking for a doubles partner
• Was friendly to everyone in attendance despite obviously juggling many balls at once

I don't want people reading this to think the tournament didn't have fun to it; I had a lot of fun, although I was getting ill from the heat as it went on later and had run out of drinks I had brought with me. It wasn't like everyone was grumpy at the event; it had many hype moments and lots of friendly people. It was just hot and had instances of bracket manipulation done by someone (who probably shouldn't be named, honestly) who didn't know any better.

Xivii you did your best with what you had available and I wish you luck in the future, but definitely be careful with the assistant TOs and lay some ground rules and don't be afraid to ask the out-of-towners for help. You did a LOT of cool stuff and most of my disappointment stems from telling others I was encouraging to attend that "it'll be good to go to a tournament with someone who actually knows how to run a tournament", because it's so rare now! Apologies for the mixups are very appreciated and accepted wholeheartedly.

 

[DtJ XeroXen]

If Indy hosts a tournament again and you feel like you're having a shortage of TOs, I'll gladly not enter anything and help you guys out.

Also I'd like to take this moment to promote another Indy tournament, Facebook Official, on the 9th of August hosted by the DtJ crew (Hilt/Composer/Spin/LoZ). It'll be great, their last event had ~70 entrants for Melee alone, and promises to have more emphasis on PM singles this time.

 

[JmanJ]

Despite what the last few posts have shown, this tournament was not a failure. The issues outlined above (PM pools and poor seeding in the melee bracket due to the circuit) actually affected a very small number of entrants. Although this does not excuse the TOing mistakes that were made, it should be noted that 90% of the attendees probably weren't even aware of these issues.

My biggest regret is not scheduling both doubles events first, that was clearly an oversight and caused the tournament to move much more slowly. As for the PM pool issue, that was clearly a mistake in retrospect, largely due to inexperience in TOing. Although I was not the one who made the mistake in missing writing the names, I did indeed pull the trigger on switching a player out of too stacked of a pool.

As for Xivii, wayyy too much was put on his plate, and the tournament WOULD have been a failure without him. As he outlined earlier, our original intention was to get a more established TO to help out. Despite a couple veteran TOs in the region showing interest, all of them ultimately bailed on us, forcing us into running a large event without much experience. Xivii did a ton of the right behind the scenes that many will never know about. I'd hate to see all of that overshadowed by a couple incorrect decisions.

 

[OldManDyl (OMD)]

Ya'll are just lucky I stayed behind the mic and didn't step into the ring for the Title Bout.
(Note: I've got on Monday Night Raw and it's really showing)
Edit: That's a lie, I suck at Smash.

 

[Wiisnake]

You do, however, consider the estimated skill levels of known players, and try to avoid putting too many potential, say, top-8 players in the same pool. This is considered good practice, because without it you wind up with death pools and a lopsided bracket. These TO's were aware of this-- that's why XeroXen wasn't put in with SneakyTako, and why you weren't put in with Village Mascot.

I was not part of this consideration because again, my name was mistakenly not copied down by the staff. I was in none of the pools. Neither was Drephen, for the same reason, though it seems like he noticed much later than I did. The following attempts to solve this problem simply created more problems, but as I said before? Every pool was already underway. There was no easy solution.

Perhaps it would have been bigger of me to simply take a refund and sit out. I would happily have done so, had I known how badly the switch would ruin Wiisnake's day. Wiisnake's a great sport, though, and didn't want my charity. He did everything he could to beat me-- and almost did-- and I admire him for that. What sucks is, though, our match didn't matter since he dropped a game to OS, and I didn't. Just another reason why top-2 was just a little too brutal for our tastes.

For real though, @ Wiisnake , if that offer to team still stands next time I see you? We're totally doing it. Little boys will wreck the world with fire and ice. It is known.

@ Xiivi :

No worries here, man. Thanks for the tournament, and for being so understanding about our delay. We'll leave an hour earlier than usual next time just in case.

Dude that offer stands, and it always stands, Ness has the fire to set up Lucas' grabs and Lucas has the Ice to chill some electric bairs.

 

[Overswarm]

You do, however, consider the estimated skill levels of known players, and try to avoid putting too many potential, say, top-8 players in the same pool. This is considered good practice, because without it you wind up with death pools and a lopsided bracket. These TO's were aware of this-- that's why XeroXen wasn't put in with SneakyTako, and why you weren't put in with Village Mascot.

You... do not do that. Or shouldn't, anyway. Not to that extent. That's bracket manipulation and bad TOing; its putting your own perception of "who is good" on everyone in attendance, including those who you do not know. These TOs certainly didn't know everyone in attendance.


Tournament theory time so everyone can lose the idea of "this guy is SUPPOSED to advance because we KNOW he'll advance, so we give him a pool that confirms our prior expectations". Doing that for an entire pool? That's bracket manipulation and, for some reason, people think it is okay. There should be question as to who should advance.

If you run tournaments, are interested in running tournaments, or interested in helping run tournaments, do yourself a favor and read this post. I have eliminated as much of the math junk as possible to make it enjoyable to read.



A tournament's best format is a Round Robin in its entirety; we simply do not have enough time for it, so instead we attempt to emulate the core aspects of a Round Robin.

Since a RR involves multiple matchups that are irrelevant (Mew2King vs someone's 5 year old brother) that end up taking up the time, our goal is to decrease the amount of those matches and increase the amount of relevant matches (Mew2King vs. Drephen).

We do this by reducing the total number of matches by having an elimination bracket. You play a round and the better player should move on, thus creating a ranking of roughly the top players.

Random seeding can end up with #1 and #2 playing together though, so we make the bracket double or triple elimination. The more rounds of elimination, the more relevant the results. In addition to this, we also "seed" our players based on their previous results, so the #1 vs. #2 shouldn't happen.

Adding more layers isn't always feasible for time and can be mentally confusing and can result in many long waits and bottlenecks. We also don't always have prior results due to odd attendance; if a top player suddenly decides to show up and gets a bad seed and plays a top seed, we've effectively manufactured the very problem we wanted to avoid with random seeding!

To solve this we use a miniature round robin -- Pools.

Pools allow us to properly seed a bracket or, in some cases, more pools.

But how do you put people in pools? The correct response is always "totally random" if you have no prior information and the pools simply seed into bracket, but there is no elimination.

But we typically have SOME information.

So when we create our pools, the number of pools should be relevant to the number of 1 seeds that will be needed. That information is typically available. If it is a bit messy ("oh we have 5 people that could be considered 1 seed... but only 4 pools!") then randomization is your friend. Chaos is fair. You're allowed to have a hard pool or a bad seed.

http://www.printyourbrackets.com/pdfbrackets/32singleseeded.pdf

If you take a look at that bracket, you see that it is separated into 4 quadrants. There should be 1 4-seed for both. This allows for those 4 "one seed" players to not play each other in bracket early on. You see it illustrated on that image by 1, 2, 3, and 4 being separated so that they are far from each other.

4 pools of 8 creates this as a possibility if you have 32 people.

You would then separate players by location and separate four individual players that have previously performed at top seed level and place them in pools and randomize the rest based on location.

That's it! Your pools are now competitive and prevent largescale messups.

It should be noted that "separating by location" is bracket manipulation, clear and simple. However, a reality of the smash community is that no one travels many hours to play their friends they play all the time. If it was commonplace people wouldn't go. It's a courtesy to everyone involved and, while it does mess up brackets and pools, it is something everyone agrees on. Large scale professional events (like MLG) should not do this. We are grassroots, so we are forced to do so.



"But what if there are EIGHT really good players, all vying for that #1 seed?! What if I randomize it and 5 of them get in one pool?!"

That's perfectly fine. If you have a crazy number like eight "top" players that are literally all at the same skill level it means you don't have enough prior information or you're going to be playing favorites by "choosing" who plays who.

Getting 5 of those into one pool might make you afraid of "messing up the seeding", but if you manually place two of them in each of the pool then you are determining the #1 and #2 seed and not letting the pools system do it. You want randomization as much as possible (not variance, randomization!) and want to manually input things veeeeeery rarely.


If you're going to manually decide the seeds for each of the pools, why do the pools at all? This is unfair to all the players who aren't considered in that "eight top players" as someone who is an up-and-comer is forced to ALWAYS be pushed into a 3rd seed scenario at best unless he beats the top players in his pool... but the top players in his pool never have the same possibility. You can have 5 people in the pool randomly vying for the 3rd seed, but only 2 vying for 1st and 2nd with not real contest. That's unfair to the other players and systematically hinders their chances at playing in a fair tournament.

That is called "bracket manipulation". It FEELS right because it fits our expectations. Mango and Mew2King in the same pool would seem weird, but it shouldn't be an impossibility. If you find yourself manually moving people, you are playing God with the bracket. If you can't physically write down all the placements at the tournament 100% of the time, you're screwing someone over.

If you can, skip pools and seed the bracket yourself.

But you can't. No one can. That's why Drephen vs. Jiano was hype in the finals -- no one knew who was going to win prior.

"But what if people are eliminated from pools? The stakes are much higher then, it isn't just a bad seed in bracket."

This is true! A 32 man bracket is often what we want, but often we have more, as in this case. So we need more pools and some will inevitably "drown in pools" so we can get that optimal 32-man bracket. If you have time, by all means, have a 64 man bracket with a bunch of byes. No one will complain if you have time.

So how do you make these pools that eliminate players prior to the bracket?

The first rule you should try to follow is never eliminate more than half the pool.

This is a safety measure that allows for variance. If you are eliminating more than half of a pool, you should be cautious. I have only done this in all-pools events where the pools are seeded based off of past performance from literally that day. I would not recommend it otherwise.

The second rule is that you should always try to have as many people in a pool as possible. This is very difficult because the more people you have in a pool, the more time it takes.

A good number in a pool is 6. This lets 3 advance and 3 be eliminated at minimum and allows many chances for redemption. I personally prefer to have top 4 advance if at all possible, although it sometimes isn't due to time.

This, combined with separating your X one seeds (one per pool) and completely randomizing the rest, reduces variance and allows for more competitive play and more accurate seeding (rather than simply copying historical seeding). Really though, separating the 1 seeds isn't a necessity; you don't HAVE to do it. We just do it because its typically common knowledge by now and easy to do because we have a TON of past performance on these players.

That said, I was a guaranteed 1 seed that got a 2nd seed and ended up getting 8th, I think. I was a favorite to place top 3 and simply... played poorly! Putting me in a pool as a 'one seed' manually manipulated the bracket in my favor; if I had a harder pool I might not have made it out of pools and it would have reflected my play. Because of the "pick a 1 seed" policy that populates the pools, I got a free ride into bracket. Not so fair for the other people.

But, statistically speaking, taking your 1 seeds and separating them one-per-pool is fairly sound and can reduce variance significantly.


But exactly how much variance DOES occur if you use my pool system?

You can check it yourself with a randomizer. Go here: https://www.random.org/lists/

type 1-16 all on separate rows, then randomize. Split them into groups of 4, like you would in a pool, with top two advance. 4 x 2 = 8, so 8 players show up. I personally just copy it into excel and organize them in groups of 4, makes it fast.

Assume optimal results would be 1 through 8 advancing, as 1 being most skilled and 16 being least skilled.

People who advance from one of my tests:
1, 2, 3, 4, 5, 7, 8, 10

Another:
1, 2, 3, 4, 5, 6, 8, 9

Yet another:
1, 2, 3, 4, 5, 6, 7, 14


Obviously not perfect, but that's pretty good for randomizing three times in a row. The trick is that each group is effectively 1/4th the players and only half are eliminated.

Here's top 32, 4 pools of 8, top 4 advance:
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 16, 17


Do it yourself, it feels like a magic trick, but it isn't.

If you are using 1/4th of the players per pool and eliminate only half, the variance from randomization on its own is almost entirely negligible as to who will make it into bracket.

Individual seeds can be difficult, yes. But that's why we have double elimination. If pools perfectly seeding everything, a bracket could be single elimination.


Keep in mind, I didn't separate my one seeds in the examples above. If I had (guaranteeing 1-4 all entering) then the variance would become even lower.

You can't always use 1/4th of the players per pool, but you should attempt to get as close as possible.


This is very basic seeding theory. There's tons of other stuff you can do to make a tournament run super smooth that isn't necessarily intuitive, but if you run a standard pools -> bracket tournament, follow my rules here and every time your bracket will have the "right" people and it will be competitive, rather than predetermined.

For people scrolling to the bottom saying TL;DR:

You don't have to manually manipulate pools and brackets to get the "right people" into bracket. You can allow randomization to do it just fine and variance is incredibly low.

Having hard pools is okay.

Saying "that sucks, so-and-so should have advanced" and manipulating the pools or bracket to allow that to occur is entirely unfair to the other players.

 

[Doctor X]

You would then separate players by location and separate four individual players that have previously performed at top seed level and place them in pools and randomize the rest based on location.

OS's post is indeed full of useful information, but note the following-- For four pools, he suggests selecting, by hand, four players that will, guaranteed, not be placed together in the same pool. The motivation for doing this is exactly what I described in my previous post. He suggests some large disagreement between us on this matter, but there isn't one.

The question is, where do we draw the line between what counts as a completely reasonable precaution and what counts as evil, evil bracket manipulation? OS says 4 players for 4 pools. I loosely estimated 8 players for... I'm not sure how many pools there were. It think 6 or 7? Whether that's a large enough disagreement to hang me from a tree or something, I'll leave that up to you guys, lol. I'll gladly defer to the community in this matter at any event I personally run in the future.

Mango and Mew2King in the same pool would seem weird, but it shouldn't be an impossibility.

I want everyone to consider for a moment the community backlash that could result from constructing such a pool. Not just from fans of M2K or Mango, but for the lower-level players in their pool who are now far less likely to advance. Imagine throwing Armada in there with them. Statistically unlikely or no, you're pissing off a lot of people, now. The fact that the pool may have emerged from random.org does not shield you from that fact.

Death pools are bad. We can all agree on this, right? That is why we take the precaution mentioned above, isn't it? I mean, I won't speak for everyone, but in my own personal experience it seems to make people happier. :\

 

[Overswarm]

OS's post is indeed full of useful information, but note the following-- For four pools, he suggests selecting, by hand, four players that will, guaranteed, not be placed together in the same pool. The motivation for doing this is exactly what I described in my previous post. His post suggests some large disagreement between us on this matter, but there isn't one.

The question is, where do we draw the line between what counts as a completely reasonable precaution and what counts as evil, evil bracket manipulation? OS says 4 players for 4 pools. I loosely estimated 8 players for... I'm not sure how many pools there were. It think 6 or 7? Whether that's a large enough disagreement to hang me from a tree or something, I'll leave that up to you guys, lol. I'll gladly defer to the community in this matter at any event I personally run in the future.

Don't pat yourself on the back. I only suggest this as a possibility, not a necessity, and only when you have a well defined history for those players AND there aren't a greater number of potential 1 seeds.

If you have 5 "top" players and 4 pools and you pick 4 as the 1 seeds, that would be bad.

If you're just kind of guessing, that would be bad.

If you pick all the players that will advance, that would be bad.

I want everyone to consider for a moment the community backlash that could result from constructing such a pool. Not just from fans of M2K or Mango, but for the lower-level players in their pool who are now far less likely to advance. Imagine throwing Armada in there with them. Statistically unlikely or no, you're pissing off a lot of people, now. The fact that the pool may have emerged from random.org does not shield you from that fact.

Death pools are bad. We can all agree on this, right? That is why we take the precaution mentioned above, isn't it? I mean, I won't speak for everyone, but in my own personal experience it seems to make people happier. :\

"Happier" is irrelevant to running a good tournament. If Mango and Mew2King end up in the same pool, thems the breaks. It happens. If Mango and M2K are the only players of that caliber in attendance then you have a very clear example here of "well defined history" and you can go ahead and split the top X to remove the slight possibility that a group of 4 could be all the one seeds. The well-defined history allows you to properly seed the pools to that extent which is the problem you're trying to solve in the first place.

But manipulating those pools to have your top X already determined because they are 'icky' is cheating. It is giving Mango and Mew2king and whoever else you consider "good" at your tournament an advantage no one else would have. That's why you use the 1/4th and half rule. It reduces variance to the point where you don't have to guess.

Think of two imaginary players - we'll call them Mars and Luis - and imagine they are of roughly equal skill. But unlike Mango and Mew2King, they are roughly equal skill for 4th seed. Why is it exactly that the predicted ONE seed gets special treatment, but the fourth seed does not? What about 3rd seed or 2nd?

Your core premise is "Oh man, think of all the lower level players in the pool who are now far less likely to advance", but that's BS. If Mango and M2K are in the same pool, another pool has lower level players that have a greater chance of success. There's no net change.

More importantly, it doesn't matter. It is not a TOs job to hand hold everyone so they get the placement they FEEL they deserve. You EARN your placement.

Bad bracket? Deal with it. Tough pool? Step your game up.

A TO rearrange a bracket and pool so that it fits into what people expect is BS. Upsets happen. The farther apart in skill, the less of an upset can occur.

When you deliberately set up pools so that the top players that you choose will more than likely get the best seeds, you completely screw over the players who aren't on your "good list". To GET on that good list you have to not only fight an uphill battle in pools, but will likely have an immediately difficult path in bracket. Everyone who is already on your "good" list gets a free ride! I don't know how anyone can think that isn't cheating.

Your end goal is not "let's make nice pools that reflect the bracket expectations", your end goal is to create pools that have conflict so that players can earn their spot.

If you go to a tournament and you see a list of players and they are clearly designed so that certain players are guaranteed to move on, the TO just decided to give a big middle finger to everyone he doesn't personally know.


If you manipulate pools and brackets to make things "look nice" and consider "fairness" to be making all pools tiered so that everyone already knows who will make it out, you are not a good TO.

If you do that, just skip pools and go straight to bracket and place people where you want.



EDIT:

Just would like to add for the other TOs:


If you have circuit points, you can use them to seed pools. You don't necessarily need any randomness at all, assuming no out-of-staters.

The way to do this is to simply arrange them by points they earned from prior placement and separate them into groups based on their points and the available number of pools, then randomize them within those groups.

 

[Drephen]

I love smash drama lol

 

[Magic Man]

I just want to say that, even with the few flaws, I had an awesome time at this tourney and I'm extremely glad I made the trip. I don't know of a single tourney that I've been to that went completely perfect, there's obviously going to be a few hiccups. I had a great time meeting and playing against new players, and hanging out with Cinci/Indy people. Always nice to see you guys again. For the TOs, just take everyone's feedback as constructive criticism, and use it to improve the next tournaments you guys host. I know Louisville will absolutely be coming back for future events, and hopefully we'll have our full group next time! Hopefully there will be a results thread soon, going to do individual shoutouts there. Really looking forward to seeing everyone again at Facebook Official!

 

[Doctor X]

Think of two imaginary players - we'll call them Mars and Luis - and imagine they are of roughly equal skill. But unlike Mango and Mew2King, they are roughly equal skill for 4th seed. Why is it exactly that the predicted ONE seed gets special treatment, but the fourth seed does not? What about 3rd seed or 2nd?

Well, first of all, I don't know if I've ever seen a consistent 4th seed player. Consistent one seeds, by their nature, tend to win their pool no matter where you put them. Unless, of course, you put two or three of them in the same pool. The reasons we avoid doing this have already been discussed.

Your core premise is "Oh man, think of all the lower level players in the pool who are now far less likely to advance", but that's BS. If Mango and M2K are in the same pool, another pool has lower level players that have a greater chance of success. There's no net change.

Overall, maybe not. Individual players matter, though. What you're saying is that one group of players can (and should) get screwed by the whims of chance, while another group are given free pass. Instead of making sure that each pool has at least one hard competitor, and challenging people to defeat that competitor, you think it's more fair if people occasionally get lucky and wind up in a free pool.

Generally, though? Not even the latter group likes it when this happens. Yeah, they get to do better in their pool, but when they look into the situation and realize what happened to the former group? They feel bad. Often times invalidated. Like they didn't really earn the spot they had.

More importantly, it doesn't matter. It is not a TOs job to hand hold everyone so they get the placement they FEEL they deserve. You EARN your placement.

Your emphasis on the word "EARN" here is strange, considering the above.

"Happier" is irrelevant to running a good tournament.

...

Yeah that's about all I got, folks. See you around. For real this time. Shoutouts will be in the results thread, of course.

 

[exdia_16]

Overswarm: Bad bracket? Deal with it. Tough pool? Step your game up.<-------------------------- what i had to do when i lost my first pools set. had fun tho

 

[Kel]

so wait, will there be a results thread?

 

[JmanJ]

so wait, will there be a results thread?

Yes, Xivii is doing it. He will be a bit delayed though due to work related stuff.

 

[Xiivi]

http://smashboards.com/threads/the-salt-shaker-results.362768/

I made an abridged one for now~