Stop the Count(ing of Rating)

@deribus said in Stop the Count(ing of Rating):

Screw you my mapping tourney is gonna have a rating cap

You know the logic can extend here too. Have 2 tiers for a mapping tournament, one for experienced mappers and one for novice. The experienced one gets the prize pool as well as the sculptor avatar, novice simply gets an avatar. You can only submit a map to one tier and the novice pool is locked to people that have never made a map before. Or you can lock it to people that never had a map accepted into a matchmaker pool. The restriction can be tuned.

I always tried to reward a winner of hearts to new dudes that made a serious effort during these anyway since it felt like new dudes deserved some recognition.

Count the rating the players had at the time of tournament annoucment maybe? Instead of the rating they have when they signup? You can still downrate yourself, but then you have to be permanently underrated.

This stops malicious deranking it stops nothing related to gaming rating caps. It also makes choosing a teammate about 10x harder for everyone looking to play.

I guess you can apply some script to find that information for some 1v1 tournament at a specific rating and it’s still fine, though in general the rule about using trueskill as a gauge for ensuring competitive games should still be done.

I'm glad you decided to raise this problem after many years of organizing Faf tournaments.

In order to divide the participants into a large number of rating groups, you need to have its own prizes for each group. In addition, this option does not exclude manipulation with the rating in order to get into a lighter group.

If you tell people the rating limits after the fact, then next time no one wants to participate. Because many will want to play in the tournament, will cancel their plans. And will be very upset when they are not allowed

It is best to run tournaments with captains. But even with the idea of captains, people will still manipulate ratings. Because the selection order will depend on the ratings of the captains and their teammates. The only option left is to give the teammates at random or just the player with the highest rated player with the lowest. But this excludes the possibility of playing with friends, limits the format of the tournament and requires a large number of top players as captains.

Yeah you need prizes, but largely this would consist of avatars with cash prizes generally going to 1500+ brackets. Doesn’t really harm anything to give out Tournament Champion avatars for a U1000 bracket.

How are you supposed to manipulate to get into a lighter group? In addition, how would this benefit you? For manipulation to be productive you would need to know the precise breaks for brackets ie is it 1800 or 2000 and you’ll also need to know which group is getting funds as well as how many funds. This relies on having omniscience and knowing exactly how many players are signing up.

Who isn’t allowed to play? The only time anyone isn’t going to be allowed to play is when extreme outliers are in the tournaments like a single 2400 signs up and then the next highest rated player is sub 2000. This is about grouping brackets so that tournaments are LESS of waste of time with dead games that were never really a question.

No, because nobody will know how many captains will be in any bracket.

1. The bracket rating is not released prior
2. It cannot be known how many people will sign up at whatever rating level
3. The TD could use rating at X date as Katharsas pointed out if this is still something you fear (I don’t)

It doesn’t require any large number of top players as captains. If a bracket has 4 competitive teams, then there are 8 players that are considered competitive at this rating range and so 4 captains.

Your solutions of highest rating with lowest are exactly what incentives rating manipulation. The reason captain picking stops rating manipulation is that it doesn’t matter if a dude deranks by 300 points, if everyone is well aware he is better than his rating then he will still get picked first. So the only deranking gain is for captains, which again have no ability to know whether they will be captains until after signups close.

@ftxcommando said in Stop the Count(ing of Rating):

If you want to make it less restrictive, have people sign up and do a captain picking procedure.

If you go with captain picking, then give people Z amount of time to practice with their teammate if they have any interest.

But it's really exhausting to see teammate tournaments relegated to the tournament being won before it begins because the real game is beating the rules and not the opponents.

Uncapped tournaments are still fine obviously, this is about the attempts to create tournaments that are "fair" and yes I am aware this requires more TD work but I don't really care.

The one question I have is how do we have events for lower-level players that want to participate in events and do it honorably? I am not ready to say we should 100% banish rating-limited events because of the "bad apples".

As far as "2300 vs 1700" being uninteresting - agreed.

Should we implement a rating threshold for events as well? I think there was an event by Swkoll in recent where not enough people could play, and 2 of the players allowed in were quite deficient from the average rating.

In equal amounts of frustrating events with rating gaps, there have been equal amounts of frustrating events with not enough players.

I would not mind hosting some heavily-restricted events myself to test this out (although that goes completely against what I promised in my PC application and likely would not stand), but if someone like Auriko 2v2 Calypso, or Pryanichek 2v2 Kings & Veterans wants to do this, that is their choice.

Are you asking "Official FAF" to stop hosting rating limit events or everyone in general?

It's a garbage format with no redeeming qualities, nobody should be doing it and nobody should be telling others to do it.

This post gives two separate ways to make fair tournaments for lower rated players that does not require rating caps. If the people that end up signing up for a tournament on X weekend are not low rated, then low rated brackets don't exist.

The only events that got canceled for lack of players were events that shot for participation from groups hovering around 1k rating. If people want more players to sign up, we have a Discord ping for tournament signups that literally nobody uses and we have a news post that can advertise tournaments. Since the Discord role is rating-blind and this restructure of tournament formats would also turn many tournaments rating-blind and instead rely on signups to craft brackets, it synergizes well.

Can we add rating divisions to the faf? Which are updated about once a month (all this can be discussed in more detail). And hold tournaments within the division or on the principle of 1 player from division X + 1 player from division Y etc? In this case, when the tournament is announced 2 weeks before the start, the player will not be able to change his division, no matter how hard he tries. And vice versa, he can be calm that even if he plays a series of successful games, he will still be able to get into the tournament.

A division system is already being implemented.

Of course, divisions will not be reflected in regular games.
By the way, it would be great if the divisions calculated the average rating by the sum of the ladder and the global

Then we could make a limitation not by the rating but by the rating of the divisions (if we assume that the 1st division is the weakest and the 5th is the strongest) In this case, for the tournament, you can make a limitation in the total division max 6 or 7 for example

@pryanichek said in Stop the Count(ing of Rating):

it would be great if the divisions calculated the average rating by the sum of the ladder and the global

Some people never play ladder, or they haven't played in years.

Other people only play ladder.

How do you average that?

If someone has 1600 global and 800 ladder, they would be treated the same as someone with 1200/1200? That would create an incentive for certain people to never join the ladder. Or an incentive to "throw" ladder games to keep the rating down so they can get a better spot in Seton's.

There is already the possibility of people throwing games to manipulate global rating. But if you're 1600/800 you might have to throw 20 global matches to get the same impact on a "combined" rating compared to throwing a single ladder match.

The best way to prevent smurfing is to make it difficult to pull off. Fewer people will do it if it's harder to do. And if people have to put more effort into it, it should be easier to catch them.

My ladder rating gone down about 100 points while my global has gone up 100. Mostly because I'm out of practice. Does that mark me as a "smurf"? If there was a combined global + ladder rating, some people might think so. And I wouldn't have an incentive to get my ladder score up. My best move would be to abandon the ladder completely. If I play and win, I hurt myself on "combined rating" for Seton's placement. If I play and lose, people will call me a smurf.

Forgive me @FtXCommando , I should have said "frustrating events with not enough players that meet the requirements but sub in less-qualified ones to make up for it."

I don’t really think subs are going to be that big of a problem. It can be handled on a case-by-case basis just as always. The bigger issue with this stuff relates to the fact there will be a closing of signups which will hit the FAF habit of signing up 2 minutes before the tournament starts hard.

This is inspired by the gene centric model of evolution as really popularized by Richard Dawkins.

1. People of any rating can join, low rated are encouraged to join in fact.
2. Each game is either a 3v3 or a 4v4 depending on player availability.
3. People are assigned to games first based on location and then randomly.
4. Each game is set to all random auto balance.
5. Being on a losing team means you lost, which in a knockout means you are out.
   Being on the winning team is less obvious, at least it means you are NOT out yet.

This system will basically select those players that played well for their rating these games.

Determining winner is somewhat more difficult as it is possible a well playing 500 rated and a 2200 rated would come out in #1 and #2 position. A 1v1 as final would be pointless.

• Selecting a single winner would require an extensive number of rounds and points counting
• You could get 3 winners by going into a knockout 3v3 finals where again each game is all random auto balance.

N-nice tourney format you've got here ;d

I'm gonna be honest, I'm really confused by the tourney format.

@archsimkat I will post it soon-ish again in its own topic with better explanation and examples.