MBR Recommended SL+RS Discussion

[Cactuar]

Okay, the thread with the current edition of the MBR Recommended is up. Lemme know what you guys think. The doubles counterpicks are up for discussion and by no means final (only a half joke on my part though). Point out anything you see that is incorrect. Typos, grammar, anything.


Things that definitely need some discussion:

Doubles stage list





Sorry for rushing this out btw, but I'm trying to get something out with enough time to let it sink in for Apex.

 

[Strong Badam]

I still feel like the random elements of Brinstar and Kongo Jungle 64 are less effectual than that of FoD or PS. I can see an argument for stalling on Kongo Jungle 64 though.

What about teams ports for RC, KJ64, and Brinstar?

 

[Redd]

I can't see any extreme cases of KJ64 stalling ever happening... unless it already has.

I still feel that in terms of randomness: Brinstar > PS > FoD > KJ64. Hmmm. Let me think on this for a while.

EDIT: I will say I'm really excited to see bans again.

 

[Strong Badam]

http://www.youtube.com/watch?v=mTNaAUJZz5k

also Pokemon Stadium should be Pokémon Stadium

 

[Aniolas]

I would like to discuss how to determine who advances from the pool.

For example:

_____A___B____C___D____E
A ___X __2-1__2-0__2-0__2-0 A: 4 sets won, 8 games won, 1 games lost
B __1-2___X___1-2__2-0__2-0 B: 2 sets won, 6 games won, 4 games lost
C __0-2__2-1___X___2-0__0-2 C: 2 sets won, 4 games won, 5 games lost
D __0-2__0-2__0-2___X___2-0 D: 1 set won, 2 games won , 6 games lost
E __0-2__0-2__2-0___0-2___X E: 1 set won, 2 games won, 6 games lost

Say top 2 advance to the next round. Obviously, A will get top seed. The second seed however.
B and C have won equaly amount of sets. B have won more games, and lost fewer. He is, the overall better player than C.

C did beat B in the head to head match.

I think, that pools should determine the overall better players, else one unlucky MU can make a slight worse player to advance, even if the other player did better in total.

I think that the order to give people seeding from the pool should be:
Sets
Game won
Game lost
Head-to-head match

If it's still a three way tie (which happens, not often but sometimes) RR with those three players.

 

[Fortress | Sveet]

I agree with aniolas 100%. Thats the system ive been using

 

[SleepyK]

I don't know how to feel about rainbow cruise as a teams stage.

 

[Cactuar]

Yay people actually read through the whole thing!

This is just a personal opinion regarding the tiebreak methods.

Spoiler

Lmao, I've looked at it from so many different angles and always have a reason that I disagree with wins/losses/headtohead instead of headtohead/wins/losses.

To summarize,
1. I have an issue with how easy it is to abuse, and I've seen it happen many many times over the years.
2. For scores like yours to happen, there is usually an upset somewhere in the pool. This is a lower level player bridging the gap and beating a high level player. In that situation, the high level player who has been upset previously won against the dude he is now tied win. The level of play in the match between those two players was much higher and far more demanding on both players, and a much greater indicator of skill imo, than the fluke upset due to whatever reason.
3. It has always felt incredibly wrong to have to tell someone they tied for 4th, but even though they beat the guy they tied with, he gets to go on because he won extra matches at some point.

I think the one posted by Aniolas is far more commonly accepted though, and I will be changing it to that.

@Strongbad: I'm not putting that gayass skidmark above the e.

 

[Fortress | Sveet]

So both players have dropped a set to an unfortunate match-up. Shouldn't it then go by who got the most wins/least losses, and if they are exactly even in score, who beat who?

 

[Cactuar]

@Sveet: I should do my edits in notepad before posting instead of using the forum like notepad lol.

My issue would have likely stemmed from the heavy influence that counterpick stages have had in the past. In this sense, players of characters that have extremely influential counterpick stages have a significant advantage due to how easy it is for them to make any set 2-1 instead of 2-0.

We are removing those stages other than FD, so it shouldn't really be an issue.

I do have a side thought in the level of play of the sets involved. To me, the set between the first and second seed is of much more importance than the set between the first and last seed. This is likely an extension to thinking in terms of quality rather than quantity.

It is much much easier to just change it to wins/losses/h2h than attempt to accommodate for those thoughts. Hence the reason I posted all this back here and encouraged thorough reading and discussion.

 

[Strong Badam]

@Strongbad: I'm not putting that gayass skidmark above the e.

cactus noooooo

I think it should be stated that all bracket matches, if possible, should be best of 5.

 

[Cactuar]

For regionals/nationals.


When the bracket is top 32.

 

[Skler]

RC is an awesome team stage. It doesn't lag or anything, so as long as it's legal it should remain a team stage.



I still think Brinstar should drown in its own Acidlava though.

 

[Redd]

I could live with that, Skler. I just want Corneria. =(

 

[Fortress | Sveet]

lets just always play on dreamland. Screw slow characters and characters that can't zone the whole stage well.

 

[Strong Badam]

sounds like a plan

 

[Skler]

A bad plan.

WHAT A ZINGER.

zing

Edit: I desire more italics.

 

[Juggleguy]

I think everything that makes people want Kongo Jungle banned in singles is a non-issue in doubles. Platform camping is an ineffective strategy against two opponent characters and the randomness of the barrel is offset by the fact that teammates can save. I actually think it's an excellent doubles stage.

 

[HyugaRicdeau]

I think that the order to give people seeding from the pool should be:
Sets
Game won
Game lost
Head-to-head match

If it's still a three way tie (which happens, not often but sometimes) RR with those three players.

Why not game win/loss ratio (instead of games won then games lost)? So a game record of 6-4 is better than 7-5.

Also, we did actually have an exact 3-way tie in pools at Genesis 2. We had the players do a one game RR, with stage decided by striking. Thankfully it was resolved after only one iteration.

 

[Fortress | Sveet]

IMO wins are more important than win/loss ratio

 

[HyugaRicdeau]

Why do you think that?

 

[Fortress | Sveet]

I dont really have numbers or anything but my overall first gut reaction tells me that.

But yeah, i think 7-5 should get out over 6-4

 

[HyugaRicdeau]

Well, that's not very convincing. We should have a reason for whichever system we decide on. Taking wins first means that a single 1-2 loss makes up for any number of 2-1 wins (dropped games). Consistency is more important than being able to take games on counterpicks, and that's what percentage achieves. Percentage also is, naively, an indicator of the likelihood to win a game.

I could also see an argument for a system that only looks at win percentage for games in 2-1 sets, i.e. having all the 2-0's cancel out first. So in our example, 6-4 would be 0-0 and 7-5 would be 1-1, and so on. In this case, 7-7 would beat 6-6 because it translates into 1-3 vs 0-2.

 

[Fortress | Sveet]

With counterpicks on the out, its much more difficult to get single game wins. You can't just pick rainbow cruise and then move on to game 3. I think taking a game off a better player in the pool is better than just not dropping games. By allowing 6-4 to advance over 7-5 you are allowing a player who took no games off anyone above him to advance over someone who did, which is something I dont agree with. I would argue that the 7-5 is more likely the better player.

 

[Cactuar]

Ugh. I hate using match counts for any reason. I just don't care about it at all.

I'm just going to make it: set wins, head to head, three way tie is a round robin until winner is determined.

**** all this match win ****.

 

[HyugaRicdeau]

With counterpicks on the out, its much more difficult to get single game wins. You can't just pick rainbow cruise and then move on to game 3. I think taking a game off a better player in the pool is better than just not dropping games. By allowing 6-4 to advance over 7-5 you are allowing a player who took no games off anyone above him to advance over someone who did, which is something I dont agree with. I would argue that the 7-5 is more likely the better player.

But if you take a game off but drop one also, this makes you more inconsistent. Plus what you say about counterpicks depends a good deal on the matchup and personal preference/style.

Which is why I'm now thinking I could actually get behind a system in which the individual games is ignored entirely. The problem I have with using head-to-head as a tiebreak is that there is no a priori reason to weigh that particular set as worth more than any other set when you're trying to determine who advances. We want to advance the player who has the best chance of going farther in the bracket, so you could argue that actually the player who beat the highest ranked person in the pool should win the tiebreak, and if that's still tied then it goes to 1 game rematch. Head-to-head is the easy solution, yes, but that doesn't mean it's the best.

 

[KirbyKaze]

Ugh. I hate using match counts for any reason. I just don't care about it at all.

I'm just going to make it: set wins, head to head, three way tie is a round robin until winner is determined.

**** all this match win ****.

If this was FB, I would "like" it.

Not that anyone would be able to see it, of course.

 

[Cactuar]

But if you take a game off but drop one also, this makes you more inconsistent. Plus what you say about counterpicks depends a good deal on the matchup and personal preference/style.

Which is why I'm now thinking I could actually get behind a system in which the individual games is ignored entirely. The problem I have with using head-to-head as a tiebreak is that there is no a priori reason to weigh that particular set as worth more than any other set when you're trying to determine who advances. We want to advance the player who has the best chance of going farther in the bracket, so you could argue that actually the player who beat the highest ranked person in the pool should win the tiebreak, and if that's still tied then it goes to 1 game rematch. Head-to-head is the easy solution, yes, but that doesn't mean it's the best.

This is pretty much what I was thinking as well. Whoever won the set against the higher seed player continues. Which, in many cases, will be the other person in the tie.

 

[HyugaRicdeau]

This is pretty much what I was thinking as well. Whoever won the set against the higher seed player continues. Which, in many cases, will be the other person in the tie.

Yeah, that's true (Aniolas' example is actually an exception). I'm thinking about implementing this but I still have reservations about it: from my argument earlier, you could argue that the person who beats the higher ranked player but loses to a lower ranked one is more inconsistent and therefore shouldn't advance over the other guy.

I dunno, every tiebreak system seems to have some flaw in it. I think what we need to remember is that the point of pools is to find the players most likely to advance further in the bracket, so we should be thinking on that when we decide which tiebreak system is best.

 

[Fortress | Sveet]

A series with the score of 2-1 generally means a close set; either player could take a game off the other. In equal set wins, the player with more game wins had closer sets with the people he lost too, while more game losses means he had closer sets with the people he won over. Following the priority of ranking, set wins is obvious; it sorts by who won the most overall. If that is a tie, game wins is next meaning the player with closer sets against people ranked higher is ranked higher. If that is also a tie, game losses is next meaning the player who lost less to players ranked lower is ranked higher. If all of that is a tie, the player who won the set between them is ranked higher.

The only bad thing about head to head is that it is generally mutually exclusive with who beat the higher ranked player. In the event of a tie, the person who won the head to head lost to someone the other person beat.

For going by W/L ratio, it favors people who have less close sets with people ranked below rather than people who have more close sets with people ranked above. I think that in every case for realistic pools, W/L ratio has the same outcome as judging game losses before game wins. IMO taking games off better players is a better system, but i guess its matter of preference.

 

[KirbyKaze]

2-1 could also mean the losing player just had a super ridiculous counterpick to go to.

2-1 can also very easily be ****.

 

[Fortress | Sveet]

What super ridiculous counter picks are there in the new ruleset? If you take a game off someone on one of the 6 stages that are left (not to mention if they were allowed to ban) then I think it does show that that player can take a game off the winner.

 

[Cactuar]

I'm just going to put a variety of tie-break systems in and note that TO's should select which they will be using and publicly state that in the tournament thread.


Did we ever get around to teams stages?

I'm just going to do something similar to what I did with the singles stages and give a set number of bans based on the number of counterpick stages if nothing is said.

 

[Fortress | Sveet]

I think brinstar/RC/DK64 are all fine in teams. I think pokefloats should be unbanned as well, and possibly other stages should be looked at too. Teams dynamic is much different than singles; the strategies that make those stages banned in singles doesn't work the same as it does in teams.

 

[unknown522]

2-1 could also mean the losing player just had a super ridiculous counterpick to go to.

2-1 can also very easily be ****.

What super ridiculous counter picks are there in the new ruleset? If you take a game off someone on one of the 6 stages that are left (not to mention if they were allowed to ban) then I think it does show that that player can take a game off the winner.

taking ICs to brinstar.

going to brinstar with jiggs.

FD/DL vs chaingrabbable character

 

[Cactuar]

What? Brinstar isn't legal in the new ruleset and spacies will likely never go to FD vs a chaingrabber, as best of 3 sets have 1 ban.

Looks to me like someone didn't read.

 

[KirbyKaze]

I mean, this is somewhat unrelated since we're removing these stages anyway, but this set made me laugh.

http://www.youtube.com/watch?v=yvVrGhCea1Q

 

[Fortress | Sveet]

Haha fox counterpicking brinstar

 

[Cactuar]

Good example of a low tier getting dominated on CP stages. >.>

Vist was playing pretty well too...

 

[Strong Badam]

**** rainbow cruise