RETRIBUTION - October 20th - Cincinnati, OH - MK Banned Regional

[Ori_bro]

Thank you Based Sneaky

 

[Chronodiver Lokii]

lol don't worry i know when i'm being a giant *****

which ironically is most of the time? LOOOOOOOOOOOOL

Thiiiis : P
Kidding

Good job Chi. For the size of the event and the very packed time frame, you did work. Just take what you learned and keep on improving as a TO : ) You have the experience, just keep improvin~ YOU GOT THISSSSSS
Wish I coulda went to cheer you and everyone else on, but yeah xD

Good job everyone. Saw some great matches on the stream.

And Ty do work. Great job on 2nd place x3 Yaayyyy!

 

[Overswarm]

Oh tako.

Ok first of all this example is confusing. I'm assuming a total of 4 people were playing, and that P-1 has a set record of 3-0 vs P-2 has a set record of 2-1. Clearly P-1 deserves to progress farther in the tournament, set counts are clearly more important than overall wins/losses in pools or bracket. What is your point?

My point is that set coutns are clearly more important than overall wins/losses. You see how this applies later.

Are you seriously making an arguement against using set counts vs using wins and losses? The wins/losses system is simply in place of clearing tie breakers between people who have the same set counts. It makes 0 difference in a tournament if a player wins the tournament winnning every set 2-1, 3-2, w/e. If they win the required amount of games to win each set, they win.

Most of the time, no one cares you won a game off of someone, it plays no role in determining who advances. Like I said the only time it makes a difference it when a set count is tied in pools.

Again, my exact point. Set counts matter, win/loss does not.

If your arguement is that Wins/losses is stupid and head-to-head matters more then you need to compare these two against each other, not against set counts. I would argue that wins/losses more accurately seperates ties in set counts based off of silly arguements of inui logic. While you may have lost against the player you have tied against, you need to have beaten one of the players the other guy lost too. Is that win more significant than your loss to that player? How can you judge that? The most unbaised piece of evidence is wins/losses, numerical data is more unbiased than situational calls.

You can argue this, but you'd be wrong. When you have a tie between two people you have to choose what wins are worth more. There are two intelligent ways to do this:

1) Rank the players in the pool and compare wins, and the first time a player loses a set that the other party does not, that player loses the tie breaker

2) Eliminate the other parties and do a head-to-head comparison as it is impossible to tie and they are the only two parties involved in the tie breaker


The #1 option is legitimate only in a fair environment in which every player is consistent. Pools are not; I myself went Fox in my pools some just for the hell of it, it happens all the time. More importantly, matchups are huge deal. If a Dedede player gets 1st in his pool and a Falco and Snake are tied for second, doing #1 would lead the Falco to win more often than the Snake by a large degree. This gives unfair advantages to players that are completely out of their control as pool sizes are small enough to where this variance has a large effect.

#2 is the option that has been chosen forever because it is not only impossible to tie in this format, but it assesses the situation in a vacuum. An arbitrary loss to a completely separate entity is irrelevant and this accounts for that. The decision is limited between the two players and is very cut and dry.

Let me play Carls a sad song on the worlds smallest violin. While it's true that some character, even player MUs exist in a rock-paper-scissors balance, the player that wins most often deserves to progress. If we are to doubt that the most consistant player is to advance, then what are we doing? Why even play this game if the results can't tell who is the better player?

Substantiated on what claim? Going 2-0 or 2-1 is irrelevant to your placement in the tournament. I've won several tournaments with close sets back to back to back from 2nd round on, and I've lost tournaments where I was 2-0ing everyone until I lost 1-2. There's no bonus for winning your games more convincingly and it has no bearing on how well you will do in a tournament.

Consistency only matters in terms of sets, not games. You agreed with this up above.

Pretty much in a two way tie, how do you value one win over another? To avoid such a conflict, numerical W/L are more unbaised. That is a legitimate train of thought. I'm not going to say it unfailiable, but it derserves recognition of logical merit. It's not blasphamous to consider W/L before head to head.

It is completely blasphemous and I'm surprised you'd think something like this. You value one over the other because one is consistent with the rest of the tournament and the other is an arbitrary methodology that values completely different criteria.

Sets = sets
Head to head = sets
Win/loss ratio = W/L ratio

So your two options presented in determining victors in pools is either

A)

Sets - sets- win/loss ratio

or

B)

Sets - Win/loss ratio - sets


Believing that win/loss ratio is more important than sets is a totally legitimate thing to believe, but it's an entirely different metric and changes the metagame for Smash immensely. If you believe that win/loss ratio is more important than head to head then by default you have to believe that win/loss ratio is more important than set wins.

But that's not what we judge success by. We do it by sets, not win/loss ratio. That's Swiss, not bracket or Round Robin.

The reason we use it for 3 way ties is for time purposes only. That's it. It's just the next best option we have.

Is Kassandra beating Ori more significant than Ori beating Fizzle? Fizzle obviously made it farther than Ori, isn't Fizzle a better win? Or how about we ignore all that and look at W/L first. Seems legit.

No, it is not legit. And Kassandra beating Ori is much more significant than Ori beating Fizzle because Kassandra and Ori were the only two factors in the equation and we don't value one set more than another. Adding in other variables based off of future data such as Fizzle's placement is impossible, but more importantly we have to judge each pool performance in a vacuum so that we don't arbitrarily decide who advances based off of past performance rather than their current performance.


To say that set wins matter but win/loss ratio is more important than head-to-head because it is a better way to determine who is better is a logical inconsistency. If you believe this, it should be applied to the whole pool and sets should be irrelevant. This is not how Smash is set up and this format drastically improves or worsens certain characters and drastically increases the effect counterpicking has on the metagame as dropping one game is more important with this format.

Saying that one set is worth more than another because you arbitrarily value one set over the other not only has the same problems, but it also isn't fair to the players and results in playing favorites based off of past performance. This is not acceptable in an independent tournament. If Ally came to this tournament and didn't make it out of pools because he goofed off one too many times, we don't just put him back in because "we know" how good he will do. He'd be eliminated because that's what he earned. Kassandra tied with Ori_bro and beat him in the one set that mattered, the head-to-head.

To say that Ori should have advanced because he beat Fizzle and had a better W/L ratio is to say that not only did the set between Ori and Kassandra not matter as much, but also that every other game in the pool mattered more. This is logically inconsistent with how the pool is originally assessed as it values an entirely different metric and makes some seriously faulty assumptions.

Seriously you guys, I feel like I'm re-inventing the wheel.

 

[sneakytako]

We're not argueing on the same topic.

Your saying W/L is stupid beacuse it plays no role anywhere else in the tournament format.

I'm arguing that head-to-head is stupid because it values one set win over another.

Really the best answer is to have only set wins as a pool seeding criteria and have all tiebreakers be played out. But based on time constraints it's not possible. Then the decision to rank W/L or head-to-head first is a matter of which one is more wrong, which is stupid because they are both wrong. Which is why either way is mildly acceptable. I'm not going to sit here an argue that spiders are insects vs mushrooms are plants, it's not worth it.

Convince me why head-to-head is infailable and I'll agree to your way. Which you can't because you already accepted that brawl is rock-paper-scissors with MUs sometimes.

 

[@TKbreezy]

@Ralph Cecil spotdodge topic

you definitely do spotdodge too much. LMAO a lot of times in game 1 and 2 I would just walk up to you and wait. and you'd be like OH ****! SPOTDODGE! and then i'd be like "Got'em"

you definitely play a solid snake outside of that though... also don't rapid jab a falco coming back on stage. if anything time an uptilt and kill him/do big damage.

and PS1 >_> LMAO the sneakiest C4s

 

[FF WuvS]

Video Thread!

Doubles:

Winner's R1:

- Hadesblade/Krystedez vs Player1/Kismet
- Coney/PinkFresh vs Spec/Judo

Winner's R2:

- Coney/PinkFresh vs P1/Kismet

Loser's R2:

- WTP/DLA vs Carls/SoulPech

Winner's R3 (QF):

- Zinoto/Shugo vs Raziek/Croi

Loser's R6 (QF):

- Zinoto/Shugo vs Player1/Kismet

Winner SemiFinals:

- MJG/Kain vs Coney/PinkFresh
- Zinoto/Shugo vs Seagull/BPow

Winner's Finals:

- Coney/PinkFresh vs Seagull/BPow

Loser's Finals:

- Seagull/BPow vs P1/Kismet

Grand Finals:

- Coney/PinkFresh vs Seagull/BPow

Singles:

Pools:

- Bonds vs Jet
- Bonds vs Darc
- Clowsui vs Darc
- Clowsui vs Bonds
- Luminoth vs Future
- Coney vs DLA
- Coney vs Kee
- Coney vs TJDK9
- Coney vs Chuee
- DLA vs Chuee
- Chuee vs TJDK9 (I think...)
- Kee vs Chuee
- Kee vs DLA
- Tech vs Reflex
- Deltacod vs Andrew
- Deltacod vs Carls
- Deltacod vs IkMonster
- Deltacod vs Ralph Cecil
- Deltacod vs Silver

Bracket:

R1:

- Luminoth vs Fizzle

Winner's R1:

- Coney vs Croi
- FF Ori_bro vs Krystedez.

Winner's R2:

- Kain vs BPow
- Player-1 vs Krystedez
- Deltacod vs Seagull

Winner's R3 (QF):

- Shugo vs Coney
- Seagull vs Reflex
- Kismet vs Kain

Loser's R4:

- Nicole vs BPow

Loser's R5:

- Nicole vs Fizzle

Loser's R6:

- Fizzle vs Seagull

Loser's R7:

- Seagull vs Delux

Loser's R8 (QF):

- Seagull vs Kismet

Winner's SemiFinals

- Shugo vs Kismet

Loser's SemiFinals:

- Reflex vs Seagull

Winner's Finals:

- Shugo vs MJG

Loser's Finals:

- Reflex vs MJG

Grand Finals:

Set 1:

Game 1 (Processing. Should be up by 1:40am)
Game 2 & 3
Game 4

Set 2:

Coming soon.

Misc.:

$100 MM: Zinoto vs Player-1
Delta vs Hadesblade
Playlist


That's what I have uploaded so far. Not a bad start. Be editing/uploading more throughout the day.
Thanks again to Zinoto for splicing all of doubles and a lot of pools. Very much appreciated as always :3

 

[Overswarm]

We're not argueing on the same topic.

Your saying W/L is stupid beacuse it plays no role anywhere else in the tournament format.

I'm arguing that head-to-head is stupid because it values one set win over another.

Really the best answer is to have only set wins as a pool seeding criteria and have all tiebreakers be played out. But based on time constraints it's not possible. Then the decision to rank W/L or head-to-head first is a matter of which one is more wrong, which is stupid because they are both wrong. Which is why either way is mildly acceptable. I'm not going to sit here an argue that spiders are insects vs mushrooms are plants, it's not worth it.

Convince me why head-to-head is infailable and I'll agree to your way. Which you can't because you already accepted that brawl is rock-paper-scissors with MUs sometimes.

You value the set between the two tied parties because they are the two that tied and are the only variables present. It IS a more important set.

 

[sneakytako]

They are not the only variables present, if that was so it would be a bracket match. But since it was pools other sets matter. They are ranking the entire pool against each other, even within a tie-breaker situation.

 

[SoulPech]

Watching OS and Tako argue is like watching Romney and Obama....

 

[Xatic]

Video Thread!

Doubles:

Winner's R1:

- Hadesblade/Krystedez vs Player1/Kismet
- Coney/PinkFresh vs Spec/Judo

Winner's R2:

- Coney/PinkFresh vs P1/Kismet

Loser's R2:

- WTP/DLA vs Carls/SoulPech

Winner's R3 (QF):

- Zinoto/Shugo vs Raziek/Croi

Loser's R6 (QF):

- Zinoto/Shugo vs Player1/Kismet

Winner SemiFinals:

- MJG/Kain vs Coney/PinkFresh
- Zinoto/Shugo vs Seagull/BPow

Winner's Finals:

- Coney/PinkFresh vs Seagull/BPow

Loser's Finals:

- Seagull/BPow vs P1/Kismet

Grand Finals:

- Coney/PinkFresh vs Seagull/BPow

Singles:

Pools:

- Bonds vs Jet
- Bonds vs Darc
- Clowsui vs Darc
- Clowsui vs Bonds (Still Processing. Should be done by 11:20am)


That's what I have uploaded so far. Not a bad start. Be editing/uploading more throughout the day.
Thanks again to Zinoto for splicing all of doubles and a lot of pools. Very much appreciated as always :3

I will make you an actual video thread real quick. http://www.smashboards.com/showthread.php?p=14990708#post14990708

Huge thanks for Zinoto. You were a huge help :3

 

[Overswarm]

They are not the only variables present, if that was so it would be a bracket match. But since it was pools other sets matter. They are ranking the entire pool against each other, even within a tie-breaker situation.

Why would they do that? That's what gives you a tie in the first place!

Look at this pool and the results, 6 players:

Who wins here? Who is first seed and second seed?

Using win/loss ratio, the ranking is:

Player 2 - 7
Player 1 - 5
Player 4 - 3
Player 3 - neg5
Player 5 - neg5
Player 6 - neg5

Using set counts:

Player 1 - 5
Player 2 - 4
Player 4 - 3
Player 3 - 1
Player 5 - 1
Player 6 - 1


There's a different order because of a different methodology.

Now let's say top 4 get out.

Who wins the tie?

We look at head-to-head then everyone stays in the same order. We look at win/loss ratio and suddenly the last three are still tied, but the people above them switch order.

This is just to illustrate that win/loss ratio judges a completely different metric and should not be used unless you have to do so.

Head-to-head uses the same metric.

Now let's look at this one, same as above but slightly modified:

Suddenly Player 1 is in the lead with a score of 5 and 5 set wins.

player 2 and 4 however, have something interesting going on.

Looking at SETS, you see that Player 2 is ahead by a grand total of one game win, resulting in one set win to a total of 4 sets, win/loss of 4.

Player 4 has only three set wins, but also with a win/loss of 4. They tie in win/loss ratio but one is ahead in sets, and in a head to head player 2 beat player 4.

Depending on the metric you use, a different player can win in these scenarios with minimal, irrelevant changes. The order is very important and needs to remain consistent. Do you value sets, or game wins? You can't switch back and forth; you have to pick one.

If player 2 had lost one set to player 3 and had a score of 1-2 (resulting in a total of 3 set wins, 8-7), he'd have lost a grand total of one set more to a totally irrelevant player in that pool and tied with player 4. Through head-to-head, Player 2 moves on and Player 4 does not.

With win/loss ratio, Player 4's win against some random guy who went 5-7 in his pool in this hypothetical suddenly matters more than the win against the guy who was tied with him!



Look at top 4 making it out and compare players 5 and 6. They both are the exact same, except player 5 beat player 3 soundly and player 6 beat player 5 soundly.

In head-to-head, Player 6 cleanly wins. In win/loss they tie. Does player 6 getting one more game win against any of those opponents change anything? Does player 5 getting one more game win against any of those opponents change anything?

By the same metric we use to judge every other player, player 6 is the clear winner in that tie. To say otherwise is to say that your performance against players who are no longer relevant to the pool are more important than those that ARE relevant. That's what head to head circumvents.


But what about player 3?

If you count their sets, P3, P5, and P6 all have one set win.

Player 3 > Player 6
Player 5 > Player 3
Player 6 > Player 5

In this situation, you look at head to head and see that they are all equal. Should player 3 have beaten both of them and still been in the tie, he would move on in head to head because he beat out his competition.

If you use win/loss ratio in this scenario BEFORE head-to-head, player 3 could have beaten both of them and suddenly be eliminated due to other sets where he dropped a game. This happens on BOTH ends of the spectrum when using win/loss ratio.


When you compare win/loss ratio, player 3 loses right away; in this case that is acceptable because it is a three-way tie between them with each one set win and you don't have time for rematches.

After player 3 is eliminated, you're back to player 5 and 6. If you look at win/loss ratio they are tied but in head-to-head you have player 6 clearly winning. Is a scenario possible where player 3 could be eliminated due to win/loss ratio, then player 6 moving on due to head-to-head with player 5? Absolutely.

Is that same scenario possible where player 3 has set wins over both 5 and 6, but tied for overall sets, and is eliminated due to win/loss ratio only to see player 6 move on due to head-to-head?

Absolutely.

If you do head-to-head FIRST, that situation cannot occur.

With head-to-head first you are still judging by sets, not games, and you don't have situations where a three way tie can be broken down to a two way tie and have both fixed in a different way.

No tie breaking format will ever be infallible, but win/loss ratio uses a different metric and allows for tie breaking to occur in multiple formats within the same tie. Head-to-head does not.

 

[infiniteV115]

If somebody had a 'spotdodge habit' but they were getting away with it, then it either wasn't a habit (ie they were consciously deciding to do it, not by force of habit) or the opponent didn't catch on to this habit and wasn't able to punish.

 

[SoulPech]

I rested Ralph out of an airdodge. It was glorious!

 

[TheReflexWonder]

...Or the options allowed when the opponent is not spotdodging are so powerful that a player who defaults to either spotdodge or use those powerful options in a seemingly-random fashion prevents spotdodge from adequately being abused. :/

 

[What's The Point]

Head to head is not the same metric as counting sets. A head to head being a set does not mean they are the same. Sets first is counting your wins and loses, in sets. If anything it is closer to counting games wins and loses since you are counting an overall total.


Basically you want to say it's sets, set, win/lose, but the other side is win/lose, win/lose, set.

 

[@TKbreezy]

OR I'M JUST REALLY BAD AT THIS GAME REGARDLESS BUT I CAN STILL SPOT A HABIT EVEN IF I'M NOT PUNISHING IT!

OR MAYBE THE PLANETS WEREN'T ALIGNED!

#overanalyze

 

[Exceladon City]

@Ralph Cecil spotdodge topic

you definitely do spotdodge too much. LMAO a lot of times in game 1 and 2 I would just walk up to you and wait. and you'd be like OH ****! SPOTDODGE! and then i'd be like "Got'em"

I'd do the same thing to him on wifi.

Black people play too much alike...

 

[Overswarm]

Head to head is not the same metric as counting sets. A head to head being a set does not mean they are the same. Sets first is counting your wins and loses, in sets. If anything it is close to counting games eins and loses since you are counting an overall total.

The fact that head to head is a set is exactly what that means.

Sets are not counting wins and losses at all. It is counting sets won, just as head-to-head is.

In a pool of 2 people, head-to-head and sets won are the same thing. It is because it is the same metric.

That's no opinion. That's me showing you "hey look, same thing". They scale the exact same, but head-to-head is limited to the amount of sets shown in a tie.

In a pool of 4 people with one guy winning all matches 2-1 and the other winning 2-0 and losing one 1-2, you have one guy with a score of 6-3 and the other guy with a score of 5-2. Sets is one thing and win/loss is another.

When you are at the very top level and your primary means of breaking ties gives you different overall results it is because you are using a different metric to measure success. That is the most obvious "oh, I messed up" evidence you can get.

If you would do this in pools, why wouldn't you do this in bracket? For the same reason you don't do it in pools!

Win loss ratio is not the same thing because the effect of win/loss ratio scales with the amount of players in the pool. Head-to-head does not. It is a closed system that automatically deals with most of the issues that win/loss can run into.

The effect your character has on your potential to leave the pool should not be determined solely by the size of the pool when a tie occurs. The only time win/loss should be used is when you are forced to by a complete three-way tie and those should be rare.

 

[Raziek]

You people are goddamn ******** and need to listen to OS.

 

[@TKbreezy]

I've skipped every post either one of them have made

but I auto Co-Sign with OS because of crazy bead art deal.

#TeamOverswarm

 

[Vinylic.]

You know what's funny about Tako and Overswarm debating.

Both of their avatars looks like they're laughing.

 

[Raziek]

Overswarm
is laughing at you

Well gee, imagine that.

 

[Kel]

Brawl is a game of counterpicks and the character you choose alone can affect how many games are taken off of you in a set.

Lain used to win most of his tournament matches in 2009 by losing on Rainbow or Brinstar. It was just expected (Yes, Michigan, he eventually got better with MK and you're missing the point if you take this example further than just an example). Using this pools method, any IC main or other character that gets counterpicked really hard would be at a disadvantage simply because they excel on only one type of stage. In my example above no one cared that Lain would lose on Rainbow or Brinstar- it was expected. What mattered is that he still won the match/ tournament with SETS. W/L is not counting sets.

 

[clowsui]

@tiebreaker: we'll be thinking about this a lot more before the next one, rest assured

unrelated to this issue: the next tournament will have a few surprises in store. we'll be running it after apex, naturally. let's just say with what i intend to do with this tournament i HIGHLY doubt we will end up w less than 100 attendees.

 

[Carls493]

@tiebreaker: we'll be thinking about this a lot more before the next one, rest assured

unrelated to this issue: the next tournament will have a few surprises in store. we'll be running it after apex, naturally. let's just say with what i intend to do with this tournament i HIGHLY doubt we will end up w less than 100 attendees.

Legalizing MK with a twist?

 

[sneakytako]

Why do you keep comparing total sets against W/L, H2H? (Yeah I got tired of typing the word, deal with it). No one is arguing that total set count is the first thing that should be taken into account. But when you take H2H into account, you value one set win over another, which is stupid. And you're not ranking only two people when you rank a tie breaker, you are ranking two people against a pool of players.

The only reason W/L is taken into account is to describe how dominant/close their set wins/losses were. Now granted that it is not even a good way of measuring that since it doesn't look at stocks/percents, their performance in their other pools matches *could be deemed relevant in determing pools rank. If anyone has any ideas of illustrating pools performance better I'm all ears, but this way is the best we have atm. And putting pools performance over H2H isn't unthinkable.

 

[NH Cody]

Legalizing MK with a twist?

Nah clowsui wouldn't be that evil

 

[Raziek]

Why do you keep comparing total sets against W/L, H2H? (Yeah I got tired of typing the word, deal with it). No one is arguing that total set count is the first thing that should be taken into account. But when you take H2H into account, you value one set win over another, which is stupid. And you're not ranking only two people when you rank a tie breaker, you are ranking two people against a pool of players.

The only reason W/L is taken into account is to describe how dominant/close their set wins/losses were. Now granted that it is not even a good way of measuring that since it doesn't look at stocks/percents, their performance in their other pools matches *could be deemed relevant in determing pools rank. If anyone has any ideas of illustrating pools performance better I'm all ears, but this way is the best we have atm. And putting pools performance over H2H isn't unthinkable.

HOW ARE YOU NOT GETTING THIS?!

If two people won the SAME amount of sets.

But one beat the other....

It doesn't ****ing matter if the loser won every set 2-0, he's STILL A WORSE PLAYER THAN THE GUY WHO WON EVERY SET + H2H 2-1.

It's not hard.

That one game you dropped doesn't mean **** in bracket, and shouldn't mean **** in pools. THAT is why H2H goes before W/L ratio.

 

[DtJ Composer]

You should consider joining the debate hall

 

[@TKbreezy]

How about we work on running an 80 man tournament before shooting for the stars.

 

[sneakytako]

HOW ARE YOU NOT GETTING THIS?!

If two people won the SAME amount of sets.
.

Suddenly Player 1 is in the lead with a score of 5 and 5 set wins.

player 2 and 4 however, have something interesting going on.

Looking at SETS, you see that Player 2 is ahead by a grand total of one game win, resulting in one set win to a total of 4 sets, win/loss of 4.

Player 4 has only three set wins, but also with a win/loss of 4. They tie in win/loss ratio but one is ahead in sets, and in a head to head player 2 beat player 4.

Depending on the metric you use, a different player can win in these scenarios with minimal, irrelevant changes. The order is very important and needs to remain consistent. Do you value sets, or game wins? You can't switch back and forth; you have to pick one.

L2Read bro. He's saying player 2/4 should tie if you disregard total sets.

 

[Vinylic.]

9 - 5 = 4
8 - 4 = 4

So like, yeah. Pools is a scoreboard.

I have no idea what you guys are still talking about. A tl;dr would be nice.

 

[Raziek]

His point was to illustrate the problem with valuing game wins, if I'm not mistaken.

If you value games won/lost over set count, the wrong player could make it out.

Sets first, because he who wins sets advances in tournament.

Then H2H, because if both player won 3 sets, but Player A beat Player B, Player A is the one advancing in bracket.

Games won/lost should only apply in a 3-way tie where you have equal sets and circular head to head wins.

 

[sneakytako]

Raz is always mistaken.

Besides OS is doing wins/losses wrong.

 

[Raziek]

Because he used that as a counter-example to show you why games won/lost is a bad idea.

 

[Zeton]

Bleh....Was gonna do shoutouts but nvm.

 

[sneakytako]

No no I mean he's calculating wins vs losses wrong.

You first consider wins, then losses. In his example player 2 still has the advantage, since he has 9 wins over player 4's 8. At least that's how I was explained.

 

[Vinylic.]

How about people with the most wins proceeds to the bracket. Forget loss count.


The Problem would be solved and there's need to complicate things.

 

[Raziek]

This is just to illustrate that win/loss ratio judges a completely different metric and should not be used unless you have to do so.

Head-to-head uses the same metric.

If you do head-to-head FIRST, that situation cannot occur.

With head-to-head first you are still judging by sets, not games, and you don't have situations where a three way tie can be broken down to a two way tie and have both fixed in a different way.

No tie breaking format will ever be infallible, but win/loss ratio uses a different metric and allows for tie breaking to occur in multiple formats within the same tie. Head-to-head does not.

Read his 291 again, because this is what you're not getting.

 

[Zankoku]

How about people with the most wins proceeds to the bracket. Forget loss count.


The Problem would be solved and there's need to complicate things.

Ah, simplicity, a great solution! With your proposal, in a pool of three players where each player 2-1'd one person and lost 1-2 to the other, and top 1 advances, all three make it out because they all tied for the most wins.