|
This bracket is only 2nd to GSL code a. But its seriously going to be a grueling battle through these brackets
|
opterown
Australia54784 Posts
In tennis matches and anything similar like that, there's always seeding. I think IPL did right this time. It would suck to have, for example:
HerO vs JYP, Mvp vs Puma. HerO and Puma drop to LR, where Hero drops out.
This way, such matches will occur later in the tournament.
I assume since IGN has its own internal rankings system (see: power rank), then that's what they used to seed players?
While the RO128 matches can be lopsided, the Ro64 will be much more competitive! And since these aren't really broadcast, we're not missing much...
|
On April 05 2012 00:55 VirgilSC2 wrote:Show nested quote +On April 05 2012 00:50 m0ck wrote:On April 05 2012 00:43 VirgilSC2 wrote:On April 05 2012 00:26 Cascade wrote:Wow, kindof stacked, yes! :o On April 04 2012 21:55 Bumblebee wrote:
I don't really have a refined opinion on team kills in such brackets. It's sort of just tough luck, but I wonder if there can't be done something to prevent it at least for a while or not "that many". If you look over the bracket, we - Team Liquid - kill eachother very early all over the bracket. It's just not worthwhile business for us to send so many players to such events if they're going to eliminate eachother that early. I'm not saying we can prevent teamkills all the way until the bracket finals - or we almost can, but I don't expect it. I'm just simply thinking if there isn't a fair way to draft so that we won't have 2-3 teamkills in the third round?
What are your opinions out there? Do you think it'd be fair to put what team people are on into such an equation? It would be nice in a way, if you put it last, after you spread out all your seeds and whatever ranking the tourney is using. Then you could imagine that the random placing that is left got the extra constraint of trying to not make team kills. Problem however, is that it would then be beneficial to be on a team with many other good players. For example, if Hero goes to a pretty small tourney (as favourite to win) but that doesn't have much previous seeds or ranking, most players would be very happy to be placed far away from hero, and would thus get an (imo unfair) advantage by being on liquid. And also the opposite, if another team with a lot of bad players went to the tourney, they would run a higher risk of being matched with a strong opponent. The bad players would be better of being teamless. Imo, we don't need to make the strong teams even stronger, so I do not think this should be implemented. It sucks for liquid, I feel you, but I don't think it would work in the long run to take team into account when making brackets. I would rather them take what team the player is on into account over the arbitrary "Well, this guy is good so we should start him off vs someone bad" ranking IPL used here. Just a little interesting statistic; I counted up the Ro128 Matches and using a fairly loose judgement setting on "Will this be a close match or not?" I got: 53 ridiculously lopsided matches. 11 conceivably close matches. I was also pretty generous with what I considered "conceivably close" The matches I chose: + Show Spoiler +Ridiculously Lopsided Matches
Leenock > Zephyr
Golden > Rowrin
DIMAGA > NasD
PartinG > Auderned
KiWiKaKi > IcedragonCF
Virus > Mike
Moon > MafiaCheese
Oz > Kukaracha
DdoRo > Light
GanZi > IMBAoracle
Dragon > RxNinja
Classic > Stillborn
HayprO > Woody (Already Forfeit-Win)
HerO > DJTyrant
Ret > Wind
Sleep > Blink
JYP > BearJewSlava
qxc > Astral
Sound > Garthmeister
KawaiiRice > TenryuRH
viOLet > mMmSiezE
Annyeong > zOCHL
BlinG > twttrCeaser
Curious > Truman
TT1 > InsaneRabbi
July > Sruff
MVP > Rumcake
SeleCT > LookitsIan
SaSe > mockturtle
Squirtle > Demonlord
ToD > EROhime
GhostKing > CryMoar
ChAnCe > silvaWare
Creator > CypherX
AcE > twttrNvictus
State > Wailord
Heart > Exiledz
Drewbie > PanzerRH
Bboongbboong > Papabear
Axslav > wabachaw
Jjakji > Heathen
Sheth > RxSnoWy
Crank > Prplppleatr
HuK > GoldenMule
KiLLeR > OrbraY
ViBE > Epyon
TaeJa > Detour
Killer > Dalguno
TLO > Samuraipanda
Maru > FallenAngel
CatZ > Mettaur
Zenio > Duwie
Illusion > JPEG
Conceivably Close Matches
HasHe <> NoNy
Heavens <> Flo
Terious <> Scarlett
Caliber <> Mystik
Jinro <> Spades
DeMusliM <> MoOk
Machine <> Hawk
Binski <> SyNc
ThisIsJimmy <> STX
Panther <> Ryung
Perfect <> Future
The purpose of a double-bracket qualifier is to decide the best players. Seeding players going into those brackets lessens the impact of having a good versus a bad draw. Correcting for teams makes the qualifier less fair. Which player do you move? And to which bracket? Fairness should in my opinion go before teams. Team affiliation is relatively unimportant in a single player sport. No reason to artificially heighten their importance. So why not just seed everyone randomly? Wouldn't that be the "fairest" option available?
If you assume everyone is evenly matched, then yes. It makes everything random, but it also makes your chance to qualify almost entirely luck-dependent.
If you do IPL-style seeding, you don't get unbelievably hard or unbelievably easy brackets, you get relatively balanced ones where everyone who deserves a chance of making it into pool play will be able to do it, if he plays well.
|
On April 05 2012 00:55 VirgilSC2 wrote:Show nested quote +On April 05 2012 00:50 m0ck wrote:On April 05 2012 00:43 VirgilSC2 wrote:On April 05 2012 00:26 Cascade wrote:Wow, kindof stacked, yes! :o On April 04 2012 21:55 Bumblebee wrote:
I don't really have a refined opinion on team kills in such brackets. It's sort of just tough luck, but I wonder if there can't be done something to prevent it at least for a while or not "that many". If you look over the bracket, we - Team Liquid - kill eachother very early all over the bracket. It's just not worthwhile business for us to send so many players to such events if they're going to eliminate eachother that early. I'm not saying we can prevent teamkills all the way until the bracket finals - or we almost can, but I don't expect it. I'm just simply thinking if there isn't a fair way to draft so that we won't have 2-3 teamkills in the third round?
What are your opinions out there? Do you think it'd be fair to put what team people are on into such an equation? It would be nice in a way, if you put it last, after you spread out all your seeds and whatever ranking the tourney is using. Then you could imagine that the random placing that is left got the extra constraint of trying to not make team kills. Problem however, is that it would then be beneficial to be on a team with many other good players. For example, if Hero goes to a pretty small tourney (as favourite to win) but that doesn't have much previous seeds or ranking, most players would be very happy to be placed far away from hero, and would thus get an (imo unfair) advantage by being on liquid. And also the opposite, if another team with a lot of bad players went to the tourney, they would run a higher risk of being matched with a strong opponent. The bad players would be better of being teamless. Imo, we don't need to make the strong teams even stronger, so I do not think this should be implemented. It sucks for liquid, I feel you, but I don't think it would work in the long run to take team into account when making brackets. I would rather them take what team the player is on into account over the arbitrary "Well, this guy is good so we should start him off vs someone bad" ranking IPL used here. Just a little interesting statistic; I counted up the Ro128 Matches and using a fairly loose judgement setting on "Will this be a close match or not?" I got: 53 ridiculously lopsided matches. 11 conceivably close matches. I was also pretty generous with what I considered "conceivably close" The matches I chose: + Show Spoiler +Ridiculously Lopsided Matches
Leenock > Zephyr
Golden > Rowrin
DIMAGA > NasD
PartinG > Auderned
KiWiKaKi > IcedragonCF
Virus > Mike
Moon > MafiaCheese
Oz > Kukaracha
DdoRo > Light
GanZi > IMBAoracle
Dragon > RxNinja
Classic > Stillborn
HayprO > Woody (Already Forfeit-Win)
HerO > DJTyrant
Ret > Wind
Sleep > Blink
JYP > BearJewSlava
qxc > Astral
Sound > Garthmeister
KawaiiRice > TenryuRH
viOLet > mMmSiezE
Annyeong > zOCHL
BlinG > twttrCeaser
Curious > Truman
TT1 > InsaneRabbi
July > Sruff
MVP > Rumcake
SeleCT > LookitsIan
SaSe > mockturtle
Squirtle > Demonlord
ToD > EROhime
GhostKing > CryMoar
ChAnCe > silvaWare
Creator > CypherX
AcE > twttrNvictus
State > Wailord
Heart > Exiledz
Drewbie > PanzerRH
Bboongbboong > Papabear
Axslav > wabachaw
Jjakji > Heathen
Sheth > RxSnoWy
Crank > Prplppleatr
HuK > GoldenMule
KiLLeR > OrbraY
ViBE > Epyon
TaeJa > Detour
Killer > Dalguno
TLO > Samuraipanda
Maru > FallenAngel
CatZ > Mettaur
Zenio > Duwie
Illusion > JPEG
Conceivably Close Matches
HasHe <> NoNy
Heavens <> Flo
Terious <> Scarlett
Caliber <> Mystik
Jinro <> Spades
DeMusliM <> MoOk
Machine <> Hawk
Binski <> SyNc
ThisIsJimmy <> STX
Panther <> Ryung
Perfect <> Future
The purpose of a double-bracket qualifier is to decide the best players. Seeding players going into those brackets lessens the impact of having a good versus a bad draw. Correcting for teams makes the qualifier less fair. Which player do you move? And to which bracket? Fairness should in my opinion go before teams. Team affiliation is relatively unimportant in a single player sport. No reason to artificially heighten their importance. So why not just seed everyone randomly? Wouldn't that be the "fairest" option available?
Well, I would say that for the purpose of deciding the best players, we have a fairly good idea about how to grade the players. There are no objective measures, and it's by no means perfect, but it's still possible to decide reasonably well among the top 40-50 of the bracket or so. I think it's worth the disfavor to the lower seeds (seed 50 and 80 might be fairly arbitrary) to ensure that luck-of-the-draw doesn't knock out the very best players because they meet each other too early. But it's a judgement call.
In any case, that is beside the point of changing brackets to avoid the possibility of team kills for as long as possible.
|
safe bet would be mkp taking home the massive 40K cheque ... no naniwa?
hope MC and kiwi have a good showing though :-)
|
United States5162 Posts
On April 05 2012 00:55 VirgilSC2 wrote:Show nested quote +On April 05 2012 00:50 m0ck wrote:On April 05 2012 00:43 VirgilSC2 wrote:On April 05 2012 00:26 Cascade wrote:Wow, kindof stacked, yes! :o On April 04 2012 21:55 Bumblebee wrote:
I don't really have a refined opinion on team kills in such brackets. It's sort of just tough luck, but I wonder if there can't be done something to prevent it at least for a while or not "that many". If you look over the bracket, we - Team Liquid - kill eachother very early all over the bracket. It's just not worthwhile business for us to send so many players to such events if they're going to eliminate eachother that early. I'm not saying we can prevent teamkills all the way until the bracket finals - or we almost can, but I don't expect it. I'm just simply thinking if there isn't a fair way to draft so that we won't have 2-3 teamkills in the third round?
What are your opinions out there? Do you think it'd be fair to put what team people are on into such an equation? It would be nice in a way, if you put it last, after you spread out all your seeds and whatever ranking the tourney is using. Then you could imagine that the random placing that is left got the extra constraint of trying to not make team kills. Problem however, is that it would then be beneficial to be on a team with many other good players. For example, if Hero goes to a pretty small tourney (as favourite to win) but that doesn't have much previous seeds or ranking, most players would be very happy to be placed far away from hero, and would thus get an (imo unfair) advantage by being on liquid. And also the opposite, if another team with a lot of bad players went to the tourney, they would run a higher risk of being matched with a strong opponent. The bad players would be better of being teamless. Imo, we don't need to make the strong teams even stronger, so I do not think this should be implemented. It sucks for liquid, I feel you, but I don't think it would work in the long run to take team into account when making brackets. I would rather them take what team the player is on into account over the arbitrary "Well, this guy is good so we should start him off vs someone bad" ranking IPL used here. Just a little interesting statistic; I counted up the Ro128 Matches and using a fairly loose judgement setting on "Will this be a close match or not?" I got: 53 ridiculously lopsided matches. 11 conceivably close matches. I was also pretty generous with what I considered "conceivably close" The matches I chose: + Show Spoiler +Ridiculously Lopsided Matches
Leenock > Zephyr
Golden > Rowrin
DIMAGA > NasD
PartinG > Auderned
KiWiKaKi > IcedragonCF
Virus > Mike
Moon > MafiaCheese
Oz > Kukaracha
DdoRo > Light
GanZi > IMBAoracle
Dragon > RxNinja
Classic > Stillborn
HayprO > Woody (Already Forfeit-Win)
HerO > DJTyrant
Ret > Wind
Sleep > Blink
JYP > BearJewSlava
qxc > Astral
Sound > Garthmeister
KawaiiRice > TenryuRH
viOLet > mMmSiezE
Annyeong > zOCHL
BlinG > twttrCeaser
Curious > Truman
TT1 > InsaneRabbi
July > Sruff
MVP > Rumcake
SeleCT > LookitsIan
SaSe > mockturtle
Squirtle > Demonlord
ToD > EROhime
GhostKing > CryMoar
ChAnCe > silvaWare
Creator > CypherX
AcE > twttrNvictus
State > Wailord
Heart > Exiledz
Drewbie > PanzerRH
Bboongbboong > Papabear
Axslav > wabachaw
Jjakji > Heathen
Sheth > RxSnoWy
Crank > Prplppleatr
HuK > GoldenMule
KiLLeR > OrbraY
ViBE > Epyon
TaeJa > Detour
Killer > Dalguno
TLO > Samuraipanda
Maru > FallenAngel
CatZ > Mettaur
Zenio > Duwie
Illusion > JPEG
Conceivably Close Matches
HasHe <> NoNy
Heavens <> Flo
Terious <> Scarlett
Caliber <> Mystik
Jinro <> Spades
DeMusliM <> MoOk
Machine <> Hawk
Binski <> SyNc
ThisIsJimmy <> STX
Panther <> Ryung
Perfect <> Future
The purpose of a double-bracket qualifier is to decide the best players. Seeding players going into those brackets lessens the impact of having a good versus a bad draw. Correcting for teams makes the qualifier less fair. Which player do you move? And to which bracket? Fairness should in my opinion go before teams. Team affiliation is relatively unimportant in a single player sport. No reason to artificially heighten their importance. So why not just seed everyone randomly? Wouldn't that be the "fairest" option available? Show nested quote +On April 05 2012 00:49 Purple Haze wrote:
You must get pretty upset whenever you watch...any individual sport in the world. It's not exactly a new concept to save the closer matches for later in the tournament. I'd rather get the massacres out of the way early, and enjoy the upsets.
Are the losers brackets each linked directly to their respective winners bracket, so there's no mixing of players between brackets?
In tournaments where they "save the closer matches for later", like say NCAA's March Madness Basketball tournament. (Not an individual sport, but frankly I don't watch many individual sports) The seeds in that tournament are based off previous performance in the league.....here the rankings seem* (as I'm not aware what metric was used) to be chosen based one one person, or a group of people's idea on who is good and who is bad.
March Madness is determined the same exact way. There is no system that decides a 1 seed meets x criteria while a 2 seed only meets y criteria. They look at how everyone played that year, and make a subjective decision on which teams like think are better than others.
|
On April 05 2012 00:59 opterown wrote:
In tennis matches and anything similar like that, there's always seeding. I think IPL did right this time. It would suck to have, for example:
HerO vs JYP, Mvp vs Puma. HerO and Puma drop to LR, where Hero drops out.
This way, such matches will occur later in the tournament.
I assume since IGN has its own internal rankings system (see:power rank), then that's what they used to seed players?
While the RO128 matches can be lopsided, the Ro64 will be much more competitive! And since these aren't really broadcast, we're not missing much...
I think the Power Rank is a terrible example. It usually only encompasses the Top 10 current players.
I think they MAY (I haven't looked, but it would make sense) have used North American TLPD ELOs, which is a pretty fair method of seeding, if you don't want to randomize the bracket for some reason.
|
Wow that's a hell of a stacked open bracket o.O
|
I predict full koreans from RO8 onwards.
|
opterown
Australia54784 Posts
On April 05 2012 01:07 VirgilSC2 wrote:Show nested quote +On April 05 2012 00:59 opterown wrote:
In tennis matches and anything similar like that, there's always seeding. I think IPL did right this time. It would suck to have, for example:
HerO vs JYP, Mvp vs Puma. HerO and Puma drop to LR, where Hero drops out.
This way, such matches will occur later in the tournament.
I assume since IGN has its own internal rankings system (see:power rank), then that's what they used to seed players?
While the RO128 matches can be lopsided, the Ro64 will be much more competitive! And since these aren't really broadcast, we're not missing much... I think the Power Rank is a terrible example. It usually only encompasses the Top 10 current players. I think they MAY (I haven't looked, but it would make sense) have used North American TLPD ELOs, which is a pretty fair method of seeding, if you don't want to randomize the bracket for some reason.
Haha I agree power rank itself is a bad example, but if you click around, there's a component that has computer-based rankings that they get from somewhere. That might be the source of their seeding. Anyhow, their seeding is fairly good, nothing toooo off that I can see?
I don't think they used international TLPD since it doesn't fit very well with the current rankings etc. But that's probably a decently fair method.
|
On April 05 2012 01:06 Myles wrote:
March Madness is determined the same exact way. There is no system that decides a 1 seed meets x criteria while a 2 seed only meets y criteria. They look at how everyone played that year, and make a subjective decision on which teams like think are better than others.
March Madness determines it's rankings based on current NCAA results where the top teams from a pool of all Division 1 teams are selected to play.
This is an open qualifier.
Two completely different scenarios.
|
Seeding is used in basically every sport. I don't see the problem.
|
United States5162 Posts
On April 05 2012 01:10 VirgilSC2 wrote:Show nested quote +On April 05 2012 01:06 Myles wrote:
March Madness is determined the same exact way. There is no system that decides a 1 seed meets x criteria while a 2 seed only meets y criteria. They look at how everyone played that year, and make a subjective decision on which teams like think are better than others.
March Madness determines it's rankings based on current NCAA results where the top teams from a pool of all Division 1 teams are selected to play. This is an open qualifier. Two completely different scenarios.
Very much so, but the way the select the rankings are very much the same - they look at past performance and make a subjective decision.
|
On April 05 2012 01:10 opterown wrote:Show nested quote +On April 05 2012 01:07 VirgilSC2 wrote:On April 05 2012 00:59 opterown wrote:
In tennis matches and anything similar like that, there's always seeding. I think IPL did right this time. It would suck to have, for example:
HerO vs JYP, Mvp vs Puma. HerO and Puma drop to LR, where Hero drops out.
This way, such matches will occur later in the tournament.
I assume since IGN has its own internal rankings system (see:power rank), then that's what they used to seed players?
While the RO128 matches can be lopsided, the Ro64 will be much more competitive! And since these aren't really broadcast, we're not missing much... I think the Power Rank is a terrible example. It usually only encompasses the Top 10 current players. I think they MAY (I haven't looked, but it would make sense) have used North American TLPD ELOs, which is a pretty fair method of seeding, if you don't want to randomize the bracket for some reason. Haha I agree power rank itself is a bad example, but if you click around, there's a component that has computer-based rankings that they get from somewhere. That might be the source of their seeding. Anyhow, their seeding is fairly good, nothing toooo off that I can see? I don't think they used international TLPD since it doesn't fit very well with the current rankings etc. But that's probably a decently fair method.
I'm not saying anything is wrong with the seeding, I just think something is wrong with seeding players and then not disclosing HOW they were seeded.
On April 05 2012 01:12 Myles wrote:Show nested quote +On April 05 2012 01:10 VirgilSC2 wrote:On April 05 2012 01:06 Myles wrote:
March Madness is determined the same exact way. There is no system that decides a 1 seed meets x criteria while a 2 seed only meets y criteria. They look at how everyone played that year, and make a subjective decision on which teams like think are better than others.
March Madness determines it's rankings based on current NCAA results where the top teams from a pool of all Division 1 teams are selected to play. This is an open qualifier. Two completely different scenarios. Very much so, but the way the select the rankings are very much the same - they look at past performance and make a subjective decision.
It's not a subjective decision. It's based off cold, hard numbers from Win/Loss ratios within the league.
|
On April 04 2012 20:48 ceaRshaf wrote:Show nested quote +On April 04 2012 20:32 Moneyscoop wrote:
Ahh poor HuK...i hope he'll make it through the bracket! He used his luck mana on MLG. Now: "Not enough mana to cast spell."
He's out of mana, he needs a po-
+ Show Spoiler +
|
Lol I get to play  HerO...this should be fun 
|
On April 05 2012 01:16 DJTyrant wrote:Lol I get to play HerO...this should be fun
haha, good luck!! ^^
|
United States5162 Posts
On April 05 2012 01:15 VirgilSC2 wrote:Show nested quote +On April 05 2012 01:10 opterown wrote:On April 05 2012 01:07 VirgilSC2 wrote:On April 05 2012 00:59 opterown wrote:
In tennis matches and anything similar like that, there's always seeding. I think IPL did right this time. It would suck to have, for example:
HerO vs JYP, Mvp vs Puma. HerO and Puma drop to LR, where Hero drops out.
This way, such matches will occur later in the tournament.
I assume since IGN has its own internal rankings system (see:power rank), then that's what they used to seed players?
While the RO128 matches can be lopsided, the Ro64 will be much more competitive! And since these aren't really broadcast, we're not missing much... I think the Power Rank is a terrible example. It usually only encompasses the Top 10 current players. I think they MAY (I haven't looked, but it would make sense) have used North American TLPD ELOs, which is a pretty fair method of seeding, if you don't want to randomize the bracket for some reason. Haha I agree power rank itself is a bad example, but if you click around, there's a component that has computer-based rankings that they get from somewhere. That might be the source of their seeding. Anyhow, their seeding is fairly good, nothing toooo off that I can see? I don't think they used international TLPD since it doesn't fit very well with the current rankings etc. But that's probably a decently fair method. I'm not saying anything is wrong with the seeding, I just think something is wrong with seeding players and then not disclosing HOW they were seeded. Show nested quote +On April 05 2012 01:12 Myles wrote:On April 05 2012 01:10 VirgilSC2 wrote:On April 05 2012 01:06 Myles wrote:
March Madness is determined the same exact way. There is no system that decides a 1 seed meets x criteria while a 2 seed only meets y criteria. They look at how everyone played that year, and make a subjective decision on which teams like think are better than others.
March Madness determines it's rankings based on current NCAA results where the top teams from a pool of all Division 1 teams are selected to play. This is an open qualifier. Two completely different scenarios. Very much so, but the way the select the rankings are very much the same - they look at past performance and make a subjective decision. It's not a subjective decision. It's based off cold, hard numbers from Win/Loss ratios within the league.
No it's not. They don't look at the W/L of all the teams, put them in order, and then say 'Ok, there's the rankings'. They might look at a 30-9 team and give them a higher seed then a 34-4 team because of a variety of reasons like RPI, strength of schedule, injuries, ect, all which are subjective.
|
opterown
Australia54784 Posts
On April 05 2012 01:16 DJTyrant wrote:Lol I get to play HerO...this should be fun
PvP upset man, anything can happen. You should just cram cram cram PvP and see if you can knock him to the lower bracket :p
|
On April 05 2012 01:16 DJTyrant wrote:Lol I get to play HerO...this should be fun
entire community will hate you forever if you cheese HerO to the lower brackets....but it will be hilarious!
|
|
|
|
|
|
EPT
