Drunken_masterToday's birthdays
KotD4 | Questions & Answers
-
Thread starter
Chrazini
- Start date
IRC_tomate
Mouttie
buyanjavyn
There's nothing of that sort currently, and I'm not sure we'll have that. Either way, it will be slightly misleading anyway, as the ultimate list of players is determined by who checks in on the day.
I'm not sure what you're asking for. Players are seeded based on their highest-achieved 1v1 RM rating. As mentioned above, there is no list of the players currently.
chrummiNow that I read it again, that would make sense. I'll make a post explaining this in the coming days, thank you.
DracKeN1) Handbook says that qualification stage starts 13 GMT on 1 page and 14 GMT on other page. Its the case for both days.
2) Spec delay is set as 2 mins. Are the spec/streamcheaters from the past year allowed in the tournament? Mr. Bean as an example? If so, can the spec delay be increased to 15 mins against those? Thanks!
2) Spec delay is set as 2 mins. Are the spec/streamcheaters from the past year allowed in the tournament? Mr. Bean as an example? If so, can the spec delay be increased to 15 mins against those? Thanks!
FestivusHas there been any explanation posted regarding the seeding of the 16 invited players yet?
I'm looking at them on liquipedia and there are several ties where the tiebreaker appears to be seemingly random which is a shame as there would be some good options.
Also I wonder if anyone could share some insight on the decision making of counting all tournaments for all players. What I mean by this is that if a player couldn't participate in a tournament for any reason the score seems to still count. I'm not saying this is definitively wrong, but for example in Yo's case, not being able to compete in RBW5 is also hurting him in this tournament which is unfortunate.
If the seeding was based on Points Per Tournament instead of Total Points, Yo would get an N/A for RBW5, be able to drop a 20, and have accrued a total of 130 points in 4 tournaments for an average of 32.5 per tournament.
Hera is another interesting case this might affect, and I understand this is where things could get murkier. Should he be given a N/A for TOC. It's his fault he didn't participate, but still forcing him to drop that score for a tournament he didn't participate in is not really reflective of skill and thus not (strictly in my opinion) a good option for seeding. He would then also be able to drop a 20 instead of a N/A and average 32.5 pts per tournament, same as yo.
Viper meanwhile has participated in all 6, would drop a 20 and total 150 points in 5 tournaments for an average of 30 and should be ranked lower.
If all three of them (viper, yo, hera) are kept at the tied score of 150, a tiebreaker of most recent best score would go Viper, Yo, Hera. A tiebreaker of best placements would go (all 3 have one first, only hera has a second) hera, and then yo/viper need another tiebreaker.
It does seem like Liereyy is well deserving of the 1 seed as once he drops HC4 (which with random seeding is dubious to include in the seeding to begin with) his worst performance is a 2nd place. After that things get messy really quick, so I'm just hoping there is clarity on the exact decisions for the rest of the ranks. Points per tournament seems like a more fair way to rank in my opinion, but I think the admins here could easily justify not wanting to go down that path, I guess I'm just interested in if it was considered. I just used certain players as some examples but I'm sure others would be affected to, such as ACCM.
I'm looking at them on liquipedia and there are several ties where the tiebreaker appears to be seemingly random which is a shame as there would be some good options.
Also I wonder if anyone could share some insight on the decision making of counting all tournaments for all players. What I mean by this is that if a player couldn't participate in a tournament for any reason the score seems to still count. I'm not saying this is definitively wrong, but for example in Yo's case, not being able to compete in RBW5 is also hurting him in this tournament which is unfortunate.
If the seeding was based on Points Per Tournament instead of Total Points, Yo would get an N/A for RBW5, be able to drop a 20, and have accrued a total of 130 points in 4 tournaments for an average of 32.5 per tournament.
Hera is another interesting case this might affect, and I understand this is where things could get murkier. Should he be given a N/A for TOC. It's his fault he didn't participate, but still forcing him to drop that score for a tournament he didn't participate in is not really reflective of skill and thus not (strictly in my opinion) a good option for seeding. He would then also be able to drop a 20 instead of a N/A and average 32.5 pts per tournament, same as yo.
Viper meanwhile has participated in all 6, would drop a 20 and total 150 points in 5 tournaments for an average of 30 and should be ranked lower.
If all three of them (viper, yo, hera) are kept at the tied score of 150, a tiebreaker of most recent best score would go Viper, Yo, Hera. A tiebreaker of best placements would go (all 3 have one first, only hera has a second) hera, and then yo/viper need another tiebreaker.
It does seem like Liereyy is well deserving of the 1 seed as once he drops HC4 (which with random seeding is dubious to include in the seeding to begin with) his worst performance is a 2nd place. After that things get messy really quick, so I'm just hoping there is clarity on the exact decisions for the rest of the ranks. Points per tournament seems like a more fair way to rank in my opinion, but I think the admins here could easily justify not wanting to go down that path, I guess I'm just interested in if it was considered. I just used certain players as some examples but I'm sure others would be affected to, such as ACCM.
SeahorsegallopI am pretty sure I saw MembTV explain the seeding that the average of 5 of the 6 tourney results was used to account for Yo not been able to take part in RBW 5. (This method would also benefit Hera, which is lucky for him, as I believe if you don't sign up / in on time, tough).
Interesting, I'm pretty sure the lowest scores have been dropped in the past, so didn't expect that to be the solution to the Yo/ACCM RBW5 situation etc. It doesn't seem like the best way in my opinion, as it also allows people who actually performed bad to drop a score, while others are forced to drop a score they may have done well in. Points per tournament would still be my favored approach, but as I've acknowledged its a big grey area which is why I'm hoping for a post explaining the decision making that went into these rankings. Thoughts on including HC4 at same weight as other tournaments, or at all. Reasons behind the tiebreakers etc. This stuff often times just seems not very well thought out to me, and I doubt that is the actual case, but its hard to know when its not very clear and different tournament to tournament.
In terms of the seeding for main event invites, you have a reference on liquidpedia it's self, if you want more detail in the though process it was explained several times by memb on twitch.
The only reason that 5 or 6 are used is for fairness of Yo/ACCM because they weren't able to attend redbull not of their own accord, Hers for example missed TOC by his own mistake but his lucky in this case only highest 5 count.
As for tie breakers in the seeding it's done based on higher results 1st place and such.
As for seeding in any of the Quals or rest of the tournament that's pretty clear in the handbook.
KoTD organizers have been very transparent with choices and if you look back at memb vods/YouTube videos you can find even more in depth thought processes on how well thought out the choices were.
Thanks for explanation. I've never doubted that the organizers have discussed/are transparent but I personally haven't caught it discussed on twitch and it feels weird to have the qualifier seeding discussed in the handbook clearly and not the main event.
I still don't understand the tiebreakers. Take Viper/Hera/Yo, all have one first place. Hera has one second place. Yo and Viper have 2 3rd/4ths. Based on your description, I'd have expected Hera to win the tiebreaker, as he has highest results after the 1st place each of them has, instead he is in the middle.
Jordan/MBL tie. Both have one 2nd place. Jordan has 2 3rd/4ths to MBLs 1 and wins the tiebreaker. Fits with what you said, not with Yo/Viper/Hera.
Tatoh/Vinch/ACCM tie. All have one 30pt. ACCM is only player with 2 20pt. ACCM is last between the 3 in seeding.
Sorry if I am not understanding something, but I don't see how these were done.
RorarimboI was wondering if I should go into long explanation of the system used right now or after the event. But as everything KotD related things are open and feedback is always welcome.
The system aims to provide fairly simple way to rate top 16 players. The system uses top 16 results from 6 highest prized RM and EW 1v1 tournaments during the last 12 months (KotD3, RBW3, HC4, RBW4, TOC, RBW5). Last year before KotD3 there were no notable absences to such events and the points given were simply accumulated similar to how it is done in tennis.
The difference is that in tennis there are differently ranked tournaments which reward different points starting from Grand slams and having ATP 1000, 500 and 250 levels. Here we have much more simple way of giving the tournaments the same weight. The reasoning is its hard to decide how relevant a tournament result would be to Kotd4 at this stage. Kotd3 is the same settings but the rest of the events are most recent. HC4 is RM but it has the players anonymity setting, while TOC uses a lot of closed, water and no TC start maps. RBW events are different mode but they generally provide a lot open land map games. DM tournaments on the other hand have settings considered way too different. It is a case where you draw the line. In future depending on how competitions develop a system of ranking the tournaments might be added. In this case it was decided that these tournaments have provided the highest level of competition any player would like to participate and win.
After Hera missed TOC and especially after Yo and ACCM could not make it to Germany for RBW5, it became evident that perhaps there should be something to account for players not participating in different events especially as sometimes the reasons are out of their control.
The first thought was to use average scores or “points per tournament”. This is good option at the top but the more you go down the list players start missing top 16 entrances. Then you do not know did they not sign up for certain events, if they did perhaps they pulled out, or simply they just did not perform well and failed to qualify. In the system only positive performances are counted being top16 or better. If we use only top 16 scores one quarterfinal and no other top16 entrances will be stronger than a player with two quarterfinal and one round of 16 finish for example.
Then perhaps we should use a minimum of top 16 entrances (how many), or start to use negative scores, or completely different scoring to accommodate levels beyond top 16. In that case there are several problems. One it takes big research to check who signed up for different qualifications and how they did. But may be they didnt sign up at all. Sometimes that information may become hard to access. Also the different tournament have different formats. Kotd3 had normal rounds of 64 and 32. The other tournaments have qualifications with different format some of them several in number. A level before top16 may have different number of players in the different formats. Also the same player may have different placement in different qualifications for the same tournament.
You can see why this option was dropped. If we are to use average performance value, we would need the tournaments to be run with the same format or formats that can accommodate the same levels and number of players. We are in a way lucky that all major events we have consist of defined top 16 stage (top 14 for RBW5). Plus, we would need to have much more knowledge of players who didn’t take part in events (not just ACCM and Yo missing RBW5). This is already too much work and practically means building completely different system while knowing it can be hardly consistent between tournaments.
That is why the decision taken was to just take one score away from each player. It is not perfect because players loose their worse results which is preferable than not having chance to compete. But yet it is better than having some players accumulate points from 6 events and others from 5.
When it comes to tiebreakers between players with the same score last year the tiebreaker was player ladder rating. This time it is the score players have in the last tournament. So RBW5 before TOC, before RBW4 etc. In this case Viper has 50 in RBW5, Hera has 20 and Yo has 0. It is an argument to be made that Yo last performance is 50 at TOC. But in this case it is scores from last, previous to last etc tournaments not accounting if players participated and what was the reason not to (their own fault, lack of motivation, or other reasons that prevented them).It is similar logic to not include the average scores idea.
I am aware this system as any other may be tweaked. There are other things I did not touch upon here that were considered but at the end of the day it comes to where you decide to draw the line and take certain decision. The system tries to account for the reality of tournament performances but there are certain objective limitations. Its main purpose is to avoid subjective seeding where admins decide who should be seeded where and who should be invited. Both taking tournament performances and ladder ranking aims at avoiding that. This argument is also the reason why when the method is decided and the players are ranked we avoid tweaking any further. Because trying to make it “more perfect” you now give one player advantage over the other. At this stage tweaks and changes will come for the next years tournament.
Perhaps the response would have ppl not agree with all decisions and logic but this is the thought process.
The system aims to provide fairly simple way to rate top 16 players. The system uses top 16 results from 6 highest prized RM and EW 1v1 tournaments during the last 12 months (KotD3, RBW3, HC4, RBW4, TOC, RBW5). Last year before KotD3 there were no notable absences to such events and the points given were simply accumulated similar to how it is done in tennis.
The difference is that in tennis there are differently ranked tournaments which reward different points starting from Grand slams and having ATP 1000, 500 and 250 levels. Here we have much more simple way of giving the tournaments the same weight. The reasoning is its hard to decide how relevant a tournament result would be to Kotd4 at this stage. Kotd3 is the same settings but the rest of the events are most recent. HC4 is RM but it has the players anonymity setting, while TOC uses a lot of closed, water and no TC start maps. RBW events are different mode but they generally provide a lot open land map games. DM tournaments on the other hand have settings considered way too different. It is a case where you draw the line. In future depending on how competitions develop a system of ranking the tournaments might be added. In this case it was decided that these tournaments have provided the highest level of competition any player would like to participate and win.
After Hera missed TOC and especially after Yo and ACCM could not make it to Germany for RBW5, it became evident that perhaps there should be something to account for players not participating in different events especially as sometimes the reasons are out of their control.
The first thought was to use average scores or “points per tournament”. This is good option at the top but the more you go down the list players start missing top 16 entrances. Then you do not know did they not sign up for certain events, if they did perhaps they pulled out, or simply they just did not perform well and failed to qualify. In the system only positive performances are counted being top16 or better. If we use only top 16 scores one quarterfinal and no other top16 entrances will be stronger than a player with two quarterfinal and one round of 16 finish for example.
Then perhaps we should use a minimum of top 16 entrances (how many), or start to use negative scores, or completely different scoring to accommodate levels beyond top 16. In that case there are several problems. One it takes big research to check who signed up for different qualifications and how they did. But may be they didnt sign up at all. Sometimes that information may become hard to access. Also the different tournament have different formats. Kotd3 had normal rounds of 64 and 32. The other tournaments have qualifications with different format some of them several in number. A level before top16 may have different number of players in the different formats. Also the same player may have different placement in different qualifications for the same tournament.
You can see why this option was dropped. If we are to use average performance value, we would need the tournaments to be run with the same format or formats that can accommodate the same levels and number of players. We are in a way lucky that all major events we have consist of defined top 16 stage (top 14 for RBW5). Plus, we would need to have much more knowledge of players who didn’t take part in events (not just ACCM and Yo missing RBW5). This is already too much work and practically means building completely different system while knowing it can be hardly consistent between tournaments.
That is why the decision taken was to just take one score away from each player. It is not perfect because players loose their worse results which is preferable than not having chance to compete. But yet it is better than having some players accumulate points from 6 events and others from 5.
When it comes to tiebreakers between players with the same score last year the tiebreaker was player ladder rating. This time it is the score players have in the last tournament. So RBW5 before TOC, before RBW4 etc. In this case Viper has 50 in RBW5, Hera has 20 and Yo has 0. It is an argument to be made that Yo last performance is 50 at TOC. But in this case it is scores from last, previous to last etc tournaments not accounting if players participated and what was the reason not to (their own fault, lack of motivation, or other reasons that prevented them).It is similar logic to not include the average scores idea.
I am aware this system as any other may be tweaked. There are other things I did not touch upon here that were considered but at the end of the day it comes to where you decide to draw the line and take certain decision. The system tries to account for the reality of tournament performances but there are certain objective limitations. Its main purpose is to avoid subjective seeding where admins decide who should be seeded where and who should be invited. Both taking tournament performances and ladder ranking aims at avoiding that. This argument is also the reason why when the method is decided and the players are ranked we avoid tweaking any further. Because trying to make it “more perfect” you now give one player advantage over the other. At this stage tweaks and changes will come for the next years tournament.
Perhaps the response would have ppl not agree with all decisions and logic but this is the thought process.
Last edited:
Thank you for the details! I may personally wish some things were calculated using a different method but as I've said I fully understand how these things are complex and difficult and respect the decisions. Although this certainly does compound an already bad situation for Yo. Not only do other players get to drop a bad score too, but he automatically loses the tiebreaker because of RBW5. Ouch!
RustyiestyIt seems at odds to me that a system designed to mitigate the impact of players missing events then fails to take that into account in the tiebreaker.
However, you could use something else, e.g. Tournament Elo, and get the same (and some would argue a fair) outcome of Yo as 4th seed.
If simply highest results were used, Hera would be 2nd seed, but Viper is higher on Tournament Elo as well.
I don't think these decisions will affect the tournament, but then again, no one saw the home map issue coming in Hidden Cup either until it was too late.
PS. Robo does have a tournament ranking system, which you could in theory use to seed by tournament type (1v1 ara, EW only etc.), but that is arguably even more complex.
However, you could use something else, e.g. Tournament Elo, and get the same (and some would argue a fair) outcome of Yo as 4th seed.
If simply highest results were used, Hera would be 2nd seed, but Viper is higher on Tournament Elo as well.
I don't think these decisions will affect the tournament, but then again, no one saw the home map issue coming in Hidden Cup either until it was too late.
PS. Robo does have a tournament ranking system, which you could in theory use to seed by tournament type (1v1 ara, EW only etc.), but that is arguably even more complex.
I knew Im in trouble when decided to post 1111.
There are several principles in play here. It is not "simply" to mitigate that Yo and ACCM could not make it to Germany. I do not think that a system like that should or can account for the reasons why certain player do not participate in a tournament but the simple fact they dont. In this case we have certain knowledge for Yo`s situation and understand his decision despite in theory he could make the trip and participate regardless of the overwhelming obstacles. Also we cant make case by case investigations and subjectively decide wether players have good excuses or not to miss a tournament.
Also when a rule/principe is accepted you do not know if and where there will be ties before you make the calculations. Further this principle would apply for every case of a tie not just Yo and in his case there could have been no tie at all.
Ofc something else could have been used, ladder ranking, tournament elo etc. In this case the idea to give priority to last tournament(s) is one of players "shape".
There are several principles in play here. It is not "simply" to mitigate that Yo and ACCM could not make it to Germany. I do not think that a system like that should or can account for the reasons why certain player do not participate in a tournament but the simple fact they dont. In this case we have certain knowledge for Yo`s situation and understand his decision despite in theory he could make the trip and participate regardless of the overwhelming obstacles. Also we cant make case by case investigations and subjectively decide wether players have good excuses or not to miss a tournament.
Also when a rule/principe is accepted you do not know if and where there will be ties before you make the calculations. Further this principle would apply for every case of a tie not just Yo and in his case there could have been no tie at all.
Ofc something else could have been used, ladder ranking, tournament elo etc. In this case the idea to give priority to last tournament(s) is one of players "shape".
I think its fairer to account for Yo and ACCM, in Yo's case going to Germany meant that he would not be able to stream and make money for a month. So yes he COULD go but it would be at an extreme disadvantage compared to other players there. It's the right call to not include his absence. Same with ACCM.
It's played on the newest patch. Players have been aware of this and have been able to practise using the public preview build.
SuperskinnyBLS
lecracheursagaciteTime
Your time
G M T
Your zone
AoE Live-Streams
Voobly Top 5 RM 1v1
MUHAMMEDD DE Top 5 RM 1v1
Hera 2644
_Barles_ 2639
Villese 2617
GL.TheViper 2602
ACCM |AOEbuilds.com 2594
Voobly Top 5 RM Team Game
DE Top 5 RM Teamgame
2583
痛大师 2245
HGB_AOE 2181
chaos_2_win 2009
正义的威震天 2008
Voobly Top 5 DM 1v1
Voobly Top 5 DM Teamgame
DE Top 5 Empire Wars 1v1
[aM]_MbL40C_ 2053
ACCM |AOEbuilds.com 2031
Hera 1999
Yo 1964
CDUB.dogao 1935
DE Top 5 Empire Wars TG
mYi.Sitaux 1752
__BadBoy__ 1726
ELEOS | ElNoniro 1690
Lauth3 1687
BlackRock 1668
