What is/are your preferred seeding methods?

What is/are your preferred seeding methods?


  • Total voters
    153

Netherlandsnimanoe

Knight
Bronze Supporter
Jan 15, 2014
2,118
1,399
138
23
#1
With a small discussion going on about the seeding in KotD2, I was wondering what your guys opinion on seeding are. What do you think is the most fair and accurate way to seed teams?

I'll give a bit more explanation to the options in the poll:
1. Elo based system: Could be a system where it looks at the current rating of the players at the end of sign-ups, but could also look at the elo over a longer amount of time or at the highest achieved elo or a combination of these. Example: KotD1
2. Elo + tournament history: Looks at the elo of the players like in the first option, but also looks at the recent results in similar tournaments and gives a bonus for those. Example: KotD2 seeding
3. Mostly depending on tournament history: Seeding based mostly on how well they performed in previous tournaments. Doing well in a certain tournament could mean you automatically qualify for another tournament. Example: Escape Gaming Masters 2.
4. Players seed themselves: Players send in their seedings of the other players and the average seed (sometimes removing the outliers) that players get will be their seed. Example: Clan Masters: The Final Showdown
5. Community seed: Could be the admins seeding the players or it could be a community vote or just a small team of casters/admins. Example: SY Nations Cup 2014
6. Other: Anything I might've missed.
 

NetherlandsQulala

Active Member
Jan 20, 2016
152
80
43
27
#4
since the pool of tournaments is to small most of the time, or not everyone participated,
i voted for players seeding themself and a select group (community) seeding the players,
Both are alwasy very capable of seeding, since they also take into account the past tournaments and highest, current elo rating without looking to one fixed time.

With the second maybe even the best eventualy for professionalism so a group of non playing experts seed them.
at the moment it isnt weird players seeding themselves for the best result, but if you look from an outsiders point of view its weird ofcourse :smile:
 

Germanykc_Ereon

Known Member
May 20, 2013
537
218
58
#5
Players seed + Elo + Tournament History (importance ~ 65% + 25% + 10%) + Admin Decision in close cases

As Admin, you then need to hope that all the players (who appear to not even read the rules/handbook sometimes) are actually going to seed each other. In ToA, I asked all teams' opinions on the possible Civ and Match Pools, and I got 11/23 answers.
I guess it's a different story for tournaments like KotD2 with lots of prizemoney and only the best players.
 

Finlandbuhanisson

Longswordman
May 29, 2015
302
715
108
#6
I think a combination of many things is generally a good idea. There are problems with ELO (obvious), problems with tournament history (how to decide which tournaments are relevant still for predicting this new tournament), and problems with players seeding eachother (e.g. western vs chinese players might not all know each other well enough, or teammates may be generous or, in case of aM, toxic towards each other). When you combine several things, those small problems should have less effect on the end result.

If you have an obvious past tournament result to use, I think its a solid starting point. E.g. NAC1->NAC2->NAC3. But of course you cant use some 2-year-old MOA results for seeding arabia 1v1.

Community seeding is just a bad idea. Players already know each others abilites well enough, adding random fanboys/haters to the mix doesnt accomplish anything.

But really, anything should work decently enough in a communtiy of this size. I think with KotD2 the problem is not the seeding. Its the fact that seeding is combined with the civ picking rules which further favor those who get low seeds. It should either be random brackets with the civ picking thing, or seeding but with very different civ pick rules. Top seeds should get easiest brackets, but they dont need any additional benefits to add insult to injury. One can already predict that theViper vs Hera/TaToH series will be a complete waste of time for everyone involved, because of this. Hera of TaToH might actually be more likely to win the tournament if they faced theViper on 1st round than on 3rd, which is kind of unfair to those hight-to-middle seed players.
 

Austria_balcanman

Active Member
Dec 6, 2013
114
75
33
#8
the problem is that ELO ranking here, in other sports like tennis or chess the rankings are taken only from tournament games and not from fun games, so seedings based on this rankings are pretty fair.
 
Likes: ImRiver

United KingdomFall

Halberdier
Jun 12, 2013
1,892
387
98
25
Voobly
Fallstar
View profile
Ladder
RM - 1v1
Rating
2000
Wins
333
Losses
268
Streak
4
#10
The problem with players seeding themselves and relying purely on tournament history is that it doesn't take much account of new players / recent form.

I think taking into account elo in some way is important to give the ladder an actual purpose. You can use multiple sources of information too, e.g. give players a tournament rating based 40% on rating, 40% on a player's ranking of one another and 20% of past performance in a selection of tournaments.
 

Finlandbuhanisson

Longswordman
May 29, 2015
302
715
108
#11
Also, when in doubt, its not necessary to always seed everyone. For example in 16(/32)-player tournament, I think players can easily and quite realistically peer seed the top 8(/16), and then the last 8(/16) can just be assigned to be their opponents at random. RNG is always fair for everyone, whereas all the other methods have the potential for drama and complaints. Maybe no1 disagrees that Viper is #1, but ppl might not have same concensus about whether Villese is #13 or #14 or whatever.
 

Nepalarchxeon

Well-Known Member
Jan 6, 2014
159
292
68
#12
I think a seeding system that clearly outlines how a player can get seeded higher (based on their own efforts) is the most fair. Although I maybe in minority and there maybe certain flaws in the ELO system, I voted for it.
 
Last edited:
Likes: AlphaAndOmega

SloveniaModri

Well Known Pikeman
Aug 10, 2013
159
456
78
#17
Yeah, very objective and fact oriented. No back scratching of friends and punishing the players who one dislikes.
If player's rankings of competition were open to the public aswell there'd be almost none of that. It's by far the fairest option imo.
 

Netherlandsnimanoe

Knight
Bronze Supporter
Jan 15, 2014
2,118
1,399
138
23
#18
Also, when in doubt, its not necessary to always seed everyone. For example in 16(/32)-player tournament, I think players can easily and quite realistically peer seed the top 8(/16), and then the last 8(/16) can just be assigned to be their opponents at random. RNG is always fair for everyone, whereas all the other methods have the potential for drama and complaints. Maybe no1 disagrees that Viper is #1, but ppl might not have same concensus about whether Villese is #13 or #14 or whatever.
Problem with this is that it emphasizes someone who might be #8/9 or #22/23 will be punished a lot more if he gets one lower seed compared to the usual model. And another problem might be that you could have the top 4 seeds all on the same side of the bracket, since you don't differentiate between them.
 

Netherlandsnimanoe

Knight
Bronze Supporter
Jan 15, 2014
2,118
1,399
138
23
#20
If player's rankings of competition were open to the public aswell there'd be almost none of that. It's by far the fairest option imo.
On top of that you should also remove the outliers imo, so that one or two people being generous to a teammate or giving a bad seeding to a rival don't get taken into account
 

Finlandbuhanisson

Longswordman
May 29, 2015
302
715
108
#21
Problem with this is that it emphasizes someone who might be #8/9 or #22/23 will be punished a lot more if he gets one lower seed compared to the usual model. And another problem might be that you could have the top 4 seeds all on the same side of the bracket, since you don't differentiate between them.
To clarify, I meant that the top seeds would still be in order and divided between the brackets accordingly. But yes, that other thing can still be frustrating for some ppl. Still I think the top10 or so would be very accurately, and thus fairly, rated by other pros, but after say top 20-> there may come the issues of whether everyone has played enough vs everyone (and time zone differences) that might make it less accurate. For example, does Viper or Liereyy rly play that many _serious_ games against 2.1-2.3k players to know who of them is better than someone else?
 

Finlandbuhanisson

Longswordman
May 29, 2015
302
715
108
#22
And one more prototype of an idea, to make sure everyone seeds other ppl in as fair a way as possible: How about make everyone's 1st-round opponents depend (slightly) on their own seed choices?

For example, say jordan seeded like this:
1. theviper 2. daut 3. tatoh 4. riut 5. mbl 6. nicov 7. dracont 8. liereyy 9. yo 10. vivi
Jordan would get a random opponent because he is not seeded. He woulld have average of 10% chance of facing everyone, but it would actually be more like 7% theviper 8% daut, 9% tatoh, ...., 13% yo 14% vivi, or something like that. Might sound complicated, but would be pretty easy to implement on excel or something like that.
 
Jul 30, 2017
46
82
18
#23
Although I am as pleb as you can get I'll give my opinion aswell, just to give Edie something to read. Keepo

At the current +/- family like stage of the game top players seeding themselves is absolutely fine. Atleast if it is a tournament with let's say 8+ players/teams. Better, imo, would be a mix of elo and past tournaments (if possible) (both maybe 50 %) + player seeding (50%). But a big disadvantage might be the plus in work for the admins and the long period of time player might take to send their seeding back.
The more the game grows, I feel the less viable a player seeding gets. When we are e-sport #1 things have to and will look very different obviously.
 
Likes: noobishere

Time

Your time
G M T
Your zone

Upcoming Events

Improvement Cup 3v3
Wednesday 19:00 (GMT +01:00)
Group D: Team Lunacy vs Team India
ECL East Asia 4v4
Thursday 15:00 (GMT +01:00)
ECL East Asia 4v4
NAC2 Promo: $100 Showmatch
Friday 16:00 (GMT +01:00)
BacT vs TheViper
KotD2 Quarterfinal
Saturday 16:30 (GMT +01:00)
Yo (5) vs (20) Lyx
KotD2 - Quarterfinal
Saturday 19:00 (GMT +01:00)
Liereyy (2) vs (7) MbL
KotD2 - Quarterfinal
Sunday 16:30 (GMT +01:00)
Fat Dragon (3) vs (27) Yinghua
KotD2 - Quarterfinal
Sunday 19:00 (GMT +01:00)
TheViper (1) vs (9) TaToH
NAC2 Promo: $100 Showmatch
Monday 16:00 (GMT +01:00)
F1Re vs DauT
NAC2 Promo: $100 Showmatch
Wednesday 16:00 (GMT +01:00)
BacT vs DauT
ECL East Asia 4v4
January 31st 15:00 (GMT +01:00)
ECL East Asia 4v4
NAC2 Promo: $100 Showmatch
February 1st 19:00 (GMT +01:00)
F1Re vs TheViper

Age Of Empires On Twitch

There are in total 29 streamers online
Click here for details
Age of Empires II 1127 viewers
Age of Empires II 345 viewers
Age of Empires II 167 viewers
Age of Empires II 147 viewers
Age of Empires II 59 viewers
Top