Forum
[ETF2L] Swiss Style Tournament System
Created 13th November 2014 @ 00:57
Add A Reply Pages: « Previous 1 ... 3 4 5 ... 9 Next »
Quoted from Admirable
[…]
The historic reason golden caps were introduced was because draws were more likely to occur on certain maps (Gravelpit main offender at the time).
Wasn’t gravelpit excluded from the golden cap rule for the reason that it was played “best of three” anyway?
Quoted from Kaneco
Guess if this ladder/swiss system is so perfect and magical the biggest sport in the planet would be using it after 100 years no?
Chess does it use it a lot, so yeah :D
Quoted from CanFo
Wasn’t gravelpit excluded from the golden cap rule for the reason that it was played “best of three” anyway?
It used to just be two rounds of attacking and defending, but they added a third round to eliminate draws.
The question of whether to use a Swiss system and the surrounding questions ought to be handled separately.
The first thing is the idea of a Swiss system. In large divisions it will provide for a better representation of skill and allow for a more even matchmaking to develop naturally as teams play off against each other. Now I like the sound of that if for example we were to say divisions 6 & 5 are now Swiss 4. They all group together play depending on results and the games will even out.
You do however have some points of note from my perspective.
Firstly, the first game is a completely random game. You could have a team pushing div4 Vs. a brand spanking new team who are wandering around cp_badlands wondering what choke is and why it’s supposedly so important.
This doesn’t exactly fix the problem we’re already talking about, and the problem that instigated this discussion. If teams are crying off league play because they are too soft to handle a loss then the stompings they could potentially receive in this set up stand to be way worse for at least the first game.
Second to this is the question of season length. I’ve played a couple of seasons in UGC and ignoring the bad admin experiences I can’t say I enjoyed it too much (maybe because HL). The division I played in admittedly was only slightly larger than the average ETF2L division at most double the size. if the first set of games sort the first crop of wheat from the chaff, then the second sorts out the finer elements of chaff and the third should leave just wheat in theory. That’s three weeks into a season. just shy of a month. Unless the seasons extend relative to the group sizes then we’re wasting games just making sure the “shitters” sit where they ought to and the teams should be playing the right level.
In addition if we went Swiss how would we organise the league. Would it be Open, IM and Prem? Would IM be 1/2 with Open being 3 – 6? If they went like that would we still look at the same length of season? With Prem playing 8-10 teams, IM playing 21-30, Open playing.. many. If we keep the same schedule lets say 8 weeks, Prem all play each other, IM play 1/3 of the div and Open plays a fraction based on the win ratios as planned. Then we enter the play off stages with I believe top 16 was mentioned for example. A week a game and we’ve added a month on top of the season. Do we then add another 2 weeks for promotion relegation matches? Or do we skip that.
Will this set up mean longer seasons and less of them as a result of the amalgamated leagues? Would the number of matches be consistent?
Further to this what if the first weeks include a wildcard map like Viaduct or Turbine or a new map that isnt tested? As much as we can say “learn the map, play the map it’s your own fault” it doesn’t work in practice and whilst the chances of a wildcard map being in the first week are slim at best and entirely at admin distrection such weeks are enjoyable in their current format for the potential for upset victories, lost points and drama in the league table. They’ll provide similar in a Swiss set up but isn’t the intent to have more balanced better games? Who would want a false negative flagging up teams to play better opposition than they ought to? It works against the intent of the system as far as I can see it.
I should’ve really written this out in word before presenting it but nevermind. I probably have far more thoughts on this but Im not convinced Swiss is the way to go even if it makes it easier for admins to seed provisionally and provides a protracted way to balance divs. Hell if I were to take a purely selfish point of view I’d look at div 2 this season and say look at the free ride you get if you get put into div 2A. Any of the top 3/4 in 2B could be riding that div and have proven themselves Vs. Div1 opposition a Swiss grouping would have changed things up significantly. But doesnt the end of seasons play offs sort that out? Isn’t this entire thing just to balance out the clustering at the lower end of the tables?
Would Swiss serve as a basis for div 6-5, 6-4 or as Admirable said (perhaps jovially) 6-3 at which point a they earn promotion into the upper divs (apologies for any elitist vibes)? But that would be a major hassle and not really likely worth while I presume.
I’ll end the drivel here but I could go on and on.
Quoted from Admirable
The historic reason golden caps were introduced was because draws were more likely to occur on certain maps (Gravelpit main offender at the time).
So as all maps/draws were not considered equal and the schedule can’t accommodate playing every team on every map the common consensus was to remove draws all together.
The admins at the time decided to soften the blow by awarding these golden cap points, even though it still seems to contradict the reason for introducing the golden cap in the first place.
So that’s why you reward people for “losing more slowly”. You could just as easily abolish it and recognise a draw for what it is.
Quoted from Admirable
[…]
It used to just be two rounds of attacking and defending, but they added a third round to eliminate draws.
And that 3rd round used to count as a Golden cap during AFS, but in Season 8 GPit was played bo3, without any possibility for a Golden cap.
@ Henghast’s wall of text
There is a variation of the swiss system which is aimed at reducing the number of div4 vs 100% new teams called the accelerated swiss system.
The top half of the division are given x points initiailly, without playing any matches. Then the system carries on as normal, basically this will ensure the first matches are good teams vs good teams (all in the top half) and weak teams vs weak teams (all in the bottom half).
There’s a brief explaination on the wikipedia page: http://en.wikipedia.org/wiki/Swiss-system_tournament
This would require us to seed the top half, but thats it and would be straightforward.
There is a nice season length tool Gentleman Jon gave me. With the accelerated pairings I think 8 gameweeks (between 128 and 255 teams) would be enough, especially with the length playoffs after (top 16 most likely for big groups).
We’re still unsure on the exact groupings, it would either be 4 tiers or 3. If its 3 it would be something like div6,5,4, div3,2,1 and prem. If its 4 tiers it can be div6,5 div4,3 div2,1 and prem. But obviously the more tiers the more spread out any prizes are!
Upsets can happen on any map, not just a new one. You’re suggesting all good teams ignore new maps, which is not true for the majority.
I’d say thats another positive, it means we don’t have to create sub divs (which are very hard to do if you didn’t realise by the fact that divs are usually a bit skewed). I don’t understand what you mean by clustering, if its the fact we have nearly half of our teams in div6/5 its not that. That’s not really important.
The Swiss System does not work for low numbers of teams, the highest tier would have to be Round robin (the same as it currently is). If there were 4 tiers, the second one would almost certainly not be swiss either, it would have to have a different system (16 teams round robin is obviously not viable). If there are 3 tiers the second tier could potentially also use swiss, but it’s hard to say for sure until you know exact numbers.
The names wouldn’t be Invite, Intermediate and Open probably. But naming potential tiers isn’t a priority right now.
Quoted from Henghast
[…]
Premiership, Scrubs and The Zoo
Last edited by CHERRY,
Quoted from CHERRY
[…]
Premiership, Scrubs and The Zoo
Nerds, Dweebs, Dorks ?
Quoted from Gentleman Jon
So that’s why you reward people for “losing more slowly”. You could just as easily abolish it and recognise a draw for what it is.
Do or do not. There is no try.
Quoted from Admirable
Do or do not. There is no try.
Look at what happened when he was league admin. A philosophy best suited to running a mud hut.
@Kaneco, in terms of football which you’ve obviously talked about before, the reasons they don’t use something like this are practical. You have to plan travel, accommodation, venues, equipment and in Swiss you don’t know who or where that is until you play a round. Obviously it’s totally impractical.
In the case of TF2 you need nerds sat at computers in their homes anywhere within acceptable ping range. That’s doable.
Quoted from Gentleman Jon
@Kaneco, in terms of football which you’ve obviously talked about before, the reasons they don’t use something like this are practical. You have to plan travel, accommodation, venues, equipment and in Swiss you don’t know who or where that is until you play a round. Obviously it’s totally impractical.
In the case of TF2 you need nerds sat at computers in their homes anywhere within acceptable ping range. That’s doable.
It’s much more than that…
You have the factor of preparation for a match, I know that means shit in lower divisions, but in higher divisions it helps a lot to know who and when are your opponents and which maps you will be playing against them so you can prepare in advance and if possible study their strategy for the map. Also there’s the scheduling diversity which allows teams to accommodate games much easier.
Quoted from Kaneco
It’s much more than that…
You have the factor of preparation for a match, I know that means shit in lower divisions, but in higher divisions it helps a lot to know who and when are your opponents and which maps you will be playing against them so you can prepare in advance and if possible study their strategy for the map. Also there’s the scheduling diversity which allows teams to accommodate games much easier.
Yeah the further you go up the ladder the less sense it makes, I don’t think there’s any doubt Prem should remain round robin, but somewhere there’s a point where it becomes an attractive option. That’s probably the most difficult judgement to make.
You can’t even draw on Viaduct with the current rule set?! Where is the consistency.
If we choose to operate from the stance that there are no draws, then that only leaves winning and losing, not this bastardised points bonanza.
LEGO failed to win either map tonight, yet they got awarded a point that resulted in them winning Division 1. Absolute madness :D
Last edited by Admirable,
Add A Reply Pages: « Previous 1 ... 3 4 5 ... 9 Next »
content rss