I also expect this will also continue to be a problem even after a leaderboard reset as long as you have an individual reward track is purely based on wins.
4 Likes
I suspect this might be an exceptionally bad idea.
Imagine signing up for the amateur teen boxing contest, then meeting Mike Tyson in the ring. The contest may have mentioned something about surprise guests, youâll probably still switch hobbies after getting out of hospital.
You never, ever want to pair significantly mismatched guilds unless you intend to alienate your whole player base:
1.) Itâs boring for the strong guild, unless they are only in for the winner rewards.
2.) Itâs miserable for the weak guild, not only are they bound to lose a majority if their battles, they also miss out on the rewards they would have received if there had been a fair pairing.
Making everyone the same for pairing purposes means significantly mismatched pairings will be the norm. This wonât balance out in one, or even ten Guild Wars runs, because reasonably sorting a list of several thousand entirely random entries requires many passes. Current pairings are already causing players to quit, having even more unfair pairings for the next years is unlikely to go down well.
This wonât help. Whoever pulls the short straw (by receiving an opponent much stronger than them, entirely out of their control) gets punished in multiple ways. They get battles forced on them they canât reasonably hope to win (which costs time and input devices) and theyâll have nothing to show for it (because the winner takes it all).
What would be needed for this to have any remote chance at working out is to reward participation, not winning. If half the guilds are supposed to serve as random punching bags for the other half, for many upcoming runs, there has to be some adequate compensation for volunteering to get punched. Otherwise the punching bags wonât show up at all, which will just cascade the situation to whichever guilds remain.
7 Likes
Itâs already terribly mismatched though so itâs loss mitigation at this point. This is trying to navigate the least bad.
If it were possible (and I donât think it is?) to recalculate ELO and subsequent matchups on a daily basis, it would cut the time down even further.
The other question of note-- how many Tysons are out there in this iteration of GW? On PC/Mobile, I would put the number as being surprisingly low. Former high bracket GW guilds are a mixed bag on how theyâre approaching this version. Letâs say thereâs ~32. (Edit: was going to do a bunch of math here but it would make this post way too long. Essentially following an exponent of 2 to illustrate how quickly the top would find themselves versus today)
32 guilds are going to have a rotten time assuming none of the Tysons match each other.
All the Tysons go 4/0.
Next month, the only ones who will even have a chance to face a Tyson are the ones that went 4/0 which, assuming a bell curve of performance, would be a relatively small number. The Tysons then again go 4/0.
The following month, the only ones who will have a chance to face a Tyson are the ones who 1) Also went 8/0 by beating another non-Tyson 4/0 guild, never faced a Tyson previously and assuming the same bell curve, the number gets smaller again. The Tysons then again go 4/0.
Month 3, the only ones who would have a shot at facing a Tyson would need to 1) Go 12/0 by beating a guild that was previously 8/0 (who themselves beat a 4/0 guild) with neither having faced one of the Tysons.
Ad infinitum with each subsequent set requiring more and more outlandish circumstances to come true until the top guilds naturally find themselves.
Or we leave it status quo where itâll take the high powered guilds beating the stuffing out of their opponents month over month to never make up any ground since the ELO is so inelastic.
Basically, today since the top GW guilds arenât necessarily the top trophy guilds, youâre FAR more likely to see those Tyson guilds randomly sprinkled throughout the leaderboards than if you were to pull a random guild out of a hat.
Someone is going to draw a short straw. Thereâs no avoiding that. This is strictly about minimizing the damage. Compare that to a GW Tyson that was, say, 500th in trophy count and how long it would take to find their footing and how many guilds are going to have really bad times along the way.
4 Likes
I wonder how gw looked like when it first introduced with no filters at first. Was it the same carnage like it is now?
Was before my time as I âonlyâ started in 2020. I imagine the total number of guilds was also significantly smaller at the launch of GW 1.0 so it would have theoretically taken less time to sort out a hierarchy as there wouldnât have been so much noise.
You do realise this thing happens in every sport, that has any type of competition.
Tennis. 1st seed can get amateurs
Snooker 1st seed can get amateurs
Soccer prem league can get non league.
Any type of athletic event. Big countries like Germany can compete against Samoa.
This is classed as real competition, the underdog always has a chance.
The only place in the world, competition isnât encouraged is the USA .
No relegation, no losers.
In NFL,NBA,NHL and baseball.
All one top league, closed off, to anyone else.
So I can understand your point if your American, if your European, Asian African or oceanic. I donât understand it tho.
A youtuber, who canât rip a paper bag, an amatuer.
Called fake Paul, beat Tyson 
.
Itâs. One. Metric. JFC. Weâve been begging this company to take our feedback into account and then when someone comes along to ask what we think - this is what they get?
4 Likes
Some of us are old enough to remember Tyson in his primeâŚ
âŚand the allowance I wasted buying Buster Douglas Boxing on the Genesis/MegaDrive
3 Likes
Very much in favor of a reset, if the rounds after the reset can sort us to fair rankings reasonably quickly. Possibly run a Swiss-system tournament for initial ranking, then switch to ELO?
From my casual end of the rankings, the problems I see with the current model are:
- Seeding by trophy count was a very poor proxy for GW performance: it skewed the seedings strongly towards (a) longer-established guilds and (b) guilds that focused on trophies.
- Defaulting to opted-in was a mistake: weâre getting matched against a lot of dead or inactive guilds that are opted into GW but which have no GW defense or attack participation.
- Too much faith in âELO will get us to fair rankingsâ â yes, it will, but only in the very long term.
- Matchmaking has been rough and opaque; nobody knows why theyâre being matched against some opponents; there arenât brackets any more except there are but we canât see themâŚ
(The other thing that strikes me is that the ELO algorithm specified here uses only win/lose/draw, i.e. 1/0.5/0 results. Win/lose/draw is great for chess, but not necessarily the best fit for us? GW draws are very rare; and GW matches are scored. At the moment winning by 1 point has the same effect on ranking as winning by 10,000 points; this feels like we are throwing useful information away.)
10 Likes
There were brackets of size 10, so you would at worst get matched against someone 9 ranks higher. Initial placement by trophies was a pretty good compromise for existing guilds, it just didnât work out well for new guilds interested in Guild Wars. This was usually mitigated by searching for an abandoned guild, taking it over and inheriting a reasonably high rank.
2 Likes
Agree 100%âŚbut it would mean a pivot away from ELO altogether and Iâm sure part of the draw was that ELO was a âtried and trueâ (read: easily copied) formula. 
2 Likes
Our guild had kinda close matches this gw, matchmaking was way better then it was in the previous wars, no ghost or dead guilds, no top guilds, lost 3 days out of 4, but those looked fair losses. My opinion, matchmaking mechanic is on the right way, and just needs some more time and little tweaks may be, no reset. Voted for keep as it is atm. Guild postitions was 109 in the guild war roster, whatever this number means.
Would also just be cautious about confirmation bias. 2 months ago we had a similar scenario with all 4 days being close enough to feel ârightâ. This month wasnât close as all the guilds we faced were well below us and were essentially walkovers.
In each case our sample size is really limited and itâs easy to draw conclusions that may not be correct.
3 Likes
Yea Iâm old enough.
Fury,Joshua,parker,usyk . All babies compared to iron Mike.
At the age of 19 , he was a animal. Best boxer since Marciano.
1 Like
This whole thing, whilst the ELO stuff is not great, is missing the point.
The point is, the battles are so much harder than old GW because the devs have given every advantage to defenders. When teams can field an almost 24k teamscore with 3 Takshakas and a random hero, 50% of the time, maybe 80% of the time depending on what happens on turn 1, itâs a loss. Add their 15 or 20% bonus, they 1 shot 99% of the playerbase, by filing up befrore their turnâŚ
We also need to address those fundimental issues where instead of a 3-5% loss rate, itâs 50% or worse. And itâs that stuff that causes players to leave, not compete, be empty or half empty guilds.
Please review the whole feedback given, not just the ELO, it will sidetrack and if most of those changes requested are not implemented, we will see a lot less people playing in our guilds.
10 Likes
110% agree that the matchups arenât an end all be all and thereâs not going to be a single fix that will do all the things.
That said, the motivation behind a reset would be to get the guilds that are full of 3x Takshakas finding each other (and thereby avoiding the lesser equipped guilds) more quickly. 3x Takshaka is a minefield. Trying to make that minefield cover the smallest possible total area meant for just those that have something resembling an answer to 3x Takshakas would be the goal.
So again, not solving the issues as much as minimizing the collateral damage.
3 Likes
Yeah Iâm wiith you with everything re the matchups/rankings/initial ELO.
Weâve calculated a little, have some educated maths people, and itâs not ideal from the original makeup and got worse before the recent fix.
However we also need to not forgot other issues that magnify and exacerbate that ranking issue. I wonât repeat my feedback request again, but it was early in the thread. 
I know itâs potentially not the developers mindsight, but they have to consider making the experience better, and not just with rankings. By and large our rankings have been very well matched, 3,4,5 and 7 3rd time, 6, 33,1 and 24 last. My opinion, the battles were nonsense.
The last thing Iâd want tbh is actually take up the ELO only and completely miss the opportunity to attempt to fix other core issues with the released content. If itâs not a priority for the devs, they will know soon enough as players will continue to leave.
4 Likes
Option 4. Scrap the whole thing and go back to the drawing board. This guild wars took all the worst things about the last one and amplified them. Its a failed mode.
In all seriousness one of the major things that needs to be addressed is rewards changed to battles fought than won. As it is people just fighting the gate is a problem and i dont blame them when they are restricted to blighted lands against 3 takshaka lol
2 Likes
You could also, make an area, un fightable, once you beat it.
So players, once they win gate, have to move on, and so on.
Each area, can only be fought once. If itâs beaten obviously.
1 Like