Unbalanced world wars (Pairing)

Chastiefold

New member
Joined
Apr 25, 2024
Messages
14
I'm going to write the post in this section because it's not a general discussion, it's a big problem that in world war the game chooses an alliance so unbalanced with respect to ours (attached screenshot), this as I said in the support, is crazy, I would like to know what those of the other alliance think. In addition to this, I take this opportunity to say that in multiplayer battles there are also times when age bases are much higher. Do your updates really work? How can you make up for the time people throw into the world wars for something that is lost from the start? What do you do with what people leave in your game?

I clarify that there is another forum about unbalanced world war that I have read but it is not related to alliance pairing, the other post basically talks about the stars you can obtain and is in Ideas & Feature requests, which is why my post ends here because the game has some premises or bases for matching alliances, some premises that I think I don't know how it has taken into account, it is not complaining for the sake of complaining


IMG_0105.PNG
IMG_0104.PNG
 
Last edited:

IDaedalusI

BHG Server Engineer
Staff member
Joined
Jun 8, 2023
Messages
29
Hey Chastiefold,

So I was able to take a look at the specific war you took a screenshot of and ran it through some analysis tools I have, and you're right. It was a bad match.

Sadly these can and will happen due to every matchmaking system needing to trade off between finding the perfect match, and not having players wait forever. We are still consistently watching and analyzing our matchamking to find ways that we can tweak it to make bad matches less common, but we will never be able to get rid of them entirely. This likely happened because during the wait period there wasn't an alliance also queuing that was close enough in strength to yours.

However, I do want to thank you for sharing your match here. It helps us get a better idea of what we can look at to improve our matchmaking!

Also I wanna say that I've been watching Mashle recently and it's been great! Its fun to see another person who is into it too! :)
 

Chastiefold

New member
Joined
Apr 25, 2024
Messages
14
Hey Chastiefold,

So I was able to take a look at the specific war you took a screenshot of and ran it through some analysis tools I have, and you're right. It was a bad match.

Sadly these can and will happen due to every matchmaking system needing to trade off between finding the perfect match, and not having players wait forever. We are still consistently watching and analyzing our matchamking to find ways that we can tweak it to make bad matches less common, but we will never be able to get rid of them entirely. This likely happened because during the wait period there wasn't an alliance also queuing that was close enough in strength to yours.

However, I do want to thank you for sharing your match here. It helps us get a better idea of what we can look at to improve our matchmaking!

Also I wanna say that I've been watching Mashle recently and it's been great! Its fun to see another person who is into it too! :)
Thank you very much for your answer, and yes, I understand it, sometimes you feel that you can have fewer filters in a game, I honestly appreciate your time in answering me and the work behind it (and I am very happy that you liked Mashle) 🤍🌞
 

rremmy79

Well-known member
Joined
Nov 12, 2021
Messages
4
Hey Chastiefold,

So I was able to take a look at the specific war you took a screenshot of and ran it through some analysis tools I have, and you're right. It was a bad match.

Sadly these can and will happen due to every matchmaking system needing to trade off between finding the perfect match, and not having players wait forever. We are still consistently watching and analyzing our matchamking to find ways that we can tweak it to make bad matches less common, but we will never be able to get rid of them entirely. This likely happened because during the wait period there wasn't an alliance also queuing that was close enough in strength to yours.

However, I do want to thank you for sharing your match here. It helps us get a better idea of what we can look at to improve our matchmaking!

Also I wanna say that I've been watching Mashle recently and it's been great! Its fun to see another person who is into it too! :)
Thanks for clarifying.
Could there be a checkbox option when starting a new WW search for something like "Optimize matchmaking for fairness even if it takes longer".. or something like that? I, as my alliance's leader, am frustrated by continually getting unfair matchups, hurts morale.
Thanks for your time, love the game!
 

IDaedalusI

BHG Server Engineer
Staff member
Joined
Jun 8, 2023
Messages
29
Thanks for clarifying.
Could there be a checkbox option when starting a new WW search for something like "Optimize matchmaking for fairness even if it takes longer".. or something like that? I, as my alliance's leader, am frustrated by continually getting unfair matchups, hurts morale.
Thanks for your time, love the game!
We are definitely aware of the issues around unfair matchups hurting morale and are trying to explore ideas like what you've described. While I am unable to say what we can or can't implement, I am able to say I do love hearing your ideas, and I do try to bring them up where relevant! It doesn't mean I can make them happen, but I can at least try to make them heard just like @Harlems369th does :)
 

rremmy79

Well-known member
Joined
Nov 12, 2021
Messages
4
We are definitely aware of the issues around unfair matchups hurting morale and are trying to explore ideas like what you've described. While I am unable to say what we can or can't implement, I am able to say I do love hearing your ideas, and I do try to bring them up where relevant! It doesn't mean I can make them happen, but I can at least try to make them heard just like @Harlems369th does :)
Thank you, appreciate the response!
 

oddin

Approved user
Joined
May 17, 2018
Messages
1,782
We are definitely aware of the issues around unfair matchups hurting morale and are trying to explore ideas like what you've described. While I am unable to say what we can or can't implement, I am able to say I do love hearing your ideas, and I do try to bring them up where relevant! It doesn't mean I can make them happen, but I can at least try to make them heard just like @Harlems369th does :)

BHG Muet has done a full renovation to the matchmaking algorithm some years ago. Although it wasn't the perfect tool, it did work much better than the precious version and took into account many variables. The problem is that from that time, we have a ton of buildings, buffs, new additions to the game and 2 new ages (or 3? Can't remember). So this algorithm probably needs some polishing around the edges.

In addition, after 9 years, why haven't you guys created new ways of playing WW?
There are endless possibilities with a game like this.

i.e. a new mode for 15vs/20vs only (fewer line up, more alliances, less time to match). You tick a box and your alliance enters the pool. Once per day, the algorithm is doing all the matchmaking together. The teams are guaranteed a fast and interesting WW.
 

IDaedalusI

BHG Server Engineer
Staff member
Joined
Jun 8, 2023
Messages
29
BHG Muet has done a full renovation to the matchmaking algorithm some years ago. Although it wasn't the perfect tool, it did work much better than the precious version and took into account many variables. The problem is that from that time, we have a ton of buildings, buffs, new additions to the game and 2 new ages (or 3? Can't remember). So this algorithm probably needs some polishing around the edges.

In addition, after 9 years, why haven't you guys created new ways of playing WW?
There are endless possibilities with a game like this.

i.e. a new mode for 15vs/20vs only (fewer line up, more alliances, less time to match). You tick a box and your alliance enters the pool. Once per day, the algorithm is doing all the matchmaking together. The teams are guaranteed a fast and interesting WW.
We've made sure to update out internal processes to keep the matchmaking system updated as each new release goes out.

New game modes and modifications to world war can be tough (though we have thought about it a lot). As you pointed out, the game is over 9 years old, and world war is touched by most other systems in the game. Doing any massive changes to it means that we also have to update other aspects of the game too in an almost cascading effect. That doesn't mean we won't do anything, but it does mean we have to set aside more time for any world war related features as compared to almost anything else. This was something we realized for even something as seemingly small as custom wars.

Now for your matchmaking idea:

I got excited as I always do hearing new ideas for a matchmaking system. At the very least it is just a lot of fun to think about! I discussed the pros and cons of the idea with one of our designers and they pointed out that different time zones would be an issue in this case. What may be a perfect time for the alliances to match in your time zone may be the worst time for someone on the other side of the world from you, and we don't want to make it harder for anyone to participate in wars, especially just because of where they live 🙂

Another concern that I thought of is that this system effectively makes the matchmaking queue time 24 hours (since you're matching only once a day), and our average time to match is certainly shorter than that :)

To your point on the algorithm getting to do all of the matchmaking at once: That certainly is great! More alliances in matchmaking is always better for it! However, who do I start with?

If I start by matching the alliances at the high-end of the bell curve, then they'll get the most fair matches, but everyone below them gets progressively worse matches, and the alliances at the bottom of the bell curve would always get a rough time. We also have to be careful about prioritizing the higher alliances because it could result in a runaway effect where no one else can ever catch up with them, because they're always getting priority!

If I start at the bottom of the bell curve, then the high-end alliances will certainly be worse off too.

I could start by alternating between the bottom and the top, but then the average match is actually worse off.

The best bet would be to start in the middle and cater to the average alliance (of which there are more), but the average alliance's match time we are ok with, and the average match quality is pretty solid too!
 

Lezley

Member
Joined
May 17, 2024
Messages
67
We've made sure to update out internal processes to keep the matchmaking system updated as each new release goes out.

New game modes and modifications to world war can be tough (though we have thought about it a lot). As you pointed out, the game is over 9 years old, and world war is touched by most other systems in the game. Doing any massive changes to it means that we also have to update other aspects of the game too in an almost cascading effect. That doesn't mean we won't do anything, but it does mean we have to set aside more time for any world war related features as compared to almost anything else. This was something we realized for even something as seemingly small as custom wars.

Now for your matchmaking idea:

I got excited as I always do hearing new ideas for a matchmaking system. At the very least it is just a lot of fun to think about! I discussed the pros and cons of the idea with one of our designers and they pointed out that different time zones would be an issue in this case. What may be a perfect time for the alliances to match in your time zone may be the worst time for someone on the other side of the world from you, and we don't want to make it harder for anyone to participate in wars, especially just because of where they live 🙂

Another concern that I thought of is that this system effectively makes the matchmaking queue time 24 hours (since you're matching only once a day), and our average time to match is certainly shorter than that :)

To your point on the algorithm getting to do all of the matchmaking at once: That certainly is great! More alliances in matchmaking is always better for it! However, who do I start with?

If I start by matching the alliances at the high-end of the bell curve, then they'll get the most fair matches, but everyone below them gets progressively worse matches, and the alliances at the bottom of the bell curve would always get a rough time. We also have to be careful about prioritizing the higher alliances because it could result in a runaway effect where no one else can ever catch up with them, because they're always getting priority!

If I start at the bottom of the bell curve, then the high-end alliances will certainly be worse off too.

I could start by alternating between the bottom and the top, but then the average match is actually worse off.

The best bet would be to start in the middle and cater to the average alliance (of which there are more), but the average alliance's match time we are ok with, and the average match quality is pretty solid too!
So currently there's a smaller pool of prospective alliances for matching at any given time, but where do you start now? Or is each search individual when they hit the search button. I saw the previous post where you said they are thrown into the matching pool. You prob worded it better, I make it sound like a swimming pool lol
 
Last edited:

Touchpad Warrior

Approved user
Joined
May 17, 2016
Messages
51
We've made sure to update out internal processes to keep the matchmaking system updated as each new release goes out.

New game modes and modifications to world war can be tough (though we have thought about it a lot). As you pointed out, the game is over 9 years old, and world war is touched by most other systems in the game. Doing any massive changes to it means that we also have to update other aspects of the game too in an almost cascading effect. That doesn't mean we won't do anything, but it does mean we have to set aside more time for any world war related features as compared to almost anything else. This was something we realized for even something as seemingly small as custom wars.

Now for your matchmaking idea:

I got excited as I always do hearing new ideas for a matchmaking system. At the very least it is just a lot of fun to think about! I discussed the pros and cons of the idea with one of our designers and they pointed out that different time zones would be an issue in this case. What may be a perfect time for the alliances to match in your time zone may be the worst time for someone on the other side of the world from you, and we don't want to make it harder for anyone to participate in wars, especially just because of where they live 🙂

Another concern that I thought of is that this system effectively makes the matchmaking queue time 24 hours (since you're matching only once a day), and our average time to match is certainly shorter than that :)

To your point on the algorithm getting to do all of the matchmaking at once: That certainly is great! More alliances in matchmaking is always better for it! However, who do I start with?

If I start by matching the alliances at the high-end of the bell curve, then they'll get the most fair matches, but everyone below them gets progressively worse matches, and the alliances at the bottom of the bell curve would always get a rough time. We also have to be careful about prioritizing the higher alliances because it could result in a runaway effect where no one else can ever catch up with them, because they're always getting priority!

If I start at the bottom of the bell curve, then the high-end alliances will certainly be worse off too.

I could start by alternating between the bottom and the top, but then the average match is actually worse off.

The best bet would be to start in the middle and cater to the average alliance (of which there are more), but the average alliance's match time we are ok with, and the average match quality is pretty solid too!
Hi there! Since you appreciated the other guy posting SS of his mismatch, I'm in one right now so I'll just post some here also.

My current war is against an alliance ranked #57 according to DomiStats while ours is ranked #443 (I know it's not an official rank but it's what we got).

They have 26.858 Glory while we got 21.938 and our leader just told us to skip war, which is unfun, and we lose 2 days giving the enemy a free conquer chest.

You mentioned some mismatches happens so we don't wait long for a war but now because of it we are waiting 2 FULL DAYS, I don't think it would take that long to match us against somebody else, defeating the waiting argument.

Screenshot 2024-06-06 140814.png

Screenshot 2024-06-06 140931.png


Thank you, I hope you guys manage to improve this.
 

IDaedalusI

BHG Server Engineer
Staff member
Joined
Jun 8, 2023
Messages
29
So currently there's a smaller pool of prospective alliances for matching at any given time, but where do you start now? Or is each search individual when they hit the search button. I saw the previous post where you said they are thrown into the matching pool. You prob worded it better, I make it sound like a swimming pool lol
That is certainly a fair question. In our current scenario there isn't really a concept of a "start" in the same way that there might be if we only match everyone at one time each day as Oddin was suggesting.

Right now the system is always running, anyone can join at any time, and that means that anyone can match at any time. We compare every alliance in matchmaking to every other alliance in matchmaking to see if we consider them close enough for a good match. That's why a "pool" is a common term used here, it is a lot like a swimming pool! These checks are constantly done all the time, just like how people will hop in and out of a swimming pool at their own rates and their own times.
 

IDaedalusI

BHG Server Engineer
Staff member
Joined
Jun 8, 2023
Messages
29
Hi there! Since you appreciated the other guy posting SS of his mismatch, I'm in one right now so I'll just post some here also.

My current war is against an alliance ranked #57 according to DomiStats while ours is ranked #443 (I know it's not an official rank but it's what we got).

They have 26.858 Glory while we got 21.938 and our leader just told us to skip war, which is unfun, and we lose 2 days giving the enemy a free conquer chest.

You mentioned some mismatches happens so we don't wait long for a war but now because of it we are waiting 2 FULL DAYS, I don't think it would take that long to match us against somebody else, defeating the waiting argument.

View attachment 12078
View attachment 12079

Thank you, I hope you guys manage to improve this.
Hello there!

So this is an interesting scenario. We wouldn't consider you guys in a mismatch. While your glory may be far apart, it can be good to remember that glory is not a measure of strength, its a measure of ranking.

An alliance with high ranking can go to war with weaker members and therefore have a strength that is more on your level. On the other end, an alliance with lower ranking could easily punch above its own ranking because it is going to war with stronger players.

While I can't share the exact behind-the-scenes numbers for your matchup, I can say we would rate it as a C-level match. The match at the top of this thread we rated at an F-level match.

Also, because I foresee everyone wanting to know what we would rate their matches, please know that I am only one person who is also trying to work on some cool new stuff for you guys, so I won't be able to analyze every single match ;) But, I will keep an eye out for the occasional request so we can continue to spot-check our analysis tools :)
 

Touchpad Warrior

Approved user
Joined
May 17, 2016
Messages
51
Hello there!

So this is an interesting scenario. We wouldn't consider you guys in a mismatch. While your glory may be far apart, it can be good to remember that glory is not a measure of strength, its a measure of ranking.

An alliance with high ranking can go to war with weaker members and therefore have a strength that is more on your level. On the other end, an alliance with lower ranking could easily punch above its own ranking because it is going to war with stronger players.

While I can't share the exact behind-the-scenes numbers for your matchup, I can say we would rate it as a C-level match. The match at the top of this thread we rated at an F-level match.

Also, because I foresee everyone wanting to know what we would rate their matches, please know that I am only one person who is also trying to work on some cool new stuff for you guys, so I won't be able to analyze every single match ;) But, I will keep an eye out for the occasional request so we can continue to spot-check our analysis tools :)
Interesting PoV here.

One look at their museum it became clear that we were in fact punching above our level like you putted, by a lot!
Most of them were paying players with the top 10 with almost maxed museums and paid artifacts that weren't offered for tokens yet, while we ourselfs have a couple of maxed players, it was clear that we wouldn't be able to keep up by the sheer number of possible successful attacks.

I suppose the only upside here is the ego of you guys deeming our humble group worth of facing such strong competition. 😅.

You are right that in terms of ages our players were somewhat close but when it comes to museum they beat us by a landslide, which beg the question, can you shed some light on how you guys approach the artifacts? do the matchmaking consider something like "all towers" dmg/hp different than let's say "redoubt dmg/hp" or it's all the same pot? Do the regular and legendary artifacts carry different weights at all? Because as we all know, the difference is brutal between an optimized group to a casual museum.

Thank you for taking the time to answer the community and I wish you a blessed week.
 

Seek

Well-known member
Joined
Aug 23, 2022
Messages
684
I think the occasional bad match up is great for the lower leveled alliance. You have slim at a massive glory boost. Mainly the gain is the knowledge of how higher level alliance attack and defend, their troop combos, their coalition choices, their base designs and most importantly how they set up their museum. Learning from thy enemy works.

Don’t look at poor match up as hopeless war. Your learn and cost your alliance next to nothing in glory.

Thing I hate about war is planning day should be cut to 12hrs or 18hrs. Would make up for long spins and save time on our special building such as Hanger FSh and Zook towers.
 

Seek

Well-known member
Joined
Aug 23, 2022
Messages
684
Hey Chastiefold,

So I was able to take a look at the specific war you took a screenshot of and ran it through some analysis tools I have, and you're right. It was a bad match.

Sadly these can and will happen due to every matchmaking system needing to trade off between finding the perfect match, and not having players wait forever. We are still consistently watching and analyzing our matchamking to find ways that we can tweak it to make bad matches less common, but we will never be able to get rid of them entirely. This likely happened because during the wait period there wasn't an alliance also queuing that was close enough in strength to yours.

However, I do want to thank you for sharing your match here. It helps us get a better idea of what we can look at to improve our matchmaking!

Also I wanna say that I've been watching Mashle recently and it's been great! Its fun to see another person who is into it too! :)
I tip my hat to you for coming on forums and actually trying to answer questions. I personally would like your fellow staff member that is charge of Replays, Battle Reports, and Troops Lost Reports to do the same.

I know myself and 98% actual players would love a chance to ask for some status updates and timeline for actual fixes on issues with Replays, Battle Reports, and Troops lost report. We are 9 years in and problems listed abound make war matching seem as a minor issue. Thank you again for coming to forums.
 

oddin

Approved user
Joined
May 17, 2018
Messages
1,782
New game modes and modifications to world war can be tough (though we have thought about it a lot). As you pointed out, the game is over 9 years old, and world war is touched by most other systems in the game. Doing any massive changes to it means that we also have to update other aspects of the game too in an almost cascading effect. That doesn't mean we won't do anything, but it does mean we have to set aside more time for any world war related features as compared to almost anything else. This was something we realized for even something as seemingly small as custom wars.
Yes it is understood. I believe it would be like where do I start and where to finish thing eh?

Throughout the years, many players have posted their own ideas in here. Yes, we may not be able to think every little aspect of what we ask of you and the technicalities behind them, but maybe we can give you the spark that you may need to start doing something new. So, although my idea may have flaws (and thank you for pointing them out), there are workarounds.

- considering our lives are run on 8 hour cycles (8 sleep, 8 work, 8 free time), the algorithm could run on an 8 hour interval thus covering all time zones.

- the issue of where to start doing the matchmaking is something I wasn't aware of. Yet, giving it some thought, I believe it should start from the top alliances because this is where we have the extreme line ups of all auto age players with max museums etc . Then it should go downwards to say top 30-70 alliances and then top 71-150 and so on, with each bracket being larger as we go down all the way to 12,000 glory.

And why I believe that to be the best scenario? Because the mismatches occur mostly with the teams that have more Auto age players than the other team. So, making sure that these teams get their best opponent, eliminates possible bad matches for the lower teams too!!

A team ranked say #70 has very few things in common with a team that's playing at the top 20. And truth be told, lower ranked teams (i.e. ranked 300 and below), are highly dependent on how well just 1-2 strikers will perform during the war and if their very few in number defenders will actually defend. So, such teams can win and lose with equal chance against other +-2000 glory teams. Let's not forget that players in those alliances barely have dedicated O or D museum, max munitions, full purple+ council or play with the latest metagame (HTs still rule!!). In other words, if they match other teams in a wider range of glory then it's ok and not a mismatch. But if this happens within the top 100 then it is a problem. That's why the matchmaking algorithm should start with the higher ranked teams.

I really appreciate you coming here and talking with the players. That's the kind if forum threads I really enjoy giving my time to. And since you are keen on hearing new ideas, here is another:

8+ years ago, when I first started doing WWs, I always wondered how it would be if the map itself was dynamic. One team has red and the other blue. And since we have colors, why not have the territories change color as if we really conquered them ? That would create a nice view of the progress of a WW. And the specifics can be endless. You conquer a territory when you can five star it faster than your opponent or that each territory can be attacked twice by both teams. The result with the most stars and less time decides the color at the end. And since it would change to a territorial strategy game, we could have bonuses coming from keeping whole continents under our banners for X amount of time!!
Hell, we could even have random objectives popping up that would change the primary target, like capture territory X within 6 hours!

Now imagine the level of player engagement in a WW such as this

Thanks for hearing my banter
 

King Crimson

Approved user
Joined
Apr 21, 2016
Messages
1,387
8+ years ago, when I first started doing WWs, I always wondered how it would be if the map itself was dynamic. One team has red and the other blue. And since we have colors, why not have the territories change color as if we really conquered them ? That would create a nice view of the progress of a WW. And the specifics can be endless. You conquer a territory when you can five star it faster than your opponent or that each territory can be attacked twice by both teams. The result with the most stars and less time decides the color at the end. And since it would change to a territorial strategy game, we could have bonuses coming from keeping whole continents under our banners for X amount of time!!
Hell, we could even have random objectives popping up that would change the primary target, like capture territory X within 6 hours!

Now imagine the level of player engagement in a WW such as this
Now this is an event even I would be happy to participate in. (y)
 
Top