How is the matching of world wars determined

[oddin]

thanks for the clarification.

 

[Gelonia]

Come viene determinata la corrispondenza delle guerre mondiali

Con il linguaggio di cerca alleanza.

 

[Lyonz]

@IDaedalusI can you consider a maximum wait period on 50vs50 wars? As 2 alliance have tried matching 3 times now 1st was 18hrs then 15hrs and then again 18hrs both with no match…. I believe the glory,war weight,power spread is two far apart but there should be a limit to wait time as we have missed the opportunity of 2-3wars because of this.

 

[IDaedalusI]

Hey @Lyonz!

This is something that @Harlems369th has brought to my attention already. I'll reiterate that I am only a simple programmer who can not promise anything However, both @Harlems369th and I have brought it up to others as well, as something to discuss.

In the meantime, because I know 50v50s take a lot for players to set up, I will share this; That matchmaking queue will "override" an alliance that has been in it for at least 18 hours to automatically match the next closest alliance in the queue. So long as one alliance has waited for 18 hours, and there is at least one more alliance in the queue, they will match then no matter the difference in glory, power, etc.

However, if there isn't another alliance in the queue to match with, then there really isn't anything we can do. So, I personally recommend always going into that queue with another alliance you're intending to match with if you really want to do a normal war.

But, at the end of the day, if you want to specifically and immediately match with another alliance, the Custom War system is definitely the way to go. I know you guys wouldn't earn rewards, but you do get the experience of a massive 100 player war without the wait that comes from a matchmaking system!

Hopefully this information helps you guys plan accordingly!

 

[UnHoLyGhOsT]

Hey @sevenjapan,

I've answered questions around this in a few other posts on the forums, but I'll take your post as an opportunity to collate a bunch of it in one place. While I can't share every detail of our algorithm (people might take advantage of it then!). I can share some of the basics!

Step 1: When an alliance enters matchmaking the leader had to say who was joining them in war. We take this list of players and then look through their main bases. On those bases we look at things like buildings, troop tactics, munitions, etc. and their levels. We use this to assign a "war weight" to each player. I'll also note here that if a player somehow manages to not have all of their buildings on their war base, that just puts them at a disadvantage because we counted everything on their main base. War bases are just for you guys to reorganize things as you wish

Step 2: We do some math. We combine all of the war weights of each player in the alliance into a total "alliance war weight" We don't do a simple summation here because we don't want to heavily disincentivize alliances from bringing along a couple weaker players to show them the ropes.

Step 3: We do some more math. We also figure out an alliance's power spread. This is our attempt to handle the scenarios where 1 alliance may have 10 drone age players, and another alliance may have 1 player in each age (I do not recommend doing this by the way ). The first alliance will have a tight spread, while the second will have a very broad spread. I will also make a note here that while I use age in my example, we take in a multitude of variables into account here, and age is not one of them. The reason for not using age is that the combination of our other values shows us a more accurate power level than something on the integer scale of 1-16. We use this power spread to try to match alliances to similar power spreads so everyone in the war is more likely to have someone else they can be on even footing with. But, the longer you're waiting in matchmaking the less we require the power spreads to match so you can get into War in a decent time.

Step 4: We toss the alliance into matchmaking! This is where all of the alliances have their values compared and we attempt to find the best match we can in the least amount of time we can. Something you may realize though is that those are two goals that contradict each other. If we have you wait longer, then the chances of you finding a better match increase as you see more opportunities, but that leads to longer matchmaking times. While if we reduce matchmaking times then you guys get to War faster, but the overall quality of everyone's matches will go down. Its a tough balancing act!

Step 5: War!

I've simplified things down here, and maybe have shared a bit more than I have in other places but I hope this gives everyone an idea of how we go about matching two alliances. In short: glory (which I didn't mention but I believe everyone knows about by now), alliance war weight, and alliance power spread are what we look at when deciding who should War against each other.

EDIT: We look at your tactics, not your troop tactics, my b!

Cs was asked if munitions add war weight and they said that they didn’t. Is cs misinformed, because you make it sound like they do.

 

[IDaedalusI]

Hey @UnHoLyGhOsT,

Munitions do in fact add to your war weight. @Harlems369th helped me in giving our latest world war matchmaking documentation to our CS team so everyone is on the same page

 

[UnHoLyGhOsT]

Hey @UnHoLyGhOsT,

Munitions do in fact add to your war weight. @Harlems369th helped me in giving our latest world war matchmaking documentation to our CS team so everyone is on the same page

So my next question. When you assign war weight for matching. Do you include everything that’s been manufactured in a players inventory or do you only include what is equipped at the time of the war search?

 

[oddin]

So my next question. When you assign war weight for matching. Do you include everything that’s been manufactured in a players inventory or do you only include what is equipped at the time of the war search?

you are a sneaky one eh? lol

I really hope the answer is what is produced and not what is currently assigned

 

[UnHoLyGhOsT]

you are a sneaky one eh? lol

I really hope the answer is what is produced and not what is currently assigned

I’ve heard of some players dropping in war line up over removing defense munitions. Just trying to get some clarity.

 

[King Crimson]

I’ve heard of some players dropping in war line up over removing defense munitions. Just trying to get some clarity.

It's crazy that the game allows this. The game should take a snapshot of what's activated at the time of a war search - and then anything that's changed after the fact doesn't get activated for that player's attacks. Like adding buildings afterward ...

 

[nobodyknowsthetrouble]

If freezing, then please do it properly. In other words, everything (museum, council, coalitions, trained troops, airplanes, hired mercenaries)
This shifts many decisions from the preparation day to the day or the minutes/hours before a new search for war. It also removes a bit of strategy, as an attacker can no longer switch between mortars and tanks to match the enemy museum, for example - just as a defender can no longer decide to weaken a particular invasion unit based on the enemy's armies. But I personally have no problem with this; it also solves the nonsense of some players trying to hide their real museum with an empty or fake museum.

But: What is this nonsensical idea with the buildings? Then I still have 3 transporters on the eighth day due to the freeze, but because the building has already expired, I can only use 2 of them in multiplayer, but can take 2 other planes with me? How do you expect this to work? The same applies to buildings that produce tanks or helicopters, for example; in war I can use them through the freeze, but not in multiplayer. Do I then have 2 armies to manage, one multiplayer and one war? If I donate troops, from which of the two armies will this be deducted?
It's fine the way it is. A building lasts 7 days; that's predictable for everyone and therefore fair.

 

[IDaedalusI]

So my next question. When you assign war weight for matching. Do you include everything that’s been manufactured in a players inventory or do you only include what is equipped at the time of the war search?

I'm sorry guys but this question is a bit too specific for me to answer All that I can say here is that munitions are taken into account.

 

[King Crimson]

All that I can say here is that munitions are taken into account.

I think what OP wants to know is:
munitions taken into account = ALL munitions manufactured?
or
munitions taken into account = munitions assigned to the troops in the war?

Quite a simple question really ...

 

[IDaedalusI]

I think what OP wants to know is:
munitions taken into account = ALL munitions manufactured?
or
munitions taken into account = munitions assigned to the troops in the war?

Quite a simple question really ...

You're right about it being a simple question, but that doesn't mean I'm allowed to answer it

If I answer this simple, but specific, question, then it is much more likely for someone else to come along and find a way to manipulate or take advantage of our matchmaking system. At that point, our ability to improve matchmaking becomes harder because we are no longer just trying to improve the complex algorithm for normal players, we would have nefarious actors affecting our numbers as well.

No one wants that to happen...

 

[UnHoLyGhOsT]

You're right about it being a simple question, but that doesn't mean I'm allowed to answer it

If I answer this simple, but specific, question, then it is much more likely for someone else to come along and find a way to manipulate or take advantage of our matchmaking system. At that point, our ability to improve matchmaking becomes harder because we are no longer just trying to improve the complex algorithm for normal players, we would have nefarious actors affecting our numbers as well.

No one wants that to happen...

Thanks for the detailed response

 

[Seek]

Hey @sevenjapan,

I've answered questions around this in a few other posts on the forums, but I'll take your post as an opportunity to collate a bunch of it in one place. While I can't share every detail of our algorithm (people might take advantage of it then!). I can share some of the basics!

Step 1: When an alliance enters matchmaking the leader had to say who was joining them in war. We take this list of players and then look through their main bases. On those bases we look at things like buildings, troop tactics, munitions, etc. and their levels. We use this to assign a "war weight" to each player. I'll also note here that if a player somehow manages to not have all of their buildings on their war base, that just puts them at a disadvantage because we counted everything on their main base. War bases are just for you guys to reorganize things as you wish

Step 2: We do some math. We combine all of the war weights of each player in the alliance into a total "alliance war weight" We don't do a simple summation here because we don't want to heavily disincentivize alliances from bringing along a couple weaker players to show them the ropes.

Step 3: We do some more math. We also figure out an alliance's power spread. This is our attempt to handle the scenarios where 1 alliance may have 10 drone age players, and another alliance may have 1 player in each age (I do not recommend doing this by the way ). The first alliance will have a tight spread, while the second will have a very broad spread. I will also make a note here that while I use age in my example, we take in a multitude of variables into account here, and age is not one of them. The reason for not using age is that the combination of our other values shows us a more accurate power level than something on the integer scale of 1-16. We use this power spread to try to match alliances to similar power spreads so everyone in the war is more likely to have someone else they can be on even footing with. But, the longer you're waiting in matchmaking the less we require the power spreads to match so you can get into War in a decent time.

Step 4: We toss the alliance into matchmaking! This is where all of the alliances have their values compared and we attempt to find the best match we can in the least amount of time we can. Something you may realize though is that those are two goals that contradict each other. If we have you wait longer, then the chances of you finding a better match increase as you see more opportunities, but that leads to longer matchmaking times. While if we reduce matchmaking times then you guys get to War faster, but the overall quality of everyone's matches will go down. Its a tough balancing act!

Step 5: War!

I've simplified things down here, and maybe have shared a bit more than I have in other places but I hope this gives everyone an idea of how we go about matching two alliances. In short: glory (which I didn't mention but I believe everyone knows about by now), alliance war weight, and alliance power spread are what we look at when deciding who should War against each other.

EDIT: We look at your tactics, not your troop tactics, my b!

Thank you for actually coming in here and taking time to explain.

Now I did see you say that a player that does not have all his buildings on War Base is at a disadvantage this I have to point out is incorrect.
Here is why. The player that would not update War Base to hide buildings is most likely an Offense Player and by hiding the players FSH and Hangars (NightHawk) they put the defender at disadvantage cause these buildings work in war and ghosts on War Base. The defender can’t scout this or spy ends up making wrong selection of coalitions. I won’t get into how FSH should be listed as a Offensive Building.
They don’t care about having these 2 buildings their offense. So ghosting has been and will always be offensive advantage.

I can’t put in Zook tower in mid war or ghost it and have it work just saying I would love to have Zook Tower that can’t be seen

 

[King Crimson]

You're right about it being a simple question, but that doesn't mean I'm allowed to answer it

If I answer this simple, but specific, question, then it is much more likely for someone else to come along and find a way to manipulate or take advantage of our matchmaking system. At that point, our ability to improve matchmaking becomes harder because we are no longer just trying to improve the complex algorithm for normal players, we would have nefarious actors affecting our numbers as well.

No one wants that to happen...

Appreciate the reply.
So why not just make it that any changes applied after the successful matchup don't come into effect?

 

[UnHoLyGhOsT]

Appreciate the reply.
So why not just make it that any changes applied after the successful matchup don't come into effect?

Let’s not give them any ideas. Anything I have set during planning day. I want to know that it’s working. Museum, council and munitions

 

[Seek]

Let’s not give them any ideas. Anything I have set during planning day. I want to know that it’s working. Museum, council and munitions

Well I would suggest as a mid way point between unholy ghost and king.

If something is not set prior to war start and visible on war screen then that buildings effect should not be allowed in war. The defender can’t add ghosted ZooK Towers to his base and still have them defend. So stop offense from getting to use ghosted buildings in war. Either you set before war starts visible in war base or can’t use it. This also gives offense advantage on duration time. They open FSH and Hanger right before war hit while the defender has to set prior to war start cost them the full time when O player only starts timers when they are ready to hit gives them extra day use of building over real defense buildings ( FSH is just I paid to beat you building)

 

[King Crimson]

If something is not set prior to war start and visible on war screen then that buildings effect should not be allowed in war.

That's what I'm saying.