Clan War Matchmaking Improvements

The last 2 war searching has taken forever. Has anyone else experienced this. Last war it took 8 hours to find a match and so far today we are at 6.5 hours with nothing found yet. 50 vs th11 to th8s

Quote:

---

Originally Posted by Dillinger

Th match should be the base of the initial weight per member. You add weight based on each def lvl for total weight.
It's the best way to handle mm. You can recover one or two extra th10s or 11s. 7 to 1 is ridiculous.

---

I think you are describing the often suggested system where each hall weighs at minimum what a max hall one level below weighs. I long resisted such suggestions and instead supported just fixing the weights, but as time has gone one I've started to think a hammer type solution might be best since the scalpel has not worked. Let's see how this upcoming change does (and we will need to give it some time to be adjusted) before going there, though.

Quote:

---

Originally Posted by Pobilly

The last 2 war searching has taken forever. Has anyone else experienced this. Last war it took 8 hours to find a match and so far today we are at 6.5 hours with nothing found yet. 50 vs th11 to th8s

---

Because not many clan can do 50vs50 wars.

Adding weight to offense is a bad move. Using Defense as the pivot point is what has worked. And it makes sense why it does. You are applying the algorithm to only one side of two variables. Adding weight to offense will bring, perhaps is what is bringing crazy wars. The MM has been tweaked recently, and wars have never been so bad in the history of clan wars.

I'm assuming the idea behind adding weight to offense is to curb the defenseless base by giving it more weight as it has more offense than defense. Except, by adding weight to offense, you are also adding weight to the other side of the coin.

There are two types of wars: The winnable war, and the impossible to win war. Inside the winnable war, you have unfair wars. The unfair war, can be won, but they are unfair because the opposition has more chances to attack the top bases. So, yea, that needs to be fixed.

How do we fix it? Now, the MM was not spinning impossible to win wars, it was giving unfair wars because of the defenseless bases. So the MM was not broken, that is to say, it was not spinning impossible to win wars as we see it now. In short, using defense as the ruling variable, works.

The solution is so simple. Yes, you look at the offense level base has, but you don't add weight to it's offense, you take the offense level and add it to it's defense weight.

In short, the MM is fine as it was, it has been fine for over a year now. Revert to it, just add an algorithm that will look for the offense level of a base, and give it a corresponding defense weight. A defenseless TH11 with max heroes and troops, should have a minimum defense weight of a max TH10. So just add that to it's defense weight. That's all you need to do SC.

Quote:

---

Originally Posted by iBork

Adding weight to offense is a bad move. Using Defense as the pivot point is what has worked. And it makes sense why it does. You are applying the algorithm to only one side of two variables. Adding weight to offense will bring, perhaps is what is bringing crazy wars. The MM has been tweaked recently, and wars have never been so bad in the history of clan wars.

I'm assuming the idea behind adding weight to offense is to curb the defenseless base by giving it more weight as it has more offense than defense. Except, by adding weight to offense, you are also adding weight to the other side of the coin.

There are two types of wars: The winnable war, and the impossible to win war. Inside the winnable war, you have unfair wars. The unfair war, can be won, but they are unfair because the opposition has more chances to attack the top bases. So, yea, that needs to be fixed.

How do we fix it? Now, the MM was not spinning impossible to win wars, it was giving unfair wars because of the defenseless bases. So the MM was not broken, that is to say, it was not spinning impossible to win wars as we see it now. In short, using defense as the ruling variable, works.

The solution is so simple. Yes, you look at the offense level base has, but you don't add weight to it's offense, you take the offense level and add it to it's defense weight.

In short, the MM is fine as it was, it has been fine for over a year now. Revert to it, just add an algorithm that will look for the offense level of a base, and give it a corresponding defense weight. A defenseless TH11 with max heroes and troops, should have a minimum defense weight of a max TH10. So just add that to it's defense weight. That's all you need to do SC.

---

It did spin impossible to win wars and it wasn't fine in my opinion, but it might have been better than it is currently.

Quote:

---

Originally Posted by 2222

It did spin impossible to win wars and it wasn't fine in my opinion, but it might have been better than it is currently.

---

Yes, in some cases, it was spinning impossible to win wars. But the reason had to do with the defenseless base. As an example, a clan whose top base was a TH10 or less, being matched with a clan that had defenseless TH11 bases with max troops is an impossible to win war.

But that problem would be solved if the MM was revised as per what I suggested.

Quote:

---

Originally Posted by iBork

Yes, in some cases, it was spinning impossible to win wars. But the reason had to do with the defenseless base. As an example, a clan whose top base was a TH10 or less, being matched with a clan that had defenseless TH11 bases with max troops is an impossible to win war.

But that problem would be solved if the MM was revised as per what I suggested.

---

I've always thought roster engineering, whether through defenseless bases or a lot of low weight bases, was the worst kind of engineering. I've never faulted engineers for building their bases in a way that they think is good for war, but I don't feel the same about clans who intentionally manipulate the roster. I really hope the upcoming update has that problem in mind, not just matching engineers vs. engineers.

Quote:

---

Originally Posted by 2222

I've always thought roster engineering, whether through defenseless bases or a lot of low weight bases, was the worst kind of engineering. I've never faulted engineers for building their bases in a way that they think is good for war, but I don't feel the same about clans who intentionally manipulate the roster. I really hope the upcoming update has that problem in mind, not just matching engineers vs. engineers.

---

I can agree with you to a point. My view on the subject is that a clan needs to place some thought into how they upgrade their bases. We've never encounter a problem with roster engineering. Matter of fact, it was kind funny when we saw them because it didn't help them at all. But at that time, the MM was working well.

As I see it, there are two ruling factors in clans wars. The MM, and the players. Many have unrealistic expectations from the MM and believe the MM should spin good wars no matter how their bases are built. This is false. You have to "engineer" in some way.

From our experience in clans wars, we have been war'ing for over 4 years, and have tried all types on combinations. What we have learned, is that maxing a base is what brings you hard wars, no matter how perfect the MM is. To this aspect, no MM will ever favor that impossibility. This is an opinion based on trial and error.

My point is, the majority of complaints, and type of clans that people complain about, is not so much the fault of the MM, but their lack of planing, or experience, in how to build good bases for war. So this causes pressure on SC to fix it. The problem is, you can't fix something that isn't broken.

I wish SC would give players at least some kind of guidance with regards to clan wars. That's what I feel is the missing link. People like me, saying that a maxed base is the worst thing you can have in war falls on death ears. The thing is, it's true.

Quote:

---

Originally Posted by iBork

I can agree with you to a point. My view on the subject is that a clan needs to place some thought into how they upgrade their bases. We've never encounter a problem with roster engineering. Matter of fact, it was kind funny when we saw them because it didn't help them at all. But at that time, the MM was working well.

As I see it, there are two ruling factors in clans wars. The MM, and the players. Many have unrealistic expectations from the MM and believe the MM should spin good wars no matter how their bases are built. This is false. You have to "engineer" in some way.

From our experience in clans wars, we have been war'ing for over 4 years, and have tried all types on combinations. What we have learned, is that maxing a base is what brings you hard wars, no matter how perfect the MM is. To this aspect, no MM will ever favor that impossibility. This is an opinion based on trial and error.

My point is, the majority of complaints, and type of clans that people complain about, is not so much the fault of the MM, but their lack of planing, or experience, in how to build good bases for war. So this causes pressure on SC to fix it. The problem is, you can't fix something that isn't broken.

I wish SC would give players at least some kind of guidance with regards to clan wars. That's what I feel is the missing link. People like me, saying that a maxed base is the worst thing you can have in war falls on death ears. The thing is, it's true.

---

Many engineers share the opinion that it is good to be rewarded for the knowledge of how to build your base to take advantage of the matchmaker's faults. I don't think that is an unreasonable opinion, even though I don't really agree (I prefer a matchmaker that causes the outcome to be as much as possible determined by the attacking and base designing skill of the players, not what they did to their base before war). However, I do think we should acknowledge them as faults. If the matchmaker overweights something so much that it is better to not even build it at all, that is a fault. It improperly weights it. If the matchmaker is going to give advantages (a perfect matchmaker would give none in my opinion, which isn't possible) then it would seem to me to make the advantages be the type of progression path that is intuitive, without having to research how to take advantage of the matchmaker. In other words, not giving an advantage to literally not building something. That said, the knowledge of the matchmaker faults has spread so much now that this type of problem is not nearly as important to me as the problem of roster engineering.

Quote:

---

Originally Posted by 2222

Many engineers share the opinion that it is good to be rewarded for the knowledge of how to build your base to take advantage of the matchmaker's faults. I don't think that is an unreasonable opinion, even though I don't really agree (I prefer a matchmaker that causes the outcome to be as much as possible determined by the attacking and base designing skill of the players, not what they did to their base before war). However, I do think we should acknowledge them as faults. If the matchmaker overweights something so much that it is better to not even build it at all, that is a fault. It improperly weights it. If the matchmaker is going to give advantages (a perfect matchmaker would give none in my opinion, which isn't possible) then it would seem to me to make the advantages be the type of progression path that is intuitive, without having to research how to take advantage of the matchmaker. In other words, not giving an advantage to literally not building something. That said, the knowledge of the matchmaker faults has spread so much now that this type of problem is not nearly as important to me as the problem of roster engineering.

---

It's not unreasonable to want a good MM. We all agree on that. Where we differ in opinion, is that the MM can only do so much. And when you try to make fair to all types of bases, it breaks it. There is a middle ground. And that middle ground is not a moving target. It is where it is.