Psa: Stop upgrading your defenses!!!

[Johnnysks]

Unthinkable! lol

[dantoddd]

Well I don't know if Moskri is accurate or not. But, its simple to get your war weight. just check the gold stored in your storage on 'war day', multiply by 4 and add 1000. plain and simple.

[jmnilsson]

The easiest way to calculate your warweight (it is not as accurate as the spreadsheet) is to look up your gold in 1 maxed storage during warday in your warbase so not prep during prep day but only warday.

So say i have 18.000 gold in 1 of my storages during warday. Im a maxed TH9 so I have 4 maxed storages. So the way to calculate is 18.000 x 4 (amount off maxed storages) + 1000 (amount off gold in TH) = 73.000. So my warweight is 73.

If you dont have maxed storages you just accumulate all the storage gold individually + 1000.

This is an pretty accurate and easy way to calculate your warweight. There are a few other ways but it is harder to explain.

I'm well aware of adding up the gold in the storages + 1000 for the TH. Thats standard practise for my clan. That wasn't what I was referring to, I was actually asking about the spreadsheet.
As stated by other posters in the thread they see other numbers (in the gold weight) when adding defenses such as x-bows or Inferno Towers. So unless you can actually either prove it to us or give us the means to check it out ourselves, what you state is questionable at best as it goes against the commonly accepted truth.

As you come in with something new that is different from the reigning theory, the burden of proof lies at your feet from my point of view. You can have the exact numbers for the matchmaking, but how are we supposed to know you are right?

[Devjo]

I'm well aware of adding up the gold in the storages + 1000 for the TH. Thats standard practise for my clan. That wasn't what I was referring to, I was actually asking about the spreadsheet.
As stated by other posters in the thread they see other numbers (in the gold weight) when adding defenses such as x-bows or Inferno Towers. So unless you can actually either prove it to us or give us the means to check it out ourselves, what you state is questionable at best as it goes against the commonly accepted truth.

As you come in with something new that is different from the reigning theory, the burden of proof lies at your feet from my point of view. You can have the exact numbers for the matchmaking, but how are we supposed to know you are right?

Well me and my clanmates used several ways to re-calculate the stats and like i said in my previous post. The easiest way to re-calculate it was by the way I told you. There are other ways with totall loot to weight calculation. But they are harder to explain.

I was just giving my opinion that the stats are correct by my calculations. If the others feel that they arent correct maybe they can give the correct way to calculate them. Maybe I missed those posts but I just saw posts that my calculations were incorrect no posts that tell me what to change in them.

It is always easy to post that something isnt right and leave it at that. Im always open to learn and try new calculations. But if no one posts them. Not much i can do with that

[Humbob]

devjo, to disprove moskris spreadsheet is as simple as this:

take any given base of any given town hall, maybe a new th8 that just hit th9, and add a defense, changing nothing else. compare the gold weights. an inferno, level 1, as noted above, will be 4k on a base that starts in that 60's weight range or up. On a defenseless th10 I am sure it will be much, much higher.

As noted above, moskri's sheet doesn't account for non linearities, which do exist, and is just plain crazily inaccurate.

Its easily tested yourself, and the youtube video your gang put out calling mortars and wizard towers the hgihest weighted defense is... unfortunate. You could mislead alot of people about alot of things by taking that spreadsheet seriously. Didn't it also have little bombs as higher in weight than inferno towers or something?

[jmnilsson]

Well me and my clanmates used several ways to re-calculate the stats and like i said in my previous post. The easiest way to re-calculate it was by the way I told you. There are other ways with totall loot to weight calculation. But they are harder to explain.

I was just giving my opinion that the stats are correct by my calculations. If the others feel that they arent correct maybe they can give the correct way to calculate them. Maybe I missed those posts but I just saw posts that my calculations were incorrect no posts that tell me what to change in them.

It is always easy to post that something isnt right and leave it at that. Im always open to learn and try new calculations. But if no one posts them. Not much i can do with that

All I really want is a way for me to actually check for myself if your numbers are correct. So far you have not been that forthcoming of how I can get access to your spreadsheet and how I can assess their validity. This to me speaks volumes.

As for missing the posts where they stated your values where incorrect check the below quoted parts. If I´m not mistaken they all rely on adding up the resource amounts in the gold storages during Battle Day.
For suggestions on what you should change in your calculations that is quite hard for us to do as we don´t know anything about your calculations, what they look like and what they entail. We need to actually see them you know for us to being able to give suggestions on them.

Until you actually give us something to check out we can only give you the answers that you already have received, if you want actual good feedback the ball is in your corner.

That spreadsheet is TOTALLY wrong. The worst thing that ever came out in CoC. So many people based their upgrade path on that stupid thing. Not even close to accurate.

ONE level 1 inferno tower adds 4k weight. That sheet has it as 1k per level. ONE level 1 xbow is 2k weight. The point d and splash are nowhere near the numbers they say also.

It's also the case that weights are nonlinear. Moskri's numbers do not in any way account for that.

In our current war, we thought it'd be a challenge so I rented ITs. Added the expected 8k (or maybe 9k, I'm up an extra k this war, will see where I drop back to next war). Turns out they couldn't attack, like really can't attack. I had a max troops/camps 20/20 TH10 hit my 71k base (that's the with infernos weight, was 62k last war) for 40% and 0 stars!

However, somewhat humorously, a TH9 no defense account in the bottom of our map rented Xbows. Went from 15k to 28k with 2 level 1 xbows. 6.5k per level 1 Xbow! It's over 3x as much "weight" in that region of the rating formula as is typically seen in the normal TH9 region (I agree it's usually 2k for a TH9 for a level 1 Xbow)

devjo, to disprove moskris spreadsheet is as simple as this:

take any given base of any given town hall, maybe a new th8 that just hit th9, and add a defense, changing nothing else. compare the gold weights. an inferno, level 1, as noted above, will be 4k on a base that starts in that 60's weight range or up. On a defenseless th10 I am sure it will be much, much higher.

As noted above, moskri's sheet doesn't account for non linearities, which do exist, and is just plain crazily inaccurate.

Its easily tested yourself, and the youtube video your gang put out calling mortars and wizard towers the hgihest weighted defense is... unfortunate. You could mislead alot of people about alot of things by taking that spreadsheet seriously. Didn't it also have little bombs as higher in weight than inferno towers or something?

[MrKel]

I highly disagree. Level 1-2 xbows are useless. Level 3-4 are VERY useful. They take out witches, wizards, minions very quickly and from a great distance. Especially if you can leave your xbow on ground, like I do some wars at TH10, because 99.9% of the people only gowixx. HUGE difference. On top of the fact they are have over 2k hps that takes troops quite awhile to destroy, while being pummeled by other defenses.

Xbows are far too easy to out-strategize. You aren't reveling in the fact that they vaporize witches/wizards/minions -- what you're enjoying is the attackers lack of skill. Well placed tank units render them inert, as do freeze spells.

Sure, there's loads of terrible strategists in the game for your premature xbows to nom on, but I don't think they offset what a stupid upgrade choice does to war weight & placement. You might overpower bad attackers, but you're putting yourself in league with a higher tier as a consequence. It's just not worth it.

There is a pretty remote case in the game whereby getting those L3/4 xbows makes sense, and it's when you're almost MAX TH10 (with all troops/heroes upgraded) and have nothing else to spend your resources on. Until that time, they're just plain dumb.

[Johnnysks]

I think the point is moskris model is inaccurate. As many have stated, weighing seems to non linear, which moskris disclosed spreadsheet never accounted for thus unusable. As the foundation of your strategy may not be accurate (or at least lacks certainty as its been debunked many times by now), you may want to revise the confidence with which you use the data as guidance, or try to work out a model that accounts for non linearity (not easy). Good luck.

[Sservis]

Its easily tested yourself, and the youtube video your gang put out calling mortars and wizard towers the hgihest weighted defense is... unfortunate. You could mislead alot of people about alot of things by taking that spreadsheet seriously. Didn't it also have little bombs as higher in weight than inferno towers or something?

The type of test that I will believe is as follows. Note that all of this is invalid if the base's weight is not based on an order preserving transformation from the total of the weight of the individual components.

How to compare defenses to figure out which is "heavier"

1. Take a few defenseless accounts, preferably heroless, CC at level 1 also
2. Build a given defense to a given level on each defenseless account.
3. Compare war weights.

I would not trust the ratio of the weights to mean anything, but I would trust the relative weights.

Given what I know about math and the additive nature of defenses, I would be tempted to believe that

if and only if
defenseless + wizard tower 8 >= defensesless + inferno tower 3
then
wizard tower 8 >= inferno tower 3

However due to nonlinearities, I wouldn't extrapolate from say defenseless + wizard tower 8 = 20k and defenseless + inferno tower 3 = 30k that the inferno tower costs 50% more (numbers made up). I would only be willing to provisionally trust that it costs more. It could be 1% more, it could be 1000% more.

It would take a lot of work, but you could create a clan (or two or more) with many accounts and war test them to find relative weights by checking how they sort on the war map. A lot of these could be low level and fairly quick, but you're going to need a lot of high level accounts. This is something I've thought about doing from time to time at least on a small scale, but haven't yet had the patience to create and level all the accounts (or even to start the process).

CC - 5 (note that these only measure the relative cost of the upgrade from CC-1 to CC-X)
Cannon - 13 (to be fair these all need a Cannon 1 AND a Cannon of the level being tested)
Archer Tower - 13
Mortar - 8
Air Defense - 8
Wizard Tower - 8
Air Sweeper - 6
Tesla - 8
Xbow - 4
Inferno - 3
Bomb - 6
Spring Trap - 1
Giant Bomb - 4
Air Bomb - 4
Seeking Air Mine - 3
Skeleton Trap - 3
BK - 40
AQ - 40


Total - 177 (not counting walls if you want to test those too)

If you want to check what 25 walls of each level rate as, that's another 11 items. (25 might be too many or too few, if you want to check all groups of 25 (25, 50, 75 ...)) of all wall levels, it's 121 item.


If you really want to answer questions like which upgrade costs more, Item A from level M to N or Item B from level X to Y, build a base with Item A at level M and Item B at level Y and compare it to a base with Item A at level N and Item B at level X. The ordering will give you which upgrade is more expensive.


PS- If you really wanted to get specific, you could use walls somehow to equalize the weights until items swapped positions and then you'd have a wall based measure on weight distance. ie Wiz Tower 8 + 20 level 11 and 13 level 1 walls rates below Inferno Tower 3, but Wiz Tower 8 + 20 level 11 and 14 level 1 walls rates above, etc. If you figured out the wall equivalence ratios you could maybe even state that they were 10,727 or 10728 level 1 walls apart and start to measure defenses in the smallest possible unit. (level 1 walls, or maybe in terms of # of spring traps or whichever level 1 trap is lightest, I'm partial to walls because there are 275 of them)

PPS- If anyone wants to go to the trouble of creating a few hundred gmail accounts, I'll go to the trouble of starting to do this.

[Humbob]

That would be interesting, Sservis, but its beyond the scale of my interest. I'm interested largely in a general idea of matching, as like with everything in Clash, its simply subject to change.

btw I referenced your rushing comments in another thread, so I apologize ahead of time if I've butchered them.