Element Shakeup

[Karawasa]

Then let them stay sucky relative, and not absolute .

[Cisz]

Aoe slow will be nerfed, right? So all builds will move closer to each other.

[Karawasa]

Yup.

[Cisz]

As far as i can tell, there are at most 6x5x4x3x2 possible arrangements that would fullfill the requirememnts of this table:

            1 2 3 4 5 6 7 8 <- Tower numbers
(Element 1) +       + + +
(Element 2) + +       +   +
(Element 3) + + +       +
(Element 4)   + + +       +
(Element 5)     + + +   +
(Element 6)       + + +   +

We could fill in the elements into the first collum in 6x5x4x3x2=720 different ways. I would expect that some of them are equivalent, but even if we get a factor like 4 or six out of this, we are still talking three figures. O_o

Those different arangements would greatly differ in what damage types are available to what build and this can affect balance. For example, a build with wnl in it likely has no weak element, while a build like edwn is probably weak vs water and dark.

It would make sense to arrange the elements in a way that strengthens the builds w/o strong towers.

[Cisz]

I've done some further calculation, and it seems that if I want to have one of the three "no good towers" builds a set of three elements that cover each other (efd or lwn), automatically the following happens:

-2 of the three bad builds get such a set of three, one each, and the third gets no such set.

-None of the game winning towers is made out of three elements of such a set (naturally, as both sets a part of a "no good towers" build).

-The number of possible element arrangements that work with the 8 game winning towers distribution and the "three of a set" limitation is not 720 but 144. Hopefully I will find a quick way to provide a list of all this possible distributions, complete with what elements are in what build, and which game winning towers have what elements. We could chose one of those 144 for best fit with the existing towers.

[Guest mrchak]

does the 144 count duplicates due to the order of the elements?

[Cisz]

Here is the list of game winning towers with their elements. (S = aoe slow, A = armour reducer, ? = other game winning tower) The upper line of 6 zeroes is were the names of the elements belong.

Here's the resulting structure again, the bad builds are blue:

The three bad builds are:

 Element number
   1 2 3 4 5 6
1  + +   + +  
2  +   + +   +  Builds
3    + +   + +

There are 4 arrangement types for elements that give two of the bad builds 3 matching elements:

1 2 3 4 5 6
  #   # #  
#     # #  
# #     #  
# #   #

All "#" are elements of one set (say efd) and all " " are the other set (wnl).

For each of the four sets we have 3x2 posibilities to arrange the first set into the first three slots and 3x2 for the second set. So we have 4x3x2x3x2= 4x6x6 = 144 possible ways to fill in the elements.

Oh - I just noticed that I missed half of the possibilities, as there are two possibilities to chose from when deciding which set goes into which half of the build. So we still have 288. Damn.

I have hope that many of those 288 are very similiar or even identical due to rules I just didn't notice yet.

[Cisz]

MrChack just had a divine moment and channelled pure elemental essence into his creativity gland.

He suggested having a pair of two towers that get a bonus when combined, and to place that pair into one of the weak builds.

Actually it's possible to place tree of such pairs in a way that each bad build gets exactly one of them and no other build gets any bonus combo.

Here is how it works.

8 game winning towers are aranged like this:

   e1 e2 e3 e4 e5 e6 <= Elements, yet to be sorted
S1     1  1  1
S2        1  1  1
S3  1           1  1
S4  1  1            1
A1  1  1  1
A2           1  1  1
?1  1     1     1
?2     1     1     1

S = Aoe slowing towers
A = Armour reduction towers
? = Other game winning towers

This leads to the following 15 4 element builds:

    Elements                       Game winning
    in build                      towers in build

e1 e2 e3 e4 e5 e6 | Nr.| S1 S2 S3 S4 A1 A2 ?1 ?2 |   |
------------------+----+-------------------------+---+-------
 1  1  1  1       |  1 |  1  0  0  0  1  0  0  0 | 2 | S1, A1
 1  1  1     1    |  2 |  0  0  0  0  1  0  1  0 | 2 | A1, ?1
 1  1  1     1    |  3 |  0  0  0  1  1  0  0  0 | 2 | S4, A1
 1  1     1  1    |  4 |  0  0  0  0  0  0  0  0 | 0 | 
 1  1     1     1 |  5 |  0  0  0  1  0  0  0  1 | 2 | S4, ?2
 1  1     1  1    |  6 |  0  0  1  1  0  0  0  0 | 2 | S3, S4
 1     1  1  1    |  7 |  0  1  0  0  0  0  1  0 | 2 | S2, ?1
 1     1  1     1 |  8 |  0  0  0  0  0  0  0  0 | 0 | 
 1     1     1  1 |  9 |  0  0  1  0  0  0  1  0 | 2 | S3, ?1
 1     1  1  1    | 10 |  0  0  1  0  0  1  0  0 | 2 | S3, A2
    1  1  1  1    | 11 |  1  1  0  0  0  0  0  0 | 2 | S1, S2
    1  1  1     1 | 12 |  1  0  0  0  0  0  0  1 | 2 | S1, ?2
    1  1     1  1 | 13 |  0  0  0  0  0  0  0  0 | 0 |
    1     1  1  1 | 14 |  0  0  0  0  0  1  0  1 | 2 | A2, ?2
    1  1  1  1    | 15 |  0  1  0  0  0  1  0  0 | 2 | S2, A2
------------------+----+-------------------------+---+-------

The following builds are the bad builds so far:

e1 e2 e3 e4 e5 e6 | Nr.|
------------------+----+
 1  1     1  1    |  4 |
 1     1  1     1 |  8 |
    1  1     1  1 | 13 |
------------------+----+

Now we need three pairs of triples, one of the combos included in each of the three bad builds and none of the pairs showing up in any other build. The towers that are used for these pairs are automatically non-game-winning towers so far, as they are in bad builds.

The first bad build (no.4) has 4 triples:

e1 e2 e3 e4 e5 e6 
------------------
 1  1     1       
 1  1        1    
 1        1  1    
    1     1  1    
------------------

Now here is where the fun begins: No other build has more than one of those 4 towers. We can chose any two of them and call the pair an overkill combo. Or we could make the combo require 3 or all 4 of them.

Same goes for all other bad builds or in fact any of the 15 4 element builds: No pair of two triple towers exists in more than one build.

So by chosing any combination of towers, one in each of the three bad builds, and awarding them a combo bonus equal to two game winning towers, all 4 element builds would be valid.

Now I have to check the 5 and 6 element builds and make sure they don't get to strong. I expect problems here, so stay tuned for more in the next post.

(Hurries back to the build lab...)

Edit: Hmmm..

I just noticed that this idea has been suggested before in this thread, and that I had a good reason against it: Any combo that boosts a single 4 element build will show up in 2 5 element builds and add to the allready available 4 game winning towers in it.

So every 5 element build would have the equivalent of 6 game winning towers, compared to 2 of them in the 4 element builds. And yes, all of those towers are maxed in the end.

If we use brute scripting force, there is a nasty workaround: Award the bonus for the combo only if no other game winning towers and /or combos have been built on the map/are available to the player.

Something like: While (build=sucks) do (improve build).

It looks really evil and arbitrary to do such a thing.

Maybe we could make the three bad builds kind of themed (the dragon special extravaganza, the brutal battlegear build, the hobbit hunters delight..), so sticking to the 4 elements would feel less forced and loosing the special bonus would apear a bit more reasonable.. Like a special hand in mah yong. "If you limit yourself to this 4 elements, the gods will favour you."

Your opinions?

Is this a dead horse? Is having all builds being valid worth the trouble? Is a build that loses a big bonus if you add the wrong element to it to strange to be acceptable?

[Karawasa]

I believe this is a dead horse. That would overly complicate the game in my opinion.

Having 12 good and 3 bad builds is damn good performance. Especially, when "bad," will simply mean not the best option in competitive play. Take DotA, majority of heroes do not see competitive play.

[Cisz]

I hear the clear sound of reason..

Ok, off to think up some more cool new towers then.

[Guest mrchak]

MrChack just had a divine moment and channelled pure elemental essence into his creativity gland.

Guess I wasted that lumber for nothing.

I still maintain that it is fun to have towers that have cool combination effects. But I see where you're going and want there to be no benefit to have a 5-element build rather than a 4-element build.

Well.. I get another lumber in 5 turns, hopefully I will use it more wisely.