Progressing Soulless Horror
A few weeks ago, or a month or.. I don't actually remember. Either way back when we were progressing soulless horror we often heard rumors about the big walls appearing when the soulless horror, the boss, is pulled out of the middle. Actually watching this behavior and having the tanks play around with it I can say with 70% certainty, myth busted.
With this myth busted the next question is how it actually works. What's the chance of a wall forming in a certain constellation? Is there a safe or safer spot on the platform?
Analyzing and Assumption
From a game developer perspective and taking into consideration that it can't be scripted as it's too random. Additionally, walls form using pieces. The fight starts out with two pieces. Once the platform shrinks the first time the wall increases by one piece having three now. Once the soulless horror reaches 30% health the platform shrinks again and now the wall consists of up to four pieces. Additionally, the walls can only cover a certain row or column of the field. Another thing you might have noticed is that the walls snap.
Using all this information we can assume that the walls spawn in a grid. Further analyzing this assumption we notice that the grid has a size of five times five. In this grid, the walls spawn either horizontally or vertically and run down to the opposite side of their spawn point.
Combinations
So in the first phase, the walls consist of two pieces filling out five fields on four different sides.
This allows there to be 2 out of 5 possibilities for four sides with a maximum of two waves being active at the same time. This allows there to be C(5, 2) = 10 possibilities for each side. So that should be 10 * 4 = 40 possible spawn combinations.
However, as I already mentioned before the amount of wall pieces increases. For the most time of the fight, it will be three pieces. With three pieces we have C(5, 3) = 10 possibilities. Okay yeah, times four get 40 okay.
At the end of the fight, it will be four pieces. So now it's C(5, 4) * 4 = 5 * 4 = 20 possible spawns.
Worst Case Scenario
The worst-case scenario would be a big wall spawning in the middle. If we map the spawns on the grid to numbers from one to five our worst case would be two-three and three-four in the first case. These are two cases and can happen on all four sides. So 8 cases that would.. suck. So 8 out of 40 possibilities would be unfortunate in the first phase. This means it has a chance of 8 / 40 = 0.2 or 20% if we assume each possible combination has the same spawn chance.
Moving on in the second phase the absolute worst case would be two, three and four. So there are 4 in 40 chances to get a wall through the whole middle. The chance on an evenly distributed system would be 4 / 40 = 0.1 or 10%.
Last but not least with four pieces our worst-case scenarios would be any combination that contains two three and four. We have two of these for each side. So that's 2 * 4 = 8 worst-case scenarios. Again with even chances that would be an 8 / 20 = 0.4 or 40% encounter rate.
Cover Chances Concept
We don't know if it's evenly distributed though. So it leads to the following research. We have a 5x5 grid like the following table.
Now we do this. We look at any footage we can find, in my case my past streams on Youtube. We'll use the mini-map with the fixed option - which means north will stay north regardless of how our character or our camera changes - and check where the walls spawn. Each time the walls spawn we're gonna count each field that will or has been covered by these wall pieces. We're going to continue this until we've summed up all cases.
For example, this might be the table after the first wall spawn.
1
|
1
|
|||
| 1 | 1 | |||
| 1 | 1 | |||
| 1 | 1 | |||
| 1 | 1 |
And this after the second wall spawn so you should get the idea now.
1
|
1
|
|||
| 1 | 1 | |||
| 1 | 2 | 2 | 1 | 1 |
| 1 | 1 | |||
| 1 | 2 | 2 | 1 | 1 |
By doing this procedure for every case we get the following result.
1
|
1
|
|||
| 1 | 1 | |||
| 1 | 2 | 2 | 1 | 1 |
| 1 | 1 | |||
| 1 | 2 | 2 | 1 | 1 |
Great! A five times five grid with numbers in it and now? The concept behind this is simple. In the end, we get a heat map with the highest numbers showing the most dangerous positions and the smallest numbers showing the least dangerous positions. At the same time counting the number of waves we can divide each field number through the total amount of death walls to get the chance of the wall moving over this field rendering it more dangerous.
To Be Continued...
This is enough for today. I have to work tomorrow. So meanwhile I'm gonna use what time I have to analyze the wall spawns in my videos in the next days. So wait for the next post in the next weeks. Yeah, sorry it takes a whole week... but with me losing a lot of my spare time due to work and everything I can't afford to spend as much time on theorycrafting and blog writing anymore. :(
