Reverse engineering and homogenizing KF2 fleshpounds

This longread pertains to analyzing a particular monster in a PvE game Killing Floor 2 and expands on some game design principles. I’ve written the bulk of this post back in 2019 and shared among the group of CD players I’ve played with back then. The goal of this post was to analyze such a Killing Floor 2 monster as a fleshpound (or FP for short). Is a fleshpound monster design good and even if it is, was it implemented right and properly tested? In this post I’ll try my best to reverse engineer KF2 fleshpounds, establish the main monster design goals, verify if these design goals were implemented right and suggest ways to correct identified flaws.

For starters consider this video:

Note, that the fleshpound randomly rolls one of its raged state attacks, specifically, “Atk_RageLunge_V1”. The issue is this attack never lands on a moving target. The fleshpound, thus, misses and just slows itself down while exposing its head in a predictable way. No damage taken by the player.

Now consider this video:

Note, that the fleshpound randomly rolls one of its raged state attacks, specifically, “Atk_RageLunge_V3”. This attack never misses and always lands. This attack also doesn’t expose the fleshpound’s head, as the fleshpound immediately transitions from its raged state with its head partially covered by the grinders into landing the attack. The melee hit registration here is very quick, just 279 milliseconds passes between the beginning of this attack animation and the melee hit registration. Note, that the human reaction speeds, on average, are in the hundreds of milliseconds, depending on the complexity of a task that measures it, with more complex tasks taking more time to react.

Also note, that right after landing the “Atk_RageLunge_V3” attack, the fleshpound immediately transitions to a calm state attack. This is always deadly for a player with no armor and/or medic buffs, as vanilla “Hell on Earth” difficulty fleshpounds deal 68 HP damage with the “Atk_RageLunge_V3” attack and roughly that same number with the most of their calm state attacks. Naturally, the player dies to it. Despite jumping to get more momentum transferred during the raged attack to create more space and make the follow up calm attack less likely. Despite trying to bash it off to prevent the calm attack from landing. Literally, the player did nothing wrong.

Now, the question. Why the player in the first video takes no damage, gets the fleshpound slowed down and gets the fleshpound to expose the head in a predictable pattern? At no demonstrable skill or play of their own? Why the player in the second video gets wasted from 100 to 0 HP at no fault of their own? The player did nothing wrong and did everything according to the best practices and yet died from 100 to 0 to a purely random chance event.

Should the game decide the player’s fate based purely on a random chance or should it allow a skilled player a narrow path to victory out of seemingly impossible situations, if played right? Read on to find out.

I’d like to bring up this poll on KF2 reddit.

It’s easy to argue that the vast majority of redditors are casual players and their opinion should be dismissed without any further analysis. But here I’d like to point out that sometimes they are right and they are right here about EDARs. Somehow, they are right about EDARs. They are also right about scrakes and husks, which get the least hated votes, as these are 2 very well designed and very well balanced monsters. If they were rating the monsters purely based on the monster difficulty they would surely place husks and scrakes closer to the top, as both of these are very dangerous. While they are dangerous, they are fair. What if the redditors are right about fleshpounds as well here? In what way specifically the fleshpounds may be not well designed? In what way they may be unfair? To answer this question I’d like to describe their existing behavior and then identify flaws in it.

Raging FP attacks

The damage FPs deal depends on the difficulty level, but I’m going to cover here only the “Hell on Earth” difficulty. Their damage is different between the single player (SP) and multi player experience (MP), the MP attacks deal roughly 1.5 times more damage than SP ones. I don’t care much about the single player experience, I’m going to talk here about the multi player experience exclusively.

In a typical gameplay, calm FP attacks are much less important for all game scenarios than raging attacks, except for a meleeing berserker. Raging FP attacks, on the other hand, are hard to avoid and therefore must be studied well to understand them. There are just 4 raging FP attacks total with an equal chance to be rolled with 1 caveat, which is the “Atk_RagePounce_V1” has a rather long 10 seconds global cooldown to it, which means that no FP on a map can roll it for the next 10 seconds after it was rolled by at least one of them, at least in theory.

The mutliplayer “Hell on Earth” damage table is:

attack type	damage dealt per first blow	damage dealt per second blow	damage dealt per third blow	number of punches	overall damage (if taken)
“Spreading arms” attack, aka Atk_RageLunge_V3, aka “crit”	68	0	0	1	68
“Double jump with arms up & down combo” attack, aka Atk_RagePounce_V1	22	22	22	3	66
“Hug” attack aka Atk_RageLunge_V2	37	37*	0	2*	37 or 74
“One-two haymaker punches combo” aka Atk_RageLunge_V1	33	33	0	2	66

Highlights:

• There are only 4 different raging attack animations;
• The 2nd melee hit of the “Hug” attack aka Atk_RageLunge_V2 almost never registers, because of the knockback effect of the 1st hit;
• None of the “One-two haymaker punches combo” aka Atk_RageLunge_V1 hits register, if a player keeps moving away;
• There is no difference in player stance for the damage calculation, e.g. a belief some share that facing the fleshpound reduces the damage dealt is untrue;
• There’s no attack damage randomization, there’s only attack animation randomization. Once the attack animation is chosen the damage is fixed;
• Some attacks are multi-punch attacks, but if they are multi-punch, each punch does less damage compared to one-punch attacks;
• The most deadly attack is a one-punch “spreading arms” attack, aka “crit”, aka “Atk_RageLunge_V3”.

Here are all these attack animations represented as actual video animations, so players, who aren’t familiar with the game’s code can still recognize what I’m talking about here.

FP attacks problems

By that time you probably know the message I’m trying to convey here: the fleshpound attacks are inconsistent. Why is this a problem? Shouldn’t the attacks deal random damage?

No, they should do consistent damage, I’m going to expand on this later, but for now let’s just state that this game designers, and I readily admit that regarding very many things they made right decisions, wanted the damage to be consistent. Keep in mind, that the damage, both ways, by players on monsters and by monsters on players, is the most important thing in the game balance-wise. Here are a few arguments regarding this:

• All perk weapons damage is strictly consistent. They are calculated to reflect various perk bonuses and zed resistances or vulnerabilities using a somewhat complex formula, but in the end you can always rely on every weapon to inflict very specific damage to any specific zed. As an example, you always kill a clot via a headshot with 9mm, no exceptions. No random chance at play here. You can always rely on M14 to kill a gorefast via a headshot, no exceptions. And so on and on;
• All zed health is strictly consistent. For example, a 6p “Hell on Earth” fleshpound’s head always has 1716 HP. Not 1 HP more or less. This means that fleshpound takedowns take always the same amount of shots from each weapon of choice;
• Almost all other zed attacks are consistent in terms of damage, with rare exceptions, fleshpounds being one of them. E.g. 1 gorefast/fiend swing almost always does 17 HP. 1 scrake chainsaw swing almost always does 37 HP. There are some discrepancies here and there due to various factors, but the general rule holds.

Why then, you may ask, fleshpounds deal inconsistent damage, while almost all other zeds almost always do consistent and predictable damage? Well, glad you asked. Mainly, because of the lack of proper testing. The developers wanted it to be consistent, but they just failed to properly test everything and shipped an undercooked product. Let’s go over all the arguments supporting this theory.

Atk_RageLunge_V1 never lands on a moving target

Again, the video I already referred to in the beginning of this longread:

The default hypothesis for all attacks is they are designed to land. No game designer designs a weapon the main function of which is to miss shots. They are supposed to land, that’s why their bullet trajectories are straight or otherwise predictable, such, that a skilled player, relying on their aim, can predictably and consistently land shots. The same applies to monsters, their attacks are supposed to land, unless they are projectiles which take time to travel or are telegraphed in some other way to allow a skilled player some room to avoid getting hit. Why this particular attack does not land?

There’s an inherited from Unreal Engine 3 parameter used for any in-game animation called “RateScale”. The default is 1, which means that the animation is played as it was captured and imported into the game engine. There’s evidence that TWI tinkered with these, as every self respecting studio should, in order to speed up or slow down various animations, so they look good enough. “Good enough”, among other things, incorporates achieving the primary goal of attack animations, which is landing when it should. Some attacks should not land if a player engages in various counterplay that depends on the player skill, but if no counterplay, then they should land. Here’s an example of one fleshpound attack animations that was sped up 20%:

However, most likely due to the lack of testing and tight deadlines, the Atk_RageLunge_V1 was overlooked. It just almost never lands on a moving target. I have 5.5K hours in this game and dealt with fleshpounds a lot and reviewed a lot of videos. I have >3K fleshpounds killed only with demolitionist weapons in ranked sessions and it’s a small fraction of the fleshpounds I’ve killed with gunslinger or sharpshooter weapons in unranked sessions. The scenarios in which it does land are extremely rare and pretty much boil down to meleeing berserkers or a player pinned into a corner. What TWI should have done was they should’ve adjusted some attack parameters, such as the animation speed via the “RateScale”, the range of melee hit registration and so on, but they didn’t.

Atk_RageLunge_V2 has 2 melee hits, but the 2nd almost never registers

If you ever wondered how come this attack deals so insignificant damage (37 HP), here’s your answer. It’s supposed to deal 74 HP, as it has 2 melee hits, but the second one almost never registers, due to the momentum transfer from the first melee hit. You can launch the testing map and experiment yourself, it’s just how it works. Oftentimes players who put thousands of hours into playing this game aren’t aware of this peculiarity, because of how rarely this attack can make 2 melee hit registrations. But the evidence is there, first, in SDK, and then, when you learn about it, you can set up a scenario in which this 2nd hit manifests itself.

Now, the question: did TWI really put effort and designed this attack to have 2 melee impacts only for the 2nd one to register as rarely, so many hardcore players with thousands of hours put, have never experienced it? The obvious answer is no. They just didn’t test it thoroughly and shipped to production due to tight deadlines.

They should have either created more time between the 2 melee impacts or reduced the momentum transfer on the 1st one, so it doesn’t push the player beyond the melee range of the 2nd one, or increased the melee range of the attack. Most likely they overlooked the momentum of the 1st hit, because it’s common for TWI to reduce the momentum transfer on the first hits of all combo attacks, so the first hits of the combo don’t push the player out of range of the next hits. But in this case, for the first hit, the momentum transfer value is set to 1 for vanilla fleshpounds, which is the default value. Most likely they just forgot to adjust it and never properly tested.

Most combo attack damage isn’t properly rounded

Some fleshpound attacks are combo attacks, which are comprised of several swings in one single animation. In which case, the designer intended to split the total damage between the number of hits in a combo. For example, the Atk_RagePounce_V1 consists of 3 hits and each hit deals roughly 1/3 of the total damage. The Atk_RageLunge_V1 was supposed consist of 2 hits, so each hit roughly deals 1/2 of the total damage. The Atk_RageLunge_V2 was supposed consist of 2 hits, so each hit roughly deals 1/2 of the total damage. But not quite, due to the peculiarity of the rounding formula. One can’t just assign 0.5 for 2-punch combos and 0.(3) for 3-punch combos and get the total damage of the 1-punch attack which deals 68 HP, because the formula is rather complex and rounded up or down at different stages. It took me more than a month to figure out the precise values that always add up to 68 HP, TWI didn’t have that time on their hands. So they put whatever and shipped it to production.

That’s why the Atk_RageLunge_V2 deals 74 HP, instead of 68 HP, but only in theory, because the 2nd hit almost never lands in real games due to the lack of testing. What most likely happened is they have found that putting 0.5 damage modifier for each hit gives weird results below the target value, so they increased it to be 0.55, but that was an overkill. That’s also why the Atk_RagePounce_V1 deals 66 HP (and that is with 0.34 modifier for each of the 3 hits, which should have already resulted in >68 HP total if not for rounding issues), instead of 68 HP, because the rounding works such that it doesn’t add up as well, but whoever was tinkering with the numbers never got to establish what damage modifiers can produce the intended result of the composite 68 HP damage.

 

Atk_RageLunge_V2 and Atk_RagePounce_V1 land inconsistently

When you get lucky and jump at the right moments you can avoid the quoted attacks completely or partially as well, if you play a perk with sprinting bonuses and are sprinting and/or under medic buffs. This is mostly because the distance between a player and a fleshpound at which the game makes a decision to roll the attack has a random component to it. So when a fleshpound, which moves faster than a player does, catches up with the player to the attack distance, it can immediately roll the attack animation. Counterintuitively, it’s the luckiest scenario, because the fleshpound moves slower during the attack animations and may, instead of landing the hits on the player, miss them completely. However, if a fleshpound keeps sprinting past the maximum attack distance and comes very close in the face of the player and then rolls the attack animation, it almost surely lands it. Except for the Atk_RageLunge_V1, which is still slow enough to never land, even if started at a point blank range. Here’s a video demonstrating just that:

Normally, the Atk_RageLunge_V2 attack lands one 37 HP hit. But because of the medic buff speed, the player jump and the fleshpound attacking from afar, this attack didn’t register and did 0 damage.

This was rather easy to avoid with proper testing, by adjusting the attack numbers such, that the max distance at which these attacks can be rolled is smaller. This way a fleshpound closes the distance with a player enough to ensure the attack always lands, because even if the animation itself is slower, it starts very close to the player, so it doesn’t slow down the fleshpound to the point of it being out of range when the melee hit is supposed to register. But TWI had tight deadlines and didn’t have time for that.

Calm fleshpound chaining follow up attacks

Fleshpounds, on occasion, chain the calm attacks right after landing the raging attack on a player with no counterplay possible. Bringing up the video that I posted in the beginning of this longread again:

This happens rather often, safe to say that it happens every game, but due to various reasons (someone else bashing, gunslinger quick bashing, medic buffs, some armor left, zed time kicking in, etc) players don’t always die to it. I would still say that they do die in most cases when this happens. I’ve collected a number of such videos, but didn’t want to bother with the hassle of downloading the full vids from YT, cutting out the needed fragments and uploading them again, but if the argument about it arises, I’m ready to do that. On occasion, if a player doesn’t die to 2 attacks, a fleshpound can chain the 3rd one right after landing the 2nd, all the same considerations apply here as well. In fact, in the video above, this is exactly what happened, that’s why the fleshpound was killed by someone else, instead of getting decapped, because fleshpounds die immediately if their head health is depleted while they are performing any attack animation. But, hypothetically, if a player somehow survived the 2 attacks, a fleshpound could immediately chain the 3rd attack with not much counterplay possible.

Note in the above video that the player does everything right. He bashes the calm fleshpound attack to run by it, but because of the way the vanilla fleshpounds are, still gets a punch in the back.

The worst part of this phenomenon is its randomness. Being in the range of the calm attack doesn’t necessarily mean that a fleshpound rolls a calm attack animation. If the fleshpounds consistently did, they would have been killing way too many players and there would have been a major uproar to address which TWI might have fixed it, but they don’t. They either attack or not attack while in range randomly. That’s why the player on the video above turns around after getting hit in disbelief. He assumed that if he bashed it and sprinted away he wouldn’t get hit. But he did, because of the random nature of this phenomenon.

To avoid this, TWI should have either instituted a cooldown on all attacks right after landing one, for, say, 0.2 seconds. Or, just reduced the calm attacks maximum range and increased a little the momentum a player receives if hit by a raging attack. This would have created enough space to ensure that a fleshpound doesn’t chain a calm attack. But they didn’t, they shipped it as is.

Vanilla fleshpounds are less dangerous than intended and are drastically inconsistent

Because of all the referred above factors, the vanilla FPs are very inconsistent in an otherwise very consistent and predictable game. One day a player dies to vanilla fleshpounds at no fault of their own. The other day that same player doing absolutely the same gets a fleshpound rolling weak or missing attacks and survives with either no damage incurred or only about a half. Only 1/4th of all the raging vanilla fleshpound attacks land consistently and deal the intended damage. All other attacks either don’t land, or land partially, or don’t have all the hit registering as intended. This is wrong and this is not how it should be. And the worst part of it is all that is completely random, decided by the game through a roll of dice, with no player’s skill factored in whatsoever. That’s why, I believe, so many redditors hate KF2 fleshpounds.

Note, that no other monster in this game comes close to this level of inconsistency. All other monsters are much more consistent and predictable in their behavior. No matter what you think about the role of randomness in this game, you can’t deny that the fleshpounds are a major outlier in this regard from all other monsters.

On top of that ridiculous level of inconsistency you have to add their vanilla spawn rage behavior introduced by TWI in 2017 which makes them spawn enraged with a 33% chance on “Hell on Earth” difficulty, which adds to their inconsistency and unpredictability.

The fleshpound rage behavior

I haven’t played vanilla KF2 games for so long that I almost forgot about this one, but it has to be called out due to its enormous impact and stupidity. For me personally, and I know quite a few players like that, it’s easily the number one frustration with the game. Whoever at TWI came up with the idea of spawnrage is just a retard and I’m not willing to talk politely about it. This person doesn’t understand the basics of the player agency, more on this later, and is somehow carelessly put in charge of making game balance decisions. I can understand pondering this idea in the brainstorming phase. But this idea was adopted and implemented and there was nobody powerful enough in the company to stand up and say that it’s BS and should never be implemented. It’s twice retarded that TWI keeps insisting on this mechanic, despite years of criticism, instead of rolling it back.

Another issue with the fleshpound rage behavior is the rage time counter ticking regardless of the line of sight with the players and cannot be stopped by breaking the line of sight. This results in unavoidable time rages on plenty of somewhat open vanilla TWI maps and community maps officialized by TWI. Here’s an example of it on Evacuation Point:

What’s sad and annoying here is TWI got it right in KF1, where fleshpounds timerage counter could be reset by breaking the line of sight with the players or reset by baiting a calm attack and avoiding it. If this mechanic was implemented like that, the time rage on the video above would not have happened.

Inconsistent afflictions

The afflictions of interest here are stumbles, stuns, freezes and knockdowns. Fleshpounds are designed to be very hard to deal with and require a lot of finesse and team play to be handled safely. A crucial part of it is the reliability of the affliction power accrual. Unfortunately, it’s not reliable. Fleshpounds routinely ignore afflictions in the following cases:

• The damage of a weapon/grenade that is supposed to apply any affliction power causes calm fleshpounds to engage in blocking, which denies the affliction power accrual. If the damage is high enough to trigger the rage, a fleshpound transitions to blocking and then immediately transitions to rage such, that the blocking animation doesn’t get a chance to play, yet, it still gets triggered and denies the affliction power accrual;
• Any monster in this game, fleshpounds included, cannot accrue any afflictions power right at the start of any attack animation. This is easy to reproduce on a test map by timing the freeze grenade fuse with the time when a zed reaches the max distance of any of their attacks and is likely to engage in them. That’s why you can’t reliably freeze a fleshpound which is about to start a calm attack animation;
• Fleshpounds (and scrakes since we are at it) can rarely deny stuns or freezes for unknown reasons that aren’t the reasons outlined above. I’m pretty sure every experienced player had a chance to deal with this multiple times, this bug is the easiest to reproduce by running the sharpshooter’s “Ranger” skill and trying to stun scrakes or fleshpounds during zed time. The “Ranger” stun power should be guaranteed in zed time, yet, on occasion it fails to stun them despite a clearly landed headshot.

Player agency principle

I’d like to bring up this quote:

What is said here is very right. A good game design requires that the game puts a player in a hard to survive situation, yet, allows a narrow path to victory, if the game is played perfectly or close to that. This way, if a player loses, it can always be tracked down to specific mistakes the player has made. A player can review their actions, adopt a more perfect playstyle, refine it, and hope to win, if executed right. That’s one of the biggest things that maintain player retention, because the process of discovering and practicing a more perfect playstyle is long, so it guarantees improved player retention. Dying to a random chance, on the other hand, is not fun. It is of utmost importance for a game designer to ensure that each time a player loses, this happens due in full or in a significant part to player mistakes.

Incomparable runs argument

Wins and losses on any maps with vanilla fleshpound behavior are incomparable. This means that if a team A won a certain map in certain difficulty settings, while a team B lost this same map in these certain difficulty settings, it doesn’t necessarily follow from here that the team A is better than the team B. Given that fleshpounds are the most dangerous enemies in this game and deal the most crippling damage, it’s likely the case that during that particular run the team A just had lucky fleshpound attack rolls that missed or didn’t deal enough damage.

This also means that the wins are hard to reproduce even by the same team. It is common for any team to “grind” certain maps and difficulty settings. A good share of this grind is just waiting for good rolls on fleshpound attacks, so they aren’t as crippling. If a map is won, it is hard to reproduce, because the win is necessarily a result, in part, of lucky fleshpound attack rolls. It’s less likely that the next attempt has similarly lucky fleshpound attack rolls and therefore is likely to be a failure.

RPG dice argument

The idea of any random game values, such as damage values, or hit point values, as far as I could track it down, comes from the RPG or role-playing games. Initially tabletop, but then computer-based, when the advancement of the computing allowed for it.

Due to their extreme simplicity, the first tabletop games had to rely on randomness, which was provided through dice rolls. I’d like to argue that dice rolls, overall, is a bad idea, however, it was necessary back in the day due to the games simplicity. Otherwise it was trivial for human minds just to fully exhaustively calculate the game and narrow it down to a series of simple winning moves that always lead to victory no matter what. To spice the things up a bit it necessarily needed a significant random component.

Modern computer games, on the other hand, are very complex. There’s a lot of moving parts there, such, that it makes tracking down all the moving parts by a human mind a nigh impossible task. This feature should replace the randomness, at the very least, to a high enough degree. The role of randomness in modern complex games should be miniscule.

Overall, the video game industry recognizes that, so we have less and less randomness as the game design progresses from the first old RPG computer games to modern games. It wasn’t uncommon for old games to have ridiculous play in HP or damage numbers, e.g. a sword could deal 1-10 damage determined by a roll of dice, that’s 10 times difference.

Overall, TWI did a good job to have just enough randomness in Killing Floor 2. Note, that all main values in this game aren’t random. No weapon in this game deals random damage, as an example, and the damage inflicted is one of the most important game mechanics.

However, it’s not the case with fleshpounds, which are a huge outlier from everything else in this game in terms of the amount of randomness that is factored into their behavior. This is a mistake and it should be corrected.

Little aesthetical things

Vanilla fleshpounds are supposed to pound the ground 4 times during their rage animation. The 4 distinct ground pound sounds can be heard during their rage animation, however, the rage_v1 animation itself is cut abruptly before the fleshpound grinder hits the ground for the fourth time. While it doesn’t have any significance from the game mechanics standpoint, it’s still an imperfection and a violation of the aesthetics of the game world. Consider these two videos:

Note that in this vanilla fleshpound raging animation the right grinder doesn’t reach the ground when it’s supposed to hit it while the sound of the pound is heard.

Note, that in this video, a full animation is played and a fleshpound can be seen clearly pounding the ground for the fourth time, matching the sounds.

Vanilla fleshpound behavior is a mistake

I hope that all the evidence presented above is enough to convince that a good share of the vanilla Killing Floor 2 fleshpound behavior is a mistake. Most of the issues with them stem from the lack of testing (e.g. missed attacks, melee hits not registering, damage rounding modifiers, affliction accrual issues), while some of them are a result of poor game design (e.g. spawnrage, calm attack chaining, random chance to block).

This fleshpound behavior is frustrating to deal with and needs to be corrected. TWI should have done it, but they don’t play and don’t know their own game, and except for Alex Quick, they don’t have any internal talent left at this point to even start getting into this. Having been released in 2016, the game is also almost dead, with only about 800 concurrent players online on average, and TWI considers it not worth the time to address such issues. Thus, such fine tuning of the game is something the hardcore players and modders should do themselves. In order to correct this I’m introducing a mod named “Homogeneous Fleshpounds”. It is currently being tested on my HoE+ and CD servers and once properly tested I may release it as a stand alone mutator. Some of the issues were already addressed, such as the attacks that almost never land or damage rounding formula adjustments or calm attacks chaining, while others are still being worked on, such as the afflictions accruals. The spawn rage has been addressed by Blackout back in 2017 (thank you very much). The rage timer dependence on line of sight status with proper stops/resumes is still being worked on.

Overall, the homogeneous, or “homo” for short, fleshpounds are more dangerous, but at the same time much more consistent and predictable. Here’s a quick summary of the already implemented adjustments:

• Their damage is uniform or homogeneous, hence the name. They have 4 raging attacks total. Vanilla FPs do full damage with only 1 attack out of the 4 with their “Atk_RageLunge_V3” attack, which deals 68 HP. This attack has a very fast animation and always lands, there’s no escaping from it. Thus the people call it a crit attack. 1 of their attacks almost never lands, except on a stationary target, that is a 1-2 punch combo, Atk_RageLunge_V1. 2 of their remaining attacks land sometimes, depending on the sprinting speed and medic speed buffs. They are Atk_RageLunge_V2 and Atk_RagePounce_V1. With homo fleshpounds all attacks summarily deal the same uniform damage of 68 HP. Both raging and calm attacks (except for the stationary “drill” attack which has a continuous damage pattern);
• Homo fleshpounds never miss any attacks when they should land (e.g. when there’s no counterplay such as juking or abusing the map environment);
• Vanilla fleshpounds can randomly chain calm attacks right after landing a raging attack. In the case of a crit attack followed by a fast calm attack this leads to a player’s death from 100 to 0. Homogeneous fleshpounds don’t chain calm attacks, unless you are actively pushing towards them or they pinned you to some object;
• Homogeneous fleshpounds can be reliably frozen or stunned;
• Homogeneous fleshpounds don’t time rage, if the line of sight with their targeted enemy is broken, similar to the KF1 fleshpounds. The author and the initial researcher into this was Keyes, go write a +rep on his profile, although I have implemented it differently in the end. However, to make it harder to kite them endlessly, they have a separate long timer which causes them to rage when it expires;
• Homogeneous fleshpounds rage scrakes by bumping into them just once. This is done to address the “squeezing the MaxMonsters” tactic players rely on easily kiteable maps such as Nuked, by leaving scrakes alone, which results in effective reduction of MM down to a manageable number as the scrake trains occupy a good share of it without posing a threat. This measure forces the players to kill the scrakes ASAP or risk the prospect of a single raging fleshpound bumping into a scrake train and raging all of them at once;
• Homogeneous fleshpounds always play the rage animation, even if shot at before the previous attack animation was over. We play this game to shoot non-stop and any interruptions or hindrances are annoying, such as waiting precious milliseconds until the animation is over doing nothing and not being able to shoot anything else because the fleshpound body covers 4/5 of the screen. Just keep unloading at the homogeneous fleshpounds at all times.

 

As always, although I think that I’m right at how I identify and address these issues, I’m open for discussion. Make a coherent argument, prove your point with videos and numbers, and I may change my position.

A short demo showcasing all homo raging fleshpound attacks, their damage, how they land, etc. The video is a bit dated and the Atk_RagePounce_V1 deals 1 HP less damage than intended, this has already been fixed: