Oh Shit Moments - Why i love XCom :>

[Arthanor]

Oh, I fully agree. If you are doing a human vs human game, you need that 50% chance of winning otherwise the players will (rightly) cry foul and blame your balance. It is a bit more complex in rock-paper-scissors systems, but even with them, if you have "classes" that are balanced, you should have a 33% chance of seeing the one that owns you, 33% of seeing the same as you and a 50% to win that one, and a 33% to see the one you own, making for 50% chance to win any ways.

Human vs computer games generally tend to have much more skewed balances, because the computer doesn't mind getting humiliated for the human to have fun (or for the human to think he is so intelligent for beating the AI.. or just because it is easy to compensate for AI flaws by increasing numbers). I added this to go against the people who compare XCom to chess or war. XCom remains a game that is aimed at giving fun to human players. You won't win the vast majority of your games of chess by wiping out your opponent but losing only a few pieces, and you are unlikely to win (symmetric) wars with significantly less casualties than your opponent.

XCom is a very odd setting where humans are supposed to have very much the bad end of an asymmetric war (low XCom effective, crappy human soldiers and tech versus gene engineered soldiers with future-tech), but somehow everybody still expects to win engagements with low casualties. Even crazy popcorn cyberdisc squad wipes don't change the fact that over the course of a campaign, you will have gone through much fewer soldiers than the aliens. But of course, humans (players, scientists and engineers) are much more intelligent than aliens

[DeltaEpsilon]

The distribution of damage on armored targets is FAR from a normal or a uniform distribution, which are the distributions we are naturally used to and understand better.

If we talk about dice rolling then it is approaching normal distribution as dice rolls increase. The more rolls are made the more "normal" distribution is.
Let's say we have 3 dices with 6 sides which is 3d6. That means we can get from 3 to 18. There is only one way you can roll 3 or 18 which is getting respectively 1-1-1 and 6-6-6 and the chances are (1/6)^3 for both rolls. However there are three ways of getting 4 or 17: 1-1-2, 1-2-1, 2-1-1 and 6-6-5, 6-5-6, 5-6-6 and so chances are higher. The average has the most ways to get and so the chance to get the average is the highest.

Perhaps our definitions of "randomness" differ a little? I recognize and appreciate this "randomness" as being intrinsically bound to certain predefined mathematical parameters (the code), if you will.

It'd be pretty hard to describe but I'll try anyway. So I think of random in roguelike and similar games as a generator for conditions, not outcomes. Let's take an abstract example. A roguelike game generates a random level which gets a legendary weapon guarded by strong enemies. This is common. You may try to fight your way through if you feel brave enough and risk getting YASD or your may take it easy and go straight to the stairs. What you are going to do is your choice and the outcome is purely dependent on your current strength and whether you are going to win the weapon or fail miserably should mostly depend on your build and only a little luck. If you shoot at an enemy and he gets shot - he will lose some health ranging a bit. This ranging may matter when you get a last enemy with 1 HP. This is what I call good random. Take a look at XCom now. The outcome is explicitly controlled by random in a huge range. Vanilla damage formula ranges from 0% to 200% which basically means you can get a situation where you can hit an enemy hundreds of times and he won't die, albeit it is rare. But see, this is what I've been talking about. Outcome is 50% dependent on luck and 50% on your build which is a really bad ratio after all.
Actually that's the whole reason why I always play with TFTD Damage Formula. I want some damage to be done at hit.
But hey, one may ask why do I say I like that kind of random while raging over a situation made by such random. The answer is pretty simple: conditions are things that influence on outcomes. In described situation the outcome has already been clear and there is nothing I could do about that. By making such starting conditions random has pretty much determined what will happen next and that is what I don't like.
So yeah, that sucks. If some system is so complex it's output is unclear at defined inputs, but still determined within its range - this is good for me. If it outputs random mess with no pattern - it's not.
If you got shot to the head - chances are that you are going to die. If you somehow survive it - you'll probably suffer from some really bad issues. But if random decides you literally deflected the bullet with your forehead - this is just... random.

Of course random outcomes have their perks. I think anyone who played XCom long enough once had a situation where the entire squad has been killed except for one which comes and saves the day by slaughtering everything it came in contact with and I am sure many players start to treasure that very soldier as a reminder of that. This is fun but it's rare, although it's definitely worth it. Apparently this is why this guy

Interestingly, this is precisely why I love X-Com. Yes, the randomness. The fact that there's always a possibility—small as it may be—that you can lose it all (that f@cking Sonic Pulser wiping out my entire squad!). If I know I am winning a game right from the start, why would I even bother playing? Sure, plain old fun. But X-Com offers more than just fun. And out-of-control randomness is my replay value here. That fact that I can get brutally murdered even under the most systematically thought-out, under-control situations. Makes you think twice about everything and manages to keep it fresh every time. There's nothing more rewarding.

and many other love XCom. I do like that but I am more of a fan for "little-error-huge-consequences" type of situation. Even without all its randomness XCom has a pretty complex system and playing perfectly here is nearly impossible. You play as perfect as you can, but you make one small mistake which eventually grows into a large hole which destroy it. This is what I like - building something complex and watch it breaking apart. DF is one ideal example of such neverending cycle. Your perfect castle will eventually break down because it's not that perfect for all.

[Arthanor]

If we talk about dice rolling then it is approaching normal distribution as dice rolls increase. The more rolls are made the more "normal" distribution is.
Let's say we have 3 dices with 6 sides which is 3d6. That means we can get from 3 to 18. There is only one way you can roll 3 or 18 which is getting respectively 1-1-1 and 6-6-6 and the chances are (1/6)^3 for both rolls. However there are three ways of getting 4 or 17: 1-1-2, 1-2-1, 2-1-1 and 6-6-5, 6-5-6, 5-6-6 and so chances are higher. The average has the most ways to get and so the chance to get the average is the highest.

I understand how dice rolling works. I said that we understand normal (you probably can get there with lots of dice) and uniform distributions (one dice) the best, maybe I should have said that things in between (few dice) are also easy.

That has nothing to do with how XCom works though. As you said below, initial damage is a uniform distribution, between 0 and 200% (or 50 to 150% in TftD) of the listed damage. Now substract the armor (since that's what my comment actually was about), and you get a range of actual damage that looks nothing like either uniform or normal distributions. Take a weapon with 18 power, EU damage, on a target with 10 armor and 17 health. You can get: 0,0,0,0,0,0,0,0,0,0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,dead,dead,dead,dead,dead,dead,dead,dead,dead,dead.

So that's

Code: [Select]

10/36 = ~27.8% of zero
15/36 = ~41.7% of wounded but standing (16 + 10% stun damage means fallen unconscious)
11/36 = ~30.6% of dead
Decrease armor by the same amount as you increase health and you get back to a uniform distribution of damage and your hits will always do something. By opposition, increase the armor by the same amount as you decrease health and you get less and less wounded results (but still the same amount of dead), until you reach the extreme example of a target with some armor and 1 hp, which is always either 100% unarmed, or killed. By tweaking the armor value relative to the health, you can then set exactly how likely that target is to die to a given weapon, and you can also play around with damage modifiers, to make a target that is 100% to die to lasers but 100% to shrug off plasmas.

Of course, extremes are silly examples, but mutons are a great example of aliens designed to be tough but enjoyable. Compared to some modded aliens which gets lots of armor (or cyberdiscs, which have decent armor compared to early XCom damage and are likely the cause of more ragequits). All that to say that by playing hellrazor's mod which generally increases the armor of aliens, you are more likely to encounter situations where your weapons do nothing and frustrate you. By opposition, if one were to increase the health of aliens to make them more resilient, you would still require more shots to kill, but since the outcome is more reliable (you are more likely to wound, so you will either wear down an alien or kill it, without having the frustrating unkillable streak).

Anthropods by robin is a great example of adding a new enemy with more health as a challenge (and also awesome for the spitter, which has armor but only on some sides, so it is frustrating if you are stupid but rewarding if you flank it). Gazers are something else, more of a FU endgame enemy, but they are labelled as something like that. The armor thing is one of the reason I don't play on Superhuman, as I don't enjoy the boosted armor on some enemies.

For some reason, when designing new units, we tend to add more armor (because the aliens lack it so it is an easy thing to change, without really thinking of why they lack it) instead of more health. I had the same problem when designing my modded tanks, and added so much armor that tanks were mostly impervious, until a lucky shot would kill it or reduce it so low I had to stop using it (which is frustrating because randome). The answer, obviously, was instead to add health so it gets worn done more slowly and behaves in a more satisfying way.

TL;DR: XCom is random and that can be frustrating, but it is random in a way we can understand and, since we can now mod it, all we need to do is mod it in a way we like. More health, less random, more armor, more random.

[hellrazor]

I am not even going to bring a screenshot here, I get such a situation so often you can picture it in mind.
Turn 1. Two THROWERs full of goddamn HE packs and smoke grenades see 3 floaters and a cyberdisc. AT FIRST TURN. They drop a smoke grenade.
Turn 1.5. Cyberdisc decides he's smart enough to go right to the Skyranger, see the throwers and shoot them
Explosions. Explosions are beautiful.

And then there were none. Bullshit. Screw OpenXCom, I am out.

Never ever use Preprimed grenades on actuall explosives. This always ends very very badly. You always can do a grenade relay if needed, to get the explosive were it is needed.

Uhh i got my ass handed by a Muton Engineer, which decided it was a nice time to spawn next to my Skyranger on this certain Battleship Mission. The Random God had decided that this lovely and nice Muton fellow was armed with a Blaster Launcher. He was just standing 2 fields out of sight of my tank on its starting position. As usual i skipped turn 1, which resulted in, lets say in a very unpleasent experience on getting a instant Squadwipe .

But thats X-Com Baby, thats why i love it.

Btw Arthanor, could you describe why my Mod is not for you? I mean what do you do not like etc.. i am always feedback hungry

[7Saturn]

Of course, extremes are silly examples, but mutons are a great example of aliens designed to be tough but enjoyable. Compared to some modded aliens which gets lots of armor (or cyberdiscs, which have decent armor compared to early XCom damage and are likely the cause of more ragequits). All that to say that by playing hellrazor's mod which generally increases the armor of aliens, you are more likely to encounter situations where your weapons do nothing and frustrate you. By opposition, if one were to increase the health of aliens to make them more resilient, you would still require more shots to kill, but since the outcome is more reliable (you are more likely to wound, so you will either wear down an alien or kill it, without having the frustrating unkillable streak).

Anthropods by robin is a great example of adding a new enemy with more health as a challenge (and also awesome for the spitter, which has armor but only on some sides, so it is frustrating if you are stupid but rewarding if you flank it). Gazers are something else, more of a FU endgame enemy, but they are labelled as something like that. The armor thing is one of the reason I don't play on Superhuman, as I don't enjoy the boosted armor on some enemies.

For some reason, when designing new units, we tend to add more armor (because the aliens lack it so it is an easy thing to change, without really thinking of why they lack it) instead of more health. I had the same problem when designing my modded tanks, and added so much armor that tanks were mostly impervious, until a lucky shot would kill it or reduce it so low I had to stop using it (which is frustrating because randome). The answer, obviously, was instead to add health so it gets worn done more slowly and behaves in a more satisfying way.

Well, that's probably, why the UFO2000-guys decided, that armour comes first, the damage done to it is defined by the weapons damage-value and once the armour is gone in that particular place, every hit goes right to the health, reducing it equally by damage-value of the weapon. The only thing, that's chance, is the question of if the unit is hit, or not. Everything else is rather deterministic.

[hellrazor]

I totally missed out on the interesting discussion in this thread.

I am not getting any more notifications through my Emails, which sucks...

[7Saturn]

Well... I do. You doublechecked, if it's still set to notify you? Yesterday, there were some server problems (503s). Maybe something got lost?

[hellrazor]

Well... I do. You doublechecked, if it's still set to notify you? Yesterday, there were some server problems (503s). Maybe something got lost?

Na its working again now, the changed some default settings and i usually like to be spammed from the forum, with each reply to a topic i have in my notify list, Since my E-mail is only for the forum here, so

[DeltaEpsilon]

But thats X-Com Baby, thats why i love it.

And as I described above I am not satisfied by that. There should be my own error which decides the fate of my squad, not just my luckiness/unluckiness. For me, XCom should be hard because of so many conditions to be taken into account and not because random fu... oddity.

[hellrazor]

All that to say that by playing hellrazor's mod which generally increases the armor of aliens, you are more likely to encounter situations where your weapons do nothing and frustrate you. By opposition, if one were to increase the health of aliens to make them more resilient, you would still require more shots to kill, but since the outcome is more reliable (you are more likely to wound, so you will either wear down an alien or kill it, without having the frustrating unkillable streak).

I did not change the Health or Armor values of any Vanilla Aliens. I did tweak Waspites and Gazers a little bit, gave them more Health primarly and change the armor a little bit around (but no really big changes here). There are units which stats i designed myself, mainly the Armored Cyberdisc and Armored Sectopod, which do have indeed higher Armor values and other damage modifiers. But those are units which i consider to be asskicking ones, i intended them be hard to take down.
For example the Armored Cyberdisc, essentially uses the same damage Modifiers as the normal Cyberdisc, but has 70 Armor on Superhuman instead of 38, same Health. It makes this unit incredible hard to kill with anything lower then Heavy Laser (90 damage). But they do die with Heavy Plasmas very quickly (usually 1 or 2 Shots).

Armored Sectopod has his weakpoint, which is his back armor, so flanking it is the most reliable way to kill it (or Blaster bombs).

There are enemies like the Muton Elites for example which do have pretty high armor (front 70, side 60, back 50 and under 50), they also have more Health then normal Mutons and are essentially designed to be immune to Laser Weapons, but they can be killed reliably with Plasma Weapons for example.

[hellrazor]

And as I described above I am not satisfied by that. There should be my own error which decides the fate of my squad, not just my luckiness/unluckiness. For me, XCom should be hard because of so many conditions to be taken into account and not because random fu... oddity.

Everytime on of your soldiers takes a shoot, the damage this shot actually does is modifed by a randomly selected multiplier ranging from 0 up 2 (equals 0% damage and 200% damage).

Also deviation of the shoots is added randomly.  And so forth randomness is in some way a integrated part of the game.
This may lead to some very weird or painstaking scenarios but that is what this game makes so exciting, because you really never know if you will get your desired outcome. Everything can be seen as a weighted probability ingame, it depends on how you minimize those numbers unfavorable to you when possible.

[Arthanor]

I did not change the Health or Armor values of any Vanilla Aliens. I did tweak Waspites and Gazers a little bit, gave them more Health primarly and change the armor a little bit around (but no really big changes here). There are units which stats i designed myself, mainly the Armored Cyberdisc and Armored Sectopod, which do have indeed higher Armor values and other damage modifiers. But those are units which i consider to be asskicking ones, i intended them be hard to take down.
For example the Armored Cyberdisc, essentially uses the same damage Modifiers as the normal Cyberdisc, but has 70 Armor on Superhuman instead of 38, same Health. It makes this unit incredible hard to kill with anything lower then Heavy Laser (90 damage). But they do die with Heavy Plasmas very quickly (usually 1 or 2 Shots).

Armored Sectopod has his weakpoint, which is his back armor, so flanking it is the most reliable way to kill it (or Blaster bombs).

There are enemies like the Muton Elites for example which do have pretty high armor (front 70, side 60, back 50 and under 50), they also have more Health then normal Mutons and are essentially designed to be immune to Laser Weapons, but they can be killed reliably with Plasma Weapons for example.

Oh, I'm not necessarily talking about tweaking vanilla aliens, but indeed the new armored stuff (or muton elites, and I think you have an ethereal elite of some kind?). I personally know that a mod that adds significant armor to aliens (or lots of new aliens with armor, whichever way you want to spin it. Is an armored cyberdisc a cyberdisc with more armor, or a new alien with lots of armor that just happens to look like a vanilla one?) and boosts the psi-str of aliens so the player gets even more punished for not carefully sorting hundreds of soldiers to get 100 psi-strength is not a mod for me.

I don't like facing armored aliens and the "frustrating nasty rng streak". I also don't like sorting and firing soldiers as I think the game should be balanced on the average soldier (I barely fire anyone, but some will get... riskier.. roles). I also never (ab)use xp training for my soldiers. They always shoot to kill. But we all like different things, and that's fine. I'm not even really consistent since I also like Piratez, which has some armored enemies (but some definite ways of handling them as well).

Everything can be seen as a weighted probability ingame, it depends on how you minimize those numbers unfavorable to you when possible.

That, however, I very much agree with. The whole battlescape game is a weighted risk (probability * cost of even occuring) game. You need a deployment method that is the most reliable. Not necessarily the one that produces the best outcome, but the one that produces the least bad outcomes.

- Tanks (or armored high reaction soldiers so they hopefully don't get reaction fired on or survive if it happens) at the front so you don't lose them, or useless rookies/dogs/scout drones so you do lose them but you don't care.
- No pre-primed explosives except maybe at the back so they can deploy when it's safe (or chain toss it to someone in front who needs it).
- I usually follow the tank/scouts by high damage soldiers (rocket launcher + sniper) to be more likely to take care of immediate threats. Riflemen come next, and sappers last. Melee, if you use any, goes mixed with the riflemen to either rush a nearby target or help with scouting.
- All melee soldiers need good reactions to be able to close with the enemy.
- Play the rest of the game as a "Does this have good odds of success? Am I screwed if it fails?" Then decide if it is worth trying and have a plan on what to do with failures.

Minimizing the odds of bad is the first step, then comes maximizing the odds of good without increasing the odds of bad (too much). Because of our perception of the game (ex.: losing 1 soldier to kill 1 alien is a failure, even though it's a 1 to 1 "neutral" trade), minimizing the bad is really important.

[hellrazor]

Oh, I'm not necessarily talking about tweaking vanilla aliens, but indeed the new armored stuff (or muton elites, and I think you have an ethereal elite of some kind?). I personally know that a mod that adds significant armor to aliens (or lots of new aliens with armor, whichever way you want to spin it. Is an armored cyberdisc a cyberdisc with more armor, or a new alien with lots of armor that just happens to look like a vanilla one?) and boosts the psi-str of aliens so the player gets even more punished for not carefully sorting hundreds of soldiers to get 100 psi-strength is not a mod for me.

Well the aliens also need the ability to "evolve" a little bit, so you can consider those new Armored Units, them learning from us. And it is actually so that those units only make you struggle as long as you do not have access to plasma weapons, once you have them they are also relatively easy to kill, it forces the player to research plasma weapons or related topics in this direction. It is also one of the aspect i wanted work out a little more that, in certain game situation what you research and when you research it should have meaning. This will even be more evident with the changes i made towards the economy so money also plays a big role into that. On the same occasion it forces the player to a degree to research specific topics, or he will suffer badly. I wanted to give the when and where you research something give some meaning here.

The PSI settings, yes i use a higher PsiStrength on Aliens, to increase their Psi defenses. But Alien who do have Psi Skill and could do Psionic Attacks have been adjusted, so their Psi Attack capability behaves like in vanilla, but they do have higher Defenses. Ethereals, for example are immune to any Psi Attacks, while their capability to MC your soldiers is the same as in the Vanilla game. So you can stick with the 80+ Psi Strength Rule for your Soldiers (since they can be considered to be basically immune from Alien Psi attacks, as in Vanilla X-Com).

I don't like facing armored aliens and the "frustrating nasty rng streak" and I don't like sorting and firing soldiers as I think the game should be balanced on the average soldier (I barely fire anyone, but some will get... riskier.. roles). I also never (ab)use xp training for my soldiers. They always shoot to kill. But we all like different things, and that's fine. I'm not even really consistent since I also like Piratez, which has some armored enemies (but some definite ways of handling them as well).

Well the aliens i have which do have higher armor stats can actually be misused for training purposes, e.g. shooting at muton elites with Laser Rifles will give them some damage over time and eventually also kill them, but you gain lots and lots of shoots and experience. Working around that, i could lower the armor, keep the damage modifier and give them even more health (Muton Elite do have 140 HP, Muton Brain Guards 160), but in the end i would end up with alien with 240 health or so, ok they could be killed reliably with weapons like the LAser Rifle, but i was thinking a little bit ahead and wanted Laser to become "outdated" so to say.

[Arthanor]

Oh, I understand why you did it and respect it. Increasing late game difficulty is challenging and you have done some things that I quite like for that. The idea to link research with missions is a great one, especially if you can make it into more of an arms race idea, where developing lasers increases the chances of seeing laser resistant enemies, and developing plasma increases the chance of seeing plasma resistant enemies (who could be less resistant to laser, bringing them back into the game after the plasma golden age).

I also prefer to work with a different implementation of Psi than just boosting resistance (since that is equivalent to increasing armor, in that it requires no skill/change in behaviour, just to be more lucky if you want to panic/MC something). I have developed my own psi-mod for that, that isn't getting tested (or published) because of XPiratez...

Well the aliens i have which do have higher armor stats can actually be misused for training purposes, e.g. shooting at muton elites with Laser Rifles will give them some damage over time and eventually also kill them, but you gain lots and lots of shoots and experience. Working around that, i could lower the armor, keep the damage modifier and give them even more health (Muton Elite do have 140 HP, Muton Brain Guards 160), but in the end i would end up with alien with 240 health or so, ok they could be killed reliably with weapons like the LAser Rifle, but i was thinking a little bit ahead and wanted Laser to become "outdated" so to say.

I don't use training intentionally, but I understand the intention. Increasing health to compensate for lower armor is not just for training though (you can always train with a starting pistol), it is mostly for the game effect of increasing reliability of outcomes and decreasing the odds of "frustrating RNG streaks". I'd rather need to hit an alien 10 times to slowly kill him than have to hit it often enough to get in the lucky 10% that kills it. 10 times is reliable and I can plan for it and I can see my progress, 10% is fun when it works on the first shot, but my frustration when it takes 50 shots is much more intense than that fun.

Finally, why would aliens who are obviously rather technologically advanced respond to developments in human weaponry by putting on some more metal plates? That's what we did in the middle ages! Aren't our tanks now working with fancy multi-layered, slanted armor to better cope with impacts, rather than just being bigger, heavier things? Planes with decoys, flares and electronic counter-measures?

In vanilla, the aliens develop/clone/produce tougher and tougher alien species (sectoid/floaters -> snakemen -> mutons) to respond to the evolution in XCom weapons. Ballistic weapons causing problems? Send in the snakemen! Laser weapons? Mutons! They didn't armor sectoids, they sent a new, tougher race. Why not something else that goes into the tech sci-fi advanced species-vibe? Super muton with more health to represent it shrugging off pain via adrenalin injectors? A refraction/dispersion field that deflects/disperses lasers (gives a laser damage modifier, making lasers obsolete in a different way than by upping armor to the point where lasers become useless)? Or a magnetic field to deflect the charged particles that form plasma and reduce its damage (again, damage multiplier).

To me, health and modifiers are better ways of increasing the challenge because the response is adapting your plan, not bringing guns with bigger numbers or hoping for the RNG not to hate you. An extreme example: Falko (I think?) even had a mod at some point where different alien ranks had different modifiers, so you had to learn the color coding of resistances and react accordingly on any given mission, which also means soldiers had to be equipped with a wide-variety of weapons (granted, that doesn't work with making tiers obsolete). There's a lot of possibilities!

[hellrazor]

Well jeah maybe, but i made some things, to be totally honest, to annoy the shit out of the player, if he doesn't know what to do exactly. Especially with the Aliens with higher Armor in the lategame, but i am not a complete dickhead, so i made my new Racemixes in a way that whenever in the next month, their will be a new alien species showing up in full and pure form, you get some of them mixed into the alien races the month before (carrying melee Weapons and Grenades), so you can experiment or notice that they can only be killed in a specific way. So i give the player the ability to adapt himself to some degree, like changing his research from crafts or armor towards stronger weapons.