Thoughts on increased randomness

[Ol' Stinky]

In any case, one of the things that I really like about Xenonauts is it's high tolerance for failure. Sure, losing a soldier to a one-shot kill might be disappointing and even frustrating. But did it cost you the mission? Probably not. Even if it did, did it cost you the game? Almost certainly not.

Hence, I'm quite happy with the game not allowing the perfect player to play the game without taking a single casuality. I'm confident that a skilled player will win this game every time (bugs aside! No one can beat CTDs. Save maybe the devs), even if they're taking casualities due to some bad rolls.

It's not a critical issue, sure. Still, I'd rather not have "too bad, so sad" happen at all - I'd prefer it if the perfect player with atrocious luck didn't have any casualties, but instead had his best men camped in sickbay for a few weeks. The average player with the same luck would have a half-dead team and the terrible player, who I guess moves soldiers at random, would be wiped out.

Not to mention that luck sometimes helps you in a bad situation. You can't tell me that you haven't cheered on really tough mission where the last couple guys have come out on top because they got lucky and one-shotted someone across the map or the Reaper ran out of TU one tile away and you dropped him at your feet. Those are the game moments I live for.

True, but it's harder to tell if you fluked out when you nailed that harridan in one turn. When you're positioning men you're a lot more aware of the quality of cover and your troops' armour.

[dmz]

Playing on 18 Stable, I've had a fair share of soldiers being one-shot with the second alien weapon encountered in the game (I believe it's the alien plasma rifle?), despite them being in cover, having full 70+ health and wearing an untouched Jackal armour - and that always after having hit the alien repeatedly doing, I would guess, very low damage. If that happens once - crap, bad luck. But I've had it happen quite a few times, and that's when it stops being funny. Having an armour shouldn't feel almost useless (I've seen far more armored guys being one-shot than armored guys surviving a single shot) but that's more or less how it felt for me.

As such, even if I haven't tried v19.7 yet (I'll wait for the stable version, since I prefer to go for slow but long lasting games and would hate to stop because of a bug or of a new build), I'm extremely wary of the increased random percentages for damage. If armours felt almost useless before, I can only think they'd feel even worse now.

Because of this, I'd like to ask: if a weapon does 50 damage (with 25-75 being the damage range after randomization), are the chances of getting a 25, a 50 and a 75 equal? If so, I believe there is a problem.

What I'd like to see is 50 having the highest chance, with every "step" in either direction lowering the chance so that highest/lowest damage are the rarest numbers to pop up. Shit will still happen - either an alien one-shotting a soldier or a soldier just grazing an alien - but the experience will be much more consistent.

Unless of course that's already how it's being done, in which case, ignore this post.

[StellarRat]

Plasma rifle are very deadly in burst mode. I hate them. The thing is with any burst weapon the alien may get lucky and put two or three rounds into your guy. That's just about 100% fatal. On the flip once in great while I've put 3 - 5 rounds from an LMG into an alien. That's pretty fatal too. Of course at one tile you dump a whole clip into just about anyone with any weapon. The only time this really irritates me is when the alien fires a full burst he's suppressed. I'm fairly sure that's a bug.

[StellarRat]

True' date=' but it's harder to tell if you fluked out when you nailed that harridan in one turn.[/quote']Yeah, the big bunny rabbits are annoying.

[Ol' Stinky]

DMZ, I imagine it's pretty much like a dice roll, so there's no weighting involved. I've not heard anyone say differently, anyway.

The only time this really irritates me is when the alien fires a full burst he's suppressed. I'm fairly sure that's a bug.

That's the #1 public enemy as far as I'm concerned. It drives me crazy.

Edited August 19, 2013 by Ol' Stinky
Misread dmz's post, somehow.

[kabill]

Because of this, I'd like to ask: if a weapon does 50 damage (with 25-75 being the damage range after randomization), are the chances of getting a 25, a 50 and a 75 equal? If so, I believe there is a problem.

What I'd like to see is 50 having the highest chance, with every "step" in either direction lowering the chance so that highest/lowest damage are the rarest numbers to pop up. Shit will still happen - either an alien one-shotting a soldier or a soldier just grazing an alien - but the experience will be much more consistent.

Unless of course that's already how it's being done, in which case, ignore this post.

Assuming it's not already, I think I'd be content with a tends-to-centre system, too (part of me, actually, is wondering whether I wouldn't prefer it, in spite of what I've written above). A simple mean of two rolls would probably work fine for this.

@Stinky - that's fair enough. Agreeing to disagree!

[Ishantil]

Is the RNG curve flat?

[Mytheos]

If the armor ends up without a small chance of death for the first shot/s then realize the new tactic is going to be open a door, wiggle your butt, get shot, move that soldier out, move a new one in, wiggle butt get shot...and rinse and repeat until you are fairly sure the Alien is out of TUs...

Then walk in and point blank them. Which isnt good game design. (When doing something silly is the best tactic you are doing it wrong)

So always having a possible risk of death when you use a tactic designed around getting your soldier shot at for any reason is needed.

[StellarRat]

Assuming it's not already, I think I'd be content with a tends-to-centre system, too (part of me, actually, is wondering whether I wouldn't prefer it, in spite of what I've written above). A simple mean of two rolls would probably work fine for this.

@Stinky - that's fair enough. Agreeing to disagree!

I like the pure random system. A bell curve is achieved either way when more than one shot is fired during a game. But I much prefer the individual shots being random. I'd even go with a 100% variance (meaning a shot could do zero to double damage) if Chris put it in.

[StellarRat]

If the armor ends up without a small chance of death for the first shot/s then realize the new tactic is going to be open a door, wiggle your butt, get shot, move that soldier out, move a new one in, wiggle butt get shot...and rinse and repeat until you are fairly sure the Alien is out of TUs...

Then walk in and point blank them. Which isnt good game design. (When doing something silly is the best tactic you are doing it wrong)

So always having a possible risk of death when you use a tactic designed around getting your soldier shot at for any reason is needed.

100% agreement.

[Mytheos]

Yeah the RNG would need a tendency towards center, OR have a increased chance of rolling higher based on distance.

So a rifle at 20 range would be almost equal chance to hit anywhere, where as at 10 range you'd be hitting the top half of the damage range near always.

The later would be MUCH better from a sound game design and feel aspect as well as realism....but an added on tendency towards center would be the quick and easy.

Edited August 19, 2013 by Mytheos

[StellarRat]

Yeah the RNG would need a tendency towards center, OR have a increased chance of rolling higher based on distance.

So a rifle at 20 range would be almost equal chance to hit anywhere, where as at 10 range you'd be hitting the top half of the damage range near always.

I hope I interpeted your comment correctly. I don't think the game system and the math support your argument. Since we don't aim for body parts making the damage go up as you get closer is essentially same as increasing the accuracy from a math stand point. You're proposal would just add even more damage. Also, the damage already increases as you close because the shots have more energy. If you take two shots you've already mathematically established a bell curve anyway. Edited August 19, 2013 by StellarRat

[kabill]

I like the pure random system. A bell curve is achieved either way when more than one shot is fired during a game. But I much prefer the individual shots being random. I'd even go with a 100% variance (meaning a shot could do zero to double damage) if Chris put it in.

This is true. As a compromise, however, it would be an effective solution. Also, while I used to be convinced of the OG format of +/-100%, at some point I decided that being hit with a giant plasma bolt and taking no damage from it just didn't make any sense to me.

[EchoFourDelta]

It almost seems like it'd be better to use a model that delineates the "winged him"/"damn, really nailed him" on a sort of bell curve type of model. Like. say 90 percent of your shots did the weapon's base damage +/- some small percentage, with 5 percent of the shots doing what amounts to grazing them, and the other 5 percent doing something like a critical. That would both normalize damage values, and keep them from being too wacky, while at the same time incorporating the random critifail/critsuccess shots we all know and love, rather than having every shot being potentially wacky.

Edited August 19, 2013 by EchoFourDelta

[dmz]

Glad my weighted values suggestion seems to be welcome - a little less glad that the term "weighted values" escaped my mind and I had to write so much to explain it. Oh well!

Depending on how fast the chances would degrade when approaching minimum/maximum damage, I think I'd even be ok with a +/-100% range modifier - the extremes would be very (very!) unlikely anyway, even the best armored soldiers would have something to fear and the whole experience would (hopefully) still remain consistent at the same time. The rationale would be soldiers (and aliens) would be aiming at center mass, with low damage scores being limb hits or grazes and high damage scores being (up to) headshots, both things happening much less often than a "standard" shot, assuming soldiers are well trained, which they should be.

[Mytheos]

I hope I interpeted your comment correctly. I don't think the game system and the math support your argument. Since we don't aim for body parts making the damage go up as you get closer is essentially same as increasing the accuracy from a math stand point. You're proposal would just add even more damage. Also, the damage already increases as you close because the shots have more energy. If you take two shots you've already mathematically established a bell curve anyway.

Accuracy is to hit, and once you hit, damage is where you hit.

So accuracy increasing with range is logical.

And as well the "average" damage increasing or the variance decreasing with range is also logical.

The concept is at max weapons range, you would be more likely to miss hitting center and hitting an arm, so less damage has a greater chance of occurrence.

However at point blank, you arent going to be missing center very often, so the variance in damage should be less.

5 feet away you are going to hit the chest or head, and the result of shooting someone in the foot 3 times in a row should almost never happen.

*This relates to your experience of taking 3 shots at close range and you had 7,7,9 and the Alien lived, so making this situation less likely as opposed to reasonably likely is a good idea.

Of course all this is considered after the fact you hit is assumed.

Edited August 19, 2013 by Mytheos

[StellarRat]

Accuracy is to hit, and once you hit, damage is where you hit.

So accuracy increasing with range is logical.

And as well the "average" damage increasing or the variance decreasing with range is also logical.

The concept is at max weapons range, you would be more likely to miss hitting center and hitting an arm, so less damage has a greater chance of occurrence.

However at point blank, you arent going to be missing center very often, so the variance in damage should be less.

5 feet away you are going to hit the chest or head, and the result of shooting someone in the foot 3 times in a row should almost never happen.

*This relates to your experience of taking 3 shots at close range and you had 7,7,9 and the Alien lived, so making this situation less likely as opposed to reasonably likely is a good idea.

Of course all this is considered after the fact you hit is assumed.

Yes, but the game doesn't "place" hits. So, a hit's a hit. The damage will work out to the middle amount for each weapon over more than one hit. It will naturally assume a bell curve. We don't need an increasing damage amount for closeness just due to center hits. In the end the math take care of itself because you'll get more hits at closer ranges. Adding another factor just complicates the calculations. If you come up with a reason why this would be useful or change the tactics people use that might convince me. If you could select what part you were aiming for, head, leg, body (ala JA2) then your argument holds water in my book. But in JA2, those shots have different hit probabilities not a uni-probability like we have in Xenonauts. Also, one has to assume Xenonaut armor provides the most protection in the most vital areas which tends to mitigate the damage. Edited August 19, 2013 by StellarRat

[Chris]

The weighted average idea is fine, but doesn't necessarily seem an upgrade to me (just equally valid but not what we're using). It makes "special" events rarer, but then I don't think that's necessarily a good thing. Pure random provides more variation and therefore more interesting events.

[dmz]

The weighted average idea is fine, but doesn't necessarily seem an upgrade to me (just equally valid but not what we're using). It makes "special" events rarer, but then I don't think that's necessarily a good thing. Pure random provides more variation and therefore more interesting events.

I'm not sure it would necessarily be more "interesting" events. Since Xenonauts is a strategic game, I enjoy having to plan ahead, use strategies etcetera but I find extremely annoying when I consistently get soldiers one-shotted despite them having (Jackal) armor, just because I have bad luck. I'm fine with some bad luck, so weighted chances wouldn't bother me too much as those events would be kind of rare, but right now it's not some bad luck, it's consistent bad luck, which impacts negatively on my experience, simply because with such variation those extreme events happen so often it feels like my decisions don't count (or, at any rate, don't count as much as luck).

With weighted chances I may lose one soldier in Jackal armor to a single shot every once in a long while, and maybe lose a lot of them when they get hit twice due to bad placement or bad strategies or, in general, due to something that's my fault. With chances as they are, I still get punished for things that are my fault but the divide between bad strategies and bad luck is much thinner, giving way more importance to luck than it should (in my opinion) have.

[Ol' Stinky]

If the armor ends up without a small chance of death for the first shot/s then realize the new tactic is going to be open a door, wiggle your butt, get shot, move that soldier out, move a new one in, wiggle butt get shot...and rinse and repeat until you are fairly sure the Alien is out of TUs...

Then walk in and point blank them. Which isnt good game design. (When doing something silly is the best tactic you are doing it wrong)

So always having a possible risk of death when you use a tactic designed around getting your soldier shot at for any reason is needed.

You're not going to cripple multiple soldiers to drain enemy TUs. Nobody's suggesting that soldiers shouldn't take any damage, just that the price of a bad roll not be so severe.

[GizmoGomez]

Personally, I'm fine with the current damage variance. I mean, stuff happens, despite our best plans.

Even the best player with the perfect tactics should probably lose a soldier or two. I mean, think about it logically; this isn't the A-Team, people get hurt and killed, despite our best efforts. If everyone survives, then it's not quite the XCom I've come to love.

(Unless you save scum. I did that in UFO94 quite a bit, but in this I avoid it. Balancing is easier when you aren't "cheating" the balance).

Difficulty Suggestion: Perhaps the RNG for damage could be tweaked in some way for higher difficulties? Just a thought.

[cody522]

Cody - the damage variance actually represents the difference between getting shot in the foot and shot in the face. In reality that probably represents more than a + / - 50% increase in damage.

Would it be possible to add onto the accuracy skill so your more experienced troops have a higher damage output than the greenhorns?

Like say a Major can shoot between 50-70 damage

while a corporal can shoot between 40 and 70?

[Ishantil]

Glad to see some intense discussion going on here.

If memory serves, X-COM had hit placement, but I'm not sure if it had any affect on damage. Or other side effects.

I know you had to heal and there was a little healing window that came up where you healed the location.

Edited August 19, 2013 by Ishantil

[GizmoGomez]

Would it be possible to add onto the accuracy skill so your more experienced troops have a higher damage output than the greenhorns?

Like say a Major can shoot between 50-70 damage

while a corporal can shoot between 40 and 70?

I disagree; you'd be penalized for losing soldiers even more, thus, save scumming would run rampant, and Tony Stark wouldn't be able to play his game mode without ragequitting.

[StellarRat]

Glad to see some intense discussion going on here.

If memory serves, X-COM had hit placement, but I'm not sure if it had any affect on damage. Or other side effects.

I know you had to heal and there was a little healing window that came up where you healed the location.

If I remember correctly their (the OG developers) intention was to have hit placement matter, but they never got that far. The soldier diagram did show you where the damage was, but other than requiring you to heal the correct location it had no other effect in the game from what I could tell. Someone correct me if I'm wrong.