aliens

Author Topic: Heat vision formulas and statistics  (Read 7868 times)

Offline Meridian

  • Global Moderator
  • Commander
  • *****
  • Posts: 9115
    • View Profile
Heat vision formulas and statistics
« on: September 15, 2016, 12:42:50 pm »
Continuation of discussion on YouTube...

Quote
1. There is a big difference between even 90% and 100% (max value I used is 80%, by the way). Only a Sith deals in absolutes.

In practical terms there is no difference between 90% and 100%.
Can you give me an example of a scenario so that we can run the numbers on it?

Quote
2. Not really triple, since half is usually more than enough (smoke gren dropped under your feet gives 360 protection). So actually double should insulate fully against up to ~70% ThV.

In theory, ideal non-overlapping double smoke cloud from standard smoke grenade protects against approx. 85-97% ThV, depending on distance between units.

But that is extremely theoretical!
Any of the following decreases the protection significantly:
a/ enemy moving towards you (inside the smoke)
b/ enemy flanking you (around the smoke) -- this makes the most difference
c/ each turn decreases smoke density

EDIT: I removed the numbers, since they depends heavily on distance to target. But the decrease in protection can be anywhere between 0 and two thirds.

Quote
3. The world is full of psychos, and there are weapons to smoke up whole map; also player has their own ThV so can play this to their advantage.

Fair enough, I haven't seen such weapons yet.
But psychos will always find a way.
Point I am trying to make is that smoke should give some protection and/or advantage to you too... but if you say that player has their own ThV too, maybe it's already balanced... I haven't looked which armor has ThV and how much.
Majority of players are not like hellrazor (I think), in my opinion the balance should be for an average player, not for the cheesiest of the cheesiest.

Quote
4. Only few enemy factions have more than 60% ThV. Most enemies have either 0 or 40-60. They're usually mixed. So smoke does help but you gotta use hard cover too.

I still think 60% is too much (unless the player has easy-ish access to powerful-ish ThV too of course).

Already 20% gives them the ability to completely negate "mutual surprise", which is the biggest factor (in my opinion).
40% gives them a window to spend their full TUs on reaction fire before you even spot them (in most common cases).

Linked is a Google Docs sheet, which you can use to calculate the effectiveness of ThV.
It works well for turn zero and non-overlapping smoke clouds.
With overlapping smoke clouds, margin for error should be around 10% (normal 10%, not flat).
With passing turns, margin for error should be also around 10% (not flat) until 6 turns or so, then it becomes bigger (I didn't test how much exactly).

PS: for others a table to have a rough idea how much ThV guarantees to see through smoke walls with different thickness under perfect conditions:

Note: numbers differ significantly with total distance to target, this example is with 14 tiles between units

3-4 tiles: heat vision 0%
5 tiles: heat vision 19%
6 tiles: heat vision 33%
7 tiles: heat vision 42%
8 tiles: heat vision 50%
9 tiles: heat vision 55%
10 tiles: heat vision 60%

Please note that these percentages guarantee visibility even under perfect conditions.
Under imperfect (i.e. normal) conditions, these numbers drop significantly as described above.
« Last Edit: September 15, 2016, 02:47:26 pm by Meridian »

Offline Dioxine

  • Commander
  • *****
  • Posts: 5460
  • punk not dead
    • View Profile
    • Nocturnal Productions
Re: Heat vision formulas and statistics
« Reply #1 on: September 15, 2016, 08:47:20 pm »
Thank you for the theory; this will help a lot with actual testing. Obviously the first version I published is rough and based more on hunches than detailed analysis (as I had no access to said theory). However it doesn't mean I didn't try to strike some balance. Smoke games are still in if the player uses ThV armors, or absurd amounts of smoke are used (which I can easily see happening, with all crew starting with pre-primed smoke, not even mentioning specialized smoke weapons, like Blizzard). Otherwise smoke protects only against selection of enemy crews; note that most enemies don't have ThV.

The difference between 90 and 100 is indeed slight, because of 40 tile vision limit; I meant that 100 gives total transparency, while 90 is eventually stopped by enough smoke. With that in mind, highest ThV value I'm using is 80.

As for the cheese balance: while I agree, the cheesiest of the cheesiest tactics IMO (map-edge smoke-camping) is used very, very commonly. The ThV numbers I used are aimed at disrupting precisely this one.

Offline Arthanor

  • Commander
  • *****
  • Posts: 2488
  • XCom Armoury Quartermaster
    • View Profile
Re: Heat vision formulas and statistics
« Reply #2 on: September 15, 2016, 10:09:05 pm »
I'm curious: How can you make nerf smoke such that edge camping isn't efficient anymore, yet keep it useful to other circumstances (most importantly, covering disembarkation).

If you're sniping from the edge, you're not moving and can afford to apply loads and loads of smoke once. If you're moving, you always need the cover in a different place, so are applying smoke always in a different place instead of increasing the density of existing smoke. This means even lower density. If something can see through enough smoke to hamper smoke&snipe, it will easily see through the rest, no?

I'm asking out of genuine interest. I haven't played since the new psi/thermal visibility idea, just a bit with the basic camo, but all this is worrying me. Armors tend to be expensive, not quite death proof and pretty difficult to recover (ex.: advanced personal armors giving one part, which isn't much of a step towards rebuilding the whole armor). Because of this, smoke was often the best/only defense to be usable: sending a gal in crap armor + smoke and losing her was a better investment than sending an armored gal, because the reduction in death likelihood wasn't worth the rarity of parts.. This is similar to the "rookie off the skyranger/in the UFO" tactics, which I think is pretty poor form as well (if not worse than sniping).

Offline Dioxine

  • Commander
  • *****
  • Posts: 5460
  • punk not dead
    • View Profile
    • Nocturnal Productions
Re: Heat vision formulas and statistics
« Reply #3 on: September 15, 2016, 10:29:16 pm »
You can't. However, disembarkation and advancement can be done in a multitude of ways; while edge-sniping is single-minded. So you can devise new disembarkation methods (doors are ever popular) while edge-smoking is simply less foolproof now, period.

Offline Meridian

  • Global Moderator
  • Commander
  • *****
  • Posts: 9115
    • View Profile
Re: Heat vision formulas and statistics
« Reply #4 on: September 15, 2016, 10:30:21 pm »
I'm curious: How can you make nerf smoke such that edge camping isn't efficient anymore, yet keep it useful to other circumstances (most importantly, covering disembarkation).

Good that you mention it.
"Map-edge smoke-camping" is not much of an interest to me, whoever uses it is "selber schuld" ("it's their own fault" in English I guess or something like that). Such players are either too new to the game and are just trying to survive; or too cheesy to deserve our attention. It takes a lot of time to set it up and they are basically just punishing themselves.

What I would like to see being less effective than it is right now is exactly the disembarkation cover... it should hurt. "Smoke and wait 1 turn" is definitely in the "Cheesiest Top 3" and unlike the previous cheese, it takes only seconds to set up and perform. It's so damn easy and effective that even I can't help myself and use it a lot more than I'm ready to admit. And to prevent this (especially with craft which are not on ground level) even very small levels of ThV are already enough :)


Offline nrafield

  • Captain
  • ***
  • Posts: 77
    • View Profile
Re: Heat vision formulas and statistics
« Reply #5 on: September 15, 2016, 10:46:53 pm »
I wouldn't say it's cheesy.  It can be all the difference between a clean sweep and having a blaster bomb flown into Skyranger turn 2. Would that be fun?

Offline Dioxine

  • Commander
  • *****
  • Posts: 5460
  • punk not dead
    • View Profile
    • Nocturnal Productions
Re: Heat vision formulas and statistics
« Reply #6 on: September 15, 2016, 10:54:54 pm »
Isn't the ability to change a TPK into a clean sweep with a couple of clicks the very definition of cheesiness?

Offline Meridian

  • Global Moderator
  • Commander
  • *****
  • Posts: 9115
    • View Profile
Re: Heat vision formulas and statistics
« Reply #7 on: September 15, 2016, 11:02:28 pm »
Isn't the ability to change a TPK into a clean sweep with a couple of clicks the very definition of cheesiness?

Yes, it is.
Well said.

I wouldn't say it's cheesy.  It can be all the difference between a clean sweep and having a blaster bomb flown into Skyranger turn 2. Would that be fun?

On top of Dioxine's excellent comment... the chance of blaster launcher on turn 2 in vanilla Xcom is exactly 0%.
AI is allowed to use blaster launchers (and also grenades!) only from turn 3...

Offline Arthanor

  • Commander
  • *****
  • Posts: 2488
  • XCom Armoury Quartermaster
    • View Profile
Re: Heat vision formulas and statistics
« Reply #8 on: September 15, 2016, 11:09:14 pm »
hum.. I gotta say this isn't the answer I was hoping for.

Sure, smoke is overly effective on deployment, but the D-Day style of deployment is so flawed that it needs this kind of compensation. Otherwise, you're just sacrificing the first 1-4 units off the ramp for nothing every mission, and that's not fun, or realistic, XCom rookies joke (and laughing at people playing LPs) aside. I'm no expert, but it seems to me like deploying in the middle of a firefight makes little sense unless you have no alternative and heavy guns covering you or have such numbers to face the losses (ex.: D-Day, Vietnam chopper deployment under fire, although that was arguably not the preferred tactic). Do you see the SWAT drive its vans all the way into the driveway/garage of a house with crazed gunmen? No, they make a perimeter, wait for the "tank" and rush in in formation with everyone able to fight, ideally coming from different angles. APCs driving in the middle of an enemy town with troops inside? No, they deploy before fighting, indeed probably before entering the town and spread out/sweep.

In XCom, you're never ambushed, you're ambushing the enemy or responding to them. It's the same with Piratez (you are the one shooting down the shipping, or responding the pogroms). Yet every time you deploy, you are getting ambushed. Except in Piratez against a few ufos where all units stay inside.

Arguably, tanks and the merc commando are good alternatives to rookie sacrifices for Piratez (not so much for vanilla when starting against plasma weapons), and making them more relevant might not be a bad thing. But if casualties are to be increased at deployment, I hope there will be considerations made for the cost of such increased casualties. The farming necessary for end game is already pretty tedious, add an increased sink of armors and it might hamper the fun that's already dwindling by the end.

Offline Meridian

  • Global Moderator
  • Commander
  • *****
  • Posts: 9115
    • View Profile
Re: Heat vision formulas and statistics
« Reply #9 on: September 15, 2016, 11:19:41 pm »
I think there's tons of options (in piratez):
1. use craft with closed doors
2. use craft with multiple exits
3. use parrots to draw reaction fire
4. prepare already before mission to fire immediately with almost certain kill without moving on the most dangerous enemy facing you (early: powder bomb, middle: sniper or missile, late: blaster?)
5. take off if it's really bad
6. take your chances they miss
7. mushroom beer
8. mind control
9. etc. etc.
« Last Edit: September 15, 2016, 11:21:51 pm by Meridian »

Offline Dioxine

  • Commander
  • *****
  • Posts: 5460
  • punk not dead
    • View Profile
    • Nocturnal Productions
Re: Heat vision formulas and statistics
« Reply #10 on: September 15, 2016, 11:25:03 pm »
Realism discussion is a bit off-topic, but if you use a drop ship to land in midst of the enemies, then, yes, you will ALWAYS get ambushed. Normally you use speed and artillery to make that ambush more survivable, and numbers to exploit the fact you attacked a vulnerable rear position (stealth is also used, but obviously impossible in Piratez-style airborne insertion); but before breakthrough is attained, losses are always substantial. Only after these initial losses, if all goes according to plan, it's the enemy who starts to take quickly mounting losses, far ofsetting your initial losses. WWI is full of good examples of infiltration of this type, that gone either good or horribly wrong. But even an insertion against vastly inferior enemy can go wrong (Mogadishu).

SWAT isn't a good example, since they do not land&secure, they breach into a stronghold of an enemy who is already surrounded and cut off.

Offline Eddie

  • Commander
  • *****
  • Posts: 560
    • View Profile
Re: Heat vision formulas and statistics
« Reply #11 on: September 16, 2016, 03:19:00 pm »
The game can only be challanging up to a certain point for a good player because you are still playing a computer. You will always think better than the computer and find its weakness. The AI is dumb. It's actually not much of an AI, it's just a set of rules that are beeing followed.
There is always one way of playing that is the most effective. You can call it cheese or good tactic, just depends on your point of view.

You can change the game rules to make certain tactics less effective. Then players have to figure out which is the new best playing style. To me, this is the fun part. But as time passes, you will have it figured out or others tell you how to do it. And then you are right where you started.

So with these limitations in mind, I think good game design is having missions/factions that require different strategies so that gameplay is diverse. Dioxine already does a pretty good job at that. Yay ratman rodeos!

Offline Dioxine

  • Commander
  • *****
  • Posts: 5460
  • punk not dead
    • View Profile
    • Nocturnal Productions
Re: Heat vision formulas and statistics
« Reply #12 on: September 17, 2016, 09:33:26 am »
You can change the game rules to make certain tactics less effective. Then players have to figure out which is the new best playing style. To me, this is the fun part. But as time passes, you will have it figured out or others tell you how to do it. And then you are right where you started.

An argument that's only true for most part. True, the AI is dumb and you can play it, but increasing the complexity and number of factors increases the number of possible scenarios, making it less obvious how to play. It's not only about learning the tactics, it's also about choosing the right one.

Offline legionof1

  • Moderator
  • Commander
  • ***
  • Posts: 1900
  • Bullets go that way. Money comes this way.
    • View Profile
Re: Heat vision formulas and statistics
« Reply #13 on: September 18, 2016, 12:51:37 am »
The biggest issue with the new vision is maintaining parity with the aliens you face if you plan to use smoke as a cornerstone of your play. Otherwise smoke is still gona do it's theoretical job of breaking long LOS lines so the player can advance safely. Meridian's table in the original post was for 14 tiles between units. As long as ThV dosen't negate a smoke grenades initial radius it's is still a wall as far as distant AI units care.

Granted if you are at disadvantage you can't  then advance through the same smoke without risk but the scouting units are expendable or should be.

There is a point if your scouts can't be equipped with advanced ThV armor because of rebuild costs. Armors that have several intermediate objects take awhile and lots of resources. I myself don't use 90% of armor options because of the potential lost resources. I go from starting tech to guerrilla/Amazon/plate.  Once  I'm ready to launch the final assault I then build one top tier setup to kit out final team.