Fractal Softworks Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Starsector 0.95a is out! (03/26/21); Blog post: Of Slipstreams and Sensor Ghosts (09/24/21)

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Vanshilar

Pages: [1] 2 3 ... 8
1
Blog Posts / Re: Of Slipstreams and Sensor Ghosts
« on: September 24, 2021, 11:47:00 AM »
Hmm...when you were reading up on fluid dynamics, did you consider modeling the flow of a fluid inside of a pipe? It turns out that (due to boundary conditions -- flow is assumed zero where the fluid meets the walls of the pipe) you naturally get a nice effect where the flow rate is parabolic (fastest in the middle, going down to zero at the walls), at least for laminar flow. I'm sure there are some good fluid-in-pipe software and papers out there to make it easier to render, plus, that may be a nice approximation for the speed bonus that you get when inside the slipstream; that the bonus speed you get (beyond your normal speed) is parabolic, with the peak when you're in the middle of the slipstream.

Along with this, there should also be a lot of resources for modeling fluid flow inside a curved pipe. Unfortunately, I don't remember much about this in my undergrad days, other than 1) the pressure on the outside part of the "knee" or bend is higher than the pressure on the inside, since that's basically what makes the fluid curve around, and 2) one homework assignment involved doing a finite element simulation of the fluid flow in a 90 degree bend in a pipe. Everybody except me did it in Matlab since it was just a (relatively) few short lines of code. I didn't know Matlab, so I ended up doing it in...Excel. That's right, a finite element fluid flow simulator written up in Excel for a homework assignment. Sigh.

It seems like the slipstreams are essentially a bunch of curves in a 2-D plane. In which case, it may be computationally easier to model the fluid flow that way (as fluid flow inside a curved pipe), so that the velocity curves are basically pre-computed (i.e. easily determined from a few input parameters). If I remember correctly, the faster fluid flow is actually on the outside part of the knee, not on the inside, which raises an interesting decision for the player: they go faster when they're fairly close to the outer boundary, but if they go too close, then the velocity bonus decays very rapidly. So there's a natural risk vs reward decision in that case.

2
General Discussion / Re: Legion XIV > STD Legion
« on: September 18, 2021, 06:18:14 AM »
At some point some wag is going to come along and say that the vanilla version of the game should henceforth be called STD Starsector.

3
Another minor point that I didn't see mentioned (on mobile though so I perhaps missed it): When hitting residual armor, that 0.75 multiplier in 0.75*D disappears (like the 2x damage bonus of HE weapons against residual armor).

That's because it's hitting hull, not armor. So the 0.75*H applies due to residual armor (i.e. for damage reduction, it's always as if it's trying to hit armor for H / (H + A)), but the 0.75*D doesn't apply since it's not actually hitting armor, but hull instead.

4
Ahh, I see, thank you for elaborating! After double-checking - this is actually how reductions to the damage taken by armor work; they're applied to the raw damage *before* the armor's intrinsic damage reduction is calculated.

Oh I see! I was thinking of it as the game doing the D * H / (H + A) first, then * 0.75 afterward. You're saying any incoming damage is multiplied by 0.75 first, then goes through the damage reduction calculation. In which case it's actually 0.75*D * 0.75*H / (0.75*H + A). Mathematically it ends up with the extra factor of 0.75 for the H's but otherwise the same.

In this case then the desc says it's -25% armor damage taken, but it's actually much more powerful, because the *0.75 is applied twice in the numerator. That's somewhat balanced out by the 0.75 in the denominator, but not completely due to the A term. Hence in this case, an Eagle taking a harpoon hit, it actually reduced the damage by 34%, from 900 (without Impact Mitigation) to 595 (with Impact Mitigation). So it's actually pretty good. In fact -25% is actually the limiting case where armor is 0; the higher the armor value, the more Impact Mitigation does, up until the armor is 425% of the incoming damage, where the H / (H + A) "bottoms out" at 15%, and so Impact Mitigation actually reduces the damage by 41% at that point. After that it gradually decreases in effectiveness (due to the 15% minimum), until the armor is about 567% of the incoming damage, when the damage hits 15%, so it goes back to reducing the damage by 75%. For much of the range Impact Mitigation actually decreases incoming damage by around 35-40%, i.e. you actually only take around 60-65% of the damage compared with if you didn't have Impact Mitigation. (I'll post a graph later that explains this a lot better I think.)

I don't know if there's an easy way to change the desc to eliminate this ambiguity though, since against hull, this only affects the hit strength H and not the actual damage D. "-25% incoming damage against armor"? It might look similar to the current desc "-25% armor damage taken" at first glance but "taken", being the past tense, implies something applied afterward, as opposed to reducing the incoming damage. Probably needs a better wordsmith than I to see if there's a way to clarify what it does easily, or just leave it as-is.

5
Hmm, where are you getting that info? I'm pretty positive this isn't the case. I mean, maybe I'm missing something that's staring me in the face? But I'm looking at the skill definition and I don't see it.

Yeah I was surprised because it wasn't anywhere in ImpactMitigation.java (so maybe it's something in the part of the code that actually does the damage reduction calculation? but I have no idea where that is), but it's through testing.

For example, taking an Eagle (1000 base armor), assign the player character to it, no skills. I used Target Practice mod's harpoon ship to fire a bunch of harpoon missiles (750 HE damage) at my Eagle.

Without any skills, the Eagle takes 899 armor damage on the first hit, then after armor is gone, it takes 725 (or 726) damage per hit to hull. This fits with the D * H / (H + A) calculation. So far so good.

With Impact Mitigation, the Eagle takes 595 armor damage on the first hit, then after armor is gone, it takes 716 (or 717) damage per hit to hull. This doesn't fit with -25% armor damage taken, which would predict 675 armor damage on the first hit, and 725 (or 726) damage to hull. However, it fits perfectly with -25% armor damage taken and the weapon's hit strength decreasing from 1500 to 1125. Every other data point in my post above (where I tested Eagle with nothing, with Impact, or with Impact + Elite, and ship with nothing (1000 armor), heavy armor (1400 armor), or heavy armor + armored weapon mount (1500 armor)) conforms with this as well, along with my testing using Medusa with stock base armor (300) and setting the base armor to 25. I can't think of any other explanation that makes the numbers work out other than that the hit strength is also reduced by 25% if the target has Impact Mitigation. (Maybe Target Practice's harpoon-firing drones does something funky in the code if there's Impact Mitigation, but I hope not.)

Actually, just tested this using sim Onslaught. TPC, 250 energy damage, vs my Eagle. No skills, it does 49 damage on the first hit against full armor, pretty much the predicted 50 damage accounting for rounding. With Impact Mitigation, it does 29 damage. The -25% armor damage taken predicts that it should've done 37.5 damage instead, but if the hit strength is also decreased from 250 to 187.5, the predicted damage is 29.6, which fits the observed data.

I only use utility mods, no faction mods, so I don't think it's coming from any of the mods. I can give a full modlist if you get different numbers.

Ahh, interesting! I'm trying to remember if I was actively aware of that at some point, but regardless, I certainly wasn't keeping that in mind when changing the value from 150 to 50. Made a note to have a look; that's really weak.

I don't know if a residual armor of 50 is particularly weak. It means no effect for ships with a base armor of 1000 or above, but helps out all other ships once their armor is gone.

On the other hand, a residual armor of 150 is really strong, since it's "as if" the ship originally had a base armor of 3000. A pulse laser (hit strength 100) does 67% damage to a residual armor of 50, but 40% damage to a residual armor of 150. So a residual armor of 150 pretty much means the attacker needs to rely on armor-breakers like Heavy Blaster or HIL even when the armor is gone.

6
Huh, TIL. I might run a few tests on enforcers or another high armor ship to double check but those numbers seem right to me. Surprising!

I know! I didn't think to test Impact Mitigation until this thread came up. Turns there are two "hidden" effects that aren't documented in the game, on the wiki, nor anywhere else that I'm aware of:

1. Impact Mitigation reduces the weapon's hit strength by 25% for damage reduction calculation.

2. Impact Mitigation's elite effect does not directly add to residual armor. (I.e., if the base armor was 1400, then the residual armor will be 70, not 70+50 = 120, nor (1400+50)/20 = 72.5). Rather, if the residual armor is less than 50, then it'll set a minimum of 50 for the damage reduction calculation.

The latter may be important if your ship has weak armor, i.e. is a frigate for example. A Tempest for example has a base armor of 200, and thus a residual armor of 10. That means that a tactical laser, which does 75 DPS and thus a hit strength of 37.5, would do 75 * 37.5 / (37.5 + 10) = 59.2 DPS against it. But if the Tempest has Impact Mitigation, then it would do only 75 * 37.5 / (37.5 + 50) = 32.1 DPS, or 54% of the original damage (damage cut almost in half). So it can do a lot to make weak ships more survivable.

I wonder what other skills have these hidden or undocumented effects.

7
One minor point: when it comes to residual armor, the elite effect still occurs, but only up to 50 armor. Thus, the elite effect has no impact on residual armor for any ship with over 1000 armor, but for less armored ships it does do a little (probably not worthwhile in the end, but non-zero).

e.g. an armor stripped IM+ Condor has 25 armor, which would be treated as 50 for damage calculations.

Huh that's interesting! Yes I tested with a Medusa (300 base armor) against the Target Practice's harpoons (750 HE damage) and this is how it works -- it does 742 damage normally (matching the D * H / (H + A) of 750 * 1500 / (1500 + 15)), 740 damage with Impact Mitigation (matching 750 * 1125 / (1125 + 15)), and 718 damage with Impact Mitigation elite (matching 750 * 1125 / (1125 + 50)).

Incidentally when I changed the Medusa's armor to 25, the first shot against it was as if it had 75 armor, not 50. So at full armor, the elite effect adds 50, while when armor is gone, the elite effect gives a floor of 50, for damage reduction calculation purposes. I have no idea what happens once the armor starts taking damage (i.e. if it's a flat +50 or if it scales (decreases) as armor decreases), nor what happens when some armor cells are still present but others are gone (i.e. hitting hull on some cells but not others). But this at least gives what happens at full armor and when armor is gone.

this leaves me with the assumption that it just adds 50 flat armour but doesn't show it in the ui, which is uhhhhhhhhhhhhhhhh let's just say not good

Impact Mitigation doesn't actually affect the amount of armor that's present, and thus it wouldn't show up in the UI. Rather, it adjusts the numbers H and A of the D * H / (H + A) damage reduction calculation. So for example even though the elite effect says it adds 50 armor for damage reduction, your ship still has whatever base armor is shown in the UI.

8
General Discussion / Re: Build Two Items into new Paragon & Hegemony Rep
« on: September 17, 2021, 10:02:17 AM »
Edit: On another note, I was informed of an AI inspection by the Hegemony while I was writing the original post and I saw one of the options was to resist. Why not? How mad could they get? Well, I went form +100 to -50 just like that and I lost my nearly 100k a month commission. Darn that's a lot of money at this stage.

The whole thing makes me wonder if there's no way to make peace with them. Maybe I should align with Tri-Tachyon because they seem to care less about all the AI's I have in my two colonies. They're both run by the Blue AI, and every industry has some AI help. I'm not sure, but it seems to me that the Hegemony would just keep sending the AI inspections and I'd keep losing story points. There are five planets in Duzahk and one colony has a Star Fortress (largest) + High Command and the other has a Battle Station + is building high command. As soon as I have the extra credits I'll upgrade the second colony to Star Fortress as well. Maybe I should just let the Hegemony bang their heads against my superior fleets while I run some bounty and pirate missions for Tri-Tachyon and bring my rep up?

Yes, the Hegemony will keep sending AI inspections every so often when you use alpha cores. If you resist then they'll go to war with you.

However, if you turn your transponder off and intercept that AI inspection fleet in hyperspace, and kill it, then you'll only lose 5 rep, since you're an unknown fleet (with transponder off). So with good relations with them, you'll pretty much never have to actually worry about those AI inspections, as long as you have a fleet that can kill the AI inspection fleet.

9
Eh there are two different bonuses at work here (a 3rd if you include the weapon/engine damage reduction):

1. -25% armor damage taken
2. (elite effect) +50 to armor for damage reduction calculation purposes

The way damage to armor/hull works in Starsector is that it goes through a damage reduction calculation, where the incoming damage D is multiplied by H / (H + A), where H is the weapon's hit strength, and A is the target's armor. So for example, if the incoming damage is 500, has a hit strength of 500, and the target's armor is 1000, then it'll end up doing 500 * 500 / (500 + 1000) = 167 damage.

The "damage reduction calculation" refers to the H / (H + A) part of the equation. It can have a minimum value of 0.15, i.e. if the hit strength is too low, it'll always do at least 15%. Generally for projectiles, the hit strength is equal to the damage per bullet, while for beams, the hit strength is half of the damage per second (note that for burst beams, you have to look at the DPS while the beam is on, not its average DPS -- this value can be found in weapon_data.csv).

Once the armor is stripped away, there is a 5% residual armor when a weapon hits hull, meaning that for damage reduction calculation purposes, the weapon will still count as if it's hitting armor, but at 5% of the armor's original value. So for a ship with 1000 base armor, it'll count as if there is still 5% or 50 armor for the damage reduction calculation. The original damage part (i.e. D) remains unchanged. The important thing to note here is that this means the kinetic/high explosive/fragmentation bonuses/penalties vs armor still apply toward H / (H + A), even though they do not apply to D. Hence HE will still do better vs hull than kinetic or frag, assuming the base damage is the same.

Impact mitigation does as what it says "on the carton". If the armor damage would have been 400 damage, after going through all the calculations, you'll only take 75% or 300 damage instead.

However, there seems to be an additional hidden effect: the incoming weapon's hit strength is also decreased by 25%. So if the weapon's hit strength is supposed to be 400, it'll end up being 300 for the damage reduction calculation.

The elite effect also does what it says "on the carton". When calculating H / (H + A), it'll add 50 to the armor. I have no idea if it decreases as the armor decreases or not (i.e. if it scales with the % armor remaining). However, it does not apply to residual armor, as far as I can tell. Once the armor is gone and the weapon is hitting hull, I don't see a difference whether the target has the elite effect or not.

This is based on testing with a stock Eagle, with the player as officer, testing with skills ranged spec (i.e. no impact mitigation), impact mitigation, or impact mitigation + elite bonus, and with the ship having base armor (1000), heavy armor (1400), or heavy armor + armored weapon mounts (1500). It also had blast doors and reinforced bulkheads for extra hull but I don't think they affect the damage reduction calculations. The weapon used for testing is the Target Practice mod's harpoon ship, firing harpoons which do 750 HE damage each. The test results, against full armor and against bare hull (i.e. with only residual armor remaining), with the conditions of ranged spec (i.e. no impact mitigation), impact mitigation, or impact mitigation with elite bonus as the columns, were:

Nothing: 899 / 725 | 595 / 718 | 581 / 718
Heavy Armor: 775 / 716 | 501 / 706 | 491 / 706
Heavy + Armored Mount: 749 / 715 | 482 / 703 | 473 / 703

Hope it makes sense despite the formatting. Basically, what I described above fits the data perfectly; the harpoon does 1500 damage and has 1500 hit strength vs armor, which drops to 750 damage and 1500 hit strength vs hull, but once there's impact mitigation, the hit strength becomes 1125 instead. On the armor side, it's fixed as 1000/1400/1500 when full and 50/70/75 when stripped, regardless of if there's impact mitigation or its elite effect.

10
Announcements / Re: Starsector 0.95a (Released) Patch Notes
« on: September 13, 2021, 09:09:23 PM »
Hmm, that's a good idea - at least specifically about introducing officers in the tutorial (maybe the agent you contact at Derinkuyu needs a lift off the station and joins you?). Not 100% on it being a merc, worth thinking about though.

Actually a good time might be when the player goes to retrieve the ships, since that's near the end of the tutorial (since officers have upkeep), and it's when the player will feel more like they have a "fleet" of ships now.

The officer should probably have a fixed personality of steady, not sure if the skill should also be fixed, to prevent the player from getting something wonky. A timid or reckless officer might turn a new player away from getting more officers.

11
Announcements / Re: Starsector 0.95a (Released) Patch Notes
« on: September 13, 2021, 06:03:57 PM »
It just seems a bit rough when officers are now hard required or it makes the game 10x harder, but the only way to get them is play 3 separate lotteries each of which involves a significant investment of playtime which may or may not line up with what you actually want to do.
If officers are so vitally important that the game is significantly harder without them, why is there no guaranteed way of getting them regardless of what the player is doing?

Eh I think you're looking at this situation wrong.

We as fellow forum-goers are trying to tell your that officers will help your fleet immensely. Basically that it's worth your time to get them.

If you choose not to make the effort, you don't have to, but then that's on you as a player, to play the game that way. It's not up to the game to force you to get officers via some guaranteed means, you're supposed to make the effort. You can lead a horse to water and all that. Yes it's probabilistic. But the chance of you never coming across an officer is vanishingly small after a little bit of play time. I mean there's a chance of you rolling a 1 100 times in a row, but we don't really worry about that.

Also, the game does guarantee you officers, if you go out and explore. There are a guaranteed number of officers per game that you'll find in cryo pods; in fact IIRC Alex will be decreasing that amount next patch because there are currently too many (i.e. it's too easy).

If you're visiting markets, you're bound to come across officers on the comm board. If you're exploring, you're bound to come across officers in cryo pods. If you're fighting, you'll get officers from your own crew. So the game gives you officers pretty much no matter what you're doing, unless you're not visiting markets, not exploring, and not fighting, in which case, I'm not sure what else there is to do.

12
General Discussion / Re: For those who don't pilot their own ships
« on: September 10, 2021, 10:37:22 AM »
For adjusting the battle size, you can go into \starsector-core\data\config\settings.json, search for "maxBattleSize", and then set it to whatever you want. Each side can deploy up to 60% of the battle size, hence your 240 DP cap when the battle size is 400. (Note that this is just a limit on how much you can deploy at once, not the overall size of your fleet.)

Be sure to make a backup of the settings.json file first! So you can always go back to the original version if anything messes up. There are many settings that you can customize to improve your playing experience. For example, if you don't like being blinded every time a ship dies, you can look for "enableShipExplosionWhiteout" and set it to "false".


13
General Discussion / Re: IR pulse laser seem lack luster
« on: September 09, 2021, 12:23:18 PM »
and i have a hard time considering autopulse a burst gun when you take around 3 second to empty the mag
This is a sad day, for never again will I read this post for the first time.

Technically, if the target dies within 3 seconds, he's not wrong.

14
General Discussion / Re: A Veteran's Insight On Guns
« on: September 06, 2021, 01:02:22 AM »
Not to mention that it gets coolness points for being flingable with plasma burn.

Man I just tried this with a Fury and...oh man this is so wrong.

In inspiration I tried this with an Onslaught, including some standard bomb bays (just for fun) and...haha screenshot attached.

15
General Discussion / Re: A Veteran's Insight On Guns
« on: September 05, 2021, 11:35:43 PM »
I find these weapon/ship threads very interesting and I'll comment more later when I have time! But briefly, I just want to note that the Breach script does 200 damage to armor, unaffected by armor reduction; in other words it directly does 200/15 or 13.333 damage to each of the 9 inner cells and 200/30 or 6.667 damage to each of the 12 outer cells, without going through the damage reduction formula. However, it only affects armor (not hull).

So even at minimum it'll do at least 245 damage (150 HE damage, minimum 15% = 45 vs armor, plus 200) against armor. By comparison, a phase lance burst against an Onslaught (1925 armor, heh the sim one) does around 251 damage when the armor is full. So if the Breach is the first hit against said Onslaught, it's doing virtually the same damage as a phase lance shot. (However, as the armor gets lower, the extra 200 damage starts being worth less compared to a regular weapon.) So against a very "hard" target that has high armor, the Breach can be pretty good.

However, this scripted damage is only to armor, not to shield nor hull. So it's not very versatile; it's only useful during that window when the shields just went down but before the armor is stripped.

Also, this scripted effect does *not* get recorded by Combat Results (at least not the version that I have), so you can't really gauge how useful it was using the mod. You can't assume that every breach hit did this extra 200 damage since it might have hit shields or hull.

Pages: [1] 2 3 ... 8