Fractal Softworks Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Starsector 0.97a is out! (02/02/24); New blog post: Simulator Enhancements (03/13/24)

Pages: 1 2 [3] 4

Author Topic: Xyphos - The sim queen  (Read 5873 times)

Sutopia

  • Admiral
  • *****
  • Posts: 1005
    • View Profile
Re: Xyphos - The sim queen
« Reply #30 on: July 12, 2021, 09:19:16 PM »

Quote from: Hiruma
So I find something interesting
The Xypohs can deter even reckless AI from closing in - the non-Xyphos side would simply get ganged on regardless of AI assigned.
However, if you simply assign a full assault on the non-Xyphos side and just aggressive officer:

So do I get to claim mining pod is the best fighter?

Well, that tells us that reckless and full assault AI with mining pods is better than aggressive AI with Xyphos.

So I used your POG file, and stuck it into AI battles, and set the timeout to 10 seconds (at timeout all ships are set to reckless and orders are changed to full assault for both sides).

"I edited the last line of round_data.csv to be"
1,TRUE,13,1,FALSE,FALSE,TRUE,TRUE,FALSE,TRUE,FALSE,TRUE,10,FALSE,TRUE,TRUE,400000,200000,FALSE,1000,5,FALSE,15,5,4,3,2,1,2,2,1,1,0
[close]

Ran the fight three times.  One time the POG Mining Pods won with 6 losses, and the other two the Xyphos won with 4 losses and 7 losses. 

So what I think that shows is that the AI aggression level and orders is more important than the fitting differences between Mining Pods and Xyphos, and that when using the same level of AI aggression Mining Pods and Xyphos are pretty close to balanced.  The extra vents and caps the Mining Pods allow for is counter balanced by the shield piercing Ion beams shutting down weapons and engines before the shields actually flux out. 

The way it breaks one way or the other looks highly dependent on who gets the first kill and starts to snowball.
I agree.
The long bow fit with both sides reckless (no full assault) result in 50-50 and the outcome is solely determined by the initial engagement.
However, playing with long bow reckless and Xypohs aggressive results in long bow winning 3/3.

Reckless officer best officer?
Logged


Since all my mods have poor reputation, I deem my efforts unworthy thus no more updates will be made.

Vanshilar

  • Admiral
  • *****
  • Posts: 585
    • View Profile
Re: Xyphos - The sim queen
« Reply #31 on: July 12, 2021, 10:19:52 PM »

Ion pulser range is 500 and anyone who has used it knows it.

Whoops, had a brain fart and was thinking of cryoflamer when I typed it, but the point still stands, it's a short range weapon. Numbers change slightly but the main point is still the same: you criticized Xyphos's 1200-range ion beam (edit: I keep forgetting about Xyphos's Advanced Optics) as too short since it's not affected by ITU, yet you consider ion pulser's 500 (700 with ITU) good enough as a counterargument.

And I just gave you the fit: replace your HB with IP, replace your Xyphos with long bow and replace your sabot pod with harpoon pod (or reaper if you trust AI using it - I don’t).
IP is short range because you’re already fighting short range, so the “extra” range on xyphos is not needed. It’s not even a counter argument.
As OP stated, it’s not good for kiting, so your use of it in a CQC ship is somewhat legit. But again, in such scenarios a long bow can safely launch its payload without even leaving the mothership shield range, rendering long bow a superior choice.

Did you ever actually try this out in combat before proclaiming it to be better? I went and built a fleet of Furies (ion pulser, 2 harpoon pods, IR pulse lasers, longbow) and tried it out against my 2-Ordos test fleet, and the results were pretty much as expected: The longbows are fine early on but they die quickly as they're sent out, thus they're relegated to 30% irrelevancy for basically the rest of the fight. So much for them staying close by just because their mother ship has short-range weapons. The AI will spam harpoons against early weak targets, resulting in a lot of overkill and wasting them. Thus for the rest of the fight, the Furies are basically limping along with their 2 IR pulse lasers and their ion pulser. By the time peak time hit (510 seconds), the enemy still outnumber the fleet on DP and the Radiants haven't even shown up yet.

I've attached a screenshot of the moment peak time dropped to 0 seconds. You can see (lower right, bottom bar of radar) that the red bar (enemy) is still larger than the fleet (green). You can also see that all the Furies are pretty much at 30% for fighters, and no Radiants have appeared yet. Nor have the Furies clustered around the spawn point (using Xyphos + sabot pods + heavy blaster, they'll kill so fast that they're clustered around the spawn point by the time Brilliants start showing up en masse, basically killing ships as they spawn in). So it can't even handle 2 Ordos fleets, much less 3.

You can see this play out on a small scale by trying it in simulation under AI control against an Astral. Xyphos/sabot wins (eventually the Astral runs out of squalls, and fighters to an extent, Fury gradually approaching all the while, eventually Fury gets close enough to sabot the Astral directly and then it's dead), while Longbow/Harpoon basically runs down to 30% early on and eventually dies because it doesn't have enough firepower.

And btw an officer can has both tech skills, idk why you would mention it.

If you're doing it via Officer Training, then you're giving up at least one of Systems Expertise/Missile Spec (very important for flagship depending on which it is), Special Modifications (extra s-mod for all ships), and/or Reliability Engineering (+5% damage dealt, -5% damage taken, +5% speed, etc. for flagship). Perhaps you don't care enough about your flagship to optimize it when possible, but that's giving up a lot when the flagship is the single most important ship in the fleet.

If you're doing it with 5 officer skills, then you're giving up at least one of Target Analysis (extra damage), Shield Modulation (-20% damage taken by shields), Missile Specialization (+100% missile ammo capacity), or Reliability Engineering (+5% damage dealt, -5% damage taken, +5% speed, etc.). I don't see how any of these is less important than +15% range, which actually ends up being +11% overall on a cruiser with ITU.

I admit it is hard to intuit what people are doing with incomplete loadout descriptions.  I'm for example assuming ITU and shield conversion, but people could arguably skip those.  Also, linked or unlinked Harpoons + Sabots?  These all can have fairly substantial impacts. 

My Fury setup is:
Weapons: 1 Heavy Blaster, 2 Sabot Pods (linked), 2 IR Pulse Lasers, 1 Xyphos (if you have Cryoblaster, I highly recommend that over Heavy Blaster)
Built-in hullmods: Expanded Missile Racks, Hardened Shields, Integrated Targeting Unit
Other mods: Shield Conversion - Front, Solar Shielding, Converted Hangar
11 Capacitors, 10 Vents (although I still experiment with the exact mix from time to time)
Officer skills: Target Analysis (Elite), Shield Modulation, Missile Specialization, Energy Weapon Mastery, Reliability Engineering

My skills are (all elite where applicable):
Combat: 1L 2L 3L 4L 5R
Leadership: 1L 2L 3L
Technology: 1L 2R 3L 4L 5L
Industry: 1R 2R

I use 7 of them. Usually my fleet will also have a flagship and an "other" ship that's somewhat "bigger" or more important than the Furies. My flagship has typically been Aurora or Doom, whereas the other ship depends on whatever I feel like trying out or seeing how the AI handles it (Champion, Aurora, Doom, Onslaught, Legion, etc., I've tried out a lot of different things). A bunch of different loadouts work. But the main backbone of the fleet is the Furies. Or Falcon (P)'s using a similar setup (Xyphos + sabot pods), but Fury has better finisher in the heavy blaster (or cryoblaster when you get it). Messing around with doing it with a regular Falcon now, which won't get the fun sabot spam, but has ballistics to help make up for it. This fleet handles 2 Ordos fleets fine. If I use a Doom and switch out the heavy blasters on the Furies for cryoblasters, then it handles 3 Ordos fleets.

Converted hangar favors support fighters since they nearly never take losses.
However, they lose out to non-converted hangar fits most of the time thus voiding the comparison to long bows in the first place. Ah, I took the bait.

Eh, you're the one who brought up using longbows instead, so I guess you took your own bait. You still haven't come up with a loadout that's better than Xyphos.

Honestly I don’t know what we’re discussing about anymore.

The original statement was about Xyphos not good in any scenarios as there’s always something better.
So far there are two challengers: Xyphos Odyssey and C_Xyphos Fury, the later is dedicated for Remnant farming.
I have not commented on the former atm. Uh, not until this reply.
For the latter I think all I need to do is use a non-converted hangar fit in campaign that do better to beat the argument. A fury vs fury fit proves nothing as it’s not the same as the use case the challenger is arguing for. The reason I brought up long bow is due to Odyssey - I tried making a good Xyphos Odyssey fit but always find long bow with the exact same fit superior to Xyphos, thus made the bold assumption that converted hangar will be the same - it’s not the same due to the loss of speed and increased damage taken. I also did not account for shield sizes - Odyssey obviously has a much better coverage to protect long bows.

All I need to do now is prove that converted hangar cringe?

No. The original statement is you claiming that Xyphos is overpriced and only good in simulation. I claim it's the opposite, i.e. you need to take it out to full fleet-on-fleet combat to see it shine. You won't really see its worth in simulation nor against simple, easy fleets. My position is based on testing out a bunch of different fleet setups (ships, weapons, etc.) against full [REDACTED] fleets; in other words, practical, full fleet combat use. I went through the logic behind the use of the Xyphos. I explained the logic for testing against [REDACTED], that if you have a fleet that can handle [REDACTED] (in this case, 2 full Ordos fleets at once), then it can handle anything easier than that -- which means pretty much the entire game, barring possibly a couple of specific exceptions. (I've never tried it against that phase fleet, because I never encountered it while doing the quest, and I've finished the quest now so I can't go back and test it out.) So it's not "dedicated for Remnant farming", it's "can handle everything up to and including Remnant farming", a very big difference. Nobody is interested in finding out whether A or B is better for killing a pirate Lasher, because there are a million ways to do it. I explained that Xyphos is the key to making this loadout strategy work, and why. That's the relevance of using Xyphos. And this is based on "real world" experience, so to speak.

If you're able to find a fleet loadout that's better, I'm all ears. I'm always looking for a better fleet setup. Converted hangar or not doesn't matter. Fury or not doesn't matter. Fury as a platform is just for convenience, multiple ships can work.

Edit: forgot the screenshot. Edit2: Keeping forgetting that Xyphos ion beam is actually 1200 range due to Advanced Optics.

[attachment deleted by admin]
« Last Edit: July 13, 2021, 02:12:09 AM by Vanshilar »
Logged

Lucky33

  • Admiral
  • *****
  • Posts: 894
    • View Profile
Re: Xyphos - The sim queen
« Reply #32 on: July 13, 2021, 12:08:41 AM »

Have you tried to implement your Xyphos setup on the Eagles?
Logged

Yunru

  • Admiral
  • *****
  • Posts: 1560
    • View Profile
Re: Xyphos - The sim queen
« Reply #33 on: July 13, 2021, 12:38:00 AM »

If you're doing it via Officer Training, then you're giving up at least one of Systems Expertise/Missile Spec (very important for flagship depending on which it is)
Sorry but no, the officer skills are the single least important for a player to take when dealing with AI vs. AI.

Outside of player control, you just use an officer.

Vanshilar

  • Admiral
  • *****
  • Posts: 585
    • View Profile
Re: Xyphos - The sim queen
« Reply #34 on: July 13, 2021, 01:22:24 AM »

Have you tried to implement your Xyphos setup on the Eagles?

Nope, haven't tried yet, but I suspect Eagles will end up being a bit too slow to chase down frigates, leading to them being spread out too much. I may try it out after the next patch, while re-evaluating Fury, Falcon (P), Falcon, and other ships for combat effectiveness.

Sorry but no, the officer skills are the single least important for a player to take when dealing with AI vs. AI.

Not sure what you mean, this is fleet on fleet combat in campaign, not sim, yes the player is controlling one of the ships. That's why I took the whole combat line; the flagship usually does around 30-40% of the damage of the whole fleet, far more than any other ship, and is pivotal to the outcome of the battle.

When I'm using an Aurora I get Missile Specialization for more antimatter missiles, when I'm using a Doom I get Systems Expertise for the buff to mines. Either of those is more important than getting my officers another skill.
Logged

Hiruma Kai

  • Admiral
  • *****
  • Posts: 878
    • View Profile
Re: Xyphos - The sim queen
« Reply #35 on: July 13, 2021, 10:05:36 AM »

I agree.
The long bow fit with both sides reckless (no full assault) result in 50-50 and the outcome is solely determined by the initial engagement.
However, playing with long bow reckless and Xypohs aggressive results in long bow winning 3/3.

Reckless officer best officer?

It is a good point.  Going Reckless plus the Full assault command helps the Longbow Furies capitalize on that first wave.

Reckless officers are often the best choice for winning a balanced AI vs AI match.  Also if you want ships to better follow your orders to engage despite heavy opposition.  Less so if you it's a 2 to 1 scenario and the objective is to minimize casualties for the side with twice the firepower.

All told, I'm definitely seeing Xyphos provide value in these engagements, at least on par with their OP cost compared to other fighters, which is what you'd want in a balanced game I think.  Mining pods work by being distractions and absorbing firepower.  Longbows provide an initial burst of more kinetic damage.  Xyphos knock out weapons and engines over the long haul while also shooting down the occasional missile.

I happen to like these approximate mirror matches with the item under test being the only real change (i.e. Xyphos vs Longbow vs Mining Pod), since the signal to noise tends to be higher in the results.  Essentially which won is your metric.   Otherwise you have to start keeping track of things like kill speed, hull condition, CR remaining, crew losses, and so forth to determine which is "better".  It also adds more variables.  Are Xyphos weak or say, Safety Overrides or Expanded Missile racks too strong?  That was way more effort than I wanted to put in.

If you're doing it via Officer Training, then you're giving up at least one of Systems Expertise/Missile Spec (very important for flagship depending on which it is)
Sorry but no, the officer skills are the single least important for a player to take when dealing with AI vs. AI.

Outside of player control, you just use an officer.

The one who brought up AI vs AI was me.  Mostly because accounting for player skill is nearly impossible in a repeatable, testable fashion across different players.

I saw a testable hypothesis, i.e. Longbows, sabot, harpoon, Ion pulsar is better than Xyphos, Heavy blaster, sabot x2, and decided to run some repeatable tests that others could also do if they wanted.  I happen to like data when discussing a balance point (i.e. Xyphos are the "most useless").

Admittedly, the data at best could show Xyphos are not useless.  Even if they didn't perform as well in a near mirror match, that's just saying something about two particular setups.  Essentially saying there's a better loadout in the campaign is kind of meaningless if you want to say a particular fighter wing is "most useless".  One might imagine the parameter space of fleet setups in the game, running through all permutations of ship choices, loadouts, skills, and tactics employed by the player.  If one had a ranking methodology (although with a game as complex as Starsector it's hard to imagine what a clean ranking methodology might be), one might imagine one particular setup being the best in that parameter space, and everything else performing worse.  The existence of one hypothetical best setup doesn't imply that everything less effective is the most useless.  It doesn't even imply that they are useless in a more general sense.

I suppose if you compared a setup with Xyphos to without Xyphos (and leaving those OP unused) could prove that they're worse than nothing, but I'm pretty sure the Xyphos win that fight every time (perhaps with losses due to localized missile saturation in these particular setups) but eventually the Ion shield piercing would matter.  Similarly if Vanshilar took the Xyphos off the Furies (and no other changes) and found the same level of success against Ordos, you might be able to say the fleet is already strong enough without the fighters, but given the Xyphos are the major source of ion damage for that setup, I'm fairly certain it would in fact perform significantly worse.
Logged

rabbistern

  • Ensign
  • *
  • Posts: 32
    • View Profile
Re: Xyphos - The sim queen
« Reply #36 on: July 14, 2021, 05:48:42 PM »

im seeing lots of back and forth arguing here, and i dont quite understand as good as all of the complaints. there isnt just a single ship with fighter bays in the game you know, and weapons are of different value depending on what ship theyre on. simply put, i dont see how anything but longbows is even in the competition for an energy-based arsenal, as in most hightech ships like the odyssey. if the ship doesnt allow the mounting of energy weapons, i will happily prefer ballistic kinetics and take xyphoses on my legion for suppression, thats all i can say
Logged

SafariJohn

  • Admiral
  • *****
  • Posts: 3010
    • View Profile
Re: Xyphos - The sim queen
« Reply #37 on: July 14, 2021, 06:56:59 PM »

I would say the original assertion, that Xyphos isn't effective in real battles, has been thoroughly disproven. The Xyphos vs Longbow vs whatever debate seems to show Xyphos, et al are close in effectiveness, generally speaking.
Logged

Vanshilar

  • Admiral
  • *****
  • Posts: 585
    • View Profile
Re: Xyphos - The sim queen
« Reply #38 on: July 15, 2021, 11:23:47 PM »

It is not about Xyphos really. Fury has one of the best flux density per DP. It is like small Radiant. Sabots provide flux free anti shield damage, boosting ship's efficiency even further. It also has anti-weapon capabilities since it has EMP damage component.

Partially, but the strongest thing about the Fury in this case is that it can get 360 shields, thus it dies less from AI stupidity (less error-prone). Falcon (P) with Xyphos/sabot can also work as well (in fact, offensively Falcon (P) works better, because it can use up to 4 sabot pods), but the problem is that the AI tends to leave its engines exposed and die from getting hit in the rear, which would have been completely avoidable with more competent AI. (I'm not talking about from Salamanders etc., since each Fury has 2 Xyphos protecting it; I'm talking about it'll try to strafe a target, thus maneuvering its engines toward another enemy ship, and then die to that other enemy ship.) See attached screenshot of this in action. Thus I spend a lot of time trying to rescue whichever Falcon (P) is sticking its engines toward the enemy ships, and at some point I'm not able to get to them all. I can save them all while piloting Doom (since I can phase and avoid ships in the way, plus it's way faster and can use mines as a distraction), but it's much harder with Aurora. Fury with front shield conversion means its engines stay protected, so I don't need to do as much babysitting. Its failure mode instead is that it tends to plasma burn straight into hulks etc., which flames it out, usually while spinning helplessly straight into the enemy fleet, leading to its death. I don't think the AI considers whether or not hulks are in the way for plasma burns, at least based on how often I see this happening.

Xyphos does not provide much to the mix. Not for their cost.

Again, if anybody can provide a better build for handling 2 full Ordos fleets at once, I'm all ears.

I suppose if you compared a setup with Xyphos to without Xyphos (and leaving those OP unused) could prove that they're worse than nothing, but I'm pretty sure the Xyphos win that fight every time (perhaps with losses due to localized missile saturation in these particular setups) but eventually the Ion shield piercing would matter.  Similarly if Vanshilar took the Xyphos off the Furies (and no other changes) and found the same level of success against Ordos, you might be able to say the fleet is already strong enough without the fighters, but given the Xyphos are the major source of ion damage for that setup, I'm fairly certain it would in fact perform significantly worse.

Yes I tested various builds with Xyphos for a while before using it, so that comparison has already been made (namely, "before I tried out Xyphos"). It was basically the key to where my fleet could go from handling 2 Ordos fleets to handling 3 Ordos fleets, by switching to Xyphos from non-Xyphos Furies.

Although different players have different amounts of skill, human players are not capricious in battle; they are generally following a set of rules as well (in terms of evaluating which target to go after, when to engage or disengage, etc.). In many ways the human player is actually easier to account for than the AI, because how the AI operates is pretty "opaque" for most players. Any time you give a ship a command, it is essentially because the ship's AI is not behaving the way you'd like it to; any time the ship does something else despite the player giving it a command (such as when you tell it to capture an objective but it runs off chasing a target, or you tell it to gather somewhere and it runs off chasing a target, etc.), it is essentially because the ship's AI is disregarding the player's intent for its own reasons, or the player doesn't understand how the commands work. In either case, the AI is opaque -- the player can't directly determine why it behaves the way it does. The player, by contrast, knows exactly why he is taking an action, regardless of whether or not the outcome is the one he wanted. Thus it's easier to account for and analyze -- and improve upon.

Success methodology is pretty straightforward. Whether or not the player fleet wins the battle, or, since each battle has randomized elements, the probability of the player fleet winning the battle. Since I'm actually playing the battle, I can observe AI behavior, but that's basically for failure mode analysis, i.e. why the AI messed up.

-----

As an aside, if Xyphos is the key to handling multiple Ordos fleets simultaneously, does this mean that Xyphos might actually be overpowered? I don't think so. The 38 OP used to equip the Xyphos is a big sacrifice (and as it should be for a warship to get fighter capability). Relying on sabot pods, and having them last the entire battle, means getting officers with Missile Specialization as well as taking expanded missile racks as a hullmod. So the Furies have very little OP remaining for caps/vents, and is somewhat overfluxed. It doesn't have any PD of its own. Basically, the fit is extremely tight. It probably wouldn't have the OP had I gone the Radiant route instead. So like all good game design, it forces the player to make some hard choices -- the player is forced to give up some good things in order to use this build.

Nor am I claiming that this is somehow the "best" build. Just that I haven't found any that are better. Again, I'm all ears if somebody comes up with a better build! There are plenty of possible builds that I haven't tested. For example, back when I tested Dominators, regular Falcons, etc., was before I started using Xyphos. I also haven't tested Apogee, which like the Fury has 360 shields. Onslaught (XIV) with Xyphos works very well as a wrecking ball, and provides good long-range artillery as well as close-range fighting and tanking with its mjolnirs and sabot pods, although I think people usually recommend other missiles for its medium missile slots. When fighting with the Furies, the Onslaught (XIV) needs unstable injector just to keep up, plus the Furies usually end up in the front or on either side so it's okay to leave the Onslaught's rear exposed (and the Xyphos will take care of any stray missiles), and forgo any armor/hull hullmods -- just rely on shields basically. So there are plenty of possibilities out there.

Edit: Forgot screenshot again. Here you can see Falcon 7 strafing to the left to surround a Fulgent that's about to die, completely oblivious to the fact that it's getting pummeled in its engines by a 5-autopulse Radiant and about to die. This happens quite often even when the Falcon (P) has extended shields, even though it doesn't have it here.

[attachment deleted by admin]
« Last Edit: July 15, 2021, 11:32:22 PM by Vanshilar »
Logged

Lucky33

  • Admiral
  • *****
  • Posts: 894
    • View Profile
Re: Xyphos - The sim queen
« Reply #39 on: July 16, 2021, 05:04:28 AM »

It is not about Xyphos really. Fury has one of the best flux density per DP. It is like small Radiant. Sabots provide flux free anti shield damage, boosting ship's efficiency even further. It also has anti-weapon capabilities since it has EMP damage component.

Partially, but the strongest thing about the Fury in this case is that it can get 360 shields, thus it dies less from AI stupidity (less error-prone).

The main problem of the Remnant fights is the sheer amount of firepower and mobility in the Radiant. It is capable of quickly overfluxing any kind of cruiser that happened to attract its attention. Typically it will be the sole overextended ship. In the case of ships without omnidirectional mobility system it all ends up as the difference between Radiant builds. If it is solid one the target ship will die and if it is a junk build the unfortunate target will back off. Most likely. The sad reality of ships with 50+K effective shielding (360 degree!) getting destroyed in a matter of seconds for the sole reason of moving half the hull too much forward led me to abolish any attempts of classic line tactics against Remnants. If you want stability you need Monitors as a skirmish force. That's all.

Xyphos does not provide much to the mix. Not for their cost.

Again, if anybody can provide a better build for handling 2 full Ordos fleets at once, I'm all ears.

My point was that it will not work on anything without over the top stats. Hence it is the flux/sabot density that gets the job done and not the Xyphos.
Logged

Chinno

  • Ensign
  • *
  • Posts: 13
    • View Profile
Re: Xyphos - The sim queen
« Reply #40 on: July 16, 2021, 08:03:35 PM »

It is not about Xyphos really. Fury has one of the best flux density per DP. It is like small Radiant. Sabots provide flux free anti shield damage, boosting ship's efficiency even further. It also has anti-weapon capabilities since it has EMP damage component.

Partially, but the strongest thing about the Fury in this case is that it can get 360 shields, thus it dies less from AI stupidity (less error-prone). Falcon (P) with Xyphos/sabot can also work as well (in fact, offensively Falcon (P) works better, because it can use up to 4 sabot pods), but the problem is that the AI tends to leave its engines exposed and die from getting hit in the rear, which would have been completely avoidable with more competent AI. (I'm not talking about from Salamanders etc., since each Fury has 2 Xyphos protecting it; I'm talking about it'll try to strafe a target, thus maneuvering its engines toward another enemy ship, and then die to that other enemy ship.) See attached screenshot of this in action. Thus I spend a lot of time trying to rescue whichever Falcon (P) is sticking its engines toward the enemy ships, and at some point I'm not able to get to them all. I can save them all while piloting Doom (since I can phase and avoid ships in the way, plus it's way faster and can use mines as a distraction), but it's much harder with Aurora. Fury with front shield conversion means its engines stay protected, so I don't need to do as much babysitting. Its failure mode instead is that it tends to plasma burn straight into hulks etc., which flames it out, usually while spinning helplessly straight into the enemy fleet, leading to its death. I don't think the AI considers whether or not hulks are in the way for plasma burns, at least based on how often I see this happening.

Xyphos does not provide much to the mix. Not for their cost.

Again, if anybody can provide a better build for handling 2 full Ordos fleets at once, I'm all ears.

I suppose if you compared a setup with Xyphos to without Xyphos (and leaving those OP unused) could prove that they're worse than nothing, but I'm pretty sure the Xyphos win that fight every time (perhaps with losses due to localized missile saturation in these particular setups) but eventually the Ion shield piercing would matter.  Similarly if Vanshilar took the Xyphos off the Furies (and no other changes) and found the same level of success against Ordos, you might be able to say the fleet is already strong enough without the fighters, but given the Xyphos are the major source of ion damage for that setup, I'm fairly certain it would in fact perform significantly worse.

Yes I tested various builds with Xyphos for a while before using it, so that comparison has already been made (namely, "before I tried out Xyphos"). It was basically the key to where my fleet could go from handling 2 Ordos fleets to handling 3 Ordos fleets, by switching to Xyphos from non-Xyphos Furies.

Although different players have different amounts of skill, human players are not capricious in battle; they are generally following a set of rules as well (in terms of evaluating which target to go after, when to engage or disengage, etc.). In many ways the human player is actually easier to account for than the AI, because how the AI operates is pretty "opaque" for most players. Any time you give a ship a command, it is essentially because the ship's AI is not behaving the way you'd like it to; any time the ship does something else despite the player giving it a command (such as when you tell it to capture an objective but it runs off chasing a target, or you tell it to gather somewhere and it runs off chasing a target, etc.), it is essentially because the ship's AI is disregarding the player's intent for its own reasons, or the player doesn't understand how the commands work. In either case, the AI is opaque -- the player can't directly determine why it behaves the way it does. The player, by contrast, knows exactly why he is taking an action, regardless of whether or not the outcome is the one he wanted. Thus it's easier to account for and analyze -- and improve upon.

Success methodology is pretty straightforward. Whether or not the player fleet wins the battle, or, since each battle has randomized elements, the probability of the player fleet winning the battle. Since I'm actually playing the battle, I can observe AI behavior, but that's basically for failure mode analysis, i.e. why the AI messed up.

-----

As an aside, if Xyphos is the key to handling multiple Ordos fleets simultaneously, does this mean that Xyphos might actually be overpowered? I don't think so. The 38 OP used to equip the Xyphos is a big sacrifice (and as it should be for a warship to get fighter capability). Relying on sabot pods, and having them last the entire battle, means getting officers with Missile Specialization as well as taking expanded missile racks as a hullmod. So the Furies have very little OP remaining for caps/vents, and is somewhat overfluxed. It doesn't have any PD of its own. Basically, the fit is extremely tight. It probably wouldn't have the OP had I gone the Radiant route instead. So like all good game design, it forces the player to make some hard choices -- the player is forced to give up some good things in order to use this build.

Nor am I claiming that this is somehow the "best" build. Just that I haven't found any that are better. Again, I'm all ears if somebody comes up with a better build! There are plenty of possible builds that I haven't tested. For example, back when I tested Dominators, regular Falcons, etc., was before I started using Xyphos. I also haven't tested Apogee, which like the Fury has 360 shields. Onslaught (XIV) with Xyphos works very well as a wrecking ball, and provides good long-range artillery as well as close-range fighting and tanking with its mjolnirs and sabot pods, although I think people usually recommend other missiles for its medium missile slots. When fighting with the Furies, the Onslaught (XIV) needs unstable injector just to keep up, plus the Furies usually end up in the front or on either side so it's okay to leave the Onslaught's rear exposed (and the Xyphos will take care of any stray missiles), and forgo any armor/hull hullmods -- just rely on shields basically. So there are plenty of possibilities out there.

Edit: Forgot screenshot again. Here you can see Falcon 7 strafing to the left to surround a Fulgent that's about to die, completely oblivious to the fact that it's getting pummeled in its engines by a 5-autopulse Radiant and about to die. This happens quite often even when the Falcon (P) has extended shields, even though it doesn't have it here.


Bro, I'm on your side.
beat 3 Ordo fleets is a great achievement in 095 Vanilla.
Fury is a bit OP (with its 15DP),Xyphos is pretty interestring and strong, but 38OP is so much,thus i think it's a good balance.
Logged

Vanshilar

  • Admiral
  • *****
  • Posts: 585
    • View Profile
Re: Xyphos - The sim queen
« Reply #41 on: July 16, 2021, 09:28:46 PM »

The main problem of the Remnant fights is the sheer amount of firepower and mobility in the Radiant. It is capable of quickly overfluxing any kind of cruiser that happened to attract its attention.

Yes, that's what the Xyphos is for -- the ion beams start disabling their weapons once they have hard flux (courtesy of the sabots and the Fury's other weapons), so that they don't get a chance to overwhelm a ship. That's their use case. They disable enemy weapons so that you get a chance to drive up their flux before they drive up yours. So if you're having an issue with Radiants overfluxing your ships, that's where Xyphos comes into play.

If you want stability you need Monitors as a skirmish force. That's all.

True, I haven't really tried Monitors. Mostly because they don't provide much offensive capability, whereas I tend to build a fleet in terms of kill rate per DP. But they might help relieve the pressure on the ships that are actually doing the killing, to tank for them.

My point was that it will not work on anything without over the top stats. Hence it is the flux/sabot density that gets the job done and not the Xyphos.

They both need each other. Xyphos doesn't do much damage on its own. You need other weapons to do the killing. But without Xyphos, the enemy fleet is also busy driving up your flux, so you're not going to outflux them if they're a more dangerous fleet. Xyphos is what enables you to win the flux war. This is very easy to see if you fit the Furies with the same build but without the Xyphos (put the points into cap/vent or whatever else you prefer), and then send it in against the same [REDACTED] fleet(s). They won't be able to kill the ships quickly enough because their flux gets driven up too quickly, and eventually they founder.
Logged

Hiruma Kai

  • Admiral
  • *****
  • Posts: 878
    • View Profile
Re: Xyphos - The sim queen
« Reply #42 on: July 16, 2021, 09:54:39 PM »

Anyone off hand know what the shield piercing chance for an EMP arc from an Ion beam is, and how much EMP an arc is?  Is it straight up hard flux percent at a fixed interval or the like?  I'm curious what the EMP per second is when a target ship is at, say, 25%, 50% and 75% hard flux per Ion beam.
Logged

Vanshilar

  • Admiral
  • *****
  • Posts: 585
    • View Profile
Re: Xyphos - The sim queen
« Reply #43 on: July 16, 2021, 10:31:29 PM »

Anyone off hand know what the shield piercing chance for an EMP arc from an Ion beam is, and how much EMP an arc is?  Is it straight up hard flux percent at a fixed interval or the like?  I'm curious what the EMP per second is when a target ship is at, say, 25%, 50% and 75% hard flux per Ion beam.

The data can be found by looking at the beam's info in \starsector-core\data\weapons\ionbeam.wpn file. In it, one of the lines is:

   "beamEffect":"com.fs.starfarer.api.impl.combat.IonBeamEffect",

Going into \starsector-core\starfarer.api.zip\com\fs\starfarer\api\impl\combat\IonBeamEffect.java, the chance is (I assume) the hard flux % - 10%. This can be modified by skills and stuff. It also gives the firing interval as 0.25 to 1.75, which I assume to be in seconds but not sure. The damage is the same as the ion beam, while the EMP damage is based on the flux component of the beam, whatever that means (not sure what is "beam.getDamage().getFluxComponent()").

For tachyon lance, it's similar; the firing interval is 0.2 to 0.3 (presumably seconds), pierce chance is hard flux % - 10%, but the damage is 25% of the beam damage and the EMP damage is 50% of the beam's flux component, whatever that is.
Logged

Lucky33

  • Admiral
  • *****
  • Posts: 894
    • View Profile
Re: Xyphos - The sim queen
« Reply #44 on: July 16, 2021, 10:57:46 PM »

The main problem of the Remnant fights is the sheer amount of firepower and mobility in the Radiant. It is capable of quickly overfluxing any kind of cruiser that happened to attract its attention.

Yes, that's what the Xyphos is for -- the ion beams start disabling their weapons once they have hard flux (courtesy of the sabots and the Fury's other weapons), so that they don't get a chance to overwhelm a ship. That's their use case. They disable enemy weapons so that you get a chance to drive up their flux before they drive up yours. So if you're having an issue with Radiants overfluxing your ships, that's where Xyphos comes into play.

It is "single overextended ship" against "Radiant plus some support". Nobody will even notice the very Xyphos existence. Overextended ship will get overfluxed and destroyed in a matter of seconds. Only fortress shield make those seconds into something protracted enough to react to.

If you want stability you need Monitors as a skirmish force. That's all.

True, I haven't really tried Monitors. Mostly because they don't provide much offensive capability, whereas I tend to build a fleet in terms of kill rate per DP. But they might help relieve the pressure on the ships that are actually doing the killing, to tank for them.

Yes. You order them to attack Radiants and this allows you to do things with noticeably less risks of losing your ships to concentrated fire. And it is better to lose Monitor than cruiser anyway.

My point was that it will not work on anything without over the top stats. Hence it is the flux/sabot density that gets the job done and not the Xyphos.

They both need each other. Xyphos doesn't do much damage on its own. You need other weapons to do the killing. But without Xyphos, the enemy fleet is also busy driving up your flux, so you're not going to outflux them if they're a more dangerous fleet. Xyphos is what enables you to win the flux war. This is very easy to see if you fit the Furies with the same build but without the Xyphos (put the points into cap/vent or whatever else you prefer), and then send it in against the same [REDACTED] fleet(s). They won't be able to kill the ships quickly enough because their flux gets driven up too quickly, and eventually they founder.

You need to rise the flux level for the Xyphos to be of any use in the first place. And since Remnants do use Ion beams of their own so you need to raise their flux level faster than they will do the same to you. Hence winning the flux war comes first.

Priorities:

1. Fast ship with the above average flux stats. So that it will not get destroyed by the destroyer-cruiser force even before Radiant will come into play.

2. Lotsa sabots. For a limited battle it is better to have small mounts (more burst), for a prolonged one - mediums (more ammo). Sabot's role is obvious - to raise the flux level of the enemy without sacrificing yours. The more sabots you have the more efficient your fleet is.

3. High dps armor/hull destruction. Overloading is not consistent that's why it is better to use general damage options compared to a strike ones. Honestly, even for my playership I preferred Cryoblasters to Typhoons.

4. Support. Here goes the disables.
Logged
Pages: 1 2 [3] 4