Proposal for 1.82: combat balance changes

[Ranthe]

I disagree that the core game should cater for all skill levels equally. It should be left to OXZ/OXPs to make the game harder for those who want it. Not for difficulty to be forced on all. Currently 1.80 tries to cater for both by making the safe systems for the "low skill" and the dangerous systems for the "high skill". This just has the effect of making the safe systems boring and the dangerous systems impassable to any but the elite...unless you run away, dump cargo and run away or run away while using your rear laser until there are only a couple left. I think it's depressing that 1.80 forces you into just one combat tactic against groups. Where is the thrill in just running away and sniping? No room for any other tactic. Elite was a dogfighting game, meant to be played mostly with the front laser. I don't want to see the rear laser removed from the game, I just don't think you should be forced to use it as your primary weapon. I think this situation has come about because of the proliferation of uber groups. You are one ship and you are expected to take on waves of 10+....forcing you to run away and use your rear laser. It's a horrible way of increasing difficulty for the elites. What is needed are more OXZ/OXPs like "skilled npcs". Not false solutions like bigger and bigger bad guy groups which just punishes everyone....unless you run away and rear laser snipe. And anyway, what pirates would be stupid enough to keep following you while you whittled them down to manageable. I don't think we should cut off half the ooniverse to lower skilled players. Everyone should, with some experience under their belt, get the chance to make that profitable trade run to a dangerous system....and survive. It should be tough, yes, but not terminal. I think sometimes elite players forget what it's like to chase that title. Wanting to get more kills to advance their ranking. Forcing them to stick to areas where there is hardly ever a pirate means that chase slows to a crawl, meaning that coveted title of elite is going to take them forever to reach.

<applauds> I agree completely with this.

[Redspear]

Beam laser: Make beam laser half-damage, half-heat. Having briefly tried this, it doesn't make that much difference to how much damage NPCs do - if they're hitting you, they're hitting you, and the reduced per-second damage is compensated for by them firing for longer. It gives you more chance to avoid dying in a head-on fight, but it's not a huge difference .

That, to me, makes the current beam laser configuration sound more like this:

Despite its hazardous nature it makes an ideal pirate vessel, primarily because of the speed, camouflage and high intensity Hassoni-Kruger Burst-lasers.

Any merit in making both 'versions' of the beam laser available?
Perhaps the high-intensity version has a shorter range or just costs more?... Or the other version costs less given that the "high-intensity" is what we already have...

[cim]

You are one ship and you are expected to take on waves of 10+

I just loaded BBC Elite in an emulator, and jumped to Riedquat. On approach to the planet, after one in-system jump:
- 1 Krait on scanner.
- By the time I've got it in sights and landed a couple of hits, a Cobra III has also shown up.
- I missile the Cobra III and go back to hunting the Krait, to see three further traces appear on the scanner.
- By the time I've got back to the Krait, a fourth has joined them (so six total so far)
...and at that point the combined laser fire from five ships is rather too much to deal with in the starting ship, but there would have been more along otherwise. The combat picture in the Elite manual has on the front screen all at once two Cobra IIIs, a Mamba, two Sidewinders, a Python, and a cloud of debris from a seventh ship. Probably that picture was staged in terms of being able to see them all at once, but it wasn't an unusual density for a fight.

Tried the same experiment in Oolite: one pack of eight or so pirates, which I led into a pack of bounty hunters and left to it but could have just run away from easily enough; one pack of four pirates which would have been a fairly straightforward fight in a decent ship, but in the stock Cobra III I just ran away because their fastest was a Gecko; Riedquat station. Not exactly a fun way to do it, but it shouldn't even have been possible to survive that...

Getting to the station in an Anarchy system in Elite basically involved fighting a continuous stream of pirates all the way to the planet (or, if you were paying too much attention to the tactical situation over the strategic one, all the way into deep space so the planet was a tiny dot in the distance... oops)

And Elite didn't have injectors, shield boosters, going off-lane to avoid them all, or simply being fast enough to run away from 90% of pirates either - with both grades of shield booster you can take basically twice as much damage as a stock ship, and injectors are a big (early game!) boost to ship power as well.

It should be left to OXZ/OXPs to make the game harder for those who want it.

This I disagree with. Certainly there'll always be people who want to make the game harder (or for that matter easier) than the core can practically deal with, but I do think the core should also cover as wide a skill range as possible effectively. You should be able to get to Elite without needing to install OXPs to keep at least some bits of the game challenging (and likewise without needing OXPs installed while Harmless->Competent to make the game survivable)

Budget based population:

I just realised I never answered this.

Basically at the moment if the game wants a pirate group, it adds N ships with that role, chosen entirely randomly. So you might get 6 Mambas, or you might get 6 Asps. 1.80 adjusts this a bit over 1.77 by having the light/medium/heavy subroles, but it's still very random in what you get - even more so if you add OXPs.

Budget-based population would say "Adding pirate group... budget 4000 credits" then say "Cobra III: 1500 credits.", "Mamba: 400 credits", "Mamba: 400 credits", "Cobra III gets Mostly Harmless pilot: 500 credits", "Cobra III upgraded to beam laser: 600 credits", "Mamba #1 gets ECM system: 600 credits" or it could instead say "Moray Starboat: 1200 credits.", "Moray upgraded to beam laser: 600 credits", "Moray gets ECM: 600 credits", "Moray gets aft beam laser: 1000 credits", "Moray gets Mostly Harmless pilot: 500 credits", "Moray gets extra missile: 30 credits" (ship+pilot prices completely made up, and probably buying with credits would just cause OXPers confusion in practice since almost nothing would cost what it would for the player...)

It's a very different way of doing system population, and one which wouldn't be compatible with existing ship OXPs, but would have future advantages (for instance, if there was a change like the 1.76-1.77 one which made the military laser stronger, just raise the NPC cost of it and carry on - ship OXPs still mostly work with the same balance as before)

Very definitely at the "try everything else first" end of the options list because it would be incredibly disruptive, difficult to implement, and time-consuming to set balanced prices for everything.

[Stormrider]

You should be able to get to Elite without needing to install OXPs to keep at least some bits of the game challenging (and likewise without needing OXPs installed while Harmless->Competent to make the game survivable)

Could AIs be made adjustable? If accuracy/behavior was split into dogfighting and sniping, as kanthoney suggests here, it seems like we could create sliders or at least easy, medium, and hard settings so players could adjust AI skill level to their own liking. I think this might even be possible to do now via oxp and oxp config, but I certainly don't have the coding skills to do it. I know you developers have worked very hard to create balance and I think you've really done a great job, but there is such a broad range of style/ability/hardware that it seems improbable that everyone can be satisfied. For instance I like the idea of laser dissipation but some obviously don't. Could this be made switchable or at least oxpable? I have no idea how hard this might be to do, if these ideas would require a lot of extra work for you devs to implement and support, then it may not be worth it.

I think slowing down npc reaction time is a good idea too, currently evasive action seems useless even with injectors.

[cim]

I think this might even be possible to do now via oxp and oxp config, but I certainly don't have the coding skills to do it.

There should be some adjustments in 1.82 to make OXP-modified difficulty settings a lot easier - and indeed OXP-modified styles (e.g. small numbers of skilled pilots vs large numbers of unskilled pilots) - though it'll be a little while before that's available.

For instance I like the idea of laser dissipation but some obviously don't. Could this be made switchable or at least oxpable?

OXPable laser settings have been on the requested list for a while, and if that was implemented the dissipation settings could be part of them.

The two tricky things with OXPable laser settings are:
- making it possible for OXP lasers to be fitted to ships where appropriate (in the same way that OXP missiles will sometimes show up now) without requiring the ship to be specifically defined to have it
- finding a sensible formula for heat-per-shot which doesn't make the core military laser far less good or allow silly exploits
(and a bit of internal recoding for the special properties of the Thargoid and Mining lasers, too)

[Scouseair]

I totally agree that a newbie in a stock cobra should not be able to survive an anarchy. But as I said, someone with a bit of experience, who presumably has some upgrades, should be able to try for it. I also agree that oolite is a much better game for all the additions that you guys have made. I just think it's one thing to face smaller group after smaller group, sometimes bumping into the next before you're finished with the first, as it was in Elite, 1.76 and 1.77. You had a chance to thin out the first group before the next arrived. If it got too hot, you retreated until shields and lasers recovered. It was a running war all the way to the station. It's totally another thing to meet a wall of ten guns to your one, especially with the AI's accuracy in 1.80 being much improved. It forces you into one tactic, run away and rear laser snipe. Think about it. If anarchies were that impassable no trader would ever go there! Either that or Her Imperial Majesty's space navy would rock in and thin out the pirates. Like what happened in the Caribbean in the pirate years, or is happening on a small scale off Somalia now.

Then how about an official OXZ add on. A Hard Mode that the devs can control. That way all levels can enjoy all the galaxy, like in previous incarnations, while the multiple elite veterans can have their own tougher world at 1.80 standards. I really don't like the current solution where half the galaxy is the baby pool and the other half is the Marianas Trench.

Now I understand your budget based population idea. I like it. It's like in wargaming where you decide on a points total for a battle and each side selects units up to that total, ensuring an even battle. You would have to be careful, though, as one dangerous player with a fully tricked out ship can be a different skill level from another with the same rank and ship. I like the principal, though.

Thanks Diz for the nightly build info. I will look into that.

[cim]

I really don't like the current solution where half the galaxy is the baby pool and the other half is the Marianas Trench.

No, that's certainly not how it should be ... a rather unforeseen consequence of the odds calculator, that one: there are in the mid-range systems plenty of pirate groups (and even a few in the safer systems) which the player with a bit of practice and equipment can defeat without too much trouble - twos, threes, fours - but most of them won't actually attack a Cobra III unless fired on first.

That's definitely going to get fixed.

[Scouseair]

That's good to know. Thanks Cim.

[Switeck]

there are in the mid-range systems plenty of pirate groups (and even a few in the safer systems) which the player with a bit of practice and equipment can defeat without too much trouble - twos, threes, fours - but most of them won't actually attack a Cobra III unless fired on first.

While most won't attack, many could demand cargo while charging towards you (to pick it up).

[cim]

And we're back...

What we could really do with is some volunteers, ideally ranked at Competent or below, ideally who at best aren't great shots with the aft laser yet (though we'll take Dangerous or above pilots who are finding the current game too difficult as well) who are willing to spend a lot of time - probably more time than the developers spend actually writing the code - testing out the balance, especially the early game balance.

So, when I asked this, there were quite a few volunteers. Now it's time to start the testing.

kanthoney's "reaction time" code is now in the nightly builds. I think it makes quite a difference to dogfighting and the dangerous approach to a large group, but I'm not the most reliable in these matters.

So, what I'd like you to do if you're interesting in testing (whether you volunteered earlier or not) is this:

1. If you've not yet completed the original combat survey, please do so for at least the first two scenarios.
2. Download and install the latest nightly build. We're going to be making various other changes to this build, some not related to combat, so I recommend that you take a backup of your current commander and keep it somewhere safe, or start an entirely new commander, so that you can go back to 1.80 afterwards. We make sure that 1.81/1.82 will load 1.80 save games correctly; we make no effort to ensure that 1.80 will load 1.81/1.82 save games correctly.
3. Repeat the first two scenarios of the combat survey in the nightly build (and the others, if you feel like it, but it's the first two that I'm particularly interested in) and submit your results with the same name you used for the original submission - make sure you change the new "Oolite Version" field on the form at the same time.
4. Go and play the nightly build for a bit, flying around doing whatever you normally do, see how the fighting goes, and post up your comments. If you have the "balance test" OXP I made earlier installed, please uninstall it for testing the nightly builds - the ideas in it may be coming back later, but we need to test one thing at a time.

There's no great hurry to this - I'd be hoping to get an idea of how much difference this has made over the next few weeks.

[Venator Dha]

Here's my observation from flying around with the 1.81-140904 build:

in 1.81 combat survey
Scenario 1 Sidewinder Harmless
1.81 3 kill & 3 kill at least 1 before I die
1.80 2 kill & 3 kill at least 1 before I die
Observation: survivability much higher. With 4 could survive indefinitely but not kill any. No instant kill by hostiles.

Scenario 2 Sidewinder Escort Harmless
1.81 2 kill & 3 kill at least 1 before I die
1.80 1 kill & 0 kill at least 1 before I die
Observation: survivability a little higher. Much more important to break off and dodge as soon as hit. Still instant kill when 4 or more hostiles

Overall there's a more obvious difference between the pulse and beam lasered hostiles in terms of damage dealt. Random manoeuvring gives a chance for shield regeneration.

In Ooniverse
Tested in both new started ships playing aggressively (i.e. not trying to run away but engaging head on) and my normal early ships (Cobra MkIII - Average (58), Python - Mostly Harmless (14), Mining Transporter - Poor (17)).
Observations:
The accuracy of hostile fire is poor from beyond about 7.5Km, Ok 7.5 to 5Km and good below 5Km.
The accuracy of hostile fire is differentiable between my ships now, the hostiles have difficulty hitting my Mining Transporter at anything beyond point blank range.
Being attacked by 4 hostiles did not guarantee death if escape was impossible but was still the most likely outcome. The type of ships and weapons used by the hostiles is more important now. E.g. 4 Sidewinders & Mambas with pulse lasers - I'd expect to win; 2 Sidewinders with pulse & 2 Kraits with beam - would be one hell of a fight; 4 of Kraits, Cobras, F-d-Ls with beam lasers - I'd expect to lose quickly. My weapon type (pulse or beam) played less of a difference due to the overheating of the beam.
Head on attacking 4 hostiles is still stupid but not instantly fatal. Although you still have to know what you are doing to stand a chance. Most of my aggressive new starters died when attacking 4 ships, but some survived when the hostiles were armed with pulse lasers, I would expect a new Jameson to not survive such encounters as they require making the right tactical choices.

Overall this has moved the balance to a better place. It is still difficult, especially at the start. The balance is probably close to perfect for an averagely experienced player but still too hard for someone who's only just learnt how not to crash into the station when docking.

A problem that I understand was not addressed by these changes but made testing harder is that I never encounter single or pairs of hostiles - it's not that they don't attack, they are never seen. So it was hard to judge how it felt encountering them.

[cim]

Here's my observation from flying around with the 1.81-140904 build:

Thanks. Has anyone else had time to test the aim modifications?

One interesting thing I've noticed is that while it makes the NPCs worse at shooting you, it also makes them better at dodging even without the evasion AIs enabled because they're less likely to be flying straight at you and more likely to be flying straight at the corkscrewing path where you were.

[Disembodied]

One interesting thing I've noticed is that while it makes the NPCs worse at shooting you, it also makes them better at dodging even without the evasion AIs enabled because they're less likely to be flying straight at you and more likely to be flying straight at the corkscrewing path where you were.

I haven't tried this, but that sounds like a great step forward, particularly for a game where the player is outnumbered. By making the NPCs harder to hit, the player's skill in hitting an enemy is rewarded; by making the NPCs worse shots, the player isn't so quickly overwhelmed by simple weight of numbers.

[Bogatyr]

Just popping in for a comment: I'm a lifetime (E)(Oo)lite player (who's been away a while) and had my ass handed to me like never before in 1.80 when trying to get a new Commander started out. Now I certainly don't have my dogfighting chops back to full yet, but those pirate packs just don't miss with their beams and I was reduced to rubble within seconds every time, until I decided that "run away and aft snipe" was the only way to go. While that is not bad in itself, if it's the ONLY option, it seems a bit boring. It seems like suicide to approach head-on even a single fighter with beam lasers. I used to really enjoy plowing through a pack of pirates head on, taking one out as I fly by them, before settling into a more defensive posture. That seems impossible in 1.80. Egads I can't bear to think of some of the mission OXPs where some of the ships have quad-(militaries?), they must be entirely unplayable....?

[Diziet Sma]

G'day, Bogatyr... and yes, that's exactly why we're having this discussion.. I'm just hoping that once things have been decided on, a way to implement them in 1.80 can be found, rather than having to wait for 1.82 to be released.. if we have to wait 'til then, I suspect we'll have lost a lot of potential players who decided to file Oolite in the 'too hard' basket.