Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[ArkAngel]

As saddening as it is to me that had to be fixed

[Psycho Society]

He added the Mule, like, last patch. :p

Plus the phase ships, those are fairly new.

[Sendrien]

Two items.

1. Does testing ship configurations in the simulator account for piloted-ship and fleet-wide perks from the skill tree?
2. When resolving battles automatically, does the algorithm somehow account for the increased power of the piloted ship as well as increased flux capacity and venting in the entire player fleet?

[Alex]

1. Does testing ship configurations in the simulator account for piloted-ship and fleet-wide perks from the skill tree?

Yes, but keep in mind that the "piloted ship" effects only apply to the ship if it's your flagship. So, you're flying the ship as it would be in battle, not necessarily as it would fly as if *you* were piloting it in battle.

2. When resolving battles automatically, does the algorithm somehow account for the increased power of the piloted ship as well as increased flux capacity and venting in the entire player fleet?

To a degree. I wouldn't count on it being particularly accurate, though. (For example, flux capacity and such - yes, the effects of missile specialization, or, say, the perks under Damage Control - no.)

Autoresolve is something I'm planning to take another look at in... well, not sure when, but at some point. For one, I don't think you'll be able to autoresolve battles that are anywhere near even - it just doesn't work out well.

[Sendrien]

Autoresolve is something I'm planning to take another look at in... well, not sure when, but at some point. For one, I don't think you'll be able to autoresolve battles that are anywhere near even - it just doesn't work out well.

An accurate and refined autoresolve algorithm is one of those small elements that generally go unnoticed, but when done properly, really sets the game apart in terms of quality and attention to detail. I'm really glad it's on your list, Alex! It shows the ambition you have for Starfarer to become a transformative AAA title.

[Alex]

Autoresolve is a tricky issue. If it were to correctly predict the results of a battle if the player played it out, then it would become preferable to playing the battle out manually (after all, it can't mess up - and if it could, it would be decried as "random"), and that's not a desired outcome. My current thinking is to simplify it a bit overall, in favor of more predictable (if perhaps less accurate) outcomes, and to either limit how much the player can actually use it, or to make it significantly less effective than playing out the battle - and to make that clear upfront.

[Gothars]

I'd vote for the second option. It doesn't hurt to pamper the player a bit with the thought that his great skills are needed to win even mediocre battles
But maybe that decision should wait until more stuff is in the campaign. It's not as if a battle could come between you and your real target at the moment, for there is no real target besides battle. Only when that changes we might be able to appreciate the annoyance autoresolve can spare us.

BTW, for some character dependent skills like Weapon Proprioception it makes perfect sense to not apply them to autoresolved battles.

[DNAz]

Autoresolve is a tricky issue. If it were to correctly predict the results of a battle if the player played it out, then it would become preferable to playing the battle out manually (after all, it can't mess up - and if it could, it would be decried as "random"), and that's not a desired outcome. My current thinking is to simplify it a bit overall, in favor of more predictable (if perhaps less accurate) outcomes, and to either limit how much the player can actually use it, or to make it significantly less effective than playing out the battle - and to make that clear upfront.

It doesn't completely solve it but how about showing the percentage to win?

[CrashToDesktop]

What's the battle currently based on?  Fleet Points?

If that were true, I'd change it by taking the amount of fleet points s ship is worth, and multiplying that by the percentage of Ordnance Points used.  Do the same thing for all ships and add up to total fleet point.  Might make it a bit more random.
(if that were to happen, make the vanilla variants have slightly less than 100% fleet points, for example, the Enforcers might only use 95 fleet points instead of the full 100 to make things different.  Or you can just multiply the total fleet point worth, mentioned above, by 90% to do something.

[Nooblies]

As far as auto-resolve goes, most of the frustration comes from when what you think would happen and what the auto-resolve thinks happened don't line up. As such I think that the best idea would be to make it very apparent exactly how likely you are to succeed or win, and at what cost. Of course, there could still be cases where the computer still screws you with a 70-80% chance to win (looking at you, shogun 2 auto-resolve), however if it was accurate enough to display realistic results of the auto-resolve, I feel that would be acceptable. Of course, this all depends on exactly how accurate a prediction the auto-resolve can be programmed to make.

[harrumph]

Personally, I'd like to see an auto-resolve where's almost no margin between certain victory and serous risks. If your fleet is clearly stronger, you should win a commanding victory, without losses, every time. If your fleet is not clearly stronger, auto-resolve should always involve losses; I don't like a huge range of save-scum-able possible outcomes.

Of course, that's coming from the perspective of seeing auto-resolve as just a time-saving device, which isn't necessarily everybody else's (or Alex's) vision of it.

[Sendrien]

Personally, I'd like to see an auto-resolve where's almost no margin between certain victory and serous risks. If your fleet is clearly stronger, you should win a commanding victory, without losses, every time. If your fleet is not clearly stronger, auto-resolve should always involve losses; I don't like a huge range of save-scum-able possible outcomes.

Of course, that's coming from the perspective of seeing auto-resolve as just a time-saving device, which isn't necessarily everybody else's (or Alex's) vision of it.

I think this is key. In order to avoid save-load abuse, you should not be able to "farm" different outcomes. I view it as a time-saving device. Sometimes, there are some battles that are just not worth fighting, because your forces are so superior. You should be able to rely on auto-resolve to achieve the same outcome you would, in these situations.

A good example would be:

Your fleet: A custom onslaught, a custom conquest, a custom eagle, two custom hyperions

Enemy fleet: Numerous pirate frigates and destroyers.

This should resolve to a flawless victory every time, assuming your custom builds are halfway decent.

[xenoargh]

If there's no risk, then why ever bother playing the actual combat out? 

Even those pirate ships may get lucky every once in a while.  They do in the real battles, after all; they should do so in auto-battles.

I think the save-scumming issue is pretty moot.  Save-scumming is an available choice for all combat outcomes in anything but Iron Mode. 

So long as there isn't any realistic possibility of excessively lucky good dice rolls allowing players to get unreasonable results from repeated runs through an auto-battle, it's a non-problem, especially if, on average, it's worse than just fighting it out. 

I really don't mind the concept of players doing a lot of auto-battles once they can build their own fleets and such, but at that point, trading losses with enemy fleets and returning to your base to repeat the process is a time tool (and may or may not save the player time) but it'll be largely at a net loss vs. manually fighting the battles. 

In short, auto-battle should always be seen as a time-saver that costs ships and crew, never as a good alternative to fighting a battle manually.  Otherwise once there's more to do, the game's merely about stacking your fleets to auto-kill X to get Y done, rather than mastering the core gameplay.  It's almost that bad now; take a custom fleet with high-end ships and max out travel speeds so that you can catch small fry and watch the profits roll in.

I don't mind the idea of having other fleets around to help your personal fleet win fights, etc., but at the end of the day, it needs to be better to go do the twitch dance or all the game guides will merely be fleet-stacking guides and how to get the MacGuffins, which would be a real shame.

[ThePinkPanzer]

It is literally impossible for a massive fleet / fleet of capital ships to lose against a pirate one unless they decide to help the pirates kill them.

[Gothars]

Wyvern opened a thread about this in Suggestions, I'd say lets move the discussion there: http://fractalsoftworks.com/forum/index.php?topic=5173.0