Hard Light Productions Forums

Modding, Mission Design, and Coding => The FRED Workshop => Topic started by: Vrets on May 25, 2015, 07:02:25 pm

Title: Making (good) missions is an ordeal; guidance needed
Post by: Vrets on May 25, 2015, 07:02:25 pm
My general experience has been that FREDing (non-crap missions) is hard as hell and I don't know how you people do this without going insane sometime during the testing process.

The enemy/friendly AI doesn't behave exactly the same each time through, if for no other reason than my slightly-varying actions having a butterfly effect on the AI. At any rate, I can easily be fooled into making sweeping balance changes in an environment where the slightest tweak to enemy wing composition can make the difference between a laughably weak attack and an impenetrable, almighty assault that no pilot could possibly stop. As my current mission grew to a length of 15ish minutes, I had to start thinking about what happens if the player dies or fails the mission 14 minutes in. I have to worry about the player having time to read the com messages in an atmosphere of (possibly) unrelenting stress. I have to write events for every possible outcome when the number of outcomes increases exponentially as the mission becomes longer and more complex. And the worst yet: trying to fine-tune some sequence that occurs at the very end of the mission, and having to play through to the end every time just to see the results of a tiny tweak. Oh, my god.

I'm mostly here to vent, but I do need some guidance on a few points:


I have mad respect for anyone who can make complex, fun FS2 missions.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: Lepanto on May 25, 2015, 07:41:33 pm
Oh, I'm certain most veteran FREDders still share your pain. Testing and tweaking a mission fifty bazillion times to get it juuuust right? Been there before. Piece of advice? Ask for a beta-tester or two. (I can do it if nobody else wants to.) Beta-tester feedback is invaluable in pinning down mission balance; you're likely too close to the mission (and have played it too many times) to accurately balance it for your average HLP player.

To prevent annoying last-minute deaths, especially if you want the most tense action at the end of the mission, use this Ship Save/Load script to build checkpoints into your mission. (http://www.mediafire.com/download/c97w7an9v3iew7g/Ship_SaveLoad_script.zip) Instructions can be found in the readme; as there's a script bug in 3.7.2 that'll break this script, stick with a pre-3.7.2 build while FREDding, or wait till 3.7.3 comes out. For your beta-testing sanity, you may also want to break up long missions into separately-testable sections. (http://www.hard-light.net/wiki/index.php/Tutorial_-_Mission_Infrastructure)

If you're running out of wing names, your mission is probably too big for a newbie FREDder to tackle. If you really want to have twenty wings in a mission, you can follow BP's route of adding more constellation-based wing names (http://en.wikipedia.org/wiki/88_modern_constellations), or make up your own naming scheme. Whatever you do, don't use invisible characters when naming ships/wings, as the recent message box script (which some of your players may be using) will display them, which is bad for obvious reasons.

For your FREDding (and also beta-testing) sanity, you might want to cut the number of branching outcomes in your mission, sticking with a successful route and a few failure states (each of which is just one message and an immediate RTB.)

---

...Yes, all this is super-complicated and might raise more questions than it fixes. I'm afraid FREDding is a very complex art form (though you knew that already.) Good thing we have a helpful community!

Don't give up easily. The community needs more FREDding blood than anything else.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: Axem on May 25, 2015, 07:57:15 pm
Welcome to the island of... despair.

(Actually its more of a peninsula)

Well, don't look at it in too wide of a fashion. Break everything down into digestible chunks. Without knowing exactly what your mission is about, I can't quite comment on the butterfly effect of everything. Maybe find some way to consolidate some specific outcomes into slightly more generalized ones? The more straight forward you can be, the easier it is to test and harder it is for players to break (either accidentally or intentionally). Sometimes you can't go that way and its a real pain.

And believe me I have been there. That was one of the reasons I changed the way I go about building missions. See this wiki article (http://www.hard-light.net/wiki/index.php/Tutorial_-_Mission_Infrastructure) (gaha, ninja'd by Lepanto!) I wrote last year about structuring a mission for easy play testing and using a checkpoint system so your players will not be so disgruntled when they die at the 14 minute mark.

Now to your specific questions:
It depends on the mission. I try to aim for 6-10 minutes for my missions, I try to aim for the shorter end but it never ends up that way... People will no doubt point at Delenda Est for a "really long, but enjoyable mission with lots of action" so I'll just preempt them all! I know there's a ton of others out there with not as memorable names (sorry missions). I would say the secret to an enjoyable long mission is to pace the mission out with action and quiet moments. Let the player take a breath and lay on some exposition-laced messages there. Don't throw out plot important messages during action scenes! They are so easy to miss. Just try not to stress them out.

The primary killer for long missions is the player's hull integrity, which can't regenerate at all in retail. War in Heaven got around this by letting the support ship heal you up to a certain point. There's nothing worse than getting to 2% integrity at 4 minutes in and knowing this mission could take a long while... If you're going to keep to a no-heal route, try not to have too much attack or overwhelm the player. Things with AAA beams or savage fast fighters will create a lot of aggravation in a short span of time.

You can reuse fighter wing names with the # character. Anything after # will not show up and FRED has started to adjust the name so it works on wings too. So naming a wing Cancer#2 will make the fighters be named Cancer 1#2 and Cancer 2#2. That method should work with the message script (at least I haven't seen it not work, if someone's got weird names coming up let me know!)
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: NGTM-1R on May 25, 2015, 08:43:36 pm
Quote from: Vrets on May 25, 2015, 07:02:25 pm
My general experience has been that FREDing (non-crap missions) is hard as hell and I don't know how you people do this without going insane sometime during the testing process.

Why do you think I stopped? The bar rose to the point that there was no return on the effort as far as I was concerned.

Quote from: Vrets on May 25, 2015, 07:02:25 pm
How long is too long for a mission? Can someone give me an example of a really long, but enjoyable mission with lots of action?

Before save checkpoints and the like, there was INFR1's Nemesis, or Transcend's ending sequence, or some of the missions from Homesick.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: General Battuta on May 25, 2015, 11:21:50 pm
Making missions is easy and fun with the magic of linear loop structure! As a bonus, players find missions built like this intuitively satisfying!

Making missions is hard as hell and exhausting if you use sandbox or system structures, you poor bastard.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: General Battuta on May 25, 2015, 11:22:36 pm
Players perceive 'action' and 'tension' not in how long you sustain either but in the patterns by which you alter them.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: niffiwan on May 25, 2015, 11:58:54 pm
Quote from: General Battuta on May 25, 2015, 11:21:50 pm
Making missions is easy and fun with the magic of linear loop structure! As a bonus, players find missions built like this intuitively satisfying!

That sounds really cool, but what is the linear loop structure? :D Is that the method outlined in Axem's tutorial?
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: General Battuta on May 26, 2015, 10:14:25 am
I think it's pretty much what he describes.

If you build your mission in 2-3 minute action sequences, each one escalating from the one before, with a pause or lull in between to reset your emotional state, you can playtest hard on each loop. People talk about BP missions like there was some kind of secret sauce! But most of it was just playing them so much during development.

Loop structure allows players to orient themselves. Players like to have one thing to do at a time. Give them a key objective for each loop, and a complication or wrinkle. Play it a bunch. Is this loop fun? Would it be fun if you came in with low hull? If all your wingmen were dead?

This also creates natural bottlenecks between each loop where the mission space is fairly simple. You can place checkpoints here without much trouble. The player will have a chance to repair or rearm. You can select certain outcomes from the last loop (like a ship's hull integrity, or whether you killed a certain beam) and feed them forward into the next loop. This is surprisingly important.

Say you're building a mission in which two warships duel through a series of close passes. Each pass might be a loop. If, at the end of loop 2, the player has performed badly and the friendly warship is at low hull, you can say, oh, the mission will not be fun from now on, because the player probably cannot save the friendly warship. So we should add a small optional loop in which a transport comes in to fix our friend up.

An exemplar of a long-ass mission built on loops is Delenda Est! I probably don't need to outline the loops in that mission, the rising and falling tension in each loop, and the rising tension across all the loops.

The alternative is the systems/sandbox approach. This is used a lot in BP Act 3, in the Wings of Dawn mission set inside the space colony, in the retail mission 'Into the Lion's Den'. Here, the player is given a set of systems and a set of goals, and their task is to learn the systems well enough to complete the goals. This is a much harder type of mission to design and playtest, and a much harder type to make satisfying for the player.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: General Battuta on May 26, 2015, 10:16:03 am
Build two or three minutes of gameplay. Play it over and over until you're sure it's fun. (Don't tune the difficulty too high.) Repeat until you have three or four loops, each of which cares about the outcome of the last but can't be ruined by it.

Use points of psychological attachment and visual keys to draw the player to places they need to be. A new ship warping in always draws attention. An escort target will pull the player back towards it.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: karajorma on May 26, 2015, 10:25:41 am
Quote from: Lepanto on May 25, 2015, 07:41:33 pm
Beta-tester feedback is invaluable in pinning down mission balance; you're likely too close to the mission (and have played it too many times) to accurately balance it for your average HLP player.

I agree strongly with this. When building missions I tend to not particularly care about balance during the mission building phase. What is important is that the mission works, that there are no dead ends or infinite sections where it is impossible to finish the mission.

Once the mission is working more or less, I give it to a beta-tester/stick it in SVN for the rest of the team to play and then balance based on their suggestions.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: General Battuta on May 26, 2015, 11:37:16 am
I think you should balance as you go, and then also do what Kara says. You cannot design well without a feel for how your mission actually plays.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: procdrone on May 26, 2015, 12:25:22 pm
As far as balance goes, keep in mind that starting FREDers have tough time finding betatesters and alike. And betatesting missions only by yourself is the worst thing ever. You play the missions as you designed them. But PlayerX didn't knew that bomber wing will appear just on the rear - and disable the transport. Or, you play on medium difficulty, and find your mission well balanced, but PlayerX uses hard or impossible... and those fighters are just too damn strong...

There are people who can do it themselves, but im certainly not one of them. As everyone saw, Balance was the one of main issues with my Venice Mirror, even when I had 2 beta-testers at hand.

On the other note, if any of you looks for a BETAs, im more then happy to offer my help/opinion.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: Cyborg17 on May 28, 2015, 08:58:54 pm
This is probably one of the top ten most helpful threads on here.
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: Vrets on May 28, 2015, 11:27:00 pm
This is great! I could spend hours reading posts (of any length) on FRED theory.

You all demolished my questions and inspired me to make several re-iterations of my mission*.

Spoiler:
*meaning, quietly fixing design choices that now seem woefully idiotic to me after reading all of your responses.

Quote from: NGTM-1R on May 25, 2015, 08:43:36 pm
Why do you think I stopped? The bar rose to the point that there was no return on the effort as far as I was concerned.

My greatest strength is that I have a near-zero need for appreciation!
Title: Re: Making (good) missions is an ordeal; guidance needed
Post by: Lepanto on May 29, 2015, 10:25:32 pm
Great to know that you've moved through the dark pit of despair and are now on your journey to the peak of FREDding excellence!