A Defense of Time-Based Mechanics

[Parkaboy]

There are several tasks in KSP that a first-time player would expect to take time, but instead are completed instantly. This is a deliberate game design decision from the developers. However, I believe that the rationale behind this decision should be discussed, if only due to the fact that many mods have adressed time-based mechanics, and in doing so were able to improve the experience of the game.

What kind of time-based mechanics are we talking about? Which game tasks could improve game experience by taking time to complete? Some examples are:

• Ship Building: putting together a ship takes only the time you spend in the editor, and further copies of that vessel can be created as fast as clicking a button. Shouldn't we have to wait, at least in Career Mode, for our engineers and technicians to build the craft? The Mod Kerbal Construction Time adresses this (and does many other things). The fact that there's a mod indicates a demand for it, although it would make sense to have a feature such as this something that could be toggled in the difficulties/settings panel.
  
• Experiments: all the existing experiments are instantaneous. It makes sense for stuff like sample collecting, but it does leave our kerbals without much to do while waiting for the next transfer window. Some mods have experiments that are performed over time or get science points in a steady trickle, such as ScanSat or Station Science. The fact that you can leave an experiment running in a vessel and go fly another one helps creating the illusion that the entire space program is alive and dynamic, and that the kerbals in the currently active vessel aren't the only guys working in a given time.
  
• Building Upgrades: re-constructing or upgrading a building in the space center happens in the blink of an eye. This is a bit unsettling to a player familiar with RTS mechanics, or anyone who played a SimCity game. In fact, we expect building something as large as the VAB to take time more than we expect travel between planets to take time, since popular fiction has mostly disregarded the proper size of space. In this sense, it's a big break in our sense of immersion.
  
• Researching: unlocking a new node in the tech tree also happens instantly. This means we can advance our technology in huge leaps: getting the payoff from a big science collection mission, one can unlock whole tiers of the tech tree in one sweep. Wouldn't it make more sense to have this process be slower? As far as I know, no mod has adressed this, so maybe people like the system as it is.
  
• Resource Gathering: currently we don't actually have a resource collection system, or any form of deep space refueling, but we know it is in the works. Some of the most popular mods that are resource collection to the game, such as Kethane and Karbonite, make the gathering of resources something that takes time. Given how the devs have avoided time-based mechanics, will they go the same route?
  

Before going any further, I think we should adress the motives behind the devs' decision not to have any time-based mechanics in the stock game. It's been stated several times - mostly, if I'm not mistaken, by MaxMaps in the Squadcasts - that HarvesteR chooses not to have tasks that take time to perform simply because the player can use time warp to fast forward to the task's completion. However, I'm not convinced by this justification, and I'll try to show why.

Flying in space takes time. So there's one time-based mechanic that is in fact the very soul of KSP. We don't have magic hyperdrives, wormholes or warp jumps that take our ships instantly, let's say, from Kerbin to Duna. A trip between those planets takes months. And yet we can use time warp to fast forward to its end. Does that mean that the time of the trip is meaningless, and we might as well have instantaneous travel? Of course not. In fact, the time warp makes the whole experience bearable - we don't have to wait months of real life time to get our kerbals to Duna.

So KSP already has its own temporality. Other time-based mechanics should contribute to create a sense of meaningful in-game time, even if the player can choose to fast-forward ahead. It causes a disjunction in the player's experience that the passage of time is so important to a side of the game while being completely irrelevant to another: in the time it took to get a ship from Kerbin to Duna, the whole KSC was rebuilt more than once and its technology advanced twenty years (if you have enough science and funds saved, you can even do it in the time it takes to get a ship from ground to orbit).

It's important to disclaim that, while I'm arguing for greater player immersion, I'm not adopting a defense of realism. More realistic features do not necessarily mean a better player experience, since they may add undesirable levels of complexity (game complexity should be emergent, it should happen as the result of the interaction between simple elements that can be known by the player). A good game strives to be intuitive - which is related to how we expect the game world to work, by relating it to previous experiences (such as reality, but also movies and other games).

Let's think about this issue for a moment. Leaving aside the fact that we're used to a lot of these features as they are right now, ask yourself: would the game be more interesting by having more time-based mechanics?

[Commander Jebidiah]

The resources point is in the what not to suggest list and chances are just for that one point this thread will probably be banned.

But other than that good points

[Anquietas314]

The resources point is in the what not to suggest list and chances are just for that one point this thread will probably be banned.

This thread isn't suggesting resource gathering, it's suggesting that it be a time-based mechanic when it's added (i.e. 0.91 if plans don't change), so it isn't covered by the do not suggest list. Even if it were, with everything else in the thread that's not, I don't think the mods would close the thread purely for that. They're pretty reasonable people, just like (most of) the rest of us

Researching: unlocking a new node in the tech tree also happens instantly.

Personally I think the fact you have to actually go and get the science is enough of a hassle without this (disregarding contract science). Otherwise, I'm happy with most of OP's suggestions/examples

Edited December 28, 2014 by armagheddonsgw

[Kegereneku]

I created myself a thread on the subject some time ago.

It made me come to the realization that to make the players WANT to take account of time you have to create a game mechanic that make use of it.

And not any game mechanic, it had to be something fundamentally simple but around which you would hinge most other features.

The basic was that the player must always have something to do as he wait, even if it's planning, in fact the BEST time-based mechanic is one where the player is constantly hoping to be in time (And the best developpers is the one that make you think you don't, even when you actually have).

From that, one of my idea was a try on the concept of "Periodic Budget", (and I'm thinking to make another one for new reasons)

At that point I'm wondering if the devs didn't already planned something in their secret base but can't finish it before all mechanism required for it to make sense are implemented.

Yep, sometime I prefer to bet on SQUAD's being more competent than they look.

In any case, it's still a good thing to discuss.

After reading your post, I find that you do focus too much on realism for the sake of realism.

I understand your reasoning on the "disjunction" but for players there is also such a thing as acceptable break from reality which is supported by Willing Suspension of Disbelief. It mean that even you/SQUAD cannot create a game mechanism that take account of the passage of time it is not harmful for the game experience.

On the other hand there's features known as anti-fustration, Timewarp is one of them, created because forcing upon the player meaningless wait time would be extremely detrimental for the game experience.

For example : "Experiments that take time" or "science over time"

We all know how much science something you will get out of something, that we won't have more by waiting, and that we are timewarping so much we could be doing 3 missions at once. With this in mind it is unnecessary to make people WAIT (for artificial reasons) over this.

BTW, I think that many people will confuse your example as masked suggestion. So expect people to argue against Ships Building taking time or Resources Gathering...even if that's not the subject. (just to say, I would argue against myself)

would the game be more interesting by having more time-based mechanics?

I can't really say yes, because I think we only need one, one that give a tempo to the progression of our entire space program.

And things like "Resources Gathering" would make very poor time-based mechanism (I have lost too much time already on Minecraft).

However... if we can find a way to make experiment more immersive, that's something else.

[Anquietas314]

For example : "Experiments that take time"

I agree with most of your other points, but something like the materials bay or mystery goo could be made a bit more interesting if the science gained was based on the time you ran the experiment for (e.g. to collect more space dust particles or to see how the bay/goo's contents would change over time in a particular environment). Obviously having time-based mechanics for stuff like atmosphere analysis and gravity scans would be silly since for the most part those things wouldn't be affected by running the experiment for longer (we can sweep the uncertainty stuff under the carpet ). Thermometers/seismic readings are a bit of a grey area there though - both would actually change with time (in some cases).

[Parkaboy]

Kegenereku, in no way I'm defending realism for realism's sake here. I'm not interested in too much realism; perhaps you misunderstood what I meant by "disjunction". I don't mean a break between game and reality, but a break between the two temporalities of the game: the distended time of space travel and the instantaneous time of pretty much everything else. I argue for time-based mechanics (whatever form they take; the list on the OP is not a list of suggestions, just an illustration) because I feel they'd match the temporality of the game, not because they would match reality.

[klgraham1013]

In 0.25, I played in Sandbox with a self enforced periodic budget and life support. Those two additions added new timed elements to the game, and really enhanced my enjoyment. Time really started to mean something as the game years rolled past and I had to start accounting for my income and whether I could afford that supply trip to Duna, which my colony so desperately needed.

Adding timed elements to science I find intriguing, as well.

[Commander Jebidiah]

This thread isn't suggesting resource gathering, it's suggesting that it be a time-based mechanic when it's added (i.e. 0.91 if plans don't change), so it isn't covered by the do not suggest list. Even if it were, with everything else in the thread that's not, I don't think the mods would close the thread purely for that. They're pretty reasonable people, just like (most of) the rest of us

Personally I think the fact you have to actually go and get the science is enough of a hassle without this (disregarding contract science). Otherwise, I'm happy with most of OP's suggestions/examples

Sorry for the misunderstanding saw the word resource gathering and didn't read the point.

I also wouldn't be so sure about the mods some of them can be unreasonable.

[Kesa]

There are two things that takes time in KSP.

processing and transmitting data. And no way to time warp them.

I'd love to have time based mechanics (all of these, and even more), but if we want to warp them, we also need background mechanics. Since many mods add it, I don't know if this is an issue at all, but the very thing stock ksp doesn't have seems to be background mechanics.

[WuphonsReach]

In order to do "time" right:

- Kerbal Alarm Clock would have to be a standard feature. Or some sort of in-game way to keep track of which craft needs attention and when. And what you plan on doing and when (such as transit windows).

- Science transmissions should be done in the background (queue them up, they get done over time).

- Science experiments should also be run in the background. Some might be instantaneous observations (low hanging fruit) like now. Other experiments might be longer running (orbit Duna for a week while the experiment is running).

Personally, I love KCT (Kerbal Construction Time). But it can be a PITA at the start when your first 10k cost rocket takes multiple days to build and you have nothing else to do but watch the clock. But that's mostly a balance issue, otherwise I think KCT nails it very well with having construction take time, having a queue of already built rockets, and having roll-out times. Plus if you recover your boosters (which really needs Stage Recovery mod to be stock), future builds take far less time.

Knowing that my Duna window opens in 33 days means I have to start planning in earnest to start getting rockets up into parking orbits while I wait for the window to open. It adds that sense of risk because I can't just wait until the last minute and start popping out 300t launches once per hour. I have to do a bit of mission planning up to a Kerbin year in advance.

There are some other things that could be time-based:

- Salaries / training costs

- Building maintenance (with potential impact on safety / reliability)

- Researching parts over time to make them better (lighter / stronger / faster / cheaper)

[thereaverofdarkness]

I very much support time-based mechanics and I feel that not implementing them is merely a lack of work on part of the developers for not bothering to come up with a good implementation. That's not to say they are wrong to avoid it; on the contrary, they have to manage production margins and perhaps they don't feel it's important enough to make the cut. But given the proper development, I can't see any reason to consider a time-based system to be sub-par to an instantaneous system.

unlocking a new node in the tech tree also happens instantly. This means we can advance our technology in huge leaps: getting the payoff from a big science collection mission, one can unlock whole tiers of the tech tree in one sweep. Wouldn't it make more sense to have this process be slower? As far as I know, no mod has adressed this, so maybe people like the system as it is.

I think if all of the science comes in over time, you won't ever have a situation in which the player suddenly accrues a whole lot of science at once. For instance, recovering a lot of samples and experiments from a mission: it will take the scientists time to run through all of that stuff and maximize the science gain from it. Perhaps it'll give a rapid science gain at first, and gradually slow down as we approach full understanding of a given sample. Also, there could be a mechanic in which the number of scientists and size of the lab are considerations. If you bring home a lot of samples or start to get a backlog going, you might want to make the lab bigger/more advanced and/or hire more scientists.

[kujuman]

A few points:

From the OP "The fact that there's a mod indicates a demand for it, although it would make sense to have a feature such as this something that could be toggled in the difficulties/settings panel."

This isn't really a good argument that something should be stock; vastly more users don't use the mod (or any mod or combination thereof) than do use it. Allowing it to be turned off means that it's a waste of programming time and effort. So long as Squad doesn't prevent mods from running, it makes sense for niche features to remain as mods.

I agree with a poster that without Kerbal Alarm Clock type functionality, plus things like transfer window calendars, time mechanics are bothersome. I'd additionally want something like "Skip to Epoch" if I had to use KCT or the like.

Some posters even demonstrate that players don't all want the same thing from KCT or other time mods: Stage recovery makes the game MUCH easier in career, but many time based mechanics make it harder. So what balance should Squad adopt?

In most tycoon games I enjoy, construction time is assumed away or is massively simpler than in real life. Sim City, The Sims, Roller Coaster Tycoon, Railroad Tycoon, Prison Architect, Banished, etc. The last two games have build time, but they are based on worker placement, not some set amount (don't give me that KSP is a worker placement game ). And with time mechanics come things like loans, and do we want loans in KSP? (I kinda do, actually...)

[r4pt0r]

consider gameplay. Once i get a boatload of SCIENCE i dont want to wait a week to use my new parts, thats unfun.

The important thing is to ask, "is it fun"? Does it add value to the game?

I find unnecessary waiting not fun.

Edited December 28, 2014 by r4pt0r

[pandaman]

Hi. In principle I agree with the implementation of time based mechanics in some areas...

Eg. for science it would simulate taking a series of readings over a certain time period without needing to actually click to do it. It would add a realistic element without making game play any more complex.

Construction time I think is in essence a good idea, but could prove awkward if not done correctly (not tried the mod so don't know how it works). Eg. Your ship takes X time to build. Get it on the pad and remember you forgot a battery, do you have to build it all again?

Construcion time would give purpose to having pre made 'ready to go' ships or 'parked' space planes for 'emergency' rescues etc. Especially if they can be made ready to go (refuelled etc) either on recovery or parking or in a relatively short game time.

[Jouni]

KSP is ultimately a game, where only one thing happens at once. The core gameplay is not managing a space program, but building rockets and actually flying them. You can have multiple missions going on at once, but there is no benefit in doing so.

Compare this to games like the Civilization series, where the core gameplay has always been managing things on the map. On a single turn, you may be waging a war against one of your neighbors, with major cities building new units and small cities close to the frontline building defenses. At the same time, other cities may be improving their productive capacity, your workers connect cities with roads and build farms and mines, your navy explores faraway lands, your scientists are developing new technologies, and you are yourself engaged in trade negotiations with one of your other neighbors. Everything happens at once, because otherwise you wouldn't be able to keep up with your enemies.

If we add Civilization-like management to KSP and run multiple missions at once, I'd like to have Civilization-like simplifications to actually running the missions. Instead of having to fly every ship manually, I might just assign a destination to a rocket on the launchpad, and forget it until it reaches its destination.

[Parkaboy]

Construction time I think is in essence a good idea, but could prove awkward if not done correctly (not tried the mod so don't know how it works). Eg. Your ship takes X time to build. Get it on the pad and remember you forgot a battery, do you have to build it all again?

This is actually a good reason agains the implementation of construction time for vessels: it would get in the way of the trial and error that is the soul of KSP. It's a better explanation than the whole "the player could just time warp" that's usually given. But it still leaves open the possibility of time based mechanics for other areas.

Personally, I don't feel we need research to take time. With the current system, where we need to unlock the nodes with science, unlock the parts with funds and unlock the tiers with building upgrades, the advancement on the tech tree became much more gradual and rewarding than it used to be. But we could have a few time-based experiments, and I think at least the buildings should take some time to be built, if not the rockets.

I don't think the game needs to invest too much in micro-management to make it work. I picture just leaving an experiment running or a resource being collected and while flying another ship, we get a message in the message panel saying something like "The experiment in vessel X was completed" or "Vessel Y has filled its tanks with resource z".

[Kegereneku]

A note : I don't think the developers are avoiding time-based mechanic. As pointed out feature like these would require a equivalent to Kerbal Alarm Clock before being playable, and furthermore, since time is the common denominator to everything. I understand they would prefer to introduce all feature that might depend on it to know what they have to balance.

The are probably... managing their time.

I agree with most of your other points, but something like the materials bay or mystery goo could be made a bit more interesting if the science gained was based on the time you ran the experiment for (e.g. to collect more space dust particles or to see how the bay/goo's contents would change over time in a particular environment). Obviously having time-based mechanics for stuff like atmosphere analysis and gravity scans would be silly since for the most part those things wouldn't be affected by running the experiment for longer (we can sweep the uncertainty stuff under the carpet ). Thermometers/seismic readings are a bit of a grey area there though - both would actually change with time (in some cases).

I'm nitpicking over the term but achieving something more interesting doesn't strictly speaking require a time-based mechanism. And done badly it could in fact be more detrimental than interesting. See my final note in bold down.

However all science would gain from being more interactive/immersive (if not too repetitive), to feel what we are studying even in a remote way.

Example right of the bat : using thermometer would show in the report a visual that we commonly associate to the temperature found (not that the science-report isn't funny in itself, just a idea)

(I'm totally thinking about creating a topic for that stupid idea)

Just to say, I once suggested to distinguish the whole science system into "soft science" that is being done automatically as long as the sensors are active, and "hard science" that require interaction for massive gain.

Kegenereku, in no way I'm defending realism for realism's sake here. I'm not interested in too much realism; perhaps you misunderstood what I meant by "disjunction". I don't mean a break between game and reality, but a break between the two temporalities of the game: the distended time of space travel and the instantaneous time of pretty much everything else. I argue for time-based mechanics (whatever form they take; the list on the OP is not a list of suggestions, just an illustration) because I feel they'd match the temporality of the game, not because they would match reality.

Having myself wrote a thread about time-based mechanic I know perfectly what you mean.

However I insist that (the way you phrased it at least) your argumentations on the subject is insisting too much on "thing to take time" because "it does" rather than as a gameplay mechanic.

I do not follow you much over the "temporalities" of the game, or maybe I don't get what you mean. The ability to instantaneously skip boring parts is an integral part of games which must only be changed if you have a specific gameplay or balance.

So there isn't so much a problem or lack -more like "can we improve the game"- that would require to add one/more time-based mechanism. To me it's more to give the tempo.

Launching a dozen of mission in less than a week for example is something that create a disjunction with any expected time-frame.

However, to use as an example to emphasize my point, keeping player waiting until their rocket is being built wouldn't be a sensible solution (even with time warp) as it deprive them of choice while offering nothing to do. (Also as pointed out it's impossible because player NEED last second modification)

I'm a little selling my own idea there but abstracting the passage of time through another mechanic (say Periodic Budget) without imposing things would push the player to plan a few "period / turn" in advance while keeping both choice, and a feeling of very similitude.

Even if a spaceship is built instantaneously, it doesn't matter if all in all it can retroactively be reckoned as "planned", improving the verisimilitude of the game in a non frustrating way.

On a final note : the time scale is a very very crucial parameter. KSP is a game where a Duna-mission is expected to launch, then near-immediately (from the player's perspective) be picked up ingame-years later. This is not a bug, not a flaw, this is a wanted feature. Thus, game-designer shouldn't design time-based mechanic that would require the player's attention every in-game day (in particular if a player might feel loosing something by skipping them).

[tater]

consider gameplay. Once i get a boatload of SCIENCE i dont want to wait a week to use my new parts, thats unfun.

The important thing is to ask, "is it fun"? Does it add value to the game?

I find unnecessary waiting not fun.

There is nothing about having things take time that forces waiting if it is done right. A simple set of buttons by the time counter that step time forward, discretely (jump 1 day, 2 days, 1 week, 1 month, etc). Ideally, any such Squad designed feature would drop you back to "normal" time if there was a node coming up… try to jump forward a day, but you have a node in 2 hours, so it jumps you 1:50, and pops you to the ship with the node needing attention.

This would be nice regardless, frankly.

Right now, a player can easily go from ~1955 to 2055 in maybe 100-200 days without really trying to unlock everything past just playing. In career I tend to wait for a Duna window for missions, for example, and given instant construction, etc, I have the Kerbin SoI packed with stuff by the time the launch window is there. If things got spread out with a few weeks here and there for construction, my first Duna mission in a career might be a probe instead of Duna-direct with Habs, multiple vehicles with nuke rockets, etc.

[Empiro]

I wouldn't mind having more time based mechanics, but there also needs to be greater consequences to spending time on things. Right now, you can just time warp past any delays, and the only consequence for that is missing a launch windows. The only consequence of that is the need to time warp some more...

Things like having upkeep costs (it's expensive to monitor satellites, and probes, and the ISS costs quite a bit to keep going) and life support would go a ways to add more consequence to just time-warping.

That said, one thing I did like about Kerbal Construction Time is that it made it so that I didn't max out the tech tree within the course of 3 months -- advancing science from basic rocketry to ion drives and nuclear engines should take more than just a few months.

[tater]

Yeah, I found this because I just grabbed KCT, and was gonna try it soon. Life support adds a time limit, and I have been playing with that for a while (just snacks, I like the idea of a simple abstraction of all consumable LS in one unit), and it really changes things as you need to resupply, etc.

[WanderingKid]

I am, generally, in support of this idea. Time should have consequences. However, I'm reluctant to rely on mods for it, due to the nature of mods. Both Life Support and funds consequences should be incorporated.

While not the most detailed assistances to this discussion, a general idea of reasonability; the basic pods should have 3-7 days of supplies on them. Minmus should be harder because of the time, not easier because of the gravity. However, newbies should be able to, by default, use Mun as a learning grounds before they have to incorporate further factors.

[Whirligig Girl]

Time based mechanics are actually pretty important. I hope this becomes part of the upcoming science revamp. I've mentioned in the past that it's kinda weird that all of a space program's first major achievements take place on the first day. In fact, the way time is handled in KSP is kinda sloppy. Just because you can use timewarp to jump to the end of a science experiment doesn't mean you're getting "instant science." The problem is, as of now it doesn't matter. Time-based decrease in budget (or at least reputation) if you're doing nothing but timewarping for a long period of time is probably needed so that not only do players have a station, but they should also be flying another mission during that time. Even multiple exploration missions at a time might boost the rep or funding of a space program.

Scott Manley's Interstellar Quest uses alarm clock because there's always a mission flying in space. This type of thing should be rewarded. Another thing is that Time is not an infinite resource already! The contracts system have an expiration date which must be met, so you might not be able to afford to be timewarping in a space station because you need to go put a satellite in orbit around Duna!

What I don't want, is having to wait for R&D. Also, a little sidenote, I think the button for unlocking parts should be "Develop" instead of "Research", because development is when you are actually creating a usable object. "Research" Doesn't quite sound right when you think of what engineers talk about. Engineers will usually talk about developing something more than researching it.

Edited January 13, 2015 by GregroxMun

[tater]

Good points,GregroxMun. Development would end up taking time anyway if it was done right---not sit and wait time, but actual playing the game time. Ideally to unlock a Mk1-2 pod, for example, you might be given it as part of a progressive set of missions. Maybe orbit it for X orbits of a certain altitude or greater (basically a wide speed range to hit the atmosphere at), then reenter straight from that apoapsis (sort of like the recent Orion test in RL). Not instant, but not tedious. Such a progression might include something like many of the new FP contracts, only with that part, so it might have another mission to put it in Mun orbit or something. The player gets to us the part ahead of time though, so it's hardly a hardship (of course such parts need to include related, required parts or it would be silly, so you need staging, and at least a few 2.5m parts one way or another).

I always have multiple things in the air in KSP after about the first week of the game, lol, so I always need to be careful about time warp (particularly with LS modded in).

It's also FUN to have time constraints. I had a, dunno, kraken explosion the other day (I keep having some ships just explode when I switch to them). It was a munbase core, and the hapless new recruit pilot was on a now suborbital trajectory (periapsis was ~60, but apo was not much above 75, so it was gonna decay fast if focused) in a cupola module. I had a "tug" I made (really a pretty normal CM/SM combo with the mk1-2 I use for rescue, etc) at the station nearby, and I tried to rescue him FAST, even though without focusing on him while he was in the atmosphere, nothing would ever happen to him… because it was FUN to do the fastest intercept, rather than the best intercept.

Such a generalized time-based scheme that OP suggests (or KCT, etc) allows for a space race as well, even if totally abstracted. The game could spawn in competing ships at some points in time (like it does for stranded kerbals) that would be en route tot he mun, or duna, etc… could be fun.

[rkman]

HarvesteR chooses not to have tasks that take time to perform simply because the player can use time warp to fast forward to the task's completion.

KSP in its current form encourages timewarping because it has no way to manage multiple ongoing missions (a lá Kerbal Alarm Clock). As the OP points out, we use timewarp routinely to skip to completion of (sub)tasks re planetary transfer etc.

If we could manage multiple ongoing missions then there could be time-based game mechanics without much of a temptation to timewarp to the task's completion. Instead of timewarping we just start another mission without worrying about missing important milestones in ongoing missions (the game would warn us ahead of time).

[WuphonsReach]

Something like KAC really needs to be part of the core game for the passage of time to be useful. At a minimum, automatic alarms for maneuver nodes. But many of the other KAC features are equally important.