Tech web?

[Codraroll]

So... bit of an unusual suggestion, this, but I guess it fits the scope of the forum anyway.

 

Basically, this is another attempt to address the tech tree. I'll explain it as simple as possible:

Have the same parts be unlockable in several different nodes. Design the tech tree accordingly.

This would allow the tech tree to have many more distinct "branches", to allow for a variety of different ways to progress through the tree. It would begin with multiple starting nodes, all containing roughly the same parts. An SRB, some simple tail fins, some form of manned or unmanned command module, a simple Science experiment, and the usual struts and girders. Parachute optional. From there on, different paths of progress would be laid out for the player. Stronger SRBs for more oomph, or liquid fuel engines for more precision? Manned or unmanned flight? Recovering your craft or sending it into the wide, wild yonder? Airplanes or rockets? Interplanetary flight, or LKO space stations? Advanced vacuum engines or powerful lifter engines?

Unlocking a part in one node would subsequently lower the cost of other nodes containing the same part, to the points of nodes being unlocked for free if every part in them is unlocked through other means. This way, you might accidentally unlock unmanned launch technology while working on a manned space program, or rudimentary aircraft while unlocking advanced rover parts, but that's just engineering in a nutshell. Sooner or later, you'll find yourself with the parts and tools required to do something completely different from what you intended.

Several of the branches would eventually merge in certain aspects. For instance, no matter whether you picked unmanned or manned exploration initially, you would progress through the same nodes for fuel tanks, engines, stabilizer fins, decouplers and suchlike. Propulsion technology doesn't depend as much on payload, the lower stages would have to be pretty similar no matter where there are Kerbals on board or not.

What originally gave me this idea was playing through Career mode and realizing that I had to unlock three or four different nodes before I could build a working 2.5 m lifter. You get the fuel tanks in one node, the engines in another, adapters in a third and decouplers in a fourth. What if there was a node that had all the parts you needed for a rudimentary 2.5 m rocket, but which didn't offer any variety with regards to engines nor fuel tanks? With a "tech web", this would be the case. You could rush to build a basic 2.5 m booster stage, or unlock a variety of engines, fuel tanks, adapters and decouplers node-by-node and eventually combine them into a more capable booster. Likewise, you could concentrate on aircraft if you so wished, without swimming through an ocean of rocket parts to get there. Or vice versa.

Using a "tech web" instead of a tech tree would make the research and parts unlocking aspect of the game a lot more flexible, and cater to a variety of different playstyles from the beginning. The initial node would be rather costly, say, 100 Science points, to encourage players to pick and stick with a specialization rather than spending all their points unlocking rudimentary parts for every purpose. Players would, of course, start with 100 Science points to unlock their first node, but have to earn their Science after that. That being said, since the initial nodes would all contain mostly the same parts, starting anew in a different branch would be dramatically less expensive after a while.

 

What do you think? Is this a thing in some mod tech trees already? I kind of suspect it is, but if so, I'd like to see it spread further. Laying down several paths of progression would do a lot to allow different playstyles with the same tech tree, without overwhelming the player with choices. I think it's an idea worth looking into, at least.

[KSK]

Interesting. As a thought experiment, I'll see your tech web and raise you some 'tech dust'.

Do away with a player-visible tech tree and just give every part a science cost. The player can then pick and choose parts as they see fit and unlock them in whatever order they see fit. However, there are a couple of restrictions.

The total science cost of unlocking all the parts either remains the same as the current cost for unlocking the whole science tree, or is a rather han that cost to a) compensate for the fact that most players will have a few parts that they'll just ignore and b) because, as per your suggestion, the costs for unlocking parts are related and parts get cheaper if you've already unlocked a lot of similar parts.

Note that there is a tech tree of sorts 'beneath the hood' which defines those relative costs. That could be exposed to the player if desired but for the most part the logical relationships should be fairly - well, logical - anyway. If I've unlocked a whole pile of engines, then my engineers already have a lot of experience with building new engines, so researching more engines is going to be easier. Likewise for power generation, nuclear components, structural components, science gear etc.

In practice, players won't be completely free to choose parts in any order. There'll always be some constraints placed by the need to get to a place where new science can be earned. So, most players will probably start by unlocking enough basic parts to get to the Mun and/or Minmus in whatever style of vessel they choose. After that though - it's entirely up to them.

Design your spacecraft, spec out the parts required, check how much science you need to unlock those parts, decide whether it's worth it. OK, the tech tree is no longer a teaching aid that walks the player through the various parts in order, but tutorials, the KSPedia and, if required, better part descriptions should take care of that.

 

Edited January 17, 2017 by KSK

[Codraroll]

The problem with having every part available from the get-go, in any order you want, is that the "guidance" aspect of the game design is lost. For absolute freedom, there is sandbox mode. A completely open tech "tree" would be nice for the players who know exactly what they want, but unless you're already sure which order it's best to unlock stuff in to suit your playstyle, you'd have no clear path of progression to follow.

Ideally, a feature like this should be as useful to new players as to experienced ones. The tech web would lay down a few distinct paths to follow, so as not to overwhelm a player with what to choose, and to prevent strategical mistakes. For instance, newbies thinking that unlocking all the engines or all the solar panels before anything else is a good idea. Tech trees help distinguish which nodes you can go for early on, and which are clearly meant to be left until the end-game.

I don't think letting players access every part to design a craft before unlocking them is a good idea in a specified career mode. For sandbox, sure, but career is meant to introduce players to the game gradually. It's best to start with the parts required to build a very bare-bones craft, just to teach players what each of them do (note how the first node of the current tech tree does this very elegantly), and then introduce new parts gradually. Each node is clearly presenting the player with a set of items that do one specific - known - task better/differently than the parts (s)he already has unlocked.

The problem with a tech tree is that the parts are only introduced in a single context. You get extendable ladders for the purpose of entering and exiting a Mun lander, but you also need ladders to enter and exit aircraft, and researching Mun landers to build a workable aircraft makes little sense. Likewise, you get probe cores only after you've achieved orbit (unless you know exactly what to do to get them earlier), so you can't make meaningful unmanned sounding rockets. It makes sense to group rocket fairings and aircraft wings together under the node "Aerodynamics", but their actual purposes are wildly different. A tech web would introduce the different parts at different points of progression, depending on your play style. For unmanned play, batteries would be among the first thing you'd unlock, in manned play they're early mid-game. For aircraft play, you'd unlock small orbital insertion engines (for satellites deployed from suborbital aircraft), before the first rocket lifter engine. A rover branch would give you large cargo containers before multi-Kerbal crew capsules, and so on. Putting together a tech web, and determining when to introduce the different nodes in different contexts would be a bit of a nightmare, but I think the gameplay experience would benefit in the end.

[septemberWaves]

I quite like the way the mod Engineering Tech Tree does it. It has very specialized lines of research, so you can focus on improving a particular type of technology (like SRBs) with science costs increasing for the more advanced nodes. It has some of the benefits of both of the ideas presented so far in this thread, because you can choose several possible paths of progression whilst also having a lot of freedom in exactly what you unlock (the nodes tend to be low-cost and contain few parts). Such a system does not detract from the idea of a somewhat linear progression, but it provides more versatility for science gathering. It also helps to make a playthrough with low science returns more viable, as the individual nodes tend to cost less for the tradeoff of containing fewer parts in general.

Edited January 17, 2017 by eloquentJane

[KSK]

32 minutes ago, Codraroll said:

The problem with having every part available from the get-go, in any order you want, is that the "guidance" aspect of the game design is lost. For absolute freedom, there is sandbox mode. A completely open tech "tree" would be nice for the players who know exactly what they want, but unless you're already sure which order it's best to unlock stuff in to suit your playstyle, you'd have no clear path of progression to follow.

*and lots more good stuff*

That's a fair point but I think the 'guidance' aspect is over-rated. Consider what you need for a science satellite, powered by a liquid-fuelled booster - and I'm picking that because it's a relatively complicated example given the current game mechanics. What component types do we need to include?

Engine
Fuel tank
Decoupler
Probe core
Battery
Solar panel
Science gear
Guidance system (whether than be fins, vectored engines or more reaction wheels)

That's quite a lot to take in for a new player but it also covers almost all of the part types. Figure out what those do and almost everything else you unlock in the tech tree falls within one of those component types. So once a new player has unlocked and figured out the first example of each part type, large chunks of the tech tree become obvious since they're just devoted to unlocking different examples of the same type.

So what are we looking for in a tech tree? What are the possible gameplay reasons for having one?

Progression (1).  It's an in-game goal to direct player activities. Get science points, unlock new stuff.
Progression (2)  The player starts with basic stuff and unlocks more advanced stuff.
Progression (3)   Adding a historical timeline to the game.
Guidance.     Introducing new players to game concepts
Strategy.       Forcing the player to prioritize research and make choices that impact on gameplay.

Progression(1) would be fulfilled by almost any system, whether that be the current system, your tech web or my tech dust.
Progression(2) is where my tech dust idea completely fails.
Progression (3). Kind of linked to Progression(2) and again where my idea fails. On the other hand, the stock tech tree doesn't provide a 'historical progression' either.
Guidance.  Over-rated in my opinion, as discussed above.
Strategy.    Rudimentary and annoying in the stock tech tree. Your example of building a 2.5m lifter is an excellent one. I think either of our systems offers comparable or better levels of strategy without being as irritating. Even my 'unlock any thing in any order' approach does require some prioritization and meaningful decision making. (How many engines do I need. All of them? Some of them? What order do I research them in? Is it worth going for the big stuff first (and not having enough research points to spend on other items)?

So, I'd say the only serious flaw in my tech dust idea is that it does lose the reward loop mechanic of doing science to unlock better stuff. Personally, the added freedom makes it worthwhile for me but I can see why others might disagree and prefer your tech web idea.
 

 

Edited January 17, 2017 by KSK

[Wjolcz]

47 minutes ago, Codraroll said:

The problem with having every part available from the get-go, in any order you want, is that the "guidance" aspect of the game design is lost. For absolute freedom, there is sandbox mode. A completely open tech "tree" would be nice for the players who know exactly what they want, but unless you're already sure which order it's best to unlock stuff in to suit your playstyle, you'd have no clear path of progression to follow.

Oh nooooo! NONONONO! I'M BECOMING TRIGGERED!

Let's start with the "guidance". First of all, there is none and never was. Second of all, why lock the player in a certainprogression in the name of "guidance"? Seriously, people need stop treating the new players as some sort of dumb infants that don't know how to have fun.

"Naaaaah, mate. You don't want these wheels. Grab that engine that doesn't fit any of your currently available fuel tanks instead. Trust me I've been to the Mun I know how to progress".

Back on topic: progression is one thing. I agree that more freedom and paths during the research should be the way it works. However, the problem of science grind would still be there. Tech tree and science points are permanently bound together. Poor decisions were made during the creation of career mode. A new tree won't fully fix the problem no matter how many paths and branches it has.