A more intuitive tech tree

[CaptainKipard]

Wanderfound gave me control of the OP so I can keep it updated with a summary.

Original Post by Wanderfound

0.90 is approaching, and scope completion as well. All of the basic features of the game are going to be in place.

So, we all know what this means: time to stop whinging about big-picture details, and time to start whinging about niggly bits of polish. All the stuff that's been put off for ages with "it'll be fixed in the final game"; soon is the time.

Most obviously, it's time for the tech tree and parts list to get a tune-up.

So, there's the question: how do you think the tech tree should be altered, and what extra parts do you think are essential versus those that can be left to mods?

Personally, I'd like to see probe cores and basic aircraft bits moved to much earlier in the tech tree. I understand that the focus is on manned rocketry, and I have no objection to the Mk1 capsule being there from day 1. But I think this easy change could provide a good mod-free option for those who would prefer to take a more cautious or historical approach in the early game.

As for parts...mostly just the cargo bays and fairings. A better variety of non-rocket engines (ramjets, scramjets, props) would also be nice. It would also be very good to see a much greater variety of scientific instruments; integrating Scansat would make a good start on that.

Summary of the thread

The points may conflict. I tried adding all of the most discussed or supported ideas.

The following list is based on consensus that emerged over the course of the discussion. If you disagree with any of it, then make your case by offering substantive arguments.

1. Rework the tree to give players an open experience. Rough outline below. Let players follow their own progression.
   
2. No need for a starting node. Give players some science to spend initially. more -- Game setup allows adding starting resources
   
3. Branches divided by technology.
   
4. Try to put few parts per node; only strongly related parts. Keep it intuitive.
   
5. Have successive nodes with larger/more advanced parts. more
   
6. Allow modders to dynamically append nodes. more -- Now possible with Module Manager
   
7. Create new part variable for determining part type/function. Useful for filtering. more
   
8. Little or no node interdependency.
   
9. Make really low-tech parts like ladders or thermometers available early. more
   
10. Create an in-game tech tree editor -- Squad have an internal one.
    
11. The ability to define tier placement per node. more
    
12. Customisable tech nodes / Custom images. -- DONE
    
13. Customisable tech branch backgrounds / Customisable branch labels (see image below).
    

Tech tree schematic by Sherkaner

This tree is designed around existing parts using their stats as a guide for their place in the tree. It's not an endorsement of the parts themselves, which do need to be overhauled as well.

Development thread for the tree mod based on the idea below.

Edited May 29, 2015 by Cpt. Kipard

[kriseric]

I totally agree with you about moving probe cores sooner in the tech tree.

In fact, I would encourage Squad to create a new tier of low-grade unmanned rocket guidance part. Like a gyroscope used in very early rockets of the 1950s. You could launch a rocket with it, but it would only keep the thing pointed in a direction and could not be remotely guided. This would simulate the early V-2 style rockets that logically proceeded manned flights.

Also, I think wings landing wheels and jets should come soon as well! Aircraft should be viable much sooner.

Everything else I am largely happy with.

[TMS]

Agreed on the probe cores. Still makes no sense to me, not least as they are essentially redundant.

Docking and associated RCS parts need moving into the earlier part of the game. Docking opens up a world of possibilities and its a shame one has to grind for so long before the option is available to you. After all, as a new player, recreating the two-ship Apollo solution is usually high on the agenda.

I'd like to see more station-building parts introduced with some sort of actual function. I'd love for there to be a reason to haul specific modules into orbit and for orbital construction to be desirable.

Some sort of abstracted life support would also be welcome, even as a difficulty option, in order to emphasise the difficulty and danger of space travel.

Part of me believes that a tech tree structure will be difficult to finalise before details of the aerodynamics are known. I still live in hope of a system rescale so that achieving orbit in itself is a true challenge.

Edited November 9, 2014 by TMS

[Kegereneku]

Being an ardent advocate that the Tech-tree really need to evolve...

...I agree with most that as been said (definitively NOT with that system rescale).

- probes core sooner (unmanned robot for the win), even without solar-panel.

- Structural part being available more logically. (it shouldn't look as hard to build a new fairing than revolutionize space propulsion)

It is not much a question of following history than giving the possibility to do the same technical choices. Whereas America developed bigger and larger rocket, the USSR went for a lot of smaller but more efficient engines.

However, I think we will certainly have to wait until all parts -including parts that do not exist yet- get firmly established before revising the tech-tree.

You cannot balance properly an incomplete system.

I also think a few new small parts could be interesting. (a long 0.625 fuel tank for example)

- Early they would make the first rockets launch of our Kerbal Barn Program

- Later they would increase design possibility

- All in all they would complete the continuous augmentation of scale.

[Starman4308]

Gotta agree on having a starter-tech probe core; for those who RP on KSP, it is very silly to start manned. I like NecroBones's probe from Modular Rocket Systems: while it does weigh less than the Mk. 1 pod, it's got some disadvantages. It has no top node (and, thus, no parachute until you hit radial chutes), it has strictly limited duration until you get to solar panels/batteries, and you can't carry a Kerbal for those science-rich EVA reports.

If something needs to be shifted later, the S1 SRB could be placed later, and the Mk1/Mk2 lander cans could switch places. The Mk2 is direly overweight anyways, and its use might help channel new players towards wide landers.

One additional idea is to have conical fuel tanks (such as 1.25-2.5m) added, and placed later in the tree than their unfueled structural equivalents, which would help when Squad moves to a less stupid aerodynamics model.

[mattinoz]

Well if the first VAB is an old barn then surely the first crane you have to assemble and launch your first rockets is going to be a tractor. Which is not going to be capable of anywhere near moving 5-10tonne of Kerbaled rocket.

I would think the weight limit of the "Old Barn" tech level would be in the order of 1 tonne max. In that case would agree some early probe cores are needed maybe even a whole new range of parts smaller than tiny like 0.2m parts. Still could be a lot of fun to had with those even later in the game use for gyro guided atmosphere probes from and orbital probe. Maybe there should be some inline weather/pressure sensor in this range.

[John FX]

For the early game I like the idea of just having an SRB and a probe core and getting the first science by getting new height records.

[Farex]

Another thing to tweak / balance out are the overall science points you can get. This could be done in a few different ways:

a) Reduce the science gaining heavily (not talking about difficulty settings)

Increase the science costs for researching

c) Double the science costs by first paying science to unlock a Tier, than pay science (and funds) for each part

d) Make more different tiers instead of a few ones that have 10 to 20 parts in them

e) Don't do a tree-chart, use a star-chart. This would open up specializations and force the player to make decisions and it would probably increase the overall science costs, too

Greetings!

[Wallygator]

If the objective of career mode is to guide a player through a pre-determined set of parts and missions then just stick to the tree as it stands (with all the above suggestions considered as tweaks)

If however the objective is to create non-deterministic play space (not sandbox) where the player develops a space program with as few boundaries as possible, then the tech tree should be split in many parallel threads (by major part type - ro a star as mentioned ^) and then into branches for for each specific part (for the most part - a few exceptions like ion drives or LNVs coming in a bit later. Allow the PLAYER to choose what parts are desired first. Grant 10 points and no tech available. You want the command pod, then unlock it. You want ot start with probe first then lock that.

This would allow a wider range of play experiences while avoiding game "play sequence" mods.

I think the current paradigm of the tech tree is too inhibiting.

[Farex]

If the objective of career mode is to guide a player through a pre-determined set of parts and missions then just stick to the tree as it stands (with all the above suggestions considered as tweaks)

[....]

I think the current paradigm of the tech tree is too inhibiting.

This gave me an idea. Would it be possible to select the techtree-type / different techcharts at start (just like the difficulty settings)? If yes, then there could be a Career-Mode for beginners which uses the current techtree so you can get introduced into the game and get a feeling of what parts you want and need. Then there could also be an advanced Career-Mode which uses a different techchart and doesn't guide you like the current Career-Mode.

Although I think that this would be somehow "too much" or rather pointless at the end, as we have already 3 different modes to play. Hm, this requires further thinking.

[KSK]

It would be nice to expand the tech tree a bit, so that the current Tier 0 parts coincide with the first set of building upgrades. If we're looking at a humble origins aesthetic for the Tier 1 buildings, it would be good to have some appropriately humble parts to go with them. The current Tier 0 parts are actually pretty capable, and experienced players can get to orbit with them.

How about:

The Transponder MK 1.

This was originally a very basic radio transmitter for locating and recovering rocket stages (ties in with one of the Strategies that the player can apply) and was re-purposed as a payload for early orbital flights. Extremely basic probe core that cannot be used for orbital maneuvers or controlling scientific instruments.

The RT-5 SRB.

A more primitive and less powerful version of the RT-10.

The LV-20

The first useful liquid-fueled engine. Radiatively cooled precursor to the more powerful, regeneratively cooled, LV-T series of engines. Includes guidance fins for steering.

The Mk 1 Pod

Lets bring it back, but give it a crew of one and a limited or no heatshield (if atmospheric heating becomes stock). The barest of bare bones capsules, will not survive re-entry from orbit, but lets players start with crewed flights if they wish. Does not include reaction wheels.

Aircraft.

I probably wouldn't do much with them myself but why not. Basic aircraft parts (possibly prop driven) for those that like them. My own take on this is that your rocketeering kerbals would regard aircraft development as a distraction in the early years, since everyone knows that you can't fly a plane to space. But that's just me.

Edited November 12, 2014 by KSK
Aircraft

[John FX]

I have to say that they way `Better Than Starting Manned` starts the tech tree seems right. You get a probe core and solid boosters with no way of controlling them like fins or landing them, like chutes. As FAR and DRE are recommended mods this means your first few flights are a bit... random. Sometimes you gain enough altitude to progress, sometimes not.

It seems to fit with goddard launching stuff that you can still see at the height of the flight and running to where it landed to recover the remains.

Hey, maybe we could start with no probe core at all, just an SRB and some fins...

[KSK]

I have to say that they way `Better Than Starting Manned` starts the tech tree seems right. You get a probe core and solid boosters with no way of controlling them like fins or landing them, like chutes. As FAR and DRE are recommended mods this means your first few flights are a bit... random. Sometimes you gain enough altitude to progress, sometimes not.

It seems to fit with goddard launching stuff that you can still see at the height of the flight and running to where it landed to recover the remains.

Hey, maybe we could start with no probe core at all, just an SRB and some fins...

I know some folks like that level of historical progression, but I don't think going that far back would add much to the core gameplay, especially once you've done it once or twice before. Given that random part failure is not going to be a thing, I know that my fins-and-SRB rocket is going to work fine, I know that my fins+SRB+probe core is also going to be fine and so on, all the way up to current Tier 0.

On the other hand, I think historical flights would make good tutorials.

Granted, you could use the same argument against my previous post - why bother starting with less powerful versions of the Tier 0 parts? Which is why I tried to tie them into later game developments, or use them as a way of expanding out early game options through planes and uncrewed flights.

[Wanderfound]

I know some folks like that level of historical progression, but I don't think going that far back would add much to the core gameplay, especially once you've done it once or twice before. Given that random part failure is not going to be a thing, I know that my fins-and-SRB rocket is going to work fine, I know that my fins+SRB+probe core is also going to be fine and so on, all the way up to current Tier 0.

On the other hand, I think historical flights would make good tutorials.

Granted, you could use the same argument against my previous post - why bother starting with less powerful versions of the Tier 0 parts? Which is why I tried to tie them into later game developments, or use them as a way of expanding out early game options through planes and uncrewed flights.

As with many of the "difficulty" options: most of us aren't calling for a compulsory enforcement on all players. What we'd like is switchable difficulty toggles so that everone can play the game that they want to play.

Wanna start with manned capsules? Cool, make sure that you haven't turned off the default "Apollo" switch. Wanna take a more realistic progression? Hit the "Goddard" toggle.

All cool, everybody happy.

[CaptainKipard]

Here's what I'd like to see. It's the most open way possible and it's the most intuitive. I'd like to choose things myself rather than having my hand held throughout the whole tree.

[Kegereneku]

I kind of agree with that. This is basically "Allow player to research everything" with an emphasis on researching what we need when we need it. Yet you don't have access to advanced technology with pre-requisite.

Also, I kind of prefer the "tech-explosion" than tech-tree.

An opinion :

- I would give some jet engine 1 tier sooner. (not like we are gonna get into orbit with them anyway)

- electric would be "storage" and you would have "power generation" after the smallest-probes-body

- you could flesh out with high-impulse and high-thrust sub-category (basically your SRB but then it's also by cost).

- basic science would regroup all 100% reusable science equipment.

- docking would be 1 tier after basic-propulsion (it is a game-changing tech)

and if it ever become a thing

- antenna followed by longer-range antenna (part of electric)

[CaptainKipard]

- docking would be 1 tier after basic-propulsion (it is a game-changing tech)

No, see this is exactly what's wrong with the current tree. It's forcing you to follow a path. Besides that the first docking IIRC took place during the Gemini missions; that's quite early.

Ideally Squad should create a new set of gendered docking ports, which would be available early, followed by the current parts in order of size. The gendered parts can work just fine but it would force the player to think more strategically about their station expansion plans. Failing that a simple size progression should be fine, but still the docking ports should be available in early-to-mid game. Docking could be an early sub-branch of the structural branch, because they connect structures. Not propulsion though. They have nothing to do with propulsion.

Everything else you suggested looks good.

[JimmyAgent007]

i like the tech web idea that got posted, very much like Beyond Earth. However Id like the parts to be finished before the tech tree. Once we know ALL the parts that are going into the game, get them balanced, then we can place them correctly on the tech tree/web. also a Probe Core / Manned Capsule toggle for the start screen could keep a lot of feathers from being ruffled.

[KSK]

As with many of the "difficulty" options: most of us aren't calling for a compulsory enforcement on all players. What we'd like is switchable difficulty toggles so that everone can play the game that they want to play.

Wanna start with manned capsules? Cool, make sure that you haven't turned off the default "Apollo" switch. Wanna take a more realistic progression? Hit the "Goddard" toggle.

All cool, everybody happy.

That works better than my idea of punting the historical stuff into tutorials, since the tutorials wouldn't necessarily carry over into the main game.

[KSK]

i like the tech web idea that got posted, very much like Beyond Earth. However Id like the parts to be finished before the tech tree. Once we know ALL the parts that are going into the game, get them balanced, then we can place them correctly on the tech tree/web. also a Probe Core / Manned Capsule toggle for the start screen could keep a lot of feathers from being ruffled.

Or just move Manned Capsules into their own branch on the web, rather than requiring the small probe bodies as a prerequisite.

[JimmyAgent007]

Or just move Manned Capsules into their own branch on the web, rather than requiring the small probe bodies as a prerequisite.

Ideally yes, but if they start with a tree or want to force it one way or another then we should be able to choose.

[Francois424]

e) Don't do a tree-chart, use a star-chart. This would open up specializations and force the player to make decisions and it would probably increase the overall science costs, too

I love that Idea. You could have these branches (quickly tought of) :

- Unmanned flight,

- Manned flight,

- Plane engines(w/propellers and electric propellers eventually),

- Powerful engines (ie: American),

- Efficient engines (ie: Russian),

- Scientific endeavors (ie: all science gathering parts amongst other things)

And spread the science into these branches.

As with the World of Tanks tech tree, sometimes one science of a branch could point towards another branch skipping a lot of the early researches. Say Advanced electricity might give you the magnetometer and/or electric propellers even if you haven't researched the tree up to that point yet.

I would probably go for rovers and science personally, then manned flight to get my M27 cockpit or equivalent.

I would also go with efficient engines, making launches a bit more tedious, but better range once in space.

"Star-shaped" science is a really cool idea. Make it so !

Edited November 12, 2014 by Francois424

[Laie]

The gendered parts can work just fine but it would force the player to think more strategically about their station expansion plans.

Or the player would just postpone any station-building plans until true ports become available. Why should I base serious infrastructure on one-offs that will soon be obsolete?

[Slam_Jones]

I think they've both already been mentioned, but...

- Probe cores before manned cores (or at least much sooner)

- Atmospheric jet parts, making Kerbal sciencing easier from the get-go

- Docking ports, even if you start with the smallest one

[CaptainKipard]

Or the player would just postpone any station-building plans until true ports become available. Why should I base serious infrastructure on one-offs that will soon be obsolete?

The simple fact that you thought of an alternative way of doing things makes me think it's still a good idea. I'm in favour of any decision (with caveats) that let's a player make choices.

Additionally, stock docking ports right now are available quite late, so even without the gendered docking ports what I proposed is still a vast improvement.

Also remember that even the ISS has both gendered docking ports and common berthing mechanisms, which aren't exactly like the stock docking ports, but they're kinda close.

Or just move Manned Capsules into their own branch on the web, rather than requiring the small probe bodies as a prerequisite.

I think they've both already been mentioned, but...

- Probe cores before manned cores (or at least much sooner)

I admit, an initial choice of manned vs unmanned (or both if you want) seems like the best way to go.

ALSO

I'd like to draw attention to the fact that I got a rep for that tree mockup I made, and I rarely get rep for things that aren't mod-related, so I'm guessing people quite like this. On a related note, I have never ever in my life seen anyone say they like the current tree. It seems like most people either tolerate it, or hate it.

Edited November 13, 2014 by Cpt. Kipard