A more intuitive tech tree

[sherkaner]

Well, I'll make you a deal. If he gets the mod up and going I'll give it a try. Maybe I'm being too harsh before having seen it.

Yeah, while I know where you're coming from (especially seeing the whole tree at once), I actually think you may be pleasantly surprised when you're playing through it. Each part you unlock should open up 1-3 other parts which are directly related to the one you got and so I think you'll find that it has a very organic and intuitive feel. Once it's all unlocked it looks complex, but the structure actually has some meaning and you'll have worked your way through it as you built your space program. Plus I fully expect that people may not unlock everything (at least not until later in the game when it becomes trivial)! Since the nodes are mostly per-part, your map will be very dependent on what lines of technology you found most helpful for what you want to do. An aircraft-focused program will build up a technology tree very different from a manned rocketry program. The current tree essentially forces you to unlock everything as you go -- this one definitely does not.

Of course I say all this not having played it yet, but that's how it feels to me as I build up my TechManager config to implement it.

For newbies in particular, I hope this will be less confusing. As it is, you buy a node full of a grab-bag of parts which you then have to figure out. In this tree, everything is organized functionally and so if you figure out the first part in the technology line, you'll be in a good position to understand the following parts. For this reason, I'm actually going to the trouble of writing descriptions for each node that provide a quick idea for what the new part(s) do for you and so I think this tree will help newbies build up an understanding of the game as they progress.

Edited January 7, 2015 by sherkaner

[tuckjohn]

Well the good news is that you'll have the whole game to get ready for that headache, and by the time you see the tree unlocked your confusion will be irrelevant... because you'll have unlocked it. You remember that you don't see the whole tree from the beginning, right? You'll have more than enough time to "learn" the tree. You'll be fine.

I still don't understand why you call it complex. Your idea takes complexity from one level and applies it to another level.

Complexity is Relevant!

[Alshain]

Yeah, while I know where you're coming from (especially seeing the whole tree at once), I actually think you may be pleasantly surprised when you're playing through it. Each part you unlock should open up 1-3 other parts which are directly related to the one you got and so I think you'll find that it has a very organic and intuitive feel. Once it's all unlocked it looks complex, but the structure actually has some meaning and you'll have worked your way through it as you built your space program. Plus I fully expect that people may not unlock everything (at least not until later in the game when it becomes trivial)! Since the nodes are mostly per-part, your map will be very dependent on what lines of technology you found most helpful for what you want to do. An aircraft-focused program will build up a technology tree very different from a manned rocketry program. The current tree essentially forces you to unlock everything as you go -- this one definitely does not.

Of course I say all this not having played it yet, but that's how it feels to me as I build up my TechManager config to implement it.

For newbies in particular, I hope this will be less confusing. As it is, you buy a node full of a grab-bag of parts which you then have to figure out. In this tree, everything is organized functionally and so if you figure out the first part in the technology line, you'll be in a good position to understand the following parts. For this reason, I'm actually going to the trouble of writing descriptions for each node that provide a quick idea for what the new part(s) do for you and so I think this tree will help newbies build up an understanding of the game as they progress.

Well that does sound good because I'm a plane person (really I like rockets too, but planes more) and I've agonized over the tech tree as it is now due to it's focus on early rockets.

[GreatGreengGoo]

One thing, I personally, would love to see is requirements for unlocking late game tech. More than just spend X amount of science points. For example to unlock the next node in space construction (or whatever you use to make stations parts) you actually have to go up and build a space station with X requirements and then the node becomes unlocked. Things like that. This concept could be applied to pretty much all of the science trees.

I hate the "Oh I have 5000 science points after this mission, because of the administration building strategy. Time to unlock 15 nodes"

[tuckjohn]

One thing, I personally, would love to see is requirements for unlocking late game tech. More than just spend X amount of science points. For example to unlock the next node in space construction (or whatever you use to make stations parts) you actually have to go up and build a space station with X requirements and then the node becomes unlocked. Things like that. This concept could be applied to pretty much all of the science trees.

I hate the "Oh I have 5000 science points after this mission, because of the administration building strategy. Time to unlock 15 nodes"

That's a great idea! I would recommend that you would still have to spend science after, say, landing on duna. I dont think that you should be able to unlock a node for "free" or at no science cost.

Also, here is some rep.

Edited January 9, 2015 by tuckjohn
rep

[Alshain]

One thing, I personally, would love to see is requirements for unlocking late game tech. More than just spend X amount of science points. For example to unlock the next node in space construction (or whatever you use to make stations parts) you actually have to go up and build a space station with X requirements and then the node becomes unlocked. Things like that. This concept could be applied to pretty much all of the science trees.

I hate the "Oh I have 5000 science points after this mission, because of the administration building strategy. Time to unlock 15 nodes"

The only thing is, that would go against the whole idea of letting people play the game the way they want. "You can't unlock the basic jet engine because you haven't used the rocket engine yet." Might as well just keep the stock tech tree.

[tater]

People can "play the game the way they want" in sandbox. Career needs limitations, that's the point of career. The tech tree cart is before the horse, anyway.

1. Add in failure %s for "experimental" parts. Failures should be pretty common, but most would be things like lower thrust, higher fuel consumption, less power generation, more power use, etc. (and engineers can possibly bring them up to "stock" values via EVA where appropriate)

2. Select a mission/contract (add some space program "strategies" that are actually strategies that involve spaceflight missions, like "land kerbals on the Mun.").

3. Said contract/mission ("Explore the Mun," for example as a mission) then unlocks Experimental Parts via a provisional unlock of nodes (picked by the player, and bought with "science" as now). Player selects contract(s), which have much more defined time periods to complete them (10 years is a long time when you can design, build, fly and return a mun mission in a few game hours, even with warping).

4. Player has these parts (in red perhaps, instead of the blue you see them in the VAB now) for doing their missions, but ones that are untested are "X" tech. The "X" parts can fail. New contracts would include some testing to be done before the final requirements. Landing on the Mun would require landing with some of the X parts attached to the vehicle, or no payout, for example. An pre-requirement might be to test some of the parts in Kerbin orbit, though.

5. Testing the parts, plus additional science is then required to revise the parts to be production parts (the parts we use now).

So mission drives tech development instead of landing on the Mun to get science to unlock the parts you want for landing on the Mun.

[tuckjohn]

People can "play the game the way they want" in sandbox. Career needs limitations, that's the point of career. The tech tree cart is before the horse, anyway.

1. Add in failure %s for "experimental" parts. Failures should be pretty common, but most would be things like lower thrust, higher fuel consumption, less power generation, more power use, etc. (and engineers can possibly bring them up to "stock" values via EVA where appropriate)

2. Select a mission/contract (add some space program "strategies" that are actually strategies that involve spaceflight missions, like "land kerbals on the Mun.").

3. Said contract/mission ("Explore the Mun," for example as a mission) then unlocks Experimental Parts via a provisional unlock of nodes (picked by the player, and bought with "science" as now). Player selects contract(s), which have much more defined time periods to complete them (10 years is a long time when you can design, build, fly and return a mun mission in a few game hours, even with warping).

4. Player has these parts (in red perhaps, instead of the blue you see them in the VAB now) for doing their missions, but ones that are untested are "X" tech. The "X" parts can fail. New contracts would include some testing to be done before the final requirements. Landing on the Mun would require landing with some of the X parts attached to the vehicle, or no payout, for example. An pre-requirement might be to test some of the parts in Kerbin orbit, though.

5. Testing the parts, plus additional science is then required to revise the parts to be production parts (the parts we use now).

So mission drives tech development instead of landing on the Mun to get science to unlock the parts you want for landing on the Mun.

I like the idea that the experimental parts have a high chance of failing, as of now you can get the "nuke" long before you unlock it on the tech tree.

[Alshain]

People can "play the game the way they want" in sandbox. Career needs limitations, that's the point of career. The tech tree cart is before the horse, anyway.

Your missing the point. If I want a career where everything is done with spaceplanes, I shouldn't have to build rockets to do it. The whole point is to NOT have to go to sandbox mode.

[CaptainKipard]

One thing, I personally, would love to see is requirements for unlocking late game tech. More than just spend X amount of science points. For example to unlock the next node in space construction (or whatever you use to make stations parts) you actually have to go up and build a space station with X requirements and then the node becomes unlocked. Things like that. This concept could be applied to pretty much all of the science trees.

I hate the "Oh I have 5000 science points after this mission, because of the administration building strategy. Time to unlock 15 nodes"

The only thing is, that would go against the whole idea of letting people play the game the way they want. "You can't unlock the basic jet engine because you haven't used the rocket engine yet." Might as well just keep the stock tech tree.

How about making the player use at least one part from the tech tree node in any contract, in order to research a descendant node? E.g. You must use the docking port once at least, before it lets you research the node with the senior?

It's not very grindy, and it fits with the way the current contract system works.

This would have to be made to work with any part (modded or stock) that's in the node, because I imagine modders might want to place some of their parts in the existing nodes, instead of always making new nodes.

[Starwhip]

I like this idea, especially the spaceplane rebalancing. At the moment, just like Alshain says: you need rockets first. Give us, at the very least, earlier access to the parts, like in the 18-science node splitoff.

[tater]

Your missing the point. If I want a career where everything is done with spaceplanes, I shouldn't have to build rockets to do it. The whole point is to NOT have to go to sandbox mode.

No, you're missing the point. At some reasonable level, the two technologies are interconnected. You will likely have to develop rocket engines before you can build a spaceplane as the latter needs a non-airbreather at some point.

If spaceplanes were easy to make before rockets, we'd not be waiting on the next Falcon 9 launch, there would be cheap planes in orbit already, too. A more intuitive tree means one that matches reality more.

The suggestion was to have more requirements than just spending "science" (largely gained from planetary science, and hence entirely unrelated to "tech" anyway). So to get updated jet engines, you'd have to test the earlier jets. To get hybrid jet/rocket engines, you'd likely need to thoroughly test both precursors.

[Alshain]

How about making the player use at least one part from the tech tree node in any contract, in order to research a descendant node? E.g. You must use the docking port once at least, before it lets you research the node with the senior?

It's not very grindy, and it fits with the way the current contract system works.

This would have to be made to work with any part (modded or stock) that's in the node, because I imagine modders might want to place some of their parts in the existing nodes, instead of always making new nodes.

Well, all it would be is tedium. I don't like part X and I never use it so I put a part X on the launchpad and immediately recover it without even lifting off to get past the tech tree lock out. Personally I would find the fact I had to do all that just to unlock the useful part I want to be annoying, and for what reason?

The suggestion was to have more requirements than just spending "science" (largely gained from planetary science, and hence entirely unrelated to "tech" anyway). So to get updated jet engines, you'd have to test the earlier jets. To get hybrid jet/rocket engines, you'd likely need to thoroughly test both precursors.

I understand that, but as I said, it won't add anything to the gameplay and it will just annoy people. If YOU want to do this, then do it. Nothing is stopping you from making it a self enforced rule, but I, and most likely MANY others, would simply find it an annoyance.

Edited January 9, 2015 by Alshain

[tater]

LOL. This is a thread about a more intuitive tech tree, and your response is "don;t touch the tech tree, it makes it annoying to me!"

An intuitive tree means that the relationship between older parts, and improved replacements matches real life expectations (intuition is informed by our real world expectations). So you'd expect to have to know how to make rocket engines before you can learn to hybridize jets and rockets.

How about they make the tree make sense, then YOU can mod the game, or do something to make the tech tree pointless/arbitrary/whatever?

The tech tree IS the gameplay of career. It is the reward system, planned that way or not. Having it make more sense (as it makes virtually none right now) WILL change the way things get unlocked.

[Alshain]

LOL. This is a thread about a more intuitive tech tree, and your response is "don;t touch the tech tree, it makes it annoying to me!"

It's a place for discussing suggestions. It's not a "pat me on the back for my idea" forum. Obviously Squad weighs the opinions of their customers or we would have a barn for tier 1. Whether or not they were listening to the right people doesn't change the fact they listened.

Your idea is a bad one. That is an opinion, it is my opinion, and I have the right to express it. I'll express it again: forcing lockouts on the tech tree because you haven't "used" the part before is a bad idea.

[CaptainKipard]

Maybe it's best to just leave sciencing how it is.

[tater]

It's a place for discussing suggestions. It's not a "pat me on the back for my idea" forum. Obviously Squad weighs the opinions of their customers or we would have a barn for tier 1. Whether or not they were listening to the right people doesn't change the fact they listened.

Your idea is a bad one. That is an opinion, it is my opinion, and I have the right to express it. I'll express it again: forcing lockouts on the tech tree because you haven't "used" the part before is a bad idea.

I'm fine with that. I'm not sure why you'd want to unlock a part you don't want to "use," but knock yourself out. The entire science/tech framework is flawed, IMO. It makes no sense whatsoever, so making a tree based upon unlocking with "science," is absurd, and by definition non-intuitive.

I'm still at a loss how a geological sample from the Mun helps me get better parachutes, for example. How can any tree that makes me spend "planetary science" to buy new tech be intuitive?

New player: "How to I unlock big parachutes, and advanced jet engines?"

Old timer: "Collect some rocks on the Mun. Or go to orbit. Or test a huge booster engine cluster in the middle of the ocean. Yeesh, darn noobs."

Old timer with a rational, intuitive tree: "Do some suborbital flights to test smaller chutes in the atmosphere, as well as some atmospheric science survey missions for the big chutes. For the jets you will need to have already researched the regular jets, plus some similar atmospheric science missions to the chutes since both work in air. You might be able to combine most of those experiments in a few flights."

[CaptainKipard]

I'm not sure why you'd want to unlock a part you don't want to "use," but knock yourself out.

In order to use a part right now you have to do three things.

• Unlock the relevant tech node
  
• Buy the part initially
  
• Buy the part when you build.
  

You're not really unlocking the part, just the tech node that the part is in. That's may seem a bit over the top, but I think that this process should stay the way it is, because modders may want to put some of their parts in the existing nodes, if they are very similar to one of the stock parts. Also Squad may want to add some more parts.

[sherkaner]

Old timer with a rational, intuitive tree: "Do some suborbital flights to test smaller chutes in the atmosphere, as well as some atmospheric science survey missions for the big chutes. For the jets you will need to have already researched the regular jets, plus some similar atmospheric science missions to the chutes since both work in air. You might be able to combine most of those experiments in a few flights."

We're most of the way to the intuitiveness of this with just the tree structure we've been discussing here. I like the idea of using testing as a method to R&D progress, but I don't think the science approach is inherently absurd. Sure it might be a little harder to come up with a rationalization for how landing on the moon helps research parachutes, but it's generally the "brute force" approach of throwing resources at a problem, whereas testing is more of the targeted approach. Both approaches happen in real life.

In most cases, I think ideas that give players more options (without adding enormous complexity) are a Good Thing for a game like KSP. The tree we've been discussing here goes a long way with that. For unlocking nodes within that tree, the current science approach lets the player who wants to be very contract/mission focused derive a resource from that to progress their R&D. I think adding to that system with a "science discount" for using/testing a node's prerequisite is very nice for providing another play-style option for the player who is more interested in the R&D itself than they are having to go off and do contracts/missions to get the science to buy R&D nodes, or for the player who might be a bit "stuck" and needs a way to grind up their tech to progress further on contracts.

[Louella]

I'm a low skilled player, though I've been playing Kerbal Space Program on and off for a long time, since before there was a Mun.

I agree that it doesn't make sense that digging up rocks on Kerbin allows bigger rocket engines. Although in the 'real' world, it's sometimes the case that a thing developed for one purpose, finds use in another. (wasn't the thing that makes cellphone cameras possible, originally from spy satellite programs?)

However, there still has to be ways for a player to unlock new parts, such as larger engines, using the stuff they have.

But, it's easy to make things overcomplicated.

You can divide up the tech tree in all sorts of ways, and make all sorts of arguments as to why that makes more sense.

Things like, dividing tech into "Materials science, Electrical engineering, Propulsion Systems, Sensor Technology, Structural Engineering" would be one way. But then you've got a lot of categories, which means complexity, which isn't necessarily bad, but complexity for its own sake isn't good either.

To me, I think it would make sense if there were no more than 3 kinds of "Science" data. Space flight data, atmospheric flight data, survey data.

And many parts would be able to be unlocked using more than one kind of data. And you'd probably get more than one kind of data from an experiment or report.

Like:

Jeb launches from KSC in a sub orbital rocket flight, takes a couple of crew reports, and the ship is recovered.

The mission would provide some data on space flight (from crew report on use of the rocket engine), some data on atmospheric flight (because of crew report in atmosphere), and some survey data (crew reports).

Recovery provides data on space and atmospheric flight, through examination of the materials of the recovered parts, and/or any telemetry data, and/or any data from the Tracking Station.

A Kerbal at the R&D centre thinks "We could use this data to improve our rocket design" = designs for new engines.

A second Kerbal thinks "We could do this, to improve flight through the atmosphere" = designs for wings or control surfaces.

A third Kerbal thinks "If Jeb had one of these, we could learn more" = designs for new sensors and other devices.

Meanwhile:

Bill and Bob are on a rover mission around KSC

The mission provides very little on space flight, little data on atmospheric flight, and a lot of survey data.

Recovery of the rover, and examination of the materials would be the only thing that makes sense in providing data for rocket engines or aerodynamics. Crew reports and telemetry would also provide data.

A Kerbal at the R&D centre thinks "We need better materials for building with" = some improvement in atmospheric and space flight science.

A second Kerbal thinks "We could learn more, if we had these" = more sensors and other devices

A third Kerbal thinks "If these were automated, we could do more at once, but we'd need to power them" = automation, battery technologies.

If a player needs more space flight data, then contracts for experimental part tests should be able to provide it. It would make sense for better space technologies to be unlocked through doing lots of space flight testing contracts, or for better atmospheric technologies to be unlocked through atmospheric flight contracts.

To me, that would make sense, but even that might be overly complicated.

Edited January 10, 2015 by Louella
brevity improvement

[CaptainKipard]

That's a lot of text for an idea that can be summarised as "various types of science for different types of parts". Keep it brief.

[sherkaner]

"various types of science for different types of parts"

I'm not a fan of that idea. I see that ending up constraining the player more without any obvious gain in fun/realism/whatever. It seems like at any given point in the game, you would end up with limited options for how you gather the particular type of science you need to get the particular tech you want. It just doesn't seem like it offers much for the complexity added.

[CaptainKipard]

I don't know. I feel like there should be something that makes the science system more than just collecting points. Maybe this, maybe not this but something like this.

[tater]

Complexity is gameplay. Complexity of choices. Walk out the front door of your office building at KSP. Collect science! Meh. That's the game now. Accept contract, click crew cabin, broadcast crew report, collect science. Meh.

The cart is before the horse. Science as collected in KSP is heavily weighted towards planetary science. There is no possible way to rationalize around this. Planetary science has little to do with spaceflight engineering. For spacecraft design, you need not reach the Mun, just operate the engines the right amount of time, inhabit the capsule, learn how to dock.

Either the science needs to be far more discipline specific (which will make much planetary science worthless from a reward standpoint), or unlocking nodes could be linked to more specific actions than "spend science."

And all that doesn't bother with the fact that the cart is before the horse. You are tasked with landing Jeb on the MUn, THEN you figure out how to do it, and order the parts needed.

An alternate paradigm:

1. Take "Explore the Mun" Contract. (you just did "Get to Orbit"). Contracts are perhaps unlocked BY SCIENCE (unsure, this just came to me, bear with me)...

2. Go to research, and pick the parts you think you want to complete that mission that you do not already have.

3. New "contracts" are generated for unlocking those parts. For example, You want the Mk1-2 pod, the Mk1 Landercan, the X200, Clamp-o-tron, and RCS parts. Contracts might include testing the Mk1-2 by reentering Kerbin atmosphere at 2200+ m/s after putting it in orbit. The clamp-o-tron might require a docking test (put 2 in orbit, and dock). The landercan might require landing a probe on the Mun (testing dust, yada, yada, yada). RCS might also require a docking event (both done in 1 mission). After testing, THEN you buy the parts, and complete the "real" contract.

This would allow the player to drive his program from day one. There might be "build an SSTO spaceplane" contract, which then sets you on a path to test the various required parts, so you can build the plane and go...

This would eliminate the requirement to "grind" science to get tech so you can do missions you want to do. This would have the player pick the missions, then work on the intermediate steps to get the tech to do the mission they want to do. The steps would be much the same, but it would have an intuitive path you'd be following. Why am I doing orbital missions, and testing parachutes? Because I am testing spacecraft for getting to Duna, and chutes to land there...

Edited January 10, 2015 by tater

[Baythan]

Complexity is gameplay. Complexity of choices. Walk out the front door of your office building at KSP. Collect science! Meh. That's the game now. Accept contract, click crew cabin, broadcast crew report, collect science. Meh.

The cart is before the horse. Science as collected in KSP is heavily weighted towards planetary science. There is no possible way to rationalize around this. Planetary science has little to do with spaceflight engineering. For spacecraft design, you need not reach the Mun, just operate the engines the right amount of time, inhabit the capsule, learn how to dock.

*snip*

Interesting ideas and all, but this is rather offtopic for this thread. There are other threads (some rather old, some not so old) for discussing redesigning how and why science is gathered and what it is used for. While a redesign of science is a great idea and should probably be done in conjunction with the changing of the tech tree, this thread is about changing the tech tree. Right now the tech tree is set up in a rather arbitrary system of nodes with somewhat random items in each node and was designed with the parts as balanced for Sandbox gameplay.

We are (or were) discussing ideas on how to fix the current tech tree and make it less arbitrary and provide more freedom of choice to the player. We don't want it to be really complex, but more freedom means more nodes and more targeted paths to specific parts.

I think it would be GREAT if Squad did a redesign of the tech tree, but as far as saying what is a higher priority: tech tree redesign, science system changes, or part balancing... that is all a matter of opinion. The science system is not dependent on the tree, nor is the shape of the tree dependent on the science system so either one could be done without changing the other, but changes to both would go well together. Part rebalancing would obviously change the tech tree, but a better layout of the tree itself would help with placing rebalanced parts in more intuitive nodes.

P.S. - Saying "the cart is before the horse" over and over again isn't helping if you don't give any argument about what the horse is. If you think the science system needs to be changed first, find one of the threads about the science system and try to get people to discuss it there. Many of us here want more than just this one change and we aren't saying "Hey SQUAD, drop everything and fix the tech tree before you do anything else!" We are putting forth the opinion that among the MANY things we would like to see modified in this Beta stage, the tech tree is one and here are some ideas about how it could be done. We are also active in other threads with our opinions on how THOSE things could be made better as well.