A more intuitive tech tree

[JimmyAgent007]

I really like the tech web diagram, Its a lot like what was done in Beyond Earth. For the newbies though, Id have the branches of the web depend on an initial choice between 3 options. Cockpit, Capsule, Probe. You cant pick anything else until you make that initial choice but once you do it all opens up. That way you can play the game however you want and progress how you want. As for the science system, I like the idea of contract testing of parts unlocking parts. The science collected on missions i see as 'nerd bait'. Sure the Mun sample doesnt help make a Mainsail but it might inspire or attract nerds to help make it.

[sherkaner]

As for the science system, I like the idea of contract testing of parts unlocking parts. The science collected on missions i see as 'nerd bait'. Sure the Mun sample doesnt help make a Mainsail but it might inspire or attract nerds to help make it.

Hm, how about this:

We want players to have to run a diversity of missions in order to unlock parts generally. Currently that's done via science gained either from contracts or taking science directly, which gives you some options since taking science requires you to go to different places.

But what if instead each part type (ie. branch on the explosion tree) had a sort of progress bar toward the next step, and actually using that part type in a variety of conditions directly caused that progress bar to go up. Essentially the game would be keeping track of the location, duration, g-forces, etc. a mission is experiencing and the wider variety of conditions a part is exposed over a longer period of time, the more credit you get toward the next level. Have weak little landing legs and want beefy ones? Create test craft and land it with your current legs on the flat, on a mountain, on a beach, on the moon, land it harder and harder until you figure out when they break. Want bigger a bigger rocket motor? Launch 20 missions with the ones you have, launch light things and heavy things, stick it on a space plane prototype and test it that way. This would encourage even more diversity of missions, give the player many many play style options, but still give a sense of actually "doing R&D" on the part types in question, working on branches directly rather than generally gathering a resource to buy steps with.

I still like the idea of initially unlocking the part with a test contract, which would work well with this. And the reliability of the part could perhaps steadily increase as you proceed toward the next level of each type (although perhaps that's more of a "hard mode" option in general).

I like your idea that "science" as "nerd bait" though. I can definitely see science being, say, a big multiplier on R&D, or providing other bonuses to spend it on in a more general way.

Thoughts? Already suggested and shot down?

[JimmyAgent007]

Well I dont think there should be a hard set goal, like you cant get mainsails without doing task 23 - x number of times. A progress bar that you need to fill with returned science might be a an unnecessary step. Since as it is you select the tech after you gain the science, a progress bar just lets you pick the tech before you get the science.

If you mean that the progress bar is more like a contract, and you complete objectives.

ie. Say heavy landing legs needs 100 science to unlock.

Do a part test on it and reduce the cost by 25.

Each use of the earlier landing legs reduces the cost by 5 up to a max of 20.

Part broke during use, max tolerance needs to improve. reduce cost by 5

Now it only costs 50 to unlock that you probably gathered during the missions you tested them.

So doing missions that use parts make the upgraded version easier to research but still cost science ('nerd bait' + actual use) would make more sense. As long as you couldnt spam missions on the launch pad and such.

[sherkaner]

Well I dont think there should be a hard set goal, like you cant get mainsails without doing task 23 - x number of times. A progress bar that you need to fill with returned science might be a an unnecessary step. Since as it is you select the tech after you gain the science, a progress bar just lets you pick the tech before you get the science.

No, sorry maybe I didn't explain well. Those lists of things to do were just examples, not meant to be fixed lists. The idea would be very much letting the player get the R&D experience with each part in their own way. Sure, you could grind it with lots of the same mission (with diminishing returns), but coming up with with different combinations of circumstances will get you there faster and so there would be lots of ways to R&D-up a part.

Think of this R&D experience system kind of like the science system, but with the "resource" gathered automatically (rather than with instruments) and only for part types that are actually on the craft at the time.

Then separate from that, science would be the same generalized bucket it is right now, and correspondingly would be used in more generalized ways to accelerate overall R&D or other things, rather than being able to directly buy parts access.

[sherkaner]

The more I think about it, Jimmy, I think your suggestion is actually better than mine: using parts of a certain type would simply reduce the cost to get to the next level of that type, but you would still use science points to do the unlock. It's a less revolutionary change than what I'm talking about, it provides even more options for the player, and it still "feels" more like R&D -- you can throw all your resources at getting a part directly in the lab, or you can take a more measured approach of testing before making the leap. Although I still envision the test-driven "discount" on a part level to not be fixed testing missions, but the more open-ended usage approach that I mentioned before. I'd really hate career contracts to become discount grinds, but I do like the idea that if you're using a part a bunch (for whatever purposes you prefer), it's going to make you more of an expert in that and let you get to the next level easier.

I think the per-part tech explosion tree, plus some usage-based science "discounts" on research levels, makes a really nice pairing -- sounds intuitive, open-ended, satisfying, and fun!

Edited December 3, 2014 by sherkaner

[CaptainKipard]

using parts simply reducing the cost to get to the next level of that type, but still using science to do the unlock. It's a less revolutionary change than what I'm talking about, it provides even more options for the player, and it still "feels" more like R&D

I'm totally down with this. It's not too grindy, and makes the unlocking process more immersive than simply spending points.

This would let people test parts on Kerbin the way people test rovers in deserts for example, but you would still need to travel to collect science. Looks good.

There might have to be some logic behind this system to reduce cost more if you test things offworld, and a cap for maximum tests in various environments.

Edited December 3, 2014 by Cpt. Kipard

[sherkaner]

There might have to be some logic behind this system to reduce cost more if you test things offworld, and a cap for maximum tests in various environments.

Exactly what I had in mind. Very similar to how science instruments have a concept of different locations of varying exoticness (with the game keeping track of what you've already measured), I imagine those same locations and exoticness could be directly used to scale the sort of R&D experience (and thus discount) you'd get for a certain part type. You'd get more R&D know-how from testing a decoupler in orbit than you would on the ground, and for either location you'd pretty quickly tap out what you could learn after a test or two.

Of course there might want to be some additional logic for each part type -- experience with legs would be better derived via landing, whereas experience with parachutes might have more to do with air density and g-forces. But really, even if Squad didn't feel like going that far, a basic system would be pretty neat I think.

[CaptainKipard]

Exactly what I had in mind. Very similar to how science instruments have a concept of different locations of varying exoticness (with the game keeping track of what you've already measured), I imagine those same locations and exoticness could be directly used to scale the sort of R&D experience (and thus discount) you'd get for a certain part type.

Neat. This has the added advantage of allowing the programmers to reuse code.

Of course there might want to be some additional logic for each part type -- experience with legs would be better derived via landing, whereas experience with parachutes might have more to do with air density and g-forces. But really, even if Squad didn't feel like going that far, a basic system would be pretty neat I think.

A basic system would do fine; at least something that would let modders develop it further.

[sherkaner]

Great, somebody notify Squad we've solved this one!

[JimmyAgent007]

Great, somebody notify Squad we've solved this one!

Internet Beers all around!

[ximrm]

Again, personal preference:

Blind techtree sounds nice. Halfblind.

I would like to have the option to have, let's say 5 nodes you can buy for less science, each one containing ~5 random unresearched parts. Not totally random parts, maximum two nodes ahead.

OR

A similar thing but instead of costing money, you can use spare sci-points to buy parts from a selected unlocked node. And random 3-4 parts will be bought from that technode.

I would also like a difficulty option that when you have unlocked let's say 2 nodes, but the parts you didn't buy from the first one, their entry cost gets lowered by a few percent. Could be depending on the amount of advanced nodes that have been unlocked.(~5% per node, max 30%)

Edited December 11, 2014 by ximrm

[MKI]

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.

Earlier Probe cores should appear, as they will have a progression system. I assume by the time they appear, Jeb (or who ever is getting exp) would be a better choice over StayPutnik.

But I dislike the idea of changing the game from manned flight from the start. You might ask "This isn't historical" and I say "Its the Kerbal Way!"

For the most part Kerbals are as much as the game as Kerbin, or Jool or SRB spam, or the Hype train. They should go to space first, because Iron Man mode or normal mode or easy mode, Jeb flying to space to be the first man in space is part of the game. Since Day 1 Jeb is the crazy fanatic, junk yard owning, Spaceman, badA$$, major tom-esque, Boss mode pilot that will be the first Kerbal in space (or to meet his end smiling before reaching space)

Adding probes to potentially take that away from him is not part of this game. To anyone who thinks differently is more a fan of space flight (which is fine) than the game itself.

Leave it to the mods for those who want to take things the "cautious" route, Kerbals are not cautious, they go to space flying on duck tape and crazy glue. Why? cuse thats the game like it or not, thats how the game has been, how it should be and probabaly how it will always be.

PS Historically Kerbals invented Spaceflight before electricity and probes. So yea, if your talking about our history, it doesn't apply to Kerbals and thus should be left to mods. ;D

[Kegereneku]

If you get access to probes sooner, you'll still have yourself the opportunity to send a manned mission.

So there is no reason to impose what is basically a "glorification of some intrinsic man/kerbal's power" to everybody else.

Some peoples are very very disappointed that manned flight is "so easy" to begin with. That is a legitimate concern that if it is "too easy" to reach orbit or do anything manned, making manned mission lose all punch. This is also why so many people also hope for Life-Support.

If the only difference between a Kerbonaut and a probes is that Kerbal aren't nerfed, what the point ?

To me your whole notion of "getting the Kerbals into orbit FIRST" is meaningless, Probes are not competitor, they are tools. They are meant to be expendable. while Kerbals aren't. A probe getting into orbit first is merely a placeholder for the arrival of Kerbal.

Then, after this first HISTORICAL manned flight, using unmanned vehicle everywhere is a tribute to Kerbals science. One might be satisfied by sending manned mission Once, Me I prefer to build the infrastructure to send Manned mission whenever it's really needed

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?

As discussed all along this thread, it would gain to be an open model, where you research parts as you need it, following a coherent technological progression rather than in surprise bundle which cost a lot even if you only wanted a structural piece.

What can be left to mods are any parts that aren't needed by the current gameplay. That simple. If KSP don't have Life-Support, then we don't need such part.

Edited December 13, 2014 by Kegereneku
minor correction

[sherkaner]

If you get access to probes sooner, you'll still have yourself the opportunity to send a manned mission.

So there is no reason to impose what is basically a "glorification of some intrinsic man/kerbal's power" to everybody else.

Exactly right. The strength of KSP is in its openness to letting the player build a space program the way they choose. The idea that the game is all about throwing caution to the wind and putting a kerbal atop your first rocket is no more or less valid as somebody else getting more of a thrill out of being methodical and measured, or somebody who wants to build a cartoon version of a real-life program, or somebody who wants an army of robotic probes gathering science across the system, or somebody who doesn't want to go to space in something that doesn't have wings.

I think all we're talking about here, beyond the mechanics of the tech tree itself, is ensuring that in career mode the player has as much choice as possible as early as possible in the game, and throughout their career -- while still having to work overcome challenges that feel authentic (I hesitate to use the word "realistic") and meaningful.

Edited December 13, 2014 by sherkaner

[NathanOwen]

I think larger Ion thruster and Xenon Tanks should be added. I have been having horrible part count issues any time I try to use them. The largest Ion thruster NASA ever tested was 1.5 meters so maybe that would be a good starting place. Of course the tanks are the more important thing at the moment as more xenon thrusters can always be added but the tanks are where the part counts get ridiculous pretty fast.

[Baythan]

I like the ideas presented here for a tech-star/web. Most of the issues I have with the tech tree have already been discussed, such as landing gear being so far into the tree, meaning I can land on the Mun and Minmus long before I can land an airplane (or even take off). So many things in the tree are arbitrarily nested with other random items.

I would really like to be able to research specific parts rather than groups of parts, or smaller groups of related parts that lead to better versions of those parts. I'm pretty sure there will be more parts added in Beta, but I disagree with waiting till they get added before changing the tech tree. I think a better tech tree using the current items would be much easier to add new parts to rather than waiting till there are even MORE parts to sort through to design a new tree.

[Tw1]

This would be better. Much more freedom.

[sherkaner]

I just want to say that with the beta release, I started my career over again and it makes me want the tech tree system we've been discussing here so much more even than before. The current tree feels so arbitrary and constraining -- and not just in the way of making the game "hard" (I love the difficulty of the new moderate/hard modes), but in that I feel so limited in how I can go about things. Especially with the new Fine Print contracts, I keep wanting to build small simple aircraft to do flyovers, or try out clever ways of meeting contract requirements on a tight science/cash budget. But I'm constantly feeling like the tech tree virtually forces me to take a very plodding tech tree progression of filling out the lower limbs before going any further because there's some silly utility part that I need.

Hopefully there isn't a belief that "bundling" tech tree parts as the game currently does is a game balance method. I think the per-part tech tree we've been discussing here could very easily be balanced with various costs while still giving the player the full freedom to develop technology in the direction of their choosing.

Edited December 18, 2014 by sherkaner

[CaptainKipard]

I'm just started playing KSP after the release for the first time since 0.24.2, and I absolutely agree with everything you said there. I'm pretty much forced to research every tier and every node just to quickly get to the batteries and solar power. Half the parts in those nodes are left unpurchased.

I like the difficulty from the pacing of the building upgrades and XP, but not from the tech tree.

I think what Squad wanted to do with the bundling (and the tree in general) is help players progress, but it just forces you to play how they want you to play. It has to go.

Now that we have progression coming from other mechanics (buildings, xp), the tree has absolutely no reason to stay the way it is.

Sorry for the bold. I feel this is an important point.

Edited December 18, 2014 by Cpt. Kipard

[Baythan]

I'm just started playing KSP after the release for the first time since 0.24.2, and I absolutely agree with everything you said there. I'm pretty much forced to research every tier and every node just to quickly get to the batteries and solar power. Half the parts in those nodes are left unpurchased.

I like the difficulty from the pacing of the building upgrades and XP, but not from the tech tree.

I think what Squad wanted to do with the bundling (and the tree in general) is help players progress, but it just forces you to play how they want you to play. It has to go.

Now that we have progression coming from other mechanics (buildings, xp), the tree has absolutely no reason to stay the way it is.

Sorry for the bold. I feel this is an important point.

Agreed! Now how do we present this idea to Squad in such a way that we can convince them a large number of people want this kind of change ASAP?

I'm almost certain they understand that the current tech tree just won't work in the long run, but it may be a low-priority item for them and we may not see anything done for quite some time. As it stands, I believe the changes to the buildings combined with the current tech tree make Career Mode nearly impossible, or at the very least VERY grindy for several hours in the beginning.

I'm looking for a modded tech tree right now so I can at least build some small planes early on to try to accomplish the survey missions without spending all my money on rocket fuel. I consider Stage Recovery to be a necessity for Career mode now because the cost of lost stages makes many contracts worth nothing. Gaining enough funds to upgrade the VAB and launchpad (to get over the 30-part and 18t limits) even with that mod STILL takes many hours(I haven't accomplished this yet) for me without the ability to do any 'fun' missions.

Maybe I need to watch some more advanced players do Career mode, because I can barely get to orbit and return with fewer than 30 parts.

[CaptainKipard]

Agreed! Now how do we present this idea to Squad in such a way that we can convince them a large number of people want this kind of change ASAP?

I don't know, but now that KSP is in beta, Squad will be revisiting everything they've done and improving it. If it's going to happen it'll happen between now and 1.0.

I'm looking for a modded tech tree right now so I can at least build some small planes early on to try to accomplish the survey missions without spending all my money on rocket fuel.

There's nothing like what I drew, but there's Engineering Based Tech Tree which is the closest thing.

[klgraham1013]

KSP needs to be about player choice, even in career mode. All we can do is keep harping about the tech tree and hope someone listens.

Alternate Tree Configurator worked pretty well in 0.25. I hate to rely on a mod for something that should be done in stock, but if worse comes to worse...

[Ziff]

KSP needs to be about player choice, even in career mode. All we can do is keep harping about the tech tree and hope someone listens.

Alternate Tree Configurator worked pretty well in 0.25. I hate to rely on a mod for something that should be done in stock, but if worse comes to worse...

Shuffling parts in the tech tree around is pretty easy compared to implementing all the stuff they've wanted to do lately. Don't worry it will be balanced eventually , it's probably not a priority because it works well for the moment.

[klgraham1013]

it's probably not a priority because it works well for the moment.

Sure, you can unlock nodes. If that's your definition of working well, then, yes, it does that. If your definition is adding to the game experience, I'd suggest it does the opposite of working well.

[Ziff]

Sure, you can unlock nodes. If that's your definition of working well, then, yes, it does that. If your definition is adding to the game experience, I'd suggest it does the opposite of working well.

It does add to the game experience, just not in the way you want it to. It does force players to play a certain way, starting small and getting bigger. There is definitely room for major improvement in the tree, but that doesn't make it a priority. I would rather see drag physics modeled better than they are before they fix the tech tree, because at the moment the tech tree works bug drag physics are just silly.