The tech tree progression is ridiculous

[Hodari]

1 minute ago, Veeltch said:

Sorry but I just can't stand the "tutorial mode" argument.

This is FAR from just a "tutorial" though.  That would imply something that you go through in a short time to learn the very bare minimum you need in order to play the game(which is what the actual tutorials in the game at least attempt to do).  Career mode is something that will easily take weeks or even months for a new player to get through and is something that you can play through multiple times before you've really exhausted all of the possibilities of it, learned how to do everything efficiently, and really start feeling the need for something more.

[Wjolcz]

2 minutes ago, Hodari said:

This is FAR from just a "tutorial" though.  That would imply something that you go through in a short time to learn the very bare minimum you need in order to play the game(which is what the actual tutorials in the game at least attempt to do).  Career mode is something that will easily take weeks or even months for a new player to get through and is something that you can play through multiple times before you've really exhausted all of the possibilities of it, learned how to do everything efficiently, and really start feeling the need for something more.

Oh, so it's not a tutorial because it takes longer to go through than a standard tutorial in other games. Makes sense.

[Hodari]

1 minute ago, Veeltch said:

Oh, so it's not a tutorial because it takes longer to go through than a standard tutorial in other games. Makes sense.

*facepalm*

[Pthigrivi]

It is very possible for a career mode to be both a sensible introduction to parts and concepts and still be challenging and fun for veterans. Both are in fact necessary.

I don't mind the idea of tidying up the tech tree but we'd have to recognize that linearizing things that way creates issues as well. Nukes for instance appear really far back in the Historical Progression tree. This might make sense balance-wise but ends up locking out a part thats really vital to interplanetary missions for much longer than in Stock. You've also precluded the ability to navigate and skip nodes you might not need right away. 

Edited August 8, 2016 by Pthigrivi

[Wjolcz]

3 hours ago, Pthigrivi said:

It is very possible for a career mode to be both a sensible introduction to parts and concepts and still be challenging and fun for veterans. Both are in fact necessary.

I don't mind the idea of tidying up the tech tree but we'd have to recognize that linearizing things that way creates issues as well. Nukes for instance appear really far back in the Historical Progression tree. This might make sense balance-wise but ends up locking out a part thats really vital to interplanetary missions for much longer than in Stock. You've also precluded the ability to navigate and skip nodes you might not need right away. 

NTR isn't needed for interplanetary missions, unless you are building something big. And I think skipping some nodes is actually a good thing, though not necessary.

I think a tree that has branches that are themed (probes with probes, pods with pods, electric with electric, engines with engines, wings with wings, etc) is a very good solution. Each part could also have its own node (not the wings though, there are too many of them). That would allow everyone to pick their own playstyle.

I would love to start an atmospheric SSTO-only career challenge (and I think it SHOULD be possible). I believe such a tree would allow for that (unless it would be somehow stupidly prevented by the devs). Though I guess it's also the case of how the tech research is handled. Unless that's changed there's no way to go fully reusable air-breathing SSTO career.

Edited August 8, 2016 by Veeltch

[John FX]

What I would like to see is contracts that are only possible once you have unlocked the entire tree, in a lot of games where a collection of parts/units/tech is unlocked in the course of a campaign/career most of the stages making up the game part happen after all the parts are unlocked. KSP is sorely missing an endgame in career (except sandbox of course).

A handful of planets much further out that require advanced techniques to reach and which only appear on the map after the entire tree is unlocked would improve the game greatly IMHO. This could be fitted in if one of the last parts is an orbital telescope which is required for the bodies to appear on the map.

[Superfluous J]

9 minutes ago, John FX said:

A handful of planets much further out that require advanced techniques to reach and which only appear on the map after the entire tree is unlocked would improve the game greatly IMHO.

The problem here is that planets further out don't require any advanced techniques. In fact, I personally think Moho is much harder to reach than any other world, and it's not really that hard.

Putting a planet further out only increases the time you can take making a sandwich while time warping. Nothing more.

Hiding them behind "unlock the whole tree and suddenly see them" is... well it would work I guess. But considering we found all the planets in our system over 100 years before we launched a single thing into space, I'm leery to accept it.

[tater]

1 hour ago, Pthigrivi said:

It is very possible for a career mode to be both a sensible introduction to parts and concepts and still be challenging and fun for veterans. Both are in fact necessary.

I don't mind the idea of tidying up the tech tree but we'd have to recognize that linearizing things that way creates issues as well. Nukes for instance appear really far back in the Historical Progression tree. This might make sense balance-wise but ends up locking out a part thats really vital to interplanetary missions for much longer than in Stock. You've also precluded the ability to navigate and skip nodes you might not need right away. 

I'm not familiar with that tree. By "really far back" do you mean early in the tree, cause it sounds like you mean later in the tree. Odd that this would be the case in a "historical" tree since the NTR existed before the first manned spaceflights ever happened in RL. Honestly, using reality as a benchmark, most everything in the current tech tree should be tier 0.

I'd have a much wider, flatter tree. In a perfect world, parts could then improve. LT-30X, LT-30 mk1, LT-30 mk2, and so forth. I'd have the experimental versions (marked X) actually have a failure chance. Improved versions could have slight improvements, weight reduction, etc. There might be vacuum variants of some engines (higher I_sp, lower thrust). Having a progression like this would commit the player to certain technologies, then maybe they find it cost-effective to stick with it, instead of starting with something else. Note that for engines, a stock ability to cluster engines would be awesome to have. So you've invested in 1.25m engines... just cluster them. Bam, we just had someone replicate the R-7 vs US launchers because they decided to stick with their LT-45 mk2 engines, just use more of them.

4 minutes ago, 5thHorseman said:

The problem here is that planets further out don't require any advanced techniques. In fact, I personally think Moho is much harder to reach than any other world, and it's not really that hard.

Putting a planet further out only increases the time you can take making a sandwich while time warping. Nothing more.

No it doesn't, it also increases the dv required, or do you think it takes the same craft to get to Eeloo as the Mun?

[John FX]

2 hours ago, 5thHorseman said:

The problem here is that planets further out don't require any advanced techniques. In fact, I personally think Moho is much harder to reach than any other world, and it's not really that hard.

Putting a planet further out only increases the time you can take making a sandwich while time warping. Nothing more.

Hiding them behind "unlock the whole tree and suddenly see them" is... well it would work I guess. But considering we found all the planets in our system over 100 years before we launched a single thing into space, I'm leery to accept it.

Well TBH I was just getting on the `KSP history does not have to follow ours, in fact it`s better if it does not` bandwagon and thinking in terms of gameplay.

I would prefer a fully realistic history and an empty map at the start of the game and you would have to watch a patch of sky with either a large ground based telescope or an orbital one for up to a year to spot a planet as it transits your view then get closer to see all the moons unless they are large and gravitational information, rotation etc would be somewhere in the middle of discovery.

Similar to how, while we have had telescopes for hundreds of years, clear images of Pluto have only happened in the last year and we are still discovering objects in the solar system.

A planet that is far from the plane would be much harder to spot for example.

The main point is, it would be nice for there still to be things to discover for the first time after the entire tech tree was unlocked.

[tater]

You don't need to have to invent the telescope, or invent aircraft for a good KSP career. That would be terrible, IMHO. Having a realistic "fog of war" would be fine---that you would start the game knowing what humans knew about our solar system in ~1960, and would learn more as you sent appropriate probes.

That's outside the scope of this thread, however. 

Kerbals need not have the same, exact history, but given an attempt to make KSP physics sort of realistic, the relative difficulty of coming up with spacecraft systems would also scale to reality.

[Superfluous J]

6 hours ago, tater said:

No it doesn't, it also increases the dv required, or do you think it takes the same craft to get to Eeloo as the Mun?

The EXACT same craft? No. Extremely similar craft, to the point where it doesn't really give me any new gameplay? Yes. And comparing Eeloo to Mun isn't the best as Eeloo's in the game. I'm not arguing to remove Eeloo, I'm arguing that adding more planets farther out would be too similar to it.

Jool gives new gameplay, because it has 5 (6 if you count Jool itself) destinations that are tantalizingly close to each other yet far from Kerbin, making a "Jool 5" mission an attractive idea that takes quite a bit of planning and engineering. It also allows for interesting interplay between its moons that can save (or cost) literally thousands of m/s to navigate. That's gameplay. Aptur in the New Horizons mod was a tiny planetoid that co-orbited a gas giant with - and around - its large neighbor (which happened to be Kerbin), making it the closest thing to Kerbin that I found extremely hard to get to. That's new gameplay. A half dozen planets that are all roughly the size of the half dozen planets we have now, farther out and with a moon or two each (or even 5) is not new gameplay. It's 6 places you'll go to once or twice and get bored with just like people are bored with Dres and Eeloo.

[tater]

Forget the planets, keep the same ones. The point is that there is a HUGE difference in gameplay choices even without LS for a 6.4X system vs stock. Or a 3.2x planets with 6.4x distances. Your biggest stock craft for a given job won't even come close to cutting it. The Mun takes over 5000 m/s from LKO. That's almost as much as to land on Tylo, right? No more 1.25m rockets to the Mun with crew as your 4th launch. Throw in LS and it becomes a really substantial gameplay change.

[Superfluous J]

@tater We're getting way outside the topic so I'm going to drop my side of the argument. I think we just simply find different parts of the game engaging.

[tater]

 @5thHorseman, sorry, honestly, I conflated the OPM stuff with another thread about difficulty levels, and rescaling things upwards certainly changes difficulty, which is the same as more distant worlds (more dv required, hence larger ships). I was replying in effect to your statement as if it was in the other tab I had open. My bad. 

This is I suppose actually on topic WRT the end of the tech tree, and the fact that KSP get easier as you play, not harder. So we have a game that only has any chance of failure at the very beginning, and actually becomes easier with tech improvements, and those tech improvements are literally all completed before you even have a good Duna launch window.

Unsure what could eb done with tech in this regard. Perhaps certain tech requires more specific use requirements for testing? Instead of unlocking the seismometer and using it anywhere, perhaps it doesn't work generally until after you complete a survey mission---the sites are not chosen at random, the idea is that they made the thing, and it needs to be tested on these 3 spots on the Mun to make sure it's working right. Do that mission, and it's unlocked. The grav detector might be provisional and require testing above Jool...

Dunno, it's tricky.

[ZooNamedGames]

Without going head first into the debate, I want to say this; I do feel the stock tech tree could be better designed, I do not agree in that it should be unmanned only. I feel it should have the option of both from the beginning. You should have both the Stayputnik and the standard single Kerbal capsule.

Granted, I will say, in my most recent career playthrough, that starting with the external command seat as the only manned option created some creative solutions  . Personally, I'd vote for it as the first (with Take Command as stock as well).

[tater]

The idea that manned capsules are somehow controlled at all is sort of funny. It's not like the astronauts/cosmonauts flew their craft off the pad, they were along for the ride.

[asmi]

On 07/08/2016 at 4:04 PM, tater said:

It's also part of the problem that "science" in KSP is almost entirely planetary science, and planetary science has exactly nothing to do with building spacecraft.

It's not that I disagree in general, but sometimes it does help with spacecraft design. For example check out Soviet Venera series of probes - first landers in series didn't even reach surface because there was no information about the kind of conditions they will encounter, and their "best guesses" (quite pessimistic at the time - they were built to withstand pressure up to 2.5 MPa - about 25 times normal atmospheric pressure) turned out to be totally wrong, and so subsequent probes were massively overbuilt to withstand ridiculous pressure only so they would be able to tell the story so to speak and transmit they read figures, after that following landers were built to be just tough enough to survive what at that point were known conditions for a little while. Same goes for other aspects - first landers had "sugar locks" for water landing as it was not known at the time that there is no liquid water on a surface of Venus, this feature was removed after first landers confirmed that conditions on a surface makes presence of liquid water impossible.

Of course none of this is present in KSP, but I thougth I'd mention it anyway as it's a rare RL example when planetary science findings did actually affect spacecraft design. The only thing of that nature that IS present in KSP is presence of atmosphere and its' parameters (assuming you don't know about them in advance and use ingame atmospheric pressure to get data).

[tater]

1 minute ago, asmi said:

It's not that I disagree in general, but sometimes it does help with spacecraft design. For example check out Soviet Venera series of probes - first landers in series didn't even reach surface because there was no information about the kind of conditions they will encounter, and their "best guesses" (quite pessimistic at the time - they were built to withstand pressure up to 2.5 MPa - about 25 times normal atmospheric pressure) turned out to be totally wrong, and so subsequent probes were massively overbuilt to withstand ridiculous pressure only so they would be able to tell the story so to speak and transmit they read figures, after that following landers were built to be just tough enough to survive what at that point were known conditions for a little while. Same goes for other aspects - first landers had "sugar locks" for water landing as it was not known at the time that there is no liquid water on a surface of Venus, this feature was removed after first landers confirmed that conditions on a surface makes presence of liquid water impossible.

Of course none of this is present in KSP, but I thougth I'd mention it anyway as it's a rare RL example when planetary science findings did actually affect spacecraft design. The only thing of that nature that IS present in KSP is presence of atmosphere and its' parameters (assuming you don't know about them in advance and use ingame atmospheric pressure to get data).

I agree, completely. In more detailed posts I have said much the same. heck, one of my first posts here said that I though science should be broken into 3 types, rocket science/engineering, medical/kerbal factors, and planetary/space science. The tech tree unlocking would then use a point system, but there would be 3 different point types, and different parts unlock with different combinations. The Mk 1-2 pod might take 100 rocket science points, and 50 medical, 10 planetary (engineering dominating, but with some crew factors and save science thrown in). Lander parts might include more planetary science, as might certain science instruments.

[asmi]

Also on the subject of "Career as tutorial" - this game is a sandbox which implies replayability is the name of the game (many of us here play it an-and-off for several years already!), so average KSP player is noob only for 0.1% of total game time at best. If career is indeed intended to be just a tutorial for new players (I don't believe it is, but some here do seem to think that way), they've invested one heck of development resources into feature most players don't need after their first few hours at most. Bottom line - KSP needs a real career mode if existing one is a tutorial.

[_stilgar_]

1 hour ago, asmi said:

so average KSP player is noob only for 0.1% of total game time at best. 

This is ridiculous. Average KSP player played the game for maybe 100 h... but let's say it's 1000 h. 0.1% would meant they learned the game in the first hour. 

They must be genius. It took me 10 hours to learn how to orbit properly and come back safe, next 10h to get to the Mum and land there. 10 hours to get to other planet. Same amount for rendezvous and docking. 

So I would have to play for 50 000 hours for that 0.1% to be true. But maybe I'm just stupid.

 

But if I'm not stupid, most players are in tutorial mode for 10-50 % of their game time. 

[asmi]

6 minutes ago, _stilgar_ said:

This is ridiculous. Average KSP player played the game for maybe 100 h... but let's say it's 1000 h. 0.1% would meant they learned the game in the first hour. 

I've learned the game in zero time as I was space enthusiast before that and so knew all math behind spaceflight. And there was no career mode at the time. Oh - and no docking either (I've started in 0.17). So there was nothing to learn really except to slap things together and see them explode (if you think in 1.x craft are wobbly, I wonder what you'd say about 0.17/0.18 ) Lastly, my total playtime in KSP is significantly over 1k hours. And there are quite a bit of people like me.

0.1% maybe is a bit of overstatement, but surely devs count on players spending 100's of hours in-game, so a couple of hours-long tutorial is maybe 1%. If people spend 50% of their game time in tutorial, it means they are either incredibly stupid, or game design is very very bad, or just don't care for the game (this is often a consequence of previous point).

 

Oh, and my numbers are pulled from the same "source" yours did.

Edited August 9, 2016 by asmi

[tater]

Yeah, my own experience was similar to asmi, as before KSP the only orbital mechanics I had done was on paper as physics problems (Newtonian, Keplarian, and Hamiltonian stuff). I was on the Mun before my Guinness was done  .

[Wjolcz]

On 8.08.2016 at 1:30 PM, Pthigrivi said:

 

 

8 hours ago, tater said:

I agree, completely. In more detailed posts I have said much the same. heck, one of my first posts here said that I though science should be broken into 3 types, rocket science/engineering, medical/kerbal factors, and planetary/space science. The tech tree unlocking would then use a point system, but there would be 3 different point types, and different parts unlock with different combinations. The Mk 1-2 pod might take 100 rocket science points, and 50 medical, 10 planetary (engineering dominating, but with some crew factors and save science thrown in). Lander parts might include more planetary science, as might certain science instruments.

Sounds a bit overcomplicated. I think that could be done with experiments that are related to the next mission's objective/topic of the research (need to know what kind of conditions are present when at supersonic speeds to unlock the Whiplash engine example). The less resources there are to deal with the better. I like the idea, but science points are pretty ridiculus to deal with.

Also I would like to apologize to @Pthigrivi for bringing his attention here for no reason. The quotes are mighty broken on mobile.

[Pthigrivi]

No trouble mate  Digging this discussion in lurk mode. 

[Codraroll]

Having read the thread, may I have a ramble?

 

It seems like the tech tree serves two purposes: It is the game's reward system, and the limited part selection at the start keeps players from being overwhelmed by choice. Progressing through it can be summed up as "unlock more parts to allow you to do more fun stuff, which again allows you to unlock more parts." At least conceptually, it seems to be a decent idea.

However, it seems like the problem many people have is that the tech tree starts out identically in every playthrough, and in practise you're hampered by the same limitations every single time. I can understand and share this frustration. Having to go deep into the tech tree to find parts you need for a varied play style means you can't have that variation until certain conditions are met.

Personally, I like the spirit of the beginning of the tech tree. You have a capsule, a parachute, and a rocket engine (with its own fuel, so you don't have to worry about that for your first flight). The first flight in Career is and should be a very short one, taking off from the launchpad, gaining altitude until your fuel runs out, after which you coast along on the velocity you gained (just observing this little detail is an important and effective introduction to mechanics). After a while, you stop going up and start going down. At the end of the flight, you activate the parachute to land safely, or realise you've forgotten it and watch your rocket slam into the ground. You keep unlocking more parts until you can achieve a stable orbit, and from there you can set your sights on the other bodies in the system.

I think that conceptually, any tech tree or replacement solution should encourage a gameplay progression like that. Starting with ballistic, suborbital flights, progressing with more controllable flights higher up, and eventually letting you get into orbit. It represents the learning curve of new players, as well as that of real-life space agencies, so the basic jist of it should be kept in the game. I hope that opinion isn't too controversial.

Now, some people dislike that you start by putting Kerbals in capsules and sending them away. Manned spaceflight took a while to develop in real life, after all. On the other hand, developing remote-controlled craft before manned craft is also not quite right. And in a game about Kerbals flying spaceships, you really ought to see the Kerbals from the get go (my suggestion to address this issue - having the Kerbal portrait screen show a Kerbal in Mission Control for unmanned flights - is off-topic, but I couldn't help but mention it here).

Overall, I think it is a good idea to have multiple starting points for the tech tree. You could start manned or unmanned, and keep picking strategies from there. Want to focus on satellites or go for the Mun? Will you develop heavier lifters or vacuum engines? Build a space station or better landers? I also think a "tech web" should be considered, rather than a tree. I think I have an as-of-yet unmentioned suggestion here: have the same parts appear in multiple tech tree nodes. That means starting with the "Flea" regardless of what route you pick. Develop ladders as hand-holds for space stations or for planetary landers - or just to get back into aircraft after landing on Kerbin. Get your first probe core to assist manned flight, or have it from the beginning as an alternative to manned flight entirely. Drogue chutes to stop fighter planes or slow down probes during atmospheric descent.  Developing 2.5 m fuel tanks separately because of focused research on fuel tanks, or alongside a 2.5 m engine because of focused research on a basic 2.5 m lifter? Unlocking a part in one node would remove its cost from other nodes with the same part. Maybe you would even find some nodes on the far side of the tree unlocking themselves as you get all their parts from other research paths.

As for how to pay for new nodes... I think the Science system is decent enough, but it could be made clearer that you sell the fruits of your research, to get funding for other research projects. I'm not sure if specialized experiments for specialized parts is the way to go. Then again, boiling it all down to Funds isn't that realistic either, so I'm actually not sure either way.

Also, I think some parts could feasibly be unlocked outside the tech tree. Upgrade the Tracking Station to get access to telescopes. Return surface samples from, say, Duna, to unlock a more efficient battery. Find all Monoliths (and/or the Kraken on Bop) to unlock the Kraken Drive.

 

But yeah, the gist of this post: Have multiple ways to progress through the tech tree. Have those multiple ways sometimes unlock the same parts, albeit in a different order. Your first flight should generally feature the same mission profile, with or without a Kerbal in the pilot seat.