Sign in to follow this  
Xavven

Career Mode Progression and Game Design Analysis

Recommended Posts

Here are some of my version 1.0 observations and suggestions with regards to Career mode on "Hard", default "hard" settings. My conclusion is that there are a few minor tweaks that can be made that will have a large positive effect on the progressive challenge of career mode, and smooth out some of the difficulty spikes. I will say that I should probably try this out on "Normal" difficulty to compare and see if I encounter the same difficulty spikes. Counterbalancing that is the fact that I'm already an expert player, so I may not see difficulty spikes on the lower difficulties whereas a new player might. Not sure what to say other than to compare notes with the community (you guys!)

Good early game

The first 30 minutes of gameplay seems to guide the player well, and gives the player enough money to be able to experiment and fail. This is because of the huge money influx from all the World First altitude, distance, and speed records. The Flea booster might need its thrust toned down a bit, though. I launch with 20% thrust to achieve around a 1.5 TWR on my first couple of crafts. 100% thrust yields 11 g of acceleration (too much for a beginner's first launch).

It is reasonably easy to unlock the set of 15-20 science nodes of General Rocketry, Stability, and Survivability. Game progression up to this point seems well thought out, and smooth. Once the aforementioned nodes were unlocked, however, I encountered the first difficulty spike.

The first difficulty spike

The next logical step at this point is to achieve a stable orbit, but this is best achieved with the Terrier engine in your upper stage, which is unfortunately locked away in a 45 science tier.

During beta, the old tech tree had the LV-909 in the 15 science "Survivability" node, which facilitated the player making orbit at this stage of the career progression. Instead, I found I had to use 6 FL-T200 fuel tanks on a Swivel engine as the upper stage, which is not efficient at all when the player has such a light payload in early game (usually just a Mk1 Command Pod and a parachute). While I could make it work (as an experienced player), it was a harder design challenge during a point in the game where you want the challenges to be gentle on a new player. Admittedly, I am playing this on "Hard" so what should I expect? Duh, right? I will say however that the old tech tree had the LV-909 in the right place whereas I don't believe the new tree does.

This is also about the time where tourism contracts are appearing, but the Mk1 Inline Cockpit is also sealed away in a 45 science node. Stacking two Mk1 Command Pods on top of one another just looks completely silly.

This is also about the time where Kerbin survey contracts are appearing, but the Aviation node is sealed away at 45 science, too. What's worse, some of these contracts want surface reports, which requires precision landing. Planes are harder to design than rockets, and a new player probably cannot precision land an early plane on rough terrain. Not good contracts for a beginner to take.

No, at this point in the progression, the game should probably be nudging a newbie player in the direction of getting comfortable with making orbit and getting a Mun flyby down, followed by a Mun orbit. This is where the second problem arises:

The Mun difficulty spike

There are three, maybe four key tools that a newbie needs to achieve an early-game Mun flyby successfully:

  1. The Terrier engine in the upper stage
  2. One launchpad upgrade
  3. Patched conics
  4. Maneuver nodes

Unfortunately, the player does not quite have all of these things by the time the Mun is the next target. Ironically, it is the expert player that is able to overcome an early Mun mission because s/he has memorized the correct phase angle to begin burning prograde, already knows how burning prograde will affect the shape of the orbit, and can eyeball a Mun intercept without patched conics. AND the expert player further has memorized when to burn retrograde for Mun orbital insertion, and when to burn prograde again to leave Mun's SOI retrograde with respect to Kerbin in order to successfully return. This, at a time when a newbie player needs to learn these valuable skills. As an expert player, I managed to handle this Mun flyby, but I immediately recognized the problem with this game design:

When you teach someone to ride a bike, you put the training wheels on first, and then take them off once the learner has gained sufficient skill. I have found KSP to be doing this backwards -- without patched conics and maneuver nodes it's like trying to learn how to ride a regular bike with no hands first, and once you have mastered that you get the handlebars and training wheels.

No, the building upgrades should instead act as GATES that the player must unlock before seeking greater challenges, and those gates must be timed with the player's progression. Here's what I suggest: patched conics and maneuver nodes should be immediately available in the tier 0 buildings, but only work in the Kerbin system. In order to get patched conics in the neighboring systems of Eve and Duna, the tracking station and mission control must be upgraded. There's even a good excuse to explain this: the tier 0 radar dishes on the tracking station aren't sensitive enough to work at long ranges, and upgrading them gets you patched conics farther out in the system.

Tech tree logical progression suggestions

Parts seem to be a little bit too scattered at the moment. For example, the Rockomax Brand Decoupler (2.5m decoupler) is under "General Construction", the Poodle and Skipper are in "Heavy Rocketry" without any 2.5m fuel tanks at all, and the 2.5m fuel tanks are in Fuel Systems. The problem with this is that parts from three different nodes all must be used in concert, or else none of them are individually useful at all.

Instead, I suggest thinking about what the player might be trying to achieve and grouping those parts together. For example, "the player's objective is to begin using 2.5m rockets" so the first node contains the Skipper, a 2.5m fuel tank, a 2.5m decoupler, and a 1.25m -> 2.5m adapter. The Poodle can be locked away in the next node, and the Mainsail in the next. Actually, Squad did a great job with the Aviation node in following this line of thinking, as well as the resource scanning and ISRU progression at the end of the tree.

Proper progression is most important in the early game when the player's options are limited, however. You can get away with some poor progression later on in the tech tree because by then, the player has plenty of different contracts to choose from and complete. Early to mid game, though? Whew... No, this is where it needs to be MOST polished.

Another example of poor tech tree placement is the Micro Landing Gear in the Survivability node. At that stage of the career, all of the engines the player has are physically too long, and the micro landing gear doesn't reach past the bell nozzle! I'm better off using makeshift girder landing legs, but this is not something a new player would be expected to try. Instead, they are given a counter-intuitive part.

Another example is the science lab being grouped with extendable ladders. These are in no way related from a player intent point of view.

Okay, enough tech tree bashing. Let's get back to contracts.

Rescue contracts

In order to pull off a successful Rescue from Orbit contract, the player needs:

  1. Maneuver nodes
  2. A lot of delta-v to execute rendezvous, which requires...
  3. An upgraded VAB

The first rescue contract I accepted was in a high orbit (almost at Mun level) and was inclined. This meant I needed not only 3,500 delta-v to get to LKO, but also another 2,400 to correct the orbital plane and execute the rendezvous (I tried my best to launch into the right plane from the get go but still was off by 5 degrees) . Now I'm no slouch -- I've got a manned Eve return and Jool-5 under my belt, but I honestly can't eyeball an orbital rendezvous with a ship without maneuver nodes.

In addition, early game is tough because you have limited parts to work with. If I need more delta-v in early game, I can't just bring out the big guns (Skipper to Mainsail to Rhino (KR-2L), etc.) like I usually do. Instead, I have to use clusters of Reliants (LV-T30). This means more engines, more tanks, more radial decouplers, more nosecones, and struts to keep the radial bits stable. I quickly run into the VAB part limit.

My suggestion is to wait until later to start offering rescue contracts. Perhaps the player also needs a cheaper VAB upgrade to bump the part count restriction up a little, as the first upgrade is both too expensive and too much of a step in allowed size and part count -- we need an intermediate step between them.

Okay, this is already getting too long, so I'll stop here. What do you think? Have you noticed anything similar? Or do you think I'm just on "Hard" and should shaddap about it? :)

Share this post


Link to post
Share on other sites

I think you've got your finger more or less on the pulse of the problem, actually. Quite a lot of the problem with tech trees in general is that they are designed with the most challenging components available earliest, with the parts that you can actually learn the game's systems on hidden away a few nodes off (or at worst, at the dead end of the tech tree). It's a classic shortcoming of game design: focusing too much on the numbers and not enough on the balance between skill required versus rewarding pay-off. A proper tech tree provides options that are completely reliable at the early stages, but give somewhat sub-par performance. The nodes that follow give better-performing items that require more skill to get that performance. The thing is, the Terrier is an entry-level engine for what it does: it's a low-thrust-high-ISP engine with good maneuvering that's an obvious choice for an upper stage engine. Putting it on the fourth tier completely misses this fact.

Share this post


Link to post
Share on other sites

What do you think? Have you noticed anything similar? Or do you think I'm just on "Hard" and should shaddap about it? :)

Hard vs. Easy has nothing to do with the progression of the game, so I think you're on target with the reasons for your examination. If you were whining about the amount of science you're getting, funds pressure, or something like that, it'd be a different story. There's something to be said that the LV-909 in the 45 tier and lowered science returns possibly interfering with expectations, but that's a detailed examination of exact values to determine.

I agree that the gate for conics in particular is a poor choice. Asteroids? Absolutely. Conics and Maneuver Nodes? Not so much. I like the idea of 'range' being the Tracking Center's gate. You could also gate the Tracking Center with range for usage of probe cores, though I'd personally rather see that done via antenna on the probe core itself.

Mission Control is also tied into the problem. You need to upgrade both mission control AND the Tracking Station for Maneuver Nodes. This is definitely something a newer player needs. While I know that if I burn out to roughly the right orbit after eyeballing the AN/DN and making a burn there to save dV for my Kerbal rendezvous, I'd agree a newer player wouldn't know this.

In the end, though, you're right. They're gating the tools that make it much easier to learn to play without the tools.

I haven't had the chance tonight to dig deeply into the tech tree organization, but this post made me go poke around a while. I had no idea they pushed the Micro up that far. Those are probe landing legs, usually used with the 48-7S, and would barely reach past the 909. My guess is they did it because they break the easiest, but you're absolutely correct. This is a poorly played out organization of the tree. The only thought I have about why you would use the micro's at that stage was to give your capsule some legs after it parachuted into Kerbin.

Of particular annoyance is starting with the modular girder, and not getting struts until the 45 tier as well. There are little to no uses for those girders in the very early game unless you're very experienced and playing with experimental crash cages. Of truly particular curiousity is locking the FLT-800 up along side *3* rockomax tanks and the fuel hose.

Looking this process over, I'm more and more convinced you're not meant for a Munshot until you're in the 90 tier of science. That's when you get your first solar panels... and wow you'll be avoiding tons of satellite spam until you get them (unless they fixed that little bug of the Probe Core opening it before you have everything you need to successfully complete it).

For the Tech Tree, I'm sure a thread will open up shortly on it for a complete overhaul of the config file. It's completely user editable now, in theory, so hopefully someone can rebuild it more intuitively.

Share this post


Link to post
Share on other sites
Okay, this is already getting too long, so I'll stop here. What do you think? Have you noticed anything similar? Or do you think I'm just on "Hard" and should shaddap about it? :)

No, I've been playing a career game on "normal" and I found exactly the same problems. On normal, however, all the contract money you get essentially for free allows you to upgrade the launch pad, mission control, and tracking station relatively easily. On hard, that's probably a bigger issue.

I think you are right -- the tech tree should be designed around paths to achieve goals (space planes ignored here):

1) launch a rocket, learn to fly and survive

2) achieve orbit

3) mun/minmus orbit <OR> satellites <OR> rescue/tourism

4) mun/minmus landing <OR> interplanetary probes <OR> docking/stations

5) follow the paths in 3 and 4 you didn't follow

6) Duna

7) other planets

I think the tech tree should be organized so that you can gain the tech needed for each goal in a directed manner. The player should be able to strategically choose to follow the paths for different routes that the player chooses to explore. But instead, the tech tree seems to randomly spam all the tech around so instead of working toward a strategic goal, it's just a race to get as much science as you can because you need to open up almost all the tree in order to do anything. (And then once you do, you can do everything.)

Share this post


Link to post
Share on other sites

The OP is totally correct on all accounts. The tech tree is an absolute mess. You need to unlock entire tiers in many cases before most things in the tier become useful. Other parts are absolutely invaluable but locked away in tiers that are so high that taking them limit what else you can get. All of that is damaging to an inexperienced player, and challenging to experienced ones.. but what's really annoying are the contracts. In .90 my early game was mostly spent flying around Kerbin doing survey missions with basic jets (so basic they're launched using mini solid boosters as a catapult because landing gear is several tiers out of reach). Now, those contracts are extremely difficult for even experienced players because jet engines are too weak to complete them.. and their science rewards are next to nothing (they were cut down to what, 1/20? 1/40 of what they were?). You're also left with tons of "go test this" part, often in places you just plain can't get to.. like on a kerbin escape.. when you're still working on making a Mun flyby.

Early contracts should be tutorials. They should guide the player in basic operations, not throw them into the deep end with a bag of bricks tied to their leg and ask them to build a submarine while they hold their breath.

Share this post


Link to post
Share on other sites

I might add, stacking M1 cockpits does not only look silly but is game-mechanic wise the most lightweight setup for multiple crews. Maybe they should change around mass values so that 2- and 3-manned cockpits are actually more efficient, mass-wise?

Share this post


Link to post
Share on other sites

It is odd that the harder stuff is earlier on...

The 30 part limit is a brick wall I have run into. The VAB looks spacious and a mere 30 parts looks small comparatively. So... It should be much more. Something like 60, or more. So... The 30 parts should be earlier on.

Share this post


Link to post
Share on other sites

Sounds like you've manage to pinpoint a lot of the problems with the career mode and tech tree. Personally I've been struggeling to like the current career mod. First I thought that it is mainly due to the contracts being kind of all the same and the lack of depth. But after reading your post I kind of think that the tech tree is a bigger problem. If I had a tech tree which made any sense I would probably feel better doing the contracts as I would be able to use proper parts for my rockets. I don't have any problems fulfilling most of the contracts now as I've been playing the game pretty much, but the tech tree just feels wrong.

Share this post


Link to post
Share on other sites

I didn't find achieving LKO all that difficult, actually - I managed it with boosters (strap 'em to the rocket, adjust their thrust to around 1.5-2.0 to minimize drag loss inefficiency and launch away). As for tourists, stacking the command pods seems silly but is the only reasonable solution. At some point I had 5 command pods stacked up so i could take 4 tourists to LKO and back.

First thing I upgraded was the launchpad so I could launch these silly tall rockets. As soon as possible, as wobbly things worry me greatly (it seems the tier 1 launchpad has some ground clipping, one of landing legs went right through ground).

As for science, I only have one word: experiments. Do them. A lot. I made a low tech biome hopper for this purpose alone (after exhausting the LKO and HKO science). This let me unlock several 45 nodes. Perhaps this needs to be emphasized in the game, as an inexperienced player may not be aware about this. Perhaps contracts instructing the player to visit certain biomes?

Share this post


Link to post
Share on other sites

Kerbal Engineer Redux has a hard time showing me the required deltaV of 4500+ m/s for achieving orbit. Some guy said he could do orbit with 3200+ m/s? Anyways, my trusted 0.9 setup based on the "Hammer" booster (staged 2+1, 1, 1) does show less deltaV than in 0.9 and does not achieve orbit anymore. Or is it just me having to adapt how to pilot that thing?

Share this post


Link to post
Share on other sites

I agree with the OP on all accounts. What struck me the most was the 2m engines not coming with any 2m tanks, and the light legs coming before the medium ones. They cannot reach below the engines you most likely have at this point. Also, I do not unserstand why the Round-8 is in the same node as ion propulsion.

I don't think a lot of the parts have to move around to fix the tech tree, just some.

I'm sure Squad already have a hotfix patch 1.0.1 coming. (Probably not 1.1 because the minor version is what their major patches will be all about.)

Share this post


Link to post
Share on other sites

Yes delta-V needed for orbit is much less with new aerodynamics.

Very good post on tech tree + contract issues. A few that I noticed:-

* Survey contracts are quite grindy, particularly on other worlds such as Mun requiring quite a bit of delta-V to perform all the landings, fly at altitude and plane changes.

* Be nice to get the 1.25m tanks in order of FLT400, FLT200, FLT100, FLT800 ie swapping the tiny tank FLT100 with the medium FLT400 one. The tiny one really limits early rocket designs with the part limit and doesn't really fit with the early T30 and T45 engines. The philosophy with engines has been start medium and then the small and large are engines are higher tech that unlock later, so not sure why same idea is no longer happening with 1.25 m tanks

* I agree on the LV909, needs to be a tier or two earlier, could possibly move it back to suvivability or swap it with the BACC booster in General Rocketry

* Could do with the stabilising AVT1 fin a tier earlier to help with stability.

* Tricoupler and Rockomax Decoupler + adapter have swapped nodes compared to 0.9 and I can't understand why. The tricoupler is useful for building bigger 1.25m boosters but now unlocks when you already have 2.5m tanks so it is not likely to get used in career, quadcoupler is even worse

* Fairings are quite late in the tree and in tech tree nodes otherwise containing only plane parts. They are more useful for a rocket player. I guess there could be arguments to force a rocket player to not just unlock rocket parts only nodes, but I personally prefer to choose my play style by what I unlock.

* The current tree and career doesn't seem to reward the plane player as survey contracts often have targets at too high an altitude and the plane parts don't really unlock before you are ready to go to the Mun.

Share this post


Link to post
Share on other sites
Also, I do not unserstand why the Round-8 is in the same node as ion propulsion.

Apparently we saved the ROUND-8 (against being repurposed as a xenon tank), but they forgot to move it? o.o

Share this post


Link to post
Share on other sites

+Rep for the constructive feedback.

If I could offer a slightly different perspective on the tech tree? I agree that there are a couple of odd placements (the micro-landing legs being one) but on the whole, I'm finding that I'm thinking a lot more about my tech progression now and how that ties in with the contracts I'm taking and the next few mission types I want to do. I would say this is a good thing on the whole.

As an experienced (although I wouldn't say expert) player, I didn't really notice that first difficulty spike but right now I'm seriously pondering what to do. Go for Fuel Systems and the necessary building upgrades to do crewed Munshots? Go for a cheaper but tech heavy approach, pick up Miniaturisation, Electrics and whatever that small engine tech is, and go for some early probes (and pick up my Explore Mun/Minmus contracts that way). Or focus on LKO operations, research-over-time and space tourism by beelining for Command Pods, then unlocking the Science Lab and Hitchhiker parts?

It's all trade-offs and figuring out how to balance science vs funds, which I'm actually quite enjoying.

What might be useful is for another early 'get some science' contract to go after the 'take a crew report from the launchpad' one. Something to point new players in the direction of adding a Goo pod or Materials Bay to their rockets perhaps. You can get an awful lot of early science that way which should make it easier to pick up a couple of 45 point techs, as well as being generally useful for climbing the rest of the tech tree.

Share this post


Link to post
Share on other sites
* Survey contracts are quite grindy, particularly on other worlds such as Mun requiring quite a bit of delta-V to perform all the landings, fly at altitude and plane changes.

Actually, with a polar orbit, you won't need much delta-V for plane changes. Switching back to KSC so that you can warp faster than 50x could get tedious though.

Share this post


Link to post
Share on other sites

Maybe I won't need to do a blog post and I'll just add to this thread. The OP was great. I currently just made my first orbit. I'll give my opinions when I'm further in.

Share this post


Link to post
Share on other sites

Career mode is unchanged from 0.90 in any meaningful way. The tech tree paradigm was broken to start with, and reordering it just makes it differently broken. My 3d launch was a poor launch profile from orbit, corrected with the 4th (orbital) flight. I've unlocked a few of the 45s.

In general, KSP is designed so that it is most difficult at the beginning and gets ever easier.

Share this post


Link to post
Share on other sites

The OP hit the nail on the head. There are design problems with the tech tree. Early contracts do no correlate well with available research. Survey contracts without proper plane parts are a pain. Personally I find solar panels too deep in the tech tree, especially if one consider using the LV-909 as mun transfer / landing stage, as it doesn't generate electricity. The sudden spike in difficulty after escaping atmosphere could frustrate new players a lot, especially due to the lack of patched conics. You need a lot of early upgrades to be able to build a proper rocket - launch pad, tracking station, VAB and mission control. The sense of progression on career could be better..

Share this post


Link to post
Share on other sites

I built my first plane and realized I don't have a ladder to EVA my Kerbal. How far am I from ladders? What? Why are they over there!?

Share this post


Link to post
Share on other sites
This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this