Champion xp/level curve: Needs flattening?

[KDR_11k]

So one of my problems with the champion is that you just never have enough upgrade points. Except that's not the whole story. You have too few as the game progresses but the first few nebulas give so many points while you have neither a lot of module options nor the ability to mount higher marks. As you progress you get quadratically more options to spend your points on (can use higher marks AND have more modules available) but you also get fewer and fewer points. As a result you pretty much need to decide from the first nebula on what modules you want to spend points on if you want to end up with a shot at actually using that mkV mounting ability on your battleship and either hope that you get the modules you want very early or float a ton of points in the hope that you will draw them later. Or you just accept that even your battleship will still run with mostly MkII - III modules (my battleship was level 30, is that normal?).

Also you better hope that you don't end up with incompatibilities, e.g. spending your early points on modules that won't fit on a Spire and later getting some awesome spire modules.

I think the extremely limited upgrade point supply is too much of a constraint. You already can't mount all the modules you want simply because you have a limited number of slots and some aren't compatible with the same champion types, the upgrade points just hurt and offer a lot of potential for messing up since you get them early but need them late. I'm not saying we need to get rid of them completely but giving a few more would be good (or make it feasible to get them from non-nebula situations). With knowledge you can use all your unlocked options concurrently, you don't have as many levels to unlock and if you really need more knowledge you can hack for it or just conquer a planet. The champ can only use a subset of its unlocked modules at the same time and outside of nebulas the XP gain may as well be zero. It's what, 20*mk per guard post? With 20k needed for a level after you've exhausted the nebulas? I understand that this is to prevent grinding but it feels like it's going too far into the other direction.

[chemical_art]

I admit I've tried different setups, but the basic module combo is good enough for almost every situation in a nebula. Due to the pressure of "min maxing" due to bonus exp I always end up with at least MK III's of these modules, which feels like railroading.

[TechSY730]

Agreed on both points.

The "experience needed to grow to next level over level" curve is too steep. I would love to see it "flattened" some, even if in return the exp needed for the earlier levels goes up some.
To be fair, many of games with RPG elements suffer similar overly high "acceleration" exp over level curves. It is quite tricky to find a good formula that works well with the balance of your game.

IMO, experience outside of nebulae needs to be not quite as low as it is now compared to exp in nebulas. I agree that it shouldn't be the primary source of exp, but at its current rate of gain outside of nebulae, past level 3 or so you can pretty much ignore it. It seems like that should happen at level 10-15 or so instead. (As long as exp gain per guard post and planet is independent of level, and exp gained inside of nebulae is dependent on level (albeit indirectly), such a point will happen. However, it should be happening later than it is now, either that, or the system for exp outside of nebulae needs a change.)

[TechSY730]

As for the modules and ship types thing, you could say that these unlocks and the availability of these unlocks form a "skill tree", with each "level" of the tree unlocked by nebula missions. You could also say that some modules have a "psuedo-dependency" on the champion hull types, even if the modules don't have dependencies on each other.

The problem is that before availability (aka, nebulae missions granting access to them), the "next" level of the "skill tree" is hidden. Making this worse, the "skill tree" is randomized, so you can't even rely on memorization for it. Worse still, the unlocks are somewhat randomized even across different instances of the same mission type at the same nebula count. (So the rewards for doing mission X at nebula N may be different in each game where mission X happens at nebula N, even though the set of rewards for mission X are constrained somewhat based on race).

Because of this, you can't really plan your unlocks when you are pressured to spend them.
To deal with this you can either hold off on substantial unlocks until you gain enough information about the "tree" (aka,  you complete a few nebulas and see what kinds of stuff were unlocked), and thus be left with a weaker champion for a good chunk of the game (note, in some lucky cases, you can get a good "path" after 1 or 2 nebulae, but as it is randomized, you may get a set of hulls and modules that don't play along very well at all until like 4-6 nebulae in).
Or you can choose only the modules that don't have dependencies on "unknowns", the base modules that can be mounted on any hull. (This one seems the most popular unless you get lucky with the unlock "rolls" early on)

I know that the "tree" is hidden because of the randomization (not even the game knows what order things will be availible at game seed), but that doesn't negate the negative effect on character progression (real or perceived).
I hate to say it, but I wonder if some degree of randomization needs to be taken out, and have some (or all) of the "unlock path" be visible and determined at game seed, possibly even constant across games.

Or possibly give a way to "respec", put all available modules back to Mk. I, but then you get all those points back. This will let you "back out" of a unlock path if the RNG didn't give you something that works well with it. (I guess it would scrap all modules on the champ, meaning this wouldn't be something you want to do mid battle. Maybe even make it cost something, like you have to be a 1000 shadow points and you lose all of them when re-specing)

[chemical_art]

Or possibly give a way to "respec", put all available modules back to Mk. I, but then you get all those points back. This will let you "back out" of a unlock path if the RNG didn't give you something that works well with it. (I guess it would scrap all modules on the champ, meaning this wouldn't be something you want to do mid battle. Maybe even make it cost something, like you have to be a 1000 shadow points and you lose all of them when re-specing)

I think respecing is the best method to prevent any sort of railroading.

I don't know you can balance having base modules that are strong enough to carry you through all the nebula (except the devourer, in which case you need a polarizer / impluse / doom module), while simultaneously having random modules that are great outside of the nebula but gimped hard in nebulae.
It is vaguely parallel to why having bonus ships be inherently stronger then non-bonus fleetships. Since you can't count on them "aside from your bonus pick" to make them worthwhile you make them a bit stronger.


The issue of "respecing" on any game is divisive though, from those who think it harms the game because it allows any build to be made without planning and others who say it avoids bad decisions and freedom or (in the case of AI War) for not being at the mercy of the RNG (because while you do get most modules eventually, you don't get them all, and the "when" is very important with the nature of xp)

[Pluto]

Or possibly give a way to "respec", put all available modules back to Mk. I, but then you get all those points back. This will let you "back out" of a unlock path if the RNG didn't give you something that works well with it. (I guess it would scrap all modules on the champ, meaning this wouldn't be something you want to do mid battle. Maybe even make it cost something, like you have to be a 1000 shadow points and you lose all of them when re-specing)

I like this, but I honestly believe it should cost a significant time investment.  Destroys all modules on the ship and takes 2-3 minutes.  This would prevent any 'mid-nebula' switches of full module upgrade paths, but allow for variations.

[orzelek]

I admit I've tried different setups, but the basic module combo is good enough for almost every situation in a nebula. Due to the pressure of "min maxing" due to bonus exp I always end up with at least MK III's of these modules, which feels like railroading.

For me this is the main issue. Especially when tied to the fact that nebula reward is based on efficiency. So to get better efficiency you need to spend points in modules that will help in increasing that efficiency. This makes you to skip more fun modules because they may not have direct nebula usage. And without them you would get less rewards which also means.. less new modules. And you get nice pretty circle

I agree that possibility of respec would help reduce or eliminate this problem depending on how easily said respec is available.

[KDR_11k]

I think if we're looking at respecs it might be worth going for loadout values instead of per-module-type upgrades. E.g. in Global Agenda you got 15 points to spend on your equipment with each piece being available in versions costing 1-4 points so you could pick a weaker grenade to get a stronger shield or something. This system was apparently abolished. I think this would have to be per module, not per module type so you aren't shepherded into monoculture loadouts. Of course with the linear gains from higher marks that may not really offer much interesting choice (as using 1x MkV, 2x MkII is the same as using 3x MkIII except maybe for overkill).

[Histidine]

Maybe modules could have a higher starting level when unlocked with a higher hull. Say frigates and destroyers unlock new modules at Mk I, cruisers and battleships at Mk II.

Also +1 to respec, flatten XP/level curve and/or increase out-of-nebula XP gains a bit.

[LaughingThesaurus]

+1 to respec.

Normally I hate respec systems. However, I only really hate respec systems when you have all of your knowledge from the get-go, and any bad choices you make are actually made when you're fully well informed that it's a bad choice. Congratulations, you got Ice Brand in Torchlight 2 when you went Fire Embermage. Do you want a medal for wasting skill points, knowing that it was a clear waste? But, in AI War, you don't get the skill tree. You don't know what you're doing, or what to plan your skill point distribution around. You can't actually think of a good plan, which can result in you having completely unfocused garbage skill points lying around.

However, no respec should come without a very significant cost, so if enough people like respeccing, we need to brainstorm a big deal cost to prevent it from happening literally all the time. It could be anything from AIP, or a permanent (for that game) loss of a nebula ship-building structure, the permanent sacrifice of a module line, or incurring an exp "debt" like in Space Pirates and Zombies, that must be paid off before you can gain real experience points again.

Alternatively, I also like the idea of simple point systems spent on modules. You gain more point capacity as you level, and can put on more modules of higher levels the more points you have. Modules cost their mark worth of points, maybe +1 for modules that are mark III or above, something simple like that.

[Draco18s]

+1 to respec.

Normally I hate respec systems. However, I only really hate respec systems when you have all of your knowledge from the get-go, and any bad choices you make are actually made when you're fully well informed that it's a bad choice

Agreed on both points.

The one respec system that I came to loath was the one used in Sonny 2.  Respec'ing was free, do it any time, feel free to experiment.  Fine, well, I want MY zombie to be this way.  Those other abilities are nice, but I don't want to use them.

Sorry, can't beat this enemy.  Never will.  I, the developer, expect you to fight this enemy, figure out what it's weakness is and then lose, respec, and fight again.  Rinse and repeat.

Barely into the game, I swear to god, the second enemy in the post-"tutorial" zone was neigh immune to all damage types except poison.  Except that I didn't want to play a poison-zombie.  I wanted to play the good guy and go full out heal-spec (and that's even ignoring the fact that the poison abilities sucked*).

*Do half damage forever.  Your remaining damage is applied over the next four turns.  Oh, and battles last about eight.  Why didn't I just go pure damage, again?  Oh, right, immune to basic melee...

[Lancefighter]

I can mostly see respecs in a game mechanic allowing, and REQUIRING the player to respond to each environment optimally.

In this case, you are allowing the player to respond optimally, but youre also not requiring him to do so. You can beat most of the nebulas without a plasma cannon equipped spire ship. But that doesnt mean they will be anywhere near as fast.

In this case, I kinda feel like we shouldnt really go for a straight respec. I am not honestly sure thats the right idea to go with - Instead, I want to propose something else.

I propose that we let the champion take time to research new modules. Now, I dont really expect this to REPLACE the current method of gaining xp/levels, but I feel like it could work alongside it. For instance, after a nebula, you instantly get whatever experience. You also, because you completed the nebula, your champion will, lets go with 20 minutes later, gain another experience point. And possibly a few more after that.

I'm not really sure if I want to force the player to research exactly what he wants (ie, choose research before it happens), or go with just adding to the xp point pool that we already have. The benefits to this system would be that you are benefitting the player for taking time to do things - For instance, if he takes an hour between each nebula, he will have a lot more unlocks to play with. If he doesnt, then he gets the current amount of points, plus whatever he takes the time to do.

I believe in my last game it took me about 3 hours to blitz through all the nebulas. This means that over the course of the next 6? hours of gameplay, I would continue to receive points at a fairly slow rate, up to about 27 more points total. I wouldnt particularly call this gamebreaking (It only lets my champion be more versatile later in the game. If you think thats a bad thing,why are you supporting respecs?). It mostly just means that I can afford to unlock whatever I need to, and later on still have enough points to get a few 'cool fun toys' that arent quitea s optimal.

Anyway, tl:dr - Let me gain points over time.

(alternatives to this - I've mentioned earlier some sort of additional post-battleship nebula re-scenarios might be interesting. Merely increasing xp gain through ai combat might also be a valid option. I am mostly not sure how 'infinite' people want xp gain to be)

[KDR_11k]

Free respecs are really more of a loadout system (e.g. Diablo 3 goes that way). We already have champion loadouts with the module slots which are reasonably easy to re-spec but your mark unlocks modify how effective each loadout is.

I think it might be enough to simply increase the point payout of later nebulas so you actually get those points when you need them the most. Increasing the total level would probably be good overall since that way you don't focus so sharply on a very small number of module lines.

Or perhaps upgrading module slots instead of module types?

[LaughingThesaurus]

Free respecs are really more of a loadout system (e.g. Diablo 3 goes that way). We already have champion loadouts with the module slots which are reasonably easy to re-spec but your mark unlocks modify how effective each loadout is.

I think it might be enough to simply increase the point payout of later nebulas so you actually get those points when you need them the most. Increasing the total level would probably be good overall since that way you don't focus so sharply on a very small number of module lines.

Or perhaps upgrading module slots instead of module types?

I never proposed free respecs. Respecs need to be very, very expensive in order to actually have a meaning. Typically, in a game where you're allowed to plan ahead, I will be in support of costing a level or two of the player's as the charge for respeccing skill points. In the case of AI War, there are a lot of more permanent things to cause. For instance, AI Progress for using this technology that allows the shadow vessel to 'realign' the things that it's learned... or the idea of sacrificing a lot of knowledge to be able to redistribute the rest... either in terms of actual Knowledge, or points, or different lines of modules or other unlocked things for the shadow vessel... Or you can go to a further extreme, and reduce the shadow vessel's maximum shadow energy for every respec.

If a respec system is allowed, it needs to have a stiff, extremely harsh penalty for use. In the case of AI War, we can be nicer due to the fact that the skill tree is randomized and hidden each time... or reveal the skill tree somehow and actually reevaluate whether something still needs to change after that.

[KDR_11k]

I think if you make it too harsh nobody will ever use it.