Just downloaded / installed B27 and took a look. Haven't had time to mod it up much. Here are some first impressions. This got kind of long, so for TL:DNR folks, basically I'm saying that it's better but it still doesn't feel like 1.0.
For the TL:DNR version:
1. The new Campaign is a great step forward, but it has a very obvious bug; when all of the players other than one have a negative balance, the Campaign doesn't end in victory.
Suggested fix: just fix the bug.
2. The Campaign rewards spending on defense, but not offense. This creates a lot of obvious problems; if you can afford to buy full base defenses, the AI never has a shot.
Suggested fix: the AI, if attacking, sends down a lot more troops, or uses an Orbital Bombardment at the start of play.
3. When attacking an AI base, I've found that late-game, you may as well just buy a bunch of drop ships and use them as bombardments to destroy the base's top-side defense.
I'd like to suggest that Bombs be removed as a class, because it really is a player-only thing that AIs can't use efficiently.
To replace it, give us a feature that offense can use, called Orbital Bombardment, that consists of dropping X bombs of your choice from the sky for Y cost, where X is picked in the usual menus and Y is a ramping cost, starting at base cost and multiplied by, say, 1.01 per bomb. Bombs should be dropped at random and a 10-second wait should occur between the last bomb being generated and the start of regular play. This would allow both the AI and the human player a valid route to assault late-game bases in un-modded play and take away a tool that players can abuse really heavily but that the AI can't use worth a darn.
4. I still feel like there isn't nearly enough differentiation between the heavier and lighter troops and other objects. There's nothing that's like a tank, other than dropships, and the dropships remain a cost-effective killing device, which they probably shouldn't be.
Suggested Fix: I really think balance needs a major overhaul.
There really aren't efficient choices for area defense, anti-aircraft defense, etc. to make countering the enemy's forces more interesting.
I'd like to see a viable option of buying, say, 20 crappy robots vs. 1 juiced-up Brownshirt at the same cost; pack attacks vs. Rambo micro. Right now the only argument for the more-expensive troops is that they can carry more, but they're so crippled in terms of mobility that it's not worth it, and it just doesn't give them enough force multiplier.
There's no point in having a guy with 5 weapons if he's got all of the mobility of a snail and hardly any survivability. So why is that even an option? Why not just limit troops to X objects carried or weight carried, and give them full mobility?
There's no point at all in buying the Gatling Drone or the Dreadnought, if they cannot move more than 20 feet away from your Brain because they can't jump and still have all the aCrab problems with terrain.
These are just some examples; I feel the game's chock-full of stuff that looks tantalizingly-useful atm but actually isn't. A game design should not be 90% newbie traps.
5. You can't, as attacker, buy 20 guys, pre-equipped, in 10 dropships that arrive dispersed, and have them be already on Brain-Hunter mode at the start of play, which is what you'd do IRL (well, on a different scale, but whatever, it's a game with a low ceiling).
There is way, way, way too much micro in this game atm; the AI's improved a bit with this version but it's still feeling like I'm up against a fortress with 20 guys in it but I have to first manually bury my Brain somewhere and then slowly and laboriously deploy my forces. The game just feels very slow paced right now and a lot of frustrations remain due to terrain destruction. I'd like to see a lot more macro; the Presets are a nice step in the right direction but the dropships and the you-only-have-the-Brain at the start are both frustrations. It's especially bad for the AI, which I've never seen even get close to damaging a full base, no matter how much money it has.
6. You can convert any battlefield into a deathtrap for the AI pretty easily by scattering mines in the areas it has to pick to land in. I honestly think that the terrain and the base should get reset for each battle, otherwise it's just too easy to design defenses that are almost insta-kill vs. the AI.
7. I really feel like the Dropship concept, however cute, basically doesn't work very well to give us mass combat, which is where a lot of the fun is. I'd like to see a drop-able object that spawns a Preset troop every 30 seconds or so on bot-hunt or brain-hunt mode, depending on a toggle, and a similar object inside bases. Then it's more like a RTS, with a struggle to pick up resources enough to feed our factories and more time spent on combat, less on finagling little things.
8. We need a bot-hunt mode to replace the ever-buggy Patrol. Bots should seek other bots within 1000 pixels or so and kill them.
9. The limitation of the area you can call for new support being limited to your commander really stifles play. No. 7 would address that, by giving people an object they could land that would automate assaults and create more of a feeling of a big war instead of encouraging the player to bury their Brain and then just buy a bunch of bombers until the AI's out of money and surface troops, which wouldn't take long.
10. The bases are very poorly designed for mobility inside. Given that the bots still have trouble navigating anywhere constricted, this causes a lot problems. Making the base interior dimensions a bit bigger and getting rid of the things like ladder rungs that just exist to keep bots from moving successfully... would help a lot.
11. The Campaign mode is broken up into a lot of unnecessary segments, imo. I don't really care that my base built 100 objects, and I'd like to see that kind of thing summarized in a number over each of my bases; I don't want to watch the Scan, I just want it to be Scanned when I arrive, and it's apparently impossible to actually attack multiple locations per turn, but it's possible to schedule the attacks and waste the gold.
12. When you own a location, you should have a full Scan. It's pretty silly that you're still in FOW, imo.
13. LOS to remove FOW is really weird and super-short, and it's obvious that the bots ignore it. I find myself shooting at black squares from which shots are coming all the time.
14. Actors are still self-gibbing all the time while jetting around. I wrote a Lua mobility script awhile ago that addressed that issue amongst many others. I did note that Actors can finally pass through each other; that is a big improvement.
15. I thought if you picked a side, you were going to be limited on tech you could buy? I don't see that happening over here. That needs to work correctly; amongst other things, it'll finally force the sides to get balanced properly and differentiated wisely. It's still basically like Skirmish, except the AI only buys one type of thing and the player gets anything they want. That pretty much just hoses the AI, because the factions aren't at all balanced right now.
16. The gold-digging AI remains pretty much unable to do its job effectively. The whole gold-digging mechanic is troublesome to begin with, because it involves terrain destruction and is a complex problem, but it's got to work well enough that the AI can and does deploy a lot of miners successfully or it's just too easy to run them down on resources, or it should get removed, imo.
One alternative is to just give the AI free money every second, so that it's effectively removed from having to bother with that issue. TBH, I'd rather see some sort of automated mining that simply collects gold within a couple hundred pixels to either side of a mining object, regardless of depth, so that mining is like planting a flag or controlling a resource in a typical RTS. Again, it's an issue of micromanagement taking us away from the fun part.
17. Seems like the number of Actors is still tied to the number of dropped objects. Really needs to be fixed; if 90 weapons are lying around, it should not be eating into total squad cap. I'll be modding out dropped objects staying around again when I get bored enough, but it just shouldn't work like that anyhow.
Tue Jun 12, 2012 1:48 am
xenoargh
Joined: Fri Dec 30, 2011 3:33 am Posts: 276
Re: Tested out B27... my thoughts
OK, just got done blowing away the B27 Campaign. Took bout 1.5 hours.
Lessons learned:
1. Build a preset with three riflemen in a dropship. Summon dropship. Take control and air-drop riflemen on newly-spawned AI Brain. Defense over.
2. AI bases can be a bear if they have money and you don't, but not vice versa. Static defenses < infinite Rambos.
3. Getting rid of FF and FF gibs due to movement made the game 100% more playable; props for listening to players on this.
4. The AI buys from the Presets, but apparently won't buy the custom ones a player writes.
5. The Browncoats remain OP for cost, due to powerful jump packs and strength vs. Coalition Light / Heavies. Coalition Heavies continue to be neither fish nor fowl, as they aren't significantly more durable than Lights and don't seem to have many other advantages. I'd really like to see them be a lot harder to kill, even if that means they need costs to go up significantly. Haven't played enough with their weapons yet to judge costs, but initial impression is that they're a bit pricey for what you get. If / When techs limit your choices, that may make more sense.
6. If I wasn't being pretty deliberate about just min-maxing my way to victory in a quick test run, I'd want a tank-like thing... a lot. Going to mod up some tanks for the AI to summon so that they might have a chance if I give them a lot more money.
7. The disparity of offense vs. defense described above proved totally solid. The need to bring down a lot of stuff in one buy call is a big deal.
8. Mining's pointless. Do no. 1 the first time you get raided, you have lots of gold. Take weak bases, suddenly you're well over the tipping-point.
9. Speaking of tipping-points, it's way, way, way too early. I never felt like I actually had to try very hard after the first offensive battle. The bases take some pounding but it's just time, not difficult. The AI needs to not run out of stuff so quickly.
10. Nice to see the Uber-Cannon finally lives up to its name; it's the first time I've played a build and haven't had the urge to mod it immediately.
11. There still isn't a real flamethrower that really does solve tunnel branch defenses nicely. The Browncoat version doesn't cut it, needs to bounce more and have longer effective range.
12. The Remote GL is almost ridiculously OP for taking out enemies sitting around a base.
13. Still too many redundant guns and things that aren't worth buying. Sure, the Shotgun's only 30 Gold, but it's not worth buying at all. It really should be a death-machine within its range band to make up for all of its other problems, otherwise there are so many other choices that make sense.
14. The man-portable Gatling is still pretty darn useless. I'd really rather have a realistic LMG that doesn't have such a lengthy spin-up and doesn't weigh a million pounds; nerf it by making the reload time pretty awful.
15. I want to see HMGs and cannons with long AI detect ranges and reasonable firing arcs in surface bunker installations. Make players have to scout out a good route or dig mines to get past surface defenses sometimes or buy a tank. Right now, it's basically as easy as killing their spawns as they slowly come in, then whittling away; it never feels like I'm dealing with an actual bunker.
16. Really would like to see the factory / auto-spawn devices, especially for the AI's sake. It's just not working for them to have to deliver their troops via dropship when they're having to play with really strict resource limits.
17. Autosave has used a huge amount of HD space saving terrain files for terrains that probably should get reset anyhow.
I've got to agree with most everything you said but I want to add or suggest a few things. These don't match up to your numbers (as I'm sure anyone can see by reading them). 1. Easiest way to win defense is to spread spawn markers evenly all over the map, 5 or so of them and you won't need a bunker at all.
2. If you want a bit more difficulty try playing as non-browncoats vs. 3 browncoats on the same team. Since they all get income from each territory they'll never run out of money and if you let them land for an attack they can actually make it through a decent sized bunker because browncoat heavies are really op. On the note of that, when it takes 3 shots from a revolver cannon or half a clip of a minigun or autoshotgun or just about any other weapon coalition has to offer to kill one browncoat heavy, it's a sign that they need a nerf.
3. A factory item isn't really needed (though I guess it'd make it easier to pick a ground based spawn point), the metafight mission is all a 800 line activity. Outside of the brain checking and picking it's not very filled out or complex so it wouldn't exactly be difficult for someone with the slightest bit of lua know how (and you seem to have a good deal of it) to add land based spawns somewhere. That said, a way easier way to give a bit of challenge on defense missions is set it so every few seconds i of the enemy troops dropped down have a low chance to go into brainhunt mode, meaning you'd actually have to try to keep your brain safe.
That said, a factory would be a really cool idea, especially if there was special bunker infrastructure for it (i.e. a factory that actually looks like one). It'd probably require a good deal of rebalancing though and the maps would have to be bigger (they could probably do with that now anyway, a full prebuilt base that takes up the whole map is kind of ridiculously unfair to attackers).
4. You touched on this in your first post about how crabs are useless. I think a lot of the maps (any of the ones with mountains) were poorly thought out for CC. Since crabs can't jump and even this better actor mobility isn't perfect, it limits your options greatly. And on the note of mobility, why are the jetpacks so wonkily balanced. Some actors have ridiculously long lasting jetpacks that may or may not do very little for lifting. If the aim is to have jetpacks used as a way of getting over obstacles, not as a primary method of movement then the ideal would be high lift low fuel jetpacks that can lift an incredibly overweight actor a good way then have to recharge for a fairly long while. See jumpacks in Dawn of war and most other strategy games.
5. As mentioned at the beginning, I've got to agree with most all of what you said, real tanks and real turrets would be a very welcome addition to the game.
Tue Jun 12, 2012 5:14 am
TheLastBanana
DRL Developer
Joined: Wed Dec 13, 2006 5:27 am Posts: 3138 Location: A little south and a lot west of Moscow
Re: Tested out B27... my thoughts
There's some really well-thought-out comments in here. Thanks for the feedback. I can't guarantee we'll get all of this into 1.0, but I'll try to keep some of it in mind. For the record: you can buy other factions' stuff, but their prices are quadrupled (or maybe more? Not sure. It's a large multiplier, anyway). So, for anything beyond basic pistols and stuff, it can get very expensive. The idea is that you're buying it through a black market, so it's still available, but only if you really desperately need it.
Tue Jun 12, 2012 6:19 am
NeoSeeker
Joined: Tue Jul 24, 2007 1:13 am Posts: 1183 Location: eating sock's face like a cupcake
Re: Tested out B27... my thoughts
i think this sums up nicely what's on everyone's mind.
quite possibly a proverbial elephant in the room. certain things are simply broken or don't work like they should.
i do know the coalition might be getting that EOD unit that's in lizzard's avatar. it's a god damn beast and bullet sponge. but i still agree, there isn't much of a point to using coalition heavies.
also i must say the browncoat's weapon selection kinda sucks, you guys probably know that but i just wanted to say. also some offsets like the support offset for the assault rifle sucks.
the assault rifle needs something else like an underslung grenade launcher or shotgun/slug gun
one more thing. i think that you devs def need to at least look into weapon attachments, alt fires and stuff like that. it could really help out balance.
\/this will probably never happen in CC but i heard there's a CC2 planned\/ a faction based around a single actor, a pistol sized weapon template and rifle sized weapon template. the faction could be completely based around modularity. instead of carrying around a bunch of different guns they are pre-equiped with the pistol and rifle template and instead of buying guns they buy kits or mods for their guns (basically you can turn the template guns into anything available to you through mod kits, assault rifle, grenade launcher, shotgun, etc.). also you pick and choose what kind of armor and jetpack they have
Tue Jun 12, 2012 6:42 am
xenoargh
Joined: Fri Dec 30, 2011 3:33 am Posts: 276
Re: Tested out B27... my thoughts
First, just wanted to say that none of the above's meant to be dogging B27. The game's clearly going in the right direction and feels a lot better; this is pretty much what I expected with B26 and I think it's going to get where it needs to.
Played a bit with three Browncoat AIs. Browncoats Heavies aren't dying when they lose their heads atm and they're all really, really tough.
Thematically, they'd be great, except that they're OP for costs, if we're just talking bodies.
They not only have great armor, they also get great jump packs and high strength. They could remain just like they are, but they should cost twice to three times what they do.
That said, I still beat the tar out of the AI, no problems. Can't win this game because there's some sort of test map that's causing issues; I guess I need to clean out my CC install folder.
On defense, I just took light losses during the early phases.
On offense, I just sought out the Brain with air-dropped guys with assault rifles or dug them out of bunkers.
Part of the problem is the absurd vulnerability of the Brains. Yes, I know, it's the Future and cloned Brains are flying through space to be the generals of our armies... so why is he the most vulnerable troop in the game, instead of a bit more challenging?
That said, I know that nobody wants them to be Rambos on a regular basis, so one thought I had was that maybe Brains need to be dropped in a box with a turret, well-armored, that auto-digs themselves quickly into softer ground (something like the Drone attack) so that the AI is instantly defended by dirt and can shoot anybody who drops down the hole.
To release them into the classic weak Brain, use a Lua pie control. Make the digging radius big enough to get out and not get hung up on the now-defunct turret.
Thematically it'd fit, and it would give the AI a lot more staying power; no more drop-my-brain-and-ram-him-with-dropship, hehe.
One of the big factors in terms of balance is that the AI's just not evaluating over full weapon ranges and doesn't use probing fire when it hears / sees shots. So I can just whittle it from range.
The fact that I can see the AI's position via holding down Q or E any time it's not in FOW contributes heavily to this.
I honestly think that the FOW concept needs to work differently, with AI troops hidden from player view unless they have an Actor within, say, 1200 pixels. Better yet, how's about an adjustable variable that auto-adjusts according to screen resolutions so that players with all sorts of rigs can play and have a similar experience.
IDK whether Lua's allowed to get resolution, but if yes, it can be prototyped in the current aHuman code easily enough.
Up close, the AI's pretty inhuman in CQB, really, with auto-aim that's focusing on players too quickly to do anything but trade fire. There's no point in dropping down a tunnel shaft that's guarded unless you're willing to trade guys. This is actually an improvement over B26 in that regard, but it feels wonky.
I honestly think the guns need an AI parameter loosening the engagement times and accuracy a bit, kind of like spinup times, so that the AI Lua-side code can be perfectly accurate and deadly but the AI won't just behave like a perfect turret up close. If it's adjustable per weapon as a parameter, the AI can have a pretty long engagement speed and very high accuracy with, say, sniper rifles, but a short wait but lousy accuracy with SMGs. Then the AI would feel a lot more human.
But get the AI foxed by terrain or at a distance and it's toast right now; I've lost count of how many times I've seen AI troops buried in terrain and easy meat. I just fire speculatively all the time.
Also, while I'm muttering about gameplay and such; what's the future of modding looking like for this engine? There are a lot of mods, and I think that as people test it out, mods are going to generally aim towards being more balanced for the Campaign, which I think is great.
That said, the chokepoint on creating new content is a big deal; I've modded a lot of engines and this is one of the least user-friendly in terms of making new characters, simply because editing is so time-consuming.
Any possibility of a way to build new Actors with an in-game offset / walkpath editor where we can grab / rotate a part of the model, define the points that detect collisions with the ground, test stiffness and strength and where we can set up each frame of animation and do some tests of speeds vs. animations, instead of being limited to aCrab / aHuman hard-coded stuff and laborious text-editing?
Most of that's just showing things that are already there and allowing read-write instead of just read; rotations are just reversing the math for the bones and giving us a widget that will allow us to define it numerically or use the mouse to do it visually and then write it back to the RTE when we're happy with it. Just need to draw the skeleton positions and allow for new bones to be built and defined instead of limiting it to aHuman / aCrab; instead of limiting the bones, I'd suggest just declaring that all Actors have a common set of animations, i.e., walk, jump, crouch, aim one / two handed, carry shields, etc. and custom ones that don't have things defined won't be able to do said action.
That would be cool; don't define a Crouch animation and the Actor cannot crouch, for characters where it's not appropriate.
But please let modders define how many legs / wheels / whatever and ground-collision points so that things like vehicles get a lot less buggy feeling, kind of like that demo of the new engine you folks were working on with true 3D collisions.
I don't actually think it needs to be that perfect, but it'd be nice to be able to define, say, 8 ground-collision points for a long skinny thing like a tank that would keep it from eating terrain.
Tue Jun 12, 2012 5:37 pm
MavericK
Joined: Mon Feb 25, 2008 12:05 pm Posts: 53 Location: Anacortes, WA
Re: Tested out B27... my thoughts
I agree heavily with everything you've said, both in the original posts and this last one, and I wanted to specifically address this:
xenoargh wrote:
The fact that I can see the AI's position via holding down Q or E any time it's not in FOW contributes heavily to this.
I honestly think that the FOW concept needs to work differently, with AI troops hidden from player view unless they have an Actor within, say, 1200 pixels. Better yet, how's about an adjustable variable that auto-adjusts according to screen resolutions so that players with all sorts of rigs can play and have a similar experience.
I totally agree that the FOW concept needs to be reworked in a major way. There have been numerous occasions where I could see vaguely where the enemy brain was based on the "Destroy!" marker and just fire blindly into the blackness, resulting in a "WIN!" and having never seen the enemy brain.
I don't have a big problem with scanning, or not being able to see where exactly gold is in the ground (both of which make sense to have an obscured view) but I agree that the LOS concept is fairly broken, and the fact that AI completely ignores it just makes it more or less an annoying cheat for them.
Additionally broken about the LOS concept is that you can magically "see further" simply by dragging the mouse to "aim". For sniper weapons this can looks really silly when you can "see" through a long, narrow section due to the extreme zoom. Then you can just swing the mouse around in a circle to clear out a bunch of area. Seems odd TBH.
Well that makes sense really, if you have a sniper and you scope in you'll be able to see farther in a limited area. Fog of war represents how far an actor is able to see. Thus how it works makes perfect sense for a sniper. For other guns without scopes it doesn't make sense but that's not a major problem really.
But yeah, recovering fog of war would be a good thing. In fact I don't understand why data didn't just choose to follow the rts formula of black fog and grey fog where black fog stays gone and grey fog comes back and obscures actors (or better yet shows them foggily at their last seen position) but not terrain or scene objects.
Tue Jun 12, 2012 9:16 pm
Arcalane
Joined: Sun Jan 28, 2007 10:32 pm Posts: 1609 Location: UK
Re: Tested out B27... my thoughts
xenoargh wrote:
Played a bit with three Browncoat AIs. Browncoats Heavies aren't dying when they lose their heads atm and they're all really, really tough.
Glad to know this isn't just happening to me. Kinda freaky taking a Heavy's head off with a bolter or sniper shot and seeing him just standing there with ~30 health left, firing back like nothing happened.
Might be a MOID limit thing, but I haven't tested it much myself.
Tue Jun 12, 2012 9:19 pm
Jomn
Joined: Sun Jun 13, 2010 12:40 am Posts: 327 Location: US of A
Re: Tested out B27... my thoughts
I think it should be kept, adds to the vision of toughness the browncoats have. Unstoppable warriors that even a headshot won't stop...
Tue Jun 12, 2012 9:47 pm
TheLastBanana
DRL Developer
Joined: Wed Dec 13, 2006 5:27 am Posts: 3138 Location: A little south and a lot west of Moscow
Re: Tested out B27... my thoughts
Browncoats will definitely be nerfed. We've already fixed the head bug.
Yeah, I found it weird that they didn't need heads but it's also pretty cool and I wouldn't really mind it staying as long as they end up properly balanced otherwise (price or other toughness). And I'm pretty sure it wasn't an MOID thing, it was just how the actor was coded.
Tue Jun 12, 2012 10:23 pm
MavericK
Joined: Mon Feb 25, 2008 12:05 pm Posts: 53 Location: Anacortes, WA
Re: Tested out B27... my thoughts
Bad Boy wrote:
Well that makes sense really, if you have a sniper and you scope in you'll be able to see farther in a limited area. Fog of war represents how far an actor is able to see. Thus how it works makes perfect sense for a sniper. For other guns without scopes it doesn't make sense but that's not a major problem really.
But yeah, recovering fog of war would be a good thing. In fact I don't understand why data didn't just choose to follow the rts formula of black fog and grey fog where black fog stays gone and grey fog comes back and obscures actors (or better yet shows them foggily at their last seen position) but not terrain or scene objects.
True, but then like I said you can just spin the mouse around and see in a huge radius. I guess that's "realistic" as well, but I think one issue is that if you don't have a scoped weapon your view distance seems incredibly limited compared to what would be "realistic".
I'm not entirely sure how to solve it but the way it is currently is somewhat off-putting.
Oh yeah, realism doesn't necessarily make for good gameplay, especially not in something like CC. Gameplay-wise it's kind of irritating and silly to see. It also makes scanners have a very limited usefulness (i.e. see through walls) because it's more useful to buy a sniper and use that for large open distances. That said, I'm not sure how to really fix this since all of the solutions I can think of seem worse than what we have now. I can say however that actors' view distances should be greater and have a wider arc, it's really irritating at times when you don't see something 30 degrees above you while you're looking straight.
Wed Jun 13, 2012 7:35 am
MavericK
Joined: Mon Feb 25, 2008 12:05 pm Posts: 53 Location: Anacortes, WA
Re: Tested out B27... my thoughts
Yeah, I was thinking maybe it would make sense to just reveal all terrain/objects within a radius of the player, except where view would be obstructed. That might be better gameplay-wise.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot post attachments in this forum