- October 25, 2017 at 3:49 am #15268CHILLYBUSKickstarter BackerTopics: 61
Hey all, back again with some Thoughts on Things for Build 83! After putting probably 10-15 hours into this patch cycle, I think I’ve got a decent feel of where things stand now. Let me know what you think; happy to duke it out in the street about anything here. As always these are just suggestions/thoughts and I’m not affiliated with Revival nor intend to occlude their creative vision etc etc etc:
Some thoughts on balance:
Overall in the last two patch cycles, things feel back towards the run-and-gun style of fighting that was prominent in early builds. Punishment for small mistakes can be handled just fine, but it’s a bit too easy to turn (or have turned for you) small mistakes into big ones that you get totally wiped out in. AI seems to cluster towards the player pretty rapidly, which is great for stacking devastator kills and the like, but if you’re exposed for more than a split second you’re taken to task for it. Aggressive play feels like it’s punished way more than guerrilla tactics, however pilots do need to get in the mix to keep their shield economy healthy – kind of…not hypocritical, but you get the gist.
Build 83 takes the crown from somewhere in the 40s or 50s for ‘most fun I’ve had in this game’, though not consistently. I’ve flown a few rounds with few hitscan robots in the level and it was money. So much fun! Dodging was on-point, every mistake was something I alone was responsible for, and it’s so satisfying to be able to outmaneuver things fairly. I hate to keep harping on hitscan weapons but they grind my gears something fierce. If there was a branched build without Ogres/SWyverns/Reavs/Guards I don’t think I’d play the final release if they were still in it. It’s that good to fly with them reduced or removed.
Player weapon balance is okay-to-pretty-good. I’m trying to stay sharp with everything but some stuff is definitely more effective than others. Missile Pod tracking feels slightly weakened though that might be more from AI dodges. Dodge code for even lowly Goblins on rookie can see them outmaneuver large salvos at medium range on occasion. Impulse feels a bit limp damage-wise and the Cyclone definitely needs some love. Would it be possible to shift the cyc/driller hardpoints further up towards (or in) the center of the crosshair? The cyclone is especially futile at point-blank since you have to remember to aim well above where you’d expect. Driller fire is the only thing I’ve found consistently effective on Ace+ — definitely anecdotal, but the dodge code and robot health makes everything but sustained driller fire somewhat futile. Also, can we get somewhat bigger Falcon hitboxes, blast radii, and stuns from them? I love cross-map trick shots with them sans tracker upgrade – few moments more satisfying than that fire-and-forget snap of the kill counter from down the tunnel. But with this dodge code especially I’m avoiding using them in some cases just because I can’t connect even at medium range without the actual tracker upgrade. I’ll waste 5 MXes on a single Harpy sometimes. I’m no Thunderbolt expert, but it just feels like it charges so slowly – I avoid using it because I can’t time the charges right and panic-fire it at close range like it’s a Crusher. The learning curve is really steep, and in a tactical sense I’m not totally sold that it’s particularly effective given the charge-up time (get at me Yoshi).
Hydra dodge/hide code is genius but total insanity. I can hardly handle them and SHydras on rookie most of the time, relying on a lot of tracking weapons or an explicit switch to focused driller fire to take them down. Those poor saps on the consoles are going to get totally destroyed! I was able to get a SHydra on an insane run to rush me faster than a Shredder, strafe me at point-blank as I turned with it, then get a bit in front of me before it was destroyed. -80 shields in that pass alone and it effectively ended the run! I think part of the danger from the Hydra family is they’re constantly facing you – turns away from the player more often while repositioning and retreat would make them less threatening. Behind Reavs/Guards, these are second tier in my targeting order.
Shredders have that bounce thing back! They’ll hit a wall, bounce off, and keep shredding, and sometimes don’t stop for some time after on Ace+. Scorpions are well out of the cannon fodder tier with their changes but they’re a bit too hard now – silent and very fast, with movement that’s a bit too difficult to reasonably track. Positive change with them, just a bit difficult!
Titan combat is fun and dangerous. Placing them correctly in the targeting order and dishing out punishment is a satisfying challenge.
I find it easier to seek and ram the Reavers (still don’t have the smash key bound lol) than deal with the hassle of taking damage and dumping fire on them. I’ve had more success with eating the shield loss from the explosion than sacrificing a lot of resources for a slim chance at coming away unscathed. These robots are a bit easier than in the 7x patches though – I’m not writing the run off whenever I hear them spawn anymore (more…penciling it off depending on my mood).
Ace+ is aggressive and fun…but only for kicks, since they’re much more difficult in these latest two patch cycles than those prior. I don’t see myself playing them in the long-term since they’re just too hard. Too much fire coming in for way too much damage. Insane right now is a serious bullet hell, and I wouldn’t mind seeing it magically copied into a 6th difficulty, and then the rest of the difficulties tweaked down some. Definitely a neat experience up there in the stratosphere, but it’s frustrating to play seriously.
I think Rook/Hotshot are approaching ‘great’ – it’s really just nailing the hitscan balance and/or reworking those mechanics some. Dodges are satisfying, damage is mostly fair, etc,. You can really feel the shield spawns drop from under you in infinite though — this was present in early builds too. Plateaus in the shield graph are obvious in longer runs. I’m an edge case though so that’s not something everyone will notice, but it’s definitely there!
Overall, balance is still not quite there, but when it’s only dodgeable projectiles and missiles to fly against, Overload is really, really fun. It sounds to me like the big balance reset comes after the main campaign content is finished – hoping this is helpful!
Some other things I’ve noticed:
1. Select-none is an absolute game changer for structuring upgrade chains. Great addition – countdown is way more focused and flying round after round is even more addicting!
2. Robots are still spawning near the player (even point-blank) on some levels, like Pipeline, Roundabout, and Labyrinth.
3. I get intense screen shakes for a long time (2-3s) when I fly very near some death rolls and explosions and turn the afterburners on. This is a bit awkward to deal with when I’m halfway across the room caught in a feedback loop of shakes.
4. Is it me or are upgrades weighted slightly to prefer heavier weapons, especially devs/time bombs? I had this feeling in early builds that I’d always see devastators get upgraded first, and now I’ve got the same with time bombs. I’m on the fence for having player-chosen upgrade chains, but they should definitely be 100% up to unweighted RNG.
5. The old mouse-interacts-with-stuff-below-where-it-looks-like-it-should bug is back – only happens when setting up loadouts between rounds using the ‘Play Again’ button. Fixed by going to any other screen.
6. Changes to round starts made things way more engaging! Some maps, like Roundabout, are still pretty slow to get off the ground though.
7. Super missile change is sweet – incentivizes always picking them up.
8. Cloaked robot AI changes came out perfect. Funny to see Golems/Reavs/Shredders/Scorpions randomly attack. How often does player position update in the AI? Flak/Cyc/Ref/Drill/Missile Pods get you totally hosed — I know that’s by design! Infrequent updates on position would make it more valuable of a combat tool.
9. Maybe I’ve just spent enough time using it but the OD doesn’t totally throw me anymore. I don’t have reason to use it and wish it wouldn’t affect my turn speed, but I’m not actively avoiding it anymore (and even seek it out on Countdown sometimes!)
Some other things I’d love to see changed:
1. Countdown invulnerability bug still not fixed – grab one late enough and you can pick up a ton of kills past the buzzer. Also, could we get these, cloaks, and ODs pulled from the mode entirely?? Runs at higher difficulties might be predicated on invulnerabilities (and lower ones by ODs) spawning and without them the playing field is leveled away from RNG a bit.
2. This has been threaded elsewhere, but a save-last for CM loadouts (or a small number of save slots for them!) would be awesome.
3. If I hit the escape key at the loadout screen after hitting ‘Play Again’, it starts the game with the current loadout. Should it go back to the CM menu instead?
4. There is a really strange part of the Driller-M firing cycle that really tickles your right ear with some headphones on.
5. I’ve never saved a game in singleplayer, and when I go to hit F9 Overload will freeze at the ‘Loading’ screen that pops up.
6. Probably more an optimization/drivers thing but my game crawls with more than two or so recently fielded novas or devastators, and it’s been this way since the patch that Hive was released in back some six months ago. It’ll come back up after ~5-10 seconds, but if I dump a pile of either secondary I can’t control the ship very well. I’d say it’s the nova bolt tracking and dev blast radius damage updating too often, but what do I know? I should be able to run this pretty well – I’ve got an i5-4590 and R9-380X with 16G of RAM, but I’ve got Overload on minimum settings just to keep it (and OBS) stable most of the time.
7. Countdown feels a bit inconsistent in its spawn pools – some games I turn a corner and 10 goblins are already spawned for me, other games it’s a full 10-15 seconds before I see a Kraken. With how important kill rates are in this mode, evening these pools out would suppress the temptation to fish for those perfect spawns and robot makeups.
Some thoughts on practical hitscan fire:
Before continuing, let’s define practical hitscan for the purposes of this discussion. Hitscan is anything that hits immediately in a game – you or the AI pull the trigger, the game processes the input, and the damage is applied instantly and is unavoidable. Practical hitscan is a projectile that does take time to move (driller shots), but are practically unavoidable by the player or even the AI in some instances. A conceptual framework for this would be something like a weapon is hitscan if this equation is less than zero: (projectile travel time) – (time to first render projectile + (visual response and player consideration of action) + player reaction + ship response). Ship response is critical and somewhat subjective/variable when looking at what is considered fair – it takes a significant amount of time to cover an entire length, and coded inaccuracy is often less than that.
Given Overload’s flight model and the time it takes to change directions, a lot of things can be considered hitscan, but be wary of including close-range fire, as that can be outmaneuvered with some foresight. Guardians, Ogres, SWyverns, Reavers, and SGuardians (pre-removal) could all be considered hitscan. In most engagements, given the flight speed of cyclone projectiles and range at which the player is attacked, Hydras and SHarpies could create a second tier of occasional hitscan weapons.
From what I’ve read of discussion here and in chat, most players are in agreement that the practical hitscan (PH) robots need some changes. They’re fun to fly against in small numbers, but once they appear in large numbers, deal a lot of damage late in the game, or get the kill entirely, it’s pretty frustrating and feels cheap. Reading my post-match details over the last several patch cycles, anything holding a driller deals outsize damage – just a handful of Guards/Reavs outdo piles and piles of lesser robots. Easier/harder robots aren’t so much an issue, but it’s obvious something is up when it’s primarily those two and occasionally Ogres. I’m going to go out on a limb and guess that PH robots (except perhaps Ogres) always take precedence in just about any successful targeting order. Also, correct me if I’m wrong, but Overload has 4+ PH robots, where D1 had just the one, and D2 had only three if I recall (do those ITDs count? You can see the red projectile). I know Overload is different, but I think the principle of limiting hitscan combat might be similar between all of the 6dof genre.
Let’s explore some options:
A: Limit type count/in-level count/shots-per-salvo/time-between-salvos/per-shot-damage of PH robots
This is a pretty obvious solution with a lot of moving parts, and there are a lot of avenues to take with it – just introduce a ton of limits on some robots. Limiting in-level count and shots-per-salvo of Ogres would be a massive change for early games. Even hidden mechanics we never hear about – disallow Guardian/Reaver spawns when the player has under 100 shields, or only allow one to spawn every 90 seconds…or suppress Ogres from spawning for 3 minutes and cap it at 3 active at once – might be worth exploring.
B: Changes to kinetics of bullet-based weapons, like the driller
Part of what makes driller damage so obnoxious is that it affects the player’s pathing. Even subtle changes stack quickly, and Guard/Reav shots in particular are heavy. Stunlocking is fun when the player does it to robots, not so much vice-versa. Descent handled this (if I recall) by stuttering the player’s momentum on a per-hit basis, and directional changes to the player’s aim were done with big kinetic impacts, like missiles (which weren’t too noticeable unless you were slow or still). Bringing back this old system would make target acquisition and dodging much easier and fairer.
Another change to consider is altering the opacity/brightness of driller fire – right now, it’s a brownish line and easy to lose or miss entirely. The SGuard’s projectiles were bright and easier to follow in pitched combat – one of the reasons they were easier than the regular variety!
A last – and unlikely change – is to slow PH fire down to be projectiles, but that doesn’t make much sense.
C: Remove some or all PH weapons
Cooked up in a stream dream – what if Guardians or Reavers shot weak (3-4 bolt and poor targeting) Novas? What if Reavers charged up over a few seconds for a big electric shock to the player if near enough, or Ogres shot the old D3 microwave cannon? Or SWyverns launched the old frag missiles? As mentioned above, the most fun I’ve had in this game has come from this build when there were few PH types and a low PH count in the level, and this removes them entirely. Lots of avenues to go with this one too — plenty of ideas to explore in terms of changing how these guns function entirely.
D: Introduce/improve telegraphing of PH weaponry
If players can tell when something is going to happen before it does and avoid accordingly, they are better able to react to it and battle outcomes are dependent more or entirely on ability rather than the AI. Giving that event (usually an attack) away like that is called telegraphing. Here’s a quick primer: http://www.chaoticstupid.com/enemy-attacks-and-telegraphing/
A few ideas in this vein:
1. Gun heat/classic telegraph – glow a part of the attacking robot more and more (or animate some part of it) for a second or two before firing in order to give the player time to engage or find cover. Good results predicated on the player seeing the robot early enough during this sequence.
2. Post-projectile callout – mentioned above, but visually loud tracers help figure out the line of incoming fire, if a bit too late to deal with the shots themselves. Not particularly helpful but better than the status quo.
3. Pre-projectile callout – targeting laser chases/lines up on the player, and when they connect for a second the robot shuts it off and fires. Good results again predicated on seeing this, but a long moving 1-D laser is easier to spot in a 6dof battle than a 0-D glowing robot.
4. Lock indicator – similar in visual design and placement to the red directional damage callout, a white one indicates hitscan robots that are 1-2 seconds away from firing. Wider indicators indicate closer distance, smaller ones are farther away. Simpler version is another lock sound, but this would be constant and really messy in the soundscape with a lot of Scourges/STitans…or a lot of PH robots.
5. Reverse targeting – the player’s HUD would identify PH robots that are readying to fire and highlight them (or directly call them out with an overlay like the secondary lock-on diamond from Freespace while in view.October 25, 2017 at 8:39 am #15269PersonicusParticipantTopics: 6
Very timely post. It was only yesterday evening that I was exchanging thoughts about the current build with DWD. I’ve been slogging away at Overload since mid-April and for me, you’ve basically managed to echo many, if not all, of my frustrations with B83. That said, I managed to score a somewhat respectable 122 kills on Blizzard (insane) with this build, but there’s no way that I could replicate that score or even come close by jumping in ad-hoc and playing for 30 mins or so….on the contrary, I had to grind that friggin’ RNG to be dealt a favourable hand with a somewhat forgiving bot-pool. As a side note, Blizzard is by far the most enjoyable CM map due to the ingenious wind tunnels !! More like this please. Pipeline would be my next favourite, but it’s simply too small, especially considering the amount of Golems that are unleashed (Ace and above).
I wholeheartedly support and agree with your suggestion of introducing another preset beyond Insane (Nightmare ?). Someone should take a detailed and statistical look at the Leaderboards, in particular the Insane preset on Infinite. With the exception of all “cavernous” levels (Hive, Labyrinth, Caverns), there seems to be a significant “thinning of the herd” approximately after the top 5 or 6 scores. There’s quite a large disparity…..half a dozen god-like individuals racking up seemingly impossible scores (500+), followed by the rest of the pack, but with a significant drop in kills. So perhaps a “Nightmare” preset should be introduced for the absolute elite pilots (throw everything at them, and then some), and slightly re-balance the current Ace and Insane presets accordingly. Alternatively, and most likely not really feasible at this point, would be to introduce combat Tiers (think Q3 Arena). This system would expose players to new enemies gradually based on their ability to defeat a previous tier. Of course, within each tier, presets would still need to be available i.e. rookie, hotshot etc). Again, most likely not feasible at this point, but just throwing it out there.
So yeah, overall I think B83 needs some serious tweaking (and herd thinning)…..and guys, I desperately need a new non-cavernous CM level….please and thank you !October 25, 2017 at 8:56 am #15270Luke SchneiderOverload TeamTopics: 10
Thanks for the detailed feedback. We’ve made some changes that address the major issues you’re discussing (too many PH weapons), but there are many more that won’t be in for a little while longer.
I have been so busy with finishing the entire SP campaign (working on level 14 now) that I have barely spent any time balancing/tweaking things in the past 3 months. I really enjoy tweaking/balancing things, but the priority right now is to get the entire game playable (every SP level), and THEN do more balances/tweaks.
I just wanted to let you guys know that we will be making major changes to balance, but we’ll be tweaking for the full SP campaign first, then adjust CM a bit more based on those changes. Most of the SP-centric changes will help CM, because the later SP levels (level 9 and beyond) weren’t terribly fun in their first iterations for the same reason CM isn’t fun sometimes. Too many PH bots and not enough dodging.
LukeOctober 25, 2017 at 11:38 am #15277sfischerOverload TeamTopics: 4
To add a little more to what Luke said (at least in regards to your specific points)…
1. Select-none: I’m also a fan.
2. Robots spawning near the player: There have been some tweaks & will likely continue to be tweaks, but for the most part, this is as intended. Early on in a run, you should have almost no robots spawn near you (other than ambushes). Every robot you kill, however, increases the odds of robots spawning ‘all up in your business’ – it’s coded that way – and in the hundreds of kills, most will be spawning too close for comfort. This is part of the ‘increasing difficulty of a run’.
3. Death/Explosions + Afterburners = Screenshake: I’ll look into it.
4. CM Upgrades favor heavier Secondaries: It shouldn’t, though not sure how it’s actually behaving. I’ll look into it.
5. Cursor vs cursor position in menus: Sigh – it’s back? I’ll look into it.
6. Faster CM engagement: I too think that it’s good to get it rolling faster.
7. Super missiles: I never really got them in CM before the change – now they’re all the rage.
8. Cloaked robot AI: There’s probably going to be some more tweaking, but now they seem to mostly react as we want.
9. OD Turn speed: It’s good that you’re adjusting. There is still an item on Luke’s plate about it, so it’s not definite one way or the other if something will be done about it, but the team knows it’s an issue for some.
1. Invulnerability in Countdown: It’s known & been discussed. Something will definitely be done about at least the extension issue (at some point) and we have talked about just removing those items for balance. Low priority on this in general – it will be addressed, just not sure when.
2. Saving CM Loadouts: No promises, but Mike seems receptive and was probably going to look at implementing it.
3. [ESC] on Play Again Screen: I’ll look into it.
4. Driller M audio: I’ll look into it.
5. Loading issue: I’ll look into it. It *should* be fixed with some of the save/load fixes Matt has already made since drop 7, but verifying isn’t a bad idea ;).
6. Nova slowdown: I’ll look into it. We’ve done some optimizations, but we haven’t spent too much time on the ‘play’ side of seeing issues like this, more had been spent on the code side. This should be changing in the near future, allowing us to diagnose/address stuff like this better.
7. CM spawn pools (both Infinite & Countdown use the same pools): The way the code works and the way we wanted the code to work weren’t lined up on this one. We’ve already made a change to it to alleviate this a bit, and (likely with Luke’s balance pass), we’ll probably mess with this a decent amount more. Using overly-vague terms, we thought the likelihood of being super screwed over by the robot spawn pool was rare, say like 3/100*. Upon a closer look, in some levels especially, it could be more like 30/100*. The changes already implemented should bring this closer to 8/100*.
* – all numbers made up, with no deep thought/numerical backing, to illustrate the general view on spawn pools.
And for clarity, no promises on any of these items ;).October 25, 2017 at 3:25 pm #15281YoshimitsuKickstarter BackerTopics: 57
I’m no Thunderbolt expert, but it just feels like it charges so slowly – I avoid using it because I can’t time the charges right and panic-fire it at close range like it’s a Crusher. The learning curve is really steep, and in a tactical sense I’m not totally sold that it’s particularly effective given the charge-up time (get at me Yoshi).
Gladly!!! I love the timing of Thunderbolt+ charge (The ++RT upgrade just makes it that much better). Conversely, the MX upgrade, which makes it charge faster, screws me up badly because you don’t get nearly as long a time holding the charge before you start taking damage. As you all know I’ve been using it heavily ever since it was introduced during the kickstarter campaign. In my history of playing Challenge Mode on Insane it accounts for more than 50% of all my kills including missiles, even in my recent high scoring runs with Ogres everywhere. I find it to be an amazingly effective weapon.
It does take practice to learn the rhythm but I charge the weapon while I’m out of combat and then go looking for bots while holding the charge. We get a nice long time before it starts to hurt us, much more generous than Descent. If I get surprised or face multiple bots the best tactic I’ve come up with so far though is to lead with a Hunter and then let fly with TB. The stun that the robot gets from the missile gives TB time to charge. Also 1 Hunter plus 1 TB blast kills almost anything other than the toughest supers who only need 1 additional hunter to finish off.
When it comes down to it, holding a TB charge around a corner is one of the easiest ways to kill a Hydra, at least it was until they got their huge speed boost in this drop…
I’m also a big fan of getting 3-4 kills with a single shot of primary! Usually one would need Devastators and luck for that kind of fun! 😉
Forums are currently locked.
Previous Topic: Everspace Expansion Next Topic: 2 small changes to make the post preview better