(He/Him)

New Zealand

Forum posts

Forum: Factorio

Thread: Rocket rush category

Started by: petr.adamekpetr.adamek

I have my 1h 36m run up for Rocket Rush Default Settings from a while ago, but I don't know of anyone else running Rocket Rush anything.

Even I haven't for ages, it remains on my list of things to revisit sometime. But RR DS doesn't seem to end up a notably different speedrun to Wave Defense, other than lacking what makes Wave Defense special & interesting.

Rocket Rush Set Seed Any% would end up a super interesting test of pure execution, which isn't what I like to run, but I'd definitely watch. I don't think anyone has run that ever.

However I do still like Rocket Rush Deathworld Marathon, which I got down to 2h 20m and that'll certainly get faster. There's currently no category using expensive recipes, which alongside Deathworld + all techs emphasising a different combat meta, imo makes RR DWM a more worthwhile contrast to Wave Defense.

Basically, no-one speedruns any form of Rocket Rush, but it could become a category if just a few people started running the same ruleset.

 

Forum: Factorio

Thread: Speedrunning with expensive recipes and 80% evo

Started by: SpleeSplee

So I've tinkered with Rocket Rush, and on Default Settings it doesn't seem that far from a Wave Defense run. There's a bit more steel for a silo, fine, otherwise the same three-copper-lanes Wave Defense base does about enough.

So there might be at least two interesting directions to take a Rocket Rush speedrun. Firstly, an open map settings / chosen seed deal like Any%. It'd be another race of execution & detail, but with a completely different base design. Also personal roboport mechanics.

I've gone the other way: further into resource constraints & combat, with Rocket Rush Death World Marathon. There's the novelty of speedrunning expensive recipes, the dramatic risk of Deathworlds immediately sending you big spitters, both pleasantly offset by the shortcuts & generosity of Rocket Rush.

With a few ideas but no real plan I first got 2h 49m 32s. I now have a plan, but not a design, and my most recent 2h 19m 54s is still deeply suboptimal. There's a lot of improvement left to enjoy, it's still early. But it's a viable thing to do, which is still speedrunning Factorio, but again different.

Conclusion: Rocket Rush Death World Marathon is a pretty accessible (or race-able) three hours, people play three hours of Factorio all the time.

 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

Amazing! In only about a week (?) WD is up to 8 participants, and there's barely any strategic cohesion across them. The "meta" is wide open, if it exists at all. The top 5 runs make bases that separately do:

• single giant parallel RF & LDS & RCU build in the late game
• small parallel RF & LDS builds and then a big late RCU mixed build
• RCUs then LDS then RF each in separate builds
• a weird new mixed build to cache blue chips & modules, then RF, then RCUs, then LDS, all big
• small early any%-looking mixed build for LDS and RCUs then a giant RF late

The strats are all over the place, good times. Someone makes LDS before red chips, someone else puts down pumpjacks before putting any kind of miners on copper. There's even a sushi belt around a silo. Madness. Defensively:

• one builds landmines, landmine pushes for a bit of gun damage & lots of mining productivity, mainly defends with landmines on & around walls
• one turret pushes, gets a lot of turret upgrades, a bit of mining productivity, skips landmines
• one landmine pushes for mining productivity, a lot explosive damage upgrades, and some gun tech, then landmines close around the base & lets the turrets die
• one prioritises grenades, throws them out from a moving car, a bit of old school DS worm dance too
• one does mass close landmines, ignores the wall entirely, only really pushes for mining productivity

But they also grow different, variations include:

• a pretty normal looking mall for basic intermediates then handcrafting everything
• a very normal mall making intermediates right after the smelting lanes like a sensible person
• a mostly normal mall on the side of a big green chips build
• a mall on a simple bus which keeps getting recipes reassigned
• an entirely handfed "mall" that keeps getting reassigned, let's call it "self service"

So yeah, very many options! Good times. My favourites are the runs that see the silo attacked as the rocket builds, they don't overspend on defense 😃

unique_2unique_2 and blazespinnakerblazespinnaker like this. 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

(edited: )

> " solvability is the enemy of gameplay", for you maybe, for me it is the essence of gameplay. I see Factorio as just a search for shortest path in a very large and complex graph. You see it as 'donkeyspace'.

If you'd checked the link, you'd see it's not "for me" but a specific term used by professional game designers with decades of experience on major label titles to discuss designing interesting games. Donkeyspace is a thing to be cultivated for a game to be worth playing.

> But the real solution there is PvP and not speedrunning.

That's ... no ... Dokeyspace is a think PvP games need in order to be more interesting. If you want to understand motivations, reading all of Sirlin's Playing To Win is a good start. (Edit: I think the Solvability article is part of a separate series about his other design work, it's all worthwhile & interesting.)

 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

Nice work, Macros! I was secretly hoping that it'd take people a little longer to catch up with me given I worked my time down from 1h 45m over the last two weeks. But, bam! One published run, there you are. Lift your game Splee, or get pushed out of the hot seat.

I liked your on-belt red chips buffer loop, and blue chips build. I think if you'd had the car, and not overcracked light oil, you'd have had this by quite a bit. Interesting that you also ran this entire thing from a powerplant eating just one belt of coal.

I'm still fond of the slow LDS / RF build I do, then a giant RCUs thing that catches up. I am aware, even as I do it, I'm probably overbuilding for beacons and prodmods in the late game. I'll have to look into that, and not cutting the prodmods promptly is a bad time for me. Though what I really want to do is get the copper & chips buffer in place to start the main push on the daytime of 4. That just means me being faster in the early game.

Your walking landmine push was surprisingly quick, but my current strat still commits me to turret pushing since I'm prioritising buffering 2 lanes of copper ASAP instead of oil before the push. But the landmines are definitely in the WD [Hard] strat, maybe they make it into my Normal as well.

I also like that it's tricky to tell who's "ahead" or "behind" at any point in time, as if that even makes sense with such different builds. I can't even easily steal your cool reds / blues build because it has a different role in your wider strategy. Good times. Looking forward to seeing how Cobai does things.

 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

(edited: )

> I'm targeting sub 30 minutes now for WD

Have you completed a run yet? What's your PB currently?

> It annoys me to no end there isn't a more methodical method of calculating minimax here

Bear in mind that solvability is the enemy of gameplay. Large & accessible Donkeyspace is exactly what makes for strategic variety. And yeah that's a PvP reference, but there's a shared sentiment: if you can calculate the "correct" move, the game stops being interesting, because that's all people do.

So WD Hard is a good time for everyone. The copper bottleneck doesn't matter at all, probably not even for BP fixed seed runs, the limiting factor is just defense. And, sure, yeah, that's in the name of the scenario? Good hint. Fair play.

Edit: Cobai! yeah, I'm enjoying WD, and it's fair to say that anyone's first run is pretty rough. Even my first run on Hard, see above. The key seems to be to have an awareness of where the power spikes / points of danger are. If you're actively keeping slightly ahead of the curve it's ~fine~ survivable. When you're surprised about falling behind (unavoidable on anyone's 1st run), distracted from whatever you're building, responding to a potentially lethal breach, that's when you slip further behind and spiral into despair.

Saying "I did another run but wasn't recording" is a bit bullshit, so let's say I'm highly confident that WD Hard will be viable in under two hours with only a few major adjustments to my route.

CobaiCobai likes this. 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

(edited: )

> yes, absolutely. You are agreeing with me I expect in every sense.

Maybe, it's not clear. You were saying this in the context of "you have to increase production of intermediates" and so ... we're agreeing that's incorrect?

> Though your total build will end up being much larger than one that appropriately used buffering

Sure, but again, who cares? The metric is speed. Less is faster to build than more, sure, but the important thing isn't the base but whether reaching 720 each of RF / LDS / RCU is net faster.

> Ironically, I had to remove all (mostly, not tricky stuff) the belts from the BP in order to make it happen and hand do them, which is sort of the opposite of what I normally do.

I mean, yeah, that's what I said? That's what you'd have seen me do? There's no irony, it's the fastest way, forcing new & unusual decisions is why I wanted to run WD & RR.

> Reducing entities is very important to cut down on robot placement. 2 assemblers require extra inserters, belt, power.

We're only talking about eight T3 assemblers vs sixteen T2 assemblers. Thirty extra inserters and half a stack of belt is lost in the weeds compared to what you get from the module slots. I need ~800 blue circuits in all and I save ~2k green circuits making them at 16% productivity. Don't get stuck on the wrong bottleneck.

> Getting logistics bots will be very useful

I don't think I'd consider logistics bots. Why have logibots bring you things from a mall when you could just have more crobots taking them straight to the blueprint? Which you'er importing, so you don't even need items on hand to manually place something to C&P. Logibots don't seem fastest, but someone could post a video of a run showing me otherwise.

Speaking of which, I've now got my WD run down to 1h 21m 9s, so I'm pretty happy. The 82 minute mark was my milestone (for now) for pivoting into Rocket Rush. There's gonna be a bunch more tuning, more likely a complete overhaul, I'll actually need a real steel lane this time.

 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

> If you did just waterfall (not going back and increasing intermediate production), you'd end up wasting resources on unnecessary massive unused over production at the last minute

I can't even tell what you're imagining that has this problem. Isn't this part of what buffers can address? You do one build, initially overproducing, then later you expand consumption and now it's underproducing, so you lean on the buffer. There isn't "unnecessary massive unused over production at the last minute", a well paced conclusion is marked by running the buffers to empty.

Also "massive unused over production at the last minute" doesn't even matter if it's the fastest way to get 720 each of RCUs, LDS & RF. If I hit all those numbers as early as possible, I don't really care about production rates 5 minutes later. (Quite possibly slowed to a crawl because buffers are empty.)

> Worth noting, we get 8 ASM3 to start with. Using those tactically would be wise ... PM production ... RCUs

If you ignore the module slots, one T3 assembler manufactures at a slower rate than two T2s. You're really thinking of using your only eight assemblers with 4 module slots to assign recipes that can't accept / aren't worth prodmods?

> One of the things I'll do for sure is get rid of the exoskeleton and replace it with batteries at the very earliest

I thought I'd need to do that, or at least sometimes turn off the exoskeleton, but I don't. I simply don't BP / C&P the long straight sections of belt in smelter columns or the mixed build. That's ~50% fewer objects for crobots to place, so ~50% less bot energy used, as well as being simply faster to build straight belt lines myself. That's enough.

 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

> but it'd be good to get some copper wire going as well before oil. Setting up some GC could be useful as well
> Maybe if you go down the PM1 in ass2 for RCUs, beacons would be worth it?

It doesn't look like you've even watched the run or are familiar with what I'm already doing.

So yeah, I have it down to 1h 24m 56s and now I need to rebalance the petrochem numbers. I'm faster now than I was when I planned that build, and now there isn't time to cache the solid fuel buffers needed for the RCU section. Everyone likes seeing a hard red powergrid as the rocket goes up, it's fine...

 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

> because I'm still evaluating strategies.

tbh this would be the best time to be sharing & collaborating on this stuff? Have you recorded any of them? (Also: how do you time the RF / LDS builds? I usually have small builds for those I put down immediately after petrochem, and they take longer to get their work done while I'm arranging a larger RCU mixed build that "catches up". It'd be interesting to see what routes other people take when research isn't a prerequisite.)

Rerolling & map luck are already things in other categories, rerolling for copper vs rerolling for peaceful maps is something I don't like, but ... accept. And yeah, outposting almost certainly not worth in RR / WD (except maybe oil, as necessary).

Artillery in WD is an interesting call. Currently in 1.1.32 WD [Normal] you need to destroy ~10 nests in the first ~40 minutes for turret upgrades or be overrun by mediums. Or have some other defensive plan fully online (or a rocket up?) in that 40m. Which is also why I currently like WD more than RR, because there's more to address than just "how quickly can you smelt copper?", more Donkeyspace to explore. WD (unexpectedly) gives you all the non-space science arty upgrade techs too, so sure, I'll try those out. Quite possibly easier than clearing those bases myself.

> the video requirement

... Is actually pretty important. It's not just "can you really do the thing?", it's also whether the run shows tactics/mechanics/etc interesting to both spectators and people who are getting better at Factorio. Saying "category rules optimised for youtube / twitch views" is a round-a-bout way of saying "Factorio how people want to see it played", that's what views are. Nefrums noted he didn't create all the strats he runs. The streams / videos are how we can observe skill, collaborate, learn, and get better, both for speedrunners and non-speedrunning Factorio players.

It's probably also related to why most categories don't allow blueprints, since BP imports shift a huge chunk of the skill expression out of the run, away from where it can be appreciated. Videos of runs help a community get better.

> with no videos

I didn't say no videos, I linked 4 videos, and my Youtube is right there. They help inform whether WD is sufficiently interesting to warrant recognition (I hope so?). Marathon Deathworld Gotlap is well defined, but probably nothing really new is going to happen there. Too many categories ends up detrimental if everyone is just in their own world, playing their own ruleset. The flipside of "community" is that collaboration requires a critical mass of people working on roughly the same things.

> open a thread for folks to post videos

... literally what I did there

> I think if I were to try it again, I'd get more bots maybe.

Record & post it here - unique_2 is wondering about their RR strats too. If the conversation and techniques are interesting enough to draw in more attention, that's a great way to get more people voting for, and even competing in, a potentially new category. Anti even said "they usually pass the vote if we find at least 2 active runners in a category" so ... just do that? Problem solved?

blazespinnakerblazespinnaker likes this. 

Forum: Factorio

Thread: Wave Defense and Rocket Rush (again)

Started by: SpleeSplee

(edited: )

So I see this has turned up before. Once again a new user with no existing runs appears to discuss new categories, specifically Wave Defense and Rocket Rush. Perhaps this time it goes differently, since I can already show runs of WD - my best so far is just under 90 minutes - these scenarios do some pretty weird stuff, links at the end. I was expecting RR to be fast, but I wasn't really expecting WD to end up a faster rocket than Any%. These are the sorts of surprises I enjoy.

Reading the previous threads, it does seem that WD & RR are already appreciated for their strangeness & strategic variety, just, not actively run. Disadvantages of Wave Defense and Rocket Rush as introductory / gateway categories vs Steelaxe and Gotlap:

• Not doing research isn't very Factorio
• No achievements granted in scenarios
• Freeplay is "the intended way of playing Factorio"
• Starting with a bunch of free stuff is crazy, the bases end up weird

But also, things that WD and RR do as introductory / gateway categories better than Steelaxe and/or Gotlap:

• Steelaxe and Gotlap don't teach you to kickstart a base, just burner city (though WD and RR aren't "normal" base kickstarts either)
• Burner city is necessary but not fun, WD and RR start with the fun (subjective, yes)
• WD and RR still get a rocket up, teaches buffers & mixed builds, still build petrochem, very Factorio
• Starting with a bunch of free stuff means wide strategic landscape
• Wube called it Rocket Rush so probably they want it done fast

So, I reckon they're pretty interesting & accessible, and worthwhile having alongside Steelaxe and Gotlap. All that said, the real question is a/ whether anyone cares enough to run WD or RR, and b/ whether the resulting runs do anything interesting worth watching. For (b) I offer runs of Wave Defense on Normal:

• in 1h 41m 44s
• in 1h 36m 13s
• in 1h 32m 45s
• in 1h 29m 42s

(I'm not currently recording game sound, the 1h 32m run includes some includes retrospective diagnostic narration.) I'm looking at these with a ruleset closely aligned to DS - only 1 player, random seed, no imported blueprints, timer goes from map preview to rocket launch victory message. Normal difficulty for WD, Default Settings for RR, but maybe we loop back to RR in a bit once I (or anyone else) has rockets going up.

These are certainly different to "normal" categories of Factorio runs, maybe they're even interesting.

(With appologies to Bilka and others who might be getting multiple notifications as I read the markdown docs and figure out formatting my lists correctly and making links link instead of embed.)

BilkaBilka, unique_2unique_2 and 4xel4xel like this.