Transmission: Death Crown
It’s happening – I’m going to deliver a full run of Death Crown‘s 30-level Domination mode with some strategy explainers before each level. It’ll take around two hours.
Wednesday 28 April. 9:00PM UK, 10:00PM Central Europe, 4:00PM EST. My Twitch channel can be found at twitch.tv/electrondance.
Previous Transmissions are available on the dedicated E/TX YouTube channel.
Update 29 Apr 2020. The stream has now been archived on YouTube:
Download my FREE eBook on the collapse of indie game prices an accessible and comprehensive explanation of what has happened to the market.
Sign up for the monthly Electron Dance Newsletter and follow on Twitter!
6 thoughts on “Transmission: Death Crown”
The stream last night was very interesting and I had a few general thoughts.
Death Crown seems very different to other real-time strategy games in that by showing the AI’s working/hand, it humanises it. You can see that the AI can only do one thing at a time and you can see it thinking, which is very cool. But that said, it’s unable to do what humans do–I mean, duh–but it’s unable to sell/destroy its own structures. Based on what I saw last night, tower rushing seems like one of the most important manoeuvres as part of a broader map specific strategy. I’m reminded of my old friend George who used to use sandbags in Command & Conquer to sandbag all the way across the map to the enemy base and sandbag their harvesters in, effectively stopping them making money. Boom. Sandbags were cheap and enemies only pathed around them.
Like you said in the stream, perhaps destroying structures is a level of sophistication too far for the AI. It’s interesting to me that Death Crown tries to make the AI look fallible or imperfect with the hand when other RTS games make no such effort but also then attack you on multiple fronts simultaneously, and that can feel horrible. Tooth and Tail is a really good example of this: you play it with a controller and control a single commander who wanders around the battlefield, rallying units to their position, building stuff and burrowing back to your base. A bit like Hellfront: Honeymoon. Effectively your commander is the mouse pointer. But in single-player, there’s no commander for the computer so it’s not clear how the AI behaves. They just attack from all kinds of angles at the same time without having to march a commander there like you do.
Another thing is the maps. So in Death Crown they’re handmade and the AI is perhaps the only real variable. In Tooth and Tail, they’re procedurally generated. The handmade maps mean that there are optimal strategies, perhaps even a single strategy that works best, if you can find and execute it. With proc. gen. it’s more about adapting as you explore the map so there’s only so much strategising you can do before the ‘twitchier’ reflexive damage control aspect of RTS comes in when things start to go awry.
Quick tangent into turn-based strategy: I’ve heard quite a few folk grumble when a TBS game seems more like a puzzle game because they feel like they have to use a specific strategy to beat a level or situation. I’m never quite sure how to feel about this because surely any kind of tough situation, whether engineered or emerging from natural play, becomes puzzle-like? Every mistake pushes you into a tighter spot and trying to find the route out is half the fun. The opposite to this situation is: any approach will work which… sounds more like a power fantasy sandbox than a challenge.
Anyway, I’m guessing that the more puzzle-y aspect of the Death Crown maps was a big draw for you? On top of the bite-sized RTS (that you gorged on!) 😉
I really appreciated the slideshow you had as well.
Also: since when did Twitch become the new Babestation? I logged in last night just before the stream and saw lots of ‘Just Chatting’ bollocks being recommended to me. The juxtaposition between that and Fortnite and Minecraft right next to it was weird.
Also also: I wonder how you’d fare with Death Crown against another human who could tower rush you… you AI bully 😉
Also also also: the black on white colour scheme would be too harsh on my eyes! There are loads of really nice muted ones… and some downright criminal ones.
I did catch the stream, although I had to mute after a while as it overlapped with a D&D session!
Watching the game gave me a better sense of it. Quite an interesting title. Streamlined, but certainly still very much an RTS, with that same frenetic sense of pace.
Makes me wonder if you’re tempted to do an Ouroboros Sequence of strategy games!
I thought I’d wait until I had time to respond to your comment properly. It’s now 23:04, a whole 24 hours since the stream.
While I picked up on the hand in The Watchtower article, I don’t think I made the more significant connection that the hand “humanises” the AI, makes it seem more fallible. I’m trying to rack my brain thinking of another example of this but nothing comes forth.
Tower rushing is a key tactic but it also needs to be used with care; during the early days of Domination I was burning up gold with extreme tower rushing. While I had partially hemmed in the AI, it was able to rain hell down on me, because it had spent the time developing proper infrastructure. There’s a sort of balance of the AI’s disadvantages and buffs (I think of AI War here, where the titular AI gets infinite resources, but follows strict rules in how they are applied) and, while I’ve laid out the strategy which dismantles some of the levels with efficiency, dear God it was hard-earned. So many losses on levels 10, 12, 14, for example. I’ve been working on Domination for 3 months and I still feel vulnerable on two levels.
For sure there’s “another game in there” if the AI could demolish it’s own buildings but I think it’s just fine as it is.
Don’t start talking about proc gen vs handcrafted, we’ll re-ignite a conversation about roguelikes again 🙂 But now you’ve done it, you’ve thrown in turn-based strategies too. This takes me back to how I like roguelikes but roguelike puzzles (like Lineglitch or Ending) because I don’t tend to like that search for the fancy dancesteps, the one optimal path. I like developing rules of thumb for staying alive in a roguelike, but finding a fixed path just doesn’t excite me the same way. Death Crown is more “rules of thumb” than fixed paths although I’ve certainly done a good job in nailing down some reliable battleplans. But, yeah, the AI can never be counted on to give you the exact same game each time. That’s pretty wonderful.
I have wondered about a Death Crown match with another player (via the “Versus” mode) and obviously I’m currently top-of-my-game which means it would be pretty bad for whoever I played with. However, I fear playing against someone who is as well-versed in Death Crown rules and strategies as I am may actually be dull, that the game will reduce to near-stalemates with minor errors responsible for any change in game direction, or be a contest of fast hands. And that’s not me at all.
Thanks for watching! I don’t think I’m tempted to do a strategy Ouroboros, because I don’t think I’ll ever have enough time to finish a game of AI War.
Right, it’s 23:49 now so that’s 45 minutes to write a comment. Makes a change to be speedy 🙂
Bloody hell I mean’t Cityglitch not Lineglitch, he says after sleeping.
Yeah, I know what you mean. I feel like I’ve seen the ‘enemy hand’ somewhere else but maybe I’m just thinking of Dungeon Keeper or something.
And yes, I certainly didn’t mean to downplay the skill required to conquer Domination mode! And get to the top! I’m still amazed you were able to unpack so much of what goes into playing it while playing it. I’d just be like ‘So, uh… I’m going to… Nice… Right, so this here… [long silence] Okay, we’re on to the next level.’ You’ve seen enough of me trying to talk while playing a game…
Just to be clear, it wasn’t that I was defending the skill required, but I felt like I needed to defend the game! i.e. tower rushing wasn’t a flaw. But you’re not really saying it was a flaw, either.
I’ve done an enormous amount of Death Crown training, as you know, and that meant it was easier to talk because there was some auto-pilot I could rely on. The dress rehearsal of the stream didn’t go as well but I think the dress rehearsal is always good for firming up an “improvised script” in my head. I think this is probably the best stream I’ve done though, I was still very “present” in the final minutes. I’d have liked a better score, of course, but you can’t have everything…