Tuesday, December 01, 2020

Universe 647

I can confidently say that Remembrance of Earth's Past is my favorite science fiction series that I've read in years. It's been both surprising and satisfying, insightful and thrilling, human and epic, audacious and inevitable. When I started reading The Three-Body Problem I wouldn't have imagined ending up where I did by the end of Death's End, and I feel impressed by the ride.


Throughout the series there's been a strong focus on basic research: not so much a specific invention created by a brilliant engineer, but rather, groups of scientists gradually delving into the fundamental mysteries of the universe. From the problems with particle accelerators in the first book through the strong nuclear force of the second book and various experiments with universal constants and the curvature of space in the third book, the most exciting question tends to be how the universe works rather than how we can exploit it. This research ultimately has practical applications, but the fundamental models are the most important part.

I love how the science gets more ambitious and stranger as the series goes along. It dovetails nicely with the telescoping time periods throughout the series: we advance through months, then years, then decades, then centuries, millennia, eons. Late in the series the characters describe things that seem bizarre and impossible; but, just think of how much commonly-accepted science today would have seemed bizarre and impossible in Benjamin Frankin's day. There's a humility in Liu's gaze, recognizing how inferior we are to the people we will become. I also really like how, even in this fantastic future, there are detours and dead ends: Earth civilization wastes time investigating theories that prove to be unfounded, and invest in ventures that prove fatally flawed. It feels so much like real life: Yes, science advances and gets better over the time, but not every step is a step forward, and while we're shaking our heads at the fools before us who believed the wrong thing, we should remember that they were fixing earlier problems from the fools before them, and one day our new ideas will seem just as flawed to those who understand more deeply than us.

In particular, I thought the presentation of lightspeed travel was especially interesting. Throughout The Dark Forest, I thought it interesting that, while this is a hard sci-fi series, it doesn't seem to grapple with the implications of relativity as much as other sci-fi books have, as we would expect travelers to experience less time than those who stay still. That does get addressed in Death's End, and in some ways the third book is about that problem, with some explanations that seem to contradict Einstein; but it's very earned, it doesn't feel hand-wavy, but rather a plausible idea for another sort of science. Just like we used to think that atoms were indivisible, we may learn one day that universal constants are not, in fact, constant.


Mostly some random thoughts in here:

I was a little surprised that the fall of Byzantium wasn't referenced again later in the book after that startling prologue. But we the reader can parse it: it seems like there was a fourth-dimensional collapse in our spacetime coinciding with the siege, leading to the exposure of previously hidden ways and contents. I wondered at the time whether there was anything special about that one woman in particular that allowed her access to such a state; but, given the later events on board Blue Space and Gravity, it was probably just that a portal opened in that one place inside a tower, and she happened to find it. That portal decayed like the fourth-dimensional space around Blue Space decayed. (It's interesting to think of how Earth's history might have been different if it had decayed slightly more slowly!)

I really enjoyed the characters in this book, and it felt like we had more access to Cheng Xin's thoughts and feelings than we have to previous protagonists. Her introduction seemed a little stiff, as she's mostly presented as an idealized feminine object of Yun Tianming's affection, but once she takes center stage she really earns it. Late in the book we're reminded that, thanks to her hibernation and light-speed travel, she's somehow still only in her thirties, which seems extraordinary considering all she has done and the outsized responsibilities she's held.

Similarly to how Shi Qiang bridged the first two books, Luo Ji bridges the last two. He does take a more active role in this book, which is cool; he's a fun and explicitly nuanced character, and it's interesting to see him play multiple roles over time. But it does make me freshly surprised at how Wang Miao completely vanished from the series after being a protagonist for the first book. It's not bad, and as I noted in my previous post I do like how the epic scope of the story is reinforced by all the fresh faces, but the disparity between Wang and Luo is curious.

Cixin Liu continues to innovate and surprise with the form of the novel, weaving in new elements. The first book stirred historical fiction and detective stories into hard sci-fi, while the second book used political thrillers. Death's End once again brings in a new genre from left field: this time, it's fairy tales! Roughly midway through the book, the main narrative pauses for a long and shockingly well-written fantastical story, presented as three separate fairy tales but really one continuous story about the Storyless Kingdom. The stories are so striking and so pertinent that they remained fresh and sparkling in my mind throughout the entire rest of the novel, much like how the Cultural Revolution loomed over the events of The Three Body Problem.

The fairy tales are well-written, but also very cleverly constructed, packed with allegories and clues and hints. I felt really impressed once they were revealed, much like how I feel impressed at, say, a long-laid David Mitchell revelation. The main thing that I'm left wondering at the end of the novel is the significance of the umbrella. Painting into two dimensions seems quite clear, an unheeded warning about the flattening threat. Given that, it seems like the spinning umbrella offers a solution of some sort. Spinning and rotating is significant throughout, like the watery vortex in the Norwegian sea. I wonder if, given enough time, Earth's scientists could have prepared some defense against this, or at least a way to buy more time. Perhaps curving space within the Solar System could have slowed the advancement enough to leave another way?

In other allegories, Prince Deep Water seems more clear. He was unaffected by the painter because he was already two-dimensional. That seems like another missed opportunity for escape: if Earth had somehow anticipated the strike, they could have followed the nihilistic strategy described by Yun Tianming and voluntarily reduced themselves to two dimensions. Which would have been an incredible loss, but would have guaranteed the continuation of Earth civilization.

Part of me was hoping that Bill Hines would make a reappearance in this book, as I also felt like there were some loose ends from his Escapist nam-shub as a Wallfacer. He doesn't, but I was extra-surprised to see Manuel Rey Diaz, out of everyone, providing a critical component to humanity's salvation. There's so much misdirection in this series, I love it!

The progression from Wallfacer to Swordholder was also really cool. It's probably just as well that Swordholding wasn't the main focus of the book, but I enjoyed this other sort of psychological contest. It's a game theory-y thing rather than an obfuscation thing; or rather, there's still some obfuscation since sophons can't see inside someone's mind, but it's reduced to a binary yes/no with catastrophic consequences, rather than an elaborate and hidden scheme. Anyways, the game theory part is really cool, explicitly harkening back to the Mutually Assured Destruction of the mid-20th-century, and realizing that mankind rationally chooses to place its fate on an irrational choice is hella exciting. Vindictiveness becomes a virtue, revenge a sacrament, one's capacity for evil the only plausible route to a good outcome.

There's a brutality running through the book. Over and over again, Wade is proved right and Cheng Xin wrong. A cold-hearted decision with limited casualties would have spared an entire civilization from extinction. It seems like the end of the book is on Cheng Xin's side, though. She "loses" every battle she's in: the high-stakes standoff of the Swordholders, gambling on the future with Halo City, and so on. But the ultimate long-term stakes are the heat death of the universe versus the big crunch, and only Cheng Xin's policy of love will lead to the right direction. Who knows, maybe those few kilotons of mass will be the deciding factor after all.


So, yeah! I do feel a little wistful at having finished this series and no more books to read in it. Which is kind of surprising; the second book ended on a thoroughly satisfying note, and I was wondering if the third entry would spoil that and follow a good note with a bad one. Not at all! It somehow manages to get even better and more ambitious, drawing me in to its telescoping story. It goes deep, it gets broad, but still stays personal, anchored to characters who we know and love dearly by the end. It's a really remarkable work.

Sunday, November 15, 2020

Lookit The Pretty Boids

I'm making good progress on my first Captain-difficulty Stellaris game. I'd initially planned to wait until finishing the game to write it up, but a ton has happened so far that seems noteworthy, and my brother recently encouraged an update, so I thought I'd chime in with my progress so far. I recently returned to the game after a detour with BG3 Early Access, and after a brief period of "Whoa, where am I and what am I doing?!" I've found my feet again.

This was my first game rolling a custom civ, after always playing as the United Nations of Earth on previous outings; now that I have a better feel for the game mechanics and my preferred playstyle, I felt more comfortable customizing a new approach. After downloading the free species portrait pack from Steam, I picked a really pretty peacock-ish portrait and dubbed my species the Phasianidae. I initially planned to do a pacifist run, and so chose a Pacifist ethic, which decreases Sprawl (allowing for bigger empires) and increases Stability (leading to higher production, trade, and immigration); you can't start wars on demand, but can declare an ideological war on any civ that doesn't share your ethics, which in practice is almost all of them. I also opted to be a Fanatic Xenophile, which gives you more Envoys to help keep enemies off your back, and increases Trade Value to supercharge your economy.

I'd always stuck with a Democracy in my Earth games and wanted to try something else. I eventually settled on an Oligarchy, mostly so I could run both the Meritocracy and Shadow Council civics for +10% output of Specialist and Ruler production. It's been pretty good, but I did end up kind of regretting not trying a Dictatorship or Imperial government for the extra Edict slot; I'd forgotten just how limiting it feels to only have a single edict for most of the game.

For species traits, I chose Rapid Breeders for +10% growth speed; at least in the games I've played so far, from the midgame on the single most important factor in power seems to be the number of pops you have, so I've started juicing everything I can to get more pops. I toyed with picking up a negative trait too but decided against it; I was a little tempted by Weak since armies aren't very important, especially in a Pacifist game, but it also hurts Worker Output so I passed.

As I recently noted to my brother, one challenging thing about Stellaris compared to Fall from Heaven 2 or other 4X games is just how long it takes to figure out how strong your start is. In other games, you can quickly explore your immediate surroundings and decide early on whether it's worth continuing the game or if you're better off re-rolling. In Stellaris, though, it takes many decades (in-game!) to figure out the hyperlane patterns, find nearby habitable planets, and determine who your neighbors are, how they feel and how close to you they are.

For the origin, I selected Galactic Doorstep, which sounded interesting: you spawn with a disabled Gateway in your home system, and there's a longish quest chain related to it: things coming out of it, investigating it, and gradually learning how to reactivate it. I'm glad that I did it and it was a fun change of pace, but I think I'll likely go back to Prosperous Unification in the future, or maybe switch to a new one if and when I pick up DLC for the game.

That Galactic Doorstep almost cut my run short. It's really helpful early on: you receive a nice chunk of Alloys relatively early on, which is huge in the early-early game before your industrial base gets off the ground. Eventually, though, some hostile aliens come through, and my pacifist self was not ready for them. They wiped out the minor defenses I had in the system, then started bombarding my starbase. I frantically upgraded another Outpost into a Starbase, built a shipyard, and started churning out new corvettes to counterattack. In the meantime, though, my Fleet Power had declined to 0, so all my neighbors started throwing their weight around, making demands and offering me "protection". I eventually managed to fight off those intruders, but oof, it was pretty hairy!

After the early-game benefit of free alloys and such, the really nice thing about the Galactic Doorstep is that you can permanently unlock the ordinarily-rare tech to reactivate Gateways (similar to how you unlock tech after scavenging alien debris), and after that the tech to build new Gateways. These are still very expensive techs, so I didn't research them as soon as they became available, but I still ended up with Gateways a lot quicker in this game than the previous one. One of the best things about this was that piracy has been a lot more manageable. There was another deactivated Gateway in my territory at the end of my empire, and after reactivating that a lot of my piracy went away; by the time the rest started to become a problem, I could start building new custom Gateways in other systems.

I feel like I definitely know the game a lot better this time around, and besides the importance of Gateways, another thing I think I'm doing much better this time around is starbase placement. In my previous game, I focused mostly on strategic location, building up starbases at borders and at internal chokepoints. Those seem to be almost useless, though; by the time of a late-game war, fleets are a lot more powerful than starbases; and because borders can shift so much, a starbase could become irrelevant. I also now realize how important Stability is (for keeping up production, which is what ultimately wins wars), so the Deep Space Black Site, which I initially thought was an almost useless building, is now usually the first thing I build.

My new general system is: First, only ever build (upgrade) Starbases in systems with colonies. This gives the Stability bonus, allows directly collecting Trade from the planet, lets you eventually link up with your gateway network, and provides extra defense if the colony system is ever invaded. Next, try to specialize each starbase. My first priority is typically Trade: I check to see what nearby colonies aren't yet covered by a starbase, and what claimed systems have trade value, and figure out the best place to put a Starbase and how many Trade Hubs I'll need to cover everything. (Ideally planning long-term with the 6 slots I expect to eventually have, though I'm now also much more happy to do an initial buildout and retrofit later, like starting with 2 starbases of 2 Trade Hubs each and later moving all 4 hubs to the same Starbase).

Other than Trade, I build some starbases that focus on Shipyards, then others with Anchorages. Honestly, I learned this from the AI in my last game, realizing that all the Starbases I conquered followed similar loadouts. Each Starbase then builds buildings that synergize with its primary function: Fleet Academies for Shipyards, Naval Logistics Offices for Anchorages, and Offworld Trading Company for Trade ones. Any extra buildings get filled with Resource Silos.

In this game I'm not building any combat-oriented things in my starbases like missile or gun batteries, communciation jammers, etc. Partly that's because of my pacifist leaning, but I also was pretty underwhelmed by these in my previous game: again, by the time you're fielding battleships, fleets will outperform starbases, and the enemy AI seems to be smart enough to not invade a system with a significantly stronger defense. I do think that, if piracy is a problem, it would make sense to fill out slots with Hangar Bays, and then retrofit those to Anchorages or Shipyards after turning on your mass effect relays gateways.

Like I noted above, it's hard to judge how good a starting position in Solaris is, and I lucked out in this game. I turned off the "clusters" option for Empire Placement and ended up with some more breathing room around me. In this game, I took a sort of "wave" approach to early exploration: my first Science Ship would just explore and not survey. Then a second ship would follow later and survey systems with habitable planets. This seems to work out pretty well: you can still find good planets to settle by the time you're ready to build colony ships, and you have a much better shot at figuring out the galaxy's topology.

As in all my earlier games, my first priority was to race for claiming chokepoints: trying to claim systems that would block off neighbors' expansion in my direction, letting me claim other systems behind it at my leisure. This typically follows a "snake" route, where I'll claim the shortest path between my homeworld and that chokepoint, instead of steadily expanding out in all directions. Of course, rival empires are expanding too, so it's good to have backup plans: if I realize I won't be able to reach a desired chokepoint in time, hopefully there's another one behind it that I can use.

In my particular case, I had a pretty clear priority order for expansion. First was towards Glebsig to my east, my nearest neighbor (albeit still relatively far away). I could see that he was also getting bottled up by the Elaamid Blessed Mandate to his east, so if I could contain him, he would be hobbled for the rest of the game. I poured my Influence into that push, and managed to lock him up. And then, I was reminded of a very unpleasant fact: your enemies will totally try to build outposts behind your borders! Even though it would result in non-contiguous territory for him, he sent some constructions ships through my chokepoint. Fortunately, closing borders ended that adventure pretty quickly.

UNfortunately, though, there was a way around this. As noted before, I'd completely neglected my military, dumping all my alloys into new Science Ships and Construction Ships. And so, when my Gateway invaders destroyed my fleet, I became Pathetic even to the Glebsig weaklings. They magnanimously guaranteed my independence. Then, a few years later, they canceled that guarantee. No big deal, except canceling a guarantee automatically creates a Truce between two empires... and a Truce allows both parties to ignore Closed Borders! Suddenly I had Glebsig ships sailing back into my territory again, making a beeline for an unclaimed system with habitable planets.

Fortunately I still had a construction ship in the area; because they're so slow and I was expanding in three directions, I made a total of three and would just park the ones I wasn't currently using. But it was quite a race between my ship and Glebsig's. I would charge in ahead of him and start building an Outpost just before he started. Fortunately, he would also start right after me and spend months building it, only to lose it when I finished first. There were a few times when I wasn't sure where he was going, and it was pretty intense to open up the star map, try to anticipate which hyperlane he was steering towards, and try to get there in front of him. After some time, though, I eventually claimed the necessary systems near our border and kept him bottled up inside; from what I can see, the AI will claim within a few hops of a border (including a navigable wormhole or gateway), but won't claim further than that.

My expansion southwest was the next-most-urgent. Out here was the Democratic K'Taknor Commisariat, a Fanatic Militarist Democracy. I prioritized sending an Envoy there for peaceful relations. Here, my goal for a chokepoint wasn't just to block expansion, but to keep our border as narrow as possible: Militaristic empires have much higher Border Friction than other empires, and I wasn't at all prepared for a war. They had more options for expansion than Glebsig did so they made a more leisurely march in my direction and I made more rapid progress down towards them. I ended up not pushing as far in as I could have: claiming one more system would have still given me a choke point, and a rump tail of two more systems to claim, but also would have meant 5 shared hyperlanes between us (from that chokepoint out to their other systems), unlike the 1 we ended up with.

The longest expansion, both physically and chronologically, was to the Cirrulan Nation to my southeast. The Cirrulans had a ton of space to expand, and even now are technically larger than me. Because this was such a long drive, I identified a good three or four potential chokepoints between us, but I was able to get all the way down to the closest one. This also meant securing a disabled Gateway, another nice bonus for me.

One thing I hadn't really appreciated in my earlier games was that Fallen Empires are also de-facto chokepoints: not only a natural border to keep you from expanding, but they also block rival empires from expanding through them, too. I ended with two on my borders: the Thek Olak Archivists were in my southeast, above Cirrulan but not sharing a border with them, and forming a buffer between me and Elaamid. More intriguingly, the Ishni Shard, a Fanatic Militarist fallen empire, was far to my west. This was a positively luxurious expansion opportunity, with the shortest path taking over a dozen systems between my borders and their, with more spur trails off to the side, and absolutely no other rival empires in a position to compete for it.

My expansion was helped by my Traditions. In my earlier games I'd always chosen a tree, maxed it out, taken the Ascension Perk, and then chosen a new tree. But, a lot of really great early-game bonuses are scattered between multiple trees, so in this game I mixed-and-matched between Discovery and Expansion. I think my overall order was something like this:

  1. Discovery tree unlocked (faster Anomaly research and Map the Stars edict for faster survey speed and anomaly discovery chance).
  2. Discovery: To Boldly Go for faster survey speed.
  3. Expansion tree unlocked (faster Colony development).
  4. Expansion: Reach For The Stars for cheaper Outposts.
  5. Expansion: Colonization Fever for more pops on new colonies.
  6. Expansion: A New Life for faster pop growth.
Then I finish the remaining Discovery traditions to boost my research, the One Vision perk to speed future Unity, finish the remaining Expansion traditions to help with sprawl and growth, and the Technological Ascendency perk to boost science. I'm pretty happy with how this turned out and will probably follow a similar route for future games, unless I decide to opt for a more conquering-oriented game.

After this I return to fully completing each Tradition before moving on to the next. Like my game as a whole, I initially focus on research, then on economy/infrastructure, and finally on military. There's useful stuff in every tree, and in practice there's usually a pretty clear priority for what to take next. Diplomacy becomes useful after you've met lots of other civs, Prosperity is good once your cities are large, Harmony is never essential but always useful, Domination is good once you have rulers near the level cap, and Supremacy is great if you're planning war.

So far I haven't encountered the awesome "Someone Loves Us" quest chain, but did get a really cool early-game quest chain about a secret cult of traitors. I'm not sure if it's tied to the Galactic Doorstep origin; I don't think the text ever references it, but I haven't seen it in any of my Prosperous Unification games. Anyways, it's a really well-designed quest that pushes you to explore certain systems and take on some early-game combat, and eventually rewards you with a Battleship, which can be really game-changing in the early game.

On the downside, this was my first game where the Precursors did not give me Fen Habbanis III; I wound up with a ton of Unity and a bunch of extra resources in the system, but no planet at all. It was still worth doing, but man, the loss of a Ecumenopolis stings. But, unlike my previous game where I was stuck in a small space for a long time, here I've been able to directly colonize well over a dozen planets, and so far (~2350) I have plenty of space left to peacefully expand before worrying about overpopulation. 

Speaking of expansion: I did too much, too quickly in this game. Again, I'm making a concerted effort to grow my pops as large and quickly as possible, so I've been trying to get colonies off the ground ASAP. But, in my second wave of expansion, I had three colonies all going at the same time, which really tanked my Food and Consumer Goods. I've never had trouble with Food before and it was a little jarring to have too little of it instead of way way too much. I had to deviate from my carefully planned economy, throwing random Agricultural Districts and Civilian Industries on whatever planets happened to have free building slots at the moment. It took many decades to get my economy back under control. Fortunately, the Galactic Market was available to plug the gaps: I was running a huge surplus of Minerals throughout the game and frequently hitting the storage gap, so I would sell a ton of those, buy food and supplies, and do it again in a few months. My new rule of thumb is to only build a new colony if I have less than two current colonies, and am running surpluses on both food and consumer goods.

But, I do wonder now if maybe it was good after all that I expanded so quickly. Once I had that chunk of colonies fully developed, I catapulted into first place in Population in the galaxy, and soon after steadily rose towards the top of the Relative Power ranking. That might have happened regardless, but maybe the earlier pop growth makes up for the economic pinch of that transition period.

Since then, my colonization has been more consistent and less worrisome; I'm more limited now by the Influence to add undeveloped planets to my borders, and by the time I'm ready to colonize a new planet, previous ones are nearly developed.

One thing I'm doing a little differently this time around is beelining for Gaia Worlds. In my successful Earth game, I followed a pretty normal Terraforming trajectory: terraforming uninhabited planets into Continental planets, then later retrofitting previously-settled non-Continental planets to be Continental, then restoring Mars to be Continental, and, much later, picking the World Shaper ascension perk to turn everything into Gaia. The thing is, Gaia planets are so insanely good that you really want to get them ASAP. The 100% Habitability doesn't sound like much, but they also bring an unadvertised 10% boost to Happiness (leading to higher Stability, leading to higher production), and 10% to Resource production, overall making an amazing boost to your empire. Because of this, I didn't bother terraforming any early worlds or messing with habitability modifications or buffs. World Shaper is locked behind a rare tech, so I kept an eye open for that research path in Society. Once you have it, it just costs Energy and Time, both of which I have in spades. For new planets, I've been Gaiaizing them prior to colonizing, though I don't think there would be any downside to terraforming them after.

In this game I've also been thinking a little more carefully about planet specializations. I try to decide relatively early on whether a planet will focus on industry, in which case it will build Civilian Industries, Alloy Foundries and eventually the Ministry of Production; focus on science, which will build Research Labs and eventually a Research Institute; bureaucracy, which will build Administrative offices; or, occasionally, culture, which builds Autochthon Monuments, and eventually an Auto-curating Vault. Likewise, a planet will typically focus on a single resource type, like Minerals, and eventually build a boosting building, like a Mineral Purification Plant. I tend to stick one-off buildings like Gas Refineries and Synthetic Crystal Plants onto Bureaucratic worlds since I'm not giving up a multiplicative slot when doing so; I'm now finding, though, that this makes employment on those planets more challenging at higher populations, so I'm still working on the right balance.

But, I'm also feeling better now about breaking out of those specializations and retrofitting later on. Early in the game I had earmarked my capital Emariss as a research world; but I needed to build some Alloy Forges there early on to kickstart my expansion plans, and Civilian Industries to move on to my next world. I kept those buildings around as the planet grew, but once I'd maxed out my 20 building slots, I gradually went back and changed those into Research Labs. There's a short-term unemployment hit from this, but there's plenty of time to retrain workers before proceeding. Likewise, during my colonization crunch I had an all-hands-on-deck phase of all planets building Agriculture districts willy-nilly; now that there's more than enough food again, those one-off districts are being repurposed for more productive uses, while my few agriculture specializing worlds focus on their job.

One of the interesting mechanical aspects of Stellaris is how many things in the game have fixed costs, while others have scaling. This is probably most obvious early on with Leader hiring: early in the game, getting 200 Energy seems like an almost insurmountable problem, and you have to choose very carefully who you recruit, because you'll be stuck with them for a long time. By the end game, though, leaders still cost 200 Energy (possibly modified by Policies, but still around that), which is nothing when you're storing 100k energy. So you think nothing at all about firing a leader who acquires an undesirable trait, or cycling through your pool to get a quality you want. Similarly, building stuff feels kind of expensive and time-consuming in the early game, but spending 300 Minerals to replace a district is nothing. 

I'm now past an inflection point in my game, where I've researched all non-repeating Physics and Engineering technologies and most Society ones, running high surpluses, have plenty of Admin Cap and room to expand, and well over 350 Naval Capacity. I'm pretty dominant in the Galactic Community, thanks in large part to the 5 Envoys I have there, further augmenting my already-significant Diplomatic Weight. Those envoys served me well in the early game, and I avoided any unwanted wars; the Pious Anathurian Theocracy tried to start something about 50 years ago with claiming my systems, but I buttered them back up, and now I could crush them if I wanted to. Of course, as a Pacifist empire I don't really want to.

I'm not sure exactly what my next steps will be. My mid-term goal is to conquer the Ishni Shard and take their awesome planets and adjacent systems, but their fleet power is still Overwhelming at the moment and it will take some time to get there. For quite a while I've wanted to form a Federation with the Neborite Authority, the second-strongest non-AI civ in the galaxy and fellow Fanatic Xenophiles, but they've been standoffish due to my restrictive War Policy. I do want to get in a Federation and level it up before the Crisis. Funnily enough the Pious Theocracy and Commonality of Mekon did invite me to their Federation, which is the one I was vaguely planning to fight against. I'm now wondering whether to start a Federation with some weaker Pacifist empires, which would be much less advantageous to me but a lot easier to get going and probably still beneficial in the long run.

This game will keep me busy for quite a while longer as it's only around 2370, but I'm already mulling over possible changes to make in future games. One thing I'm contemplating is specializing my species as I specialize planets. As a Xenophile empire with a lot of Migration Treaties, I have a ton of options for which species to put on new planets. Before I had Gaia worlds I would choose whichever one had the highest Habitability; now I just put my birds everywhere. But, it would be better to pick species based on their traits, like Intelligent species for research planets, Agrarian species for farming planets, and so on.

And, speaking of Habitability, one thing that I'm wondering now for my current game is whether to take on the Nonadaptive malus in order to add more beneficial traits to my main species. I believe that Nonadaptive species will still get 100% habitability on Gaia worlds, making it a basically free perk. With my Society tree basically all researched, now might be a good time to try that out.

I think that's it for now! We'll see how much longer this Pacifist ethic of mine lasts; I think I can easily bait Ishni into a war when I'm ready for it, and of course the Crisis won't care what my policy is, but time will tell whether there's any conflict with the mundane empires sharing my galaxy.

Monday, November 02, 2020

Baldur's Gate 3: Everyone Is Evil

I think I'm at a good pausing point now with my early access BG3 run. I've finished the first major quest chain, reached Level 4, and struck out with my intended love interest. I may push on to the end of the current content, but right now I'm leaning towards calling it done for this run and waiting for a major future update, or the actual final release; I'm trying not to get too invested in this particular character since I know they'll be wiping the saves at some point.

So, my overall impressions? It's definitely an Early Access experience and not as polished as the post-release Larian games I've played before, but I knew that early on. It has some bones that are really exciting, and it seems like it will scratch an itch for me. If you imagine a cross between Dragon Age: Origins and Divinity Original Sin 2, you'll have a really good idea of what this game plays like.

A lot of my current complaints are things that I'm 100% sure will get fixed in future patches. The inventory system is, uh, pretty bare-bones, without the ability to sort items; they've had that in other games, so I'm sure sorting is on its way. Rogues can't disarm traps, making the "trap disarming kits" you find useless. Most cinematics are pretty janky, but the ones in the opening sequence are great, so hopefully the rest will be updated to that quality.

Other complaints are things I'm resigned to. The looting system, the bane of my existence in nearly every RPG, is bigger and more bloated and more time-consuming than ever. Also, it's 2020, and we still don't have a way to highlight lootable containers. Both of those have been true in the other Larian games I've played so I think they're deliberate choices, just not choices I like.

And there's the stuff that's been inherent to D&D for ages, like Vancian magic. Particularly as low-level characters, your wizard can easily spend all of his spell slots in a single combat encounter, and then need to hit things with a stick for the rest of the day, or else you'll need to run back to camp and sleep to get the spells back. I actually avoided sleeping for quite a long time, since story-wise it sounded like I was on a pretty severe timer, but that turned out to not quite be the case, and I probably would have enjoyed it more if I'd indulged in camp visits more often. (There are also some critical plot events and companion conversations that can only be triggered in your camp, giving another reason for more-frequent visits.)  I do like how they've added a Short Rest option that gives some HP and recharges a few martial abilities. And the number of spell scrolls you loot are pretty generous, so I need to remind myself to make use of those when I'm out of magic for the day.

A few technical things: After dropping my graphics quality setting and disabling the Steam overlay, the game has been running much more smoothly for me. Every once in a while it will freeze for a couple of seconds, but it always recovers. Losing the Steam overlay does mean far fewer screenshots to share, sorry about that. The game looks fine at the lower setting but I do see the difference. There's a decent chance I'll upgrade my PC in the not too distant future to run Cyberpunk, and if so I'll be interested to see what BG3 looks like in all its ultra-quality, stutter-free glory.

I think that's it for complaints! On to the good stuff:

I really dig the combat. I'm not totally looped into the fan community, but the main complaint I've heard from old-school BG fans is the loss of Real-Time With Pause for a turn-based combat system. The main argument for RTWP is "it helps trash fights go so much quicker!" The main counter-argument is "It's better to just not have trash fights to begin with!" I'm solidly in the second camp, and this mostly delivers. Compared to BG1/BG2, fights are generally more challenging, unique, have distinct mechanics and goals. It isn't as finely tuned as DOS2, and it isn't quite as hard; again, some of that may be due to this being in EA, but even where it is now I'm highly enjoying the combat.

The most unique thing about combat in DOS 1/2 was environmental effects, where, like, you could fire a lightning bolt into a puddle of water and zap all the enemies inside it, or freeze it to make a slippery surface that would cause them to trip and fall, or ignite a barrel of oil to create a huge fireball, and so on. Those same mechanics are present in BG3, but it seems to be a lesser focus; fights rarely seem to be built around environmental mechanics the way they were in the DOS games. But there are a bunch of new elements that I really love. Going back to environmental things, you can dip a weapon into an effect to apply that effect; the most common example is dipping a weapon (like an arrow) into a flame source (torch, candle, firepit, etc.) to turn it into a flaming arrow that deals an extra 1D6 Fire damage. You can also damage objects in the environment; I was thrilled during one underground fight what I realized that I could shoot the stalactite high up above the cavern, causing it to crash down and deal damage to all enemies below me. The single most satisfying thing, though, is kicking enemies. Doing this pushes them back a short distance; if they're near a ledge, then they will fall down, taking fall damage based on the height they fell. You can one-shot some foes this way, as a Bonus Action! It's insanely fun.

I'll also call out exploration as a real highlight. The maps are extremely well-done, and particularly in the outside areas it's a blast to find hidden nooks and crannies, which often contain some rewarding loot. The ability to jump and climb is new to Baldur's Gate, and adds a great new dimension to the exploration aspect of the game.


The dialogues can feel a little weird sometimes. Your protagonist is mostly voiceless, which is fine; unlike other similar games, where your PC is always the main speaker in dialogues, here it is whatever character you had selected. The other person's dialogue is usually identical no matter who they're speaking to, but sometimes you'll unlock unique response lines based on your selected character's class or background. It seems like this shouldn't work, but, as you eventually realize, all of your party members share the same salient characteristic, so the dialogue generally does make sense even when it isn't delivered to "you".

One concept that I really love is the skill checks, which most often come in dialogues: instead of being a flat requirements, you roll a D20 based on the challenge rating and modified by your skill to try and pass it. So, you might need to pass a Deception check to trick a guard, or an Animal Handling check to calm a bear, or an Acrobatics check to show off your moves. The UI for this is really nice. Unfortunately, I personally am allergic to "failing" in conversations, so I almost always reload if I miss a roll, unless it's clearly a trivial flavor thing. 

I do wish that you could swap in party members for specific skill checks. Especially since, like, I can see Shadowheart standing right there next to me, so it seems like I should be able to ask her to handle that bear instead of trying it myself. Sometimes you can exit a conversation, switch to another character, start the conversation again, and then use the character who's actually suited for it; but other times you only have one shot, and if I don't like the outcome? You'd better believe that's a reload, baby.

There are also some behind-the-scenes GM rolls to see if your character noticed something. These can be Lore checks that provides you additional information during a conversation, or Perception checks while exploring that can reveal hidden objects or traps. Those are nice.


Going back to the title of this post: After playing the game for many hours, the single most striking thing to me is how pretty much everyone in your party seems to be evil. There doesn't seem to be a traditional alignment chart in this game so it isn't necessarily canonical, but that's the main thing I'm taking away from it. I'm really curious why that is. It could be a deliberate reaction to the alignments of vanilla BG2, which famously was tilted in favor of a Good-aligned party. It might be that evil characters are more interesting, or more fun to write. Or it might have been a way to pursue a more "dark" tone for this game.

Breaking it down: You have Lae'zel, who is a Gith; she may not technically be Evil since she's from the planes, but her whole thing is denying compassion to anyone, swiftly and mercilessly killing your enemies, and showing no patience for any weakness, so it's de-facto evil. Shadowheart is a cleric of Shar, which, yeah, that's pretty evil; she's less in-your-face aggressive than Lae'zel, but will strongly disapprove of any altruistic actions on your part. Astarion is a bloodsucking vampire, and an aristrocrat who despises the common people. Wyll seems like a good guy, but based on the cover art and his permanent buff status I'm going to go out on a limb and say that he consorts with demons. Gale is the one guy who doesn't seem to mind when you do good things, and the only person who I really got along with; even there, he's #mysterious and travels with mephits, and I wouldn't be slightly shocked to learn that he's tied up in something untoward.

Basically, I've really been missing Mazzy, Keldorn, Imoen, Aerie, or any of the other companions who liked helping people.

It does seem like BG3 is giving a lot of options for an evil playthrough; the major quest chain I just finished seems to have an entire alternate solution where you side with The Bad Guys and wipe out The Good Guys. I'm not usually tempted by that stuff in games, but I am impressed when developers include that much variety. It is less tempting when, as here, it's such a black-and-white good-or-evil choice; for me, it's a lot more compelling to choose between two shades of gray, or between pragmatists and idealists or some other thing. But, yeah.

I am increasingly thinking that in my next playthrough, I might just bite the bullet and roll an evil character. I've been talking about doing that for decades in Baldur's Gate and never have been able to do it. One strong incentive to do it in this game is to be able to pursue the romances. As noted above, the two lady love interests both despised me for my do-gooding ways, and I was locked out of romance options for both. Which is its own whole separate interesting thing; I've thought a lot (too much!) about romance in video games, and have written a lot about how characters should have more agency, and should respond to the actions of the player instead of automatically falling in love with them; I'm now experiencing what it's like when those characters don't approve of my actions. (Of course, it's entirely possible that there's a more long-term arc here, like Viconia in BG2; again, Early Access!)

After realizing I'd whiffed the romances, I hopped onto the wiki to see what it would have taken to pursue them, and was flummoxed to discover that there was another romance option... who will only join you if you take the evillest path of wiping out The Good Guys! It's a choice. 

One last final nag: I was pretty pleased with myself with how I handled the end of this quest chain, with some sneaking and trickery and stuff to assassinate bad-guy leaders without alerting the whole area. But, as soon as the last leader was dead, the entire area turned hostile to me. Which was annoying, since it's specifically the opposite of what everyone said would happen when the leaders had died, and the opposite of what I'd observed happen in previous fights. Fortunately, Quick Travel is available everywhere, so I was able to skip back to where I wanted.


Overall, I'm having a lot of fun. I think I'll wait a bit for the next major update and maybe try my hand at a bard or something.

I'd mentioned in my first post that BG3 doesn't really feel much like BG1 or BG2, and I still feel that way. There are occasional flashes: the music, particularly over the title menu, is nicely evocative of the franchise's sound. Of course, we aren't actually in Baldur's Gate yet, and there might be more familiar touches still to come. Even if this ends up being DOS3, though, that would still be a very good thing.

Tuesday, October 20, 2020


When I find an author whose work I enjoy, I'm tempted to read as little of their work as I can, as slowly as I can, to drag out a supply of fresh good novels as long as possible. So it's a little unusual that I would so quickly return to Liu Cixin so quickly, but here I am, just a month after finishing The Three Body Problem and wrapping up the sequel, The Dark Forest. Part of that might be because of COVID focusing my attention, part might be the sci-fi kick I've been on lately, but I think the most compelling thing is just that 3BP was a damn fine book, and I wanted to read another one.

MINI SPOILERS for The Dark Forest, MEGA SPOILERS for The Three Body Problem

I put this book on hold at the library the moment I finished the first one, but I also was a little skeptical about just how Liu would manage to follow the first book. 3BP had so many wonderful tricks that seemed impossible to repeat. Even though it was nominally a science-fiction book, it opens during the Cultural Revolution and feels like a historical fiction novel for a long time, and then like a mystery novel for quite a while. It isn't until near the end of the book that we finally learn what the Three Body Problem is and who is behind the events of the story. But now that we do know, how can the mystery be maintained? I was curious if the series would transition into a more typical sci-fi novel with space battles, or if he would try to repeat the historical and mystery aspects again.

As it turns out, he does neither, and instead segues into still another genre. Most of this book feels like a political thriller, on a global scale grander than anything this side of seveneves. Earth is nominally united against the Trisolaran threat, but old divisions still exist, and we see the tacit struggles occurring between West and East, global North and global South, between established industries and nascent ones, between various ideologies. Political capital is gathered and exploited and expended and lost as various leaders and strategies rise and fall.

There is a pretty solid break from the previous book, both in the structure and tone, and also just in the characters. Most of the major people from 3BP are missing in TDF or just have brief cameos, with Shi Qiang the one major recurring character. But it also adds to the sense of scope and scale of the story: this is an enormous crisis gripping all of humanity for centuries, so of course different people will be involved at different times.

There is a nice mystery aspect that continues in this book, though. The mystery in 3BP mostly revolved around the titular game, what was killing the scientists, and how it related to the Red Coast Base; it's a mystery about an adversary. In TDF, the mystery is about the allies. The single most creative and exciting idea in this novel is the "Wallfacers", four men who have been selected to create and carry out secret plans to thwart the Trisolaran invaders.

Doing a bit of recap for my own benefit: At the end of 3BP, we learned that the Trisolaran fleet is en route to the Solar System, with the stated goal of destroying all of humanity and settling on Earth. Their technology is vastly superior to ours, but they are concerned about mankind's fast progress: The Trisolarans have been more advanced than us for hundreds of thousands of years, but in the last 200 years we have progressed more rapidly than they had in thousands of years, and if that trend continues then by the time they arrive (around 2400 AD) humanity will have become vastly more powerful than them and easily able to crush the invasion.

It will take centuries for the Trisolarans to arrive, but thanks to their mastery of subatomic physics, they are able to manipulate events on Earth at the microscopic level. (Think quantum entanglement.) That isn't enough to take macroscopic action like assassinating a head of state, but through their sophons they are able to spy on humanity, seeing and hearing everything that takes place; and they can interfere with subatomic experiments, thwarting the efforts of particle accelerators and colliders to plumb the mysteries of neutrinos and other fundamental elements of the universe. Thanks to this, they effectively place a lock on humanity's progress: we can continue developing the existing technology we have, but entire fields of study are closed off, including quantum research (and its computational advantages), strong nuclear forces, and so on.

But, there's one thing the Trisolarans can't do: they can't peer inside our minds. They can observe what a human says and what he does, but the "why" remains a mystery to them. So, the Wallfacers exploit this weakness. They carry out their plans in plain sight, but everyone, human and trisolaran alike, is aware that there are other angles and wrinkles to those plans, and are constantly kept guessing what their true aim is.

This all ends up being incredibly fun, with wheels within wheels, as we try to guess what the various Wallfacers are up to.


Hines is probably the most interesting Wallfacer to me. When reading the detail about how Keiko's eyes flashed open as soon as she was going to sleep, I guessed that she was his Wallbreaker, though I still wasn't sure what the implications of that were. One random idea I had was that, since Luo Ji was chosen because Trisolaris was scared of him, Hines might have been chosen if the PDC know that Keiko was a member of the ETO. In this scenario, part of Hines' deception could have been feeding misinformation through Keiko back to Trisolaris. Or, another thought was that maybe Hines was the Wallbreaker all along and Keiko was secretly the Wallfacer. Of course, neither of those scenarios is true, but it was fun to think about! So many of the plot twists in this story were wild, so it felt like no possibility was out of bounds.

Even at the end of the novel, I'm still not totally clear on exactly what Hines' plan was and what its status is. Where are the sleeper agents? At first I thought that they had fully permeated society and implanted the mental seal on everyone, but that doesn't seem to be the case, given how virulently everyone reacts against Escapism near the end. And I don't think it had a major impact on the actions of the ships escaping the teardrop assault, as the main actors there were all hibernators who had gone to sleep prior to the seal's invention. It feels like this might be something coming back in the third book.

Of course, Liu ends up being the most successful Wallfacer of the four. His "spell" seemed simultaneously obvious and opaque. I was certain that some greater power would respond to his message by wiping out the planet he indicated, and that is in fact what happened. But I wasn't expecting the cosmic sociological theory behind it, which turned out to be shocking and fascinating. I'd thought that it would be some sort of higher-dimensional entity, somewhat like when the Trisolarans summon the sophons in the previous book, or if, like, a colony of ants were to form an arrow pointing at a thing they didn't like, and some giant human then casually destroyed that thing. But it turns out to be a matter of numbers rather than a matter of scale, which is really interesting to think about. After reading Liu's afterword to 3BP, I thought he was probably too pessimistic about encountering alien life, but the explanation in this book is definitely sobering, and makes one reconsider the wisdom of our constant broadcasting.

This is, of course, the Dark Forest of the title. I spent most of the book wondering what that referred to. Late in the novel, after Liu reawakens in the future, we learn about the tree structure of the underground city, so for a while I thought the "Forest" referred to the city, and mused that "Dark" could refer to it being underground, or engaged in sinister activity, or some future event that would cut out the limitless power from the city and plunge it into eternal gloom. But, yeah, the title ends up referring to the entire universe, so that's pretty cool. It's another thing that reminded me of 3BP in how late we learn about something huge that causes us to rethink the entire galaxy.

Besides the epic distance and physical scale of the book, the span of time is also really epic and intriguing. I almost immediately thought of A Canticle for Leibowitz, which also had multiple-century jumps through time and radical changes in the social and technological order. TDF has a lot more interweaving between the eras, though, thanks to hybernation technology. Though, now that I think of it, the Lazarus/Benjamin character in ACfL could perhaps be playing a similar role, but without offering a point-of-view perspective. And the hybernation of humans could have some parallels with the Trisolaran cycle of dehydration and rehydration.


Liu kept the magic going for the second book, which is really impressive, given how moved I was by the first one. I'm really curious where they'll go from here in the third book! It would be a fairly satisfying ending on its own, with a vision for the future sketched out, some cool callbacks to the previous book and the major characters being in appropriate places. But there are definitely some loose threads out there (Garden of Eden, anyone?), as well as the bigger implications of sociology, so I can see the potential for the stakes to, incredibly, get even higher. I've already put the third book on hold and am very eager to find where it goes!

Thursday, October 08, 2020

Toupée's Door

I haven't done one of these in a while, but this is just a quick note that I'm in the Early Access of Baldur's Gate 3! I'm pretty impressed by the little I've seen so far.

As a general policy, I almost always eschew Early Access for games. I think the last EA I played was Sunless Sea six years ago. My usual philosophy is that I only get one chance to play a game for the first time, and I'd like that first time to be as smooth and polished and bug-free as possible.

So why am I in EA for this? Well, BG2 is possibly my favorite game of all time (with the possible exception of Fall from Heaven 2). BG3 isn't made by BioWare; but the old late-90s BioWare doesn't exist anymore anyways. Instead, it's made by Larian Studios, who have made some of my favorite recent RPGs, including the absolutely fantastic Divinity Original Sin 2. Those two factors give me a lot more confidence than a lot of EA games do. The final deciding factor for me was that playing early would hopefully help me avoid any potential story spoilers floating around the Internet and experience it first-hand myself.

I'm just a couple of hours of playtime into it so far, so this definitely isn't a review, but some initial first impressions:

It is definitely taxing my system. I last upgraded my PC for Dragon Age 3, which coincidentally was also about 6 years ago. The GTX 970 I got then has served me very well for years, smoothly running The Witcher 3 and other higher-end games, but it's chugging for this one. I may manually adjust my settings down to try and get a smoother experience here. But I've been thinking for a while that I'll probably need an upgrade for Cyberpunk 2077 anyways, so I might start researching a new build sooner rather than later.

The impression I've gotten so far from the few previews I've seen is that this game looks like Dragon Age: Origins, and that's been borne out so far. Particularly the conversations and cinematics fit that mold: your protagonist is voiceless in dialogue (but has reactions outside it), and there's a nice level of zoomed-in detail that I associate with that game. Actual gameplay feels a lot like the DOS games, particularly movement, item interaction and looting. It's missing telekinesis, which I think helps this game feel much more grounded than the DOS games. And it still has the environmental effects as DOS, but at least so far they're less of a focus. On the other hand, it doesn't really remind me much of BG1/BG2, at least so far. Which is honestly OK. It would probably be worse for someone trying and failing to replicate it than to focus on making their own new thing good.

Character creation is awesome. One sign of a good RPG is that I burn most of my first night just on making my character. It's funny to me that I can make a better Qunari in BG3 than I could in DA3. Chargen uses slots instead of sliders, which I personally find a lot nicer and easier to use: you don't, like, adjust your brow width, you just pick what kind of head you want.

One of my perennial hobbyhorses is the presentation of gender in RPGs. RPGs are in a kind of awkward state now where you're forced to make a binary gender choice at the start of the game and play a cisgendered PC, but modern RPG worlds are increasingly filled with transgendered, nonbinary, or otherwise queer characters that your PC can meet and gawk at. The only studio that really does something cool with player gender is Failbetter Games, and they don't need to worry about 3D models or animation. At first blush, the BG3 creator is like all the others, starting with the fateful Venus-or-Mars toggle. But it's cool to see that this choice doesn't lock you out of anything! You can create a biological female with a masculine voice and a big bushy beard, or a hulking brute with lovely eyeshadow and long flowing hair.

Oh, and that hair! I would have murdered a room full of people to have this many long-hair options in DA3.

Anyways, I'm rolling with a normal (for me) cis character, but it's really cool to see this space opening up and developers allowing players a wider canvas to choose from in crafting their characters.

They are still in progress on adding classes. I think I want to play a Bard for my first full playthrough, but so far in EA I'm playing a Rogue and having a blast. I really love how many sub-classes and sub-races there are.

And... that will do it for now! I'll be very mindful of spoilers going forward, just wanted to share that I'm checking it out and having fun so far.