Humankind impressions: off to a promising start

This entry is part 1 of 2 in the series Humankind

I’ve finished two games of Humankind, the new historical 4X strategy game from Amplitude. It’s really good — better than I had expected. It’s also a lot more challenging, and I think that’s why I’m enjoying it.

Comparisons to Civilization will be inevitable — I’ll go out on a limb and say that at a design level, I think Humankind does Civ better than the most recent Civ games. While I play and enjoy Civilization VI as a “numbers go up” game where the fun is in designing and building powerhouse cities, that game’s AI is simply not aggressive or tactically competent enough to feel like a true rival. If I do lose a game of Civ VI, it’s because I did a poor job of making the numbers go up, hence allowing the computer to reach the victory conditions first.

Victory in Humankind goes to the player with the most fame points, which are awarded for reaching goals in each era. These goals relate to science, enemy units destroyed, money and influence earned, population size, number of territories, and the number of districts built. So far I find it very easy to accumulate points for science and (when at war) destroying enemies, and harder to accomplish the other objectives.

By contrast, playing on Empire difficulty (5 out of 7 — I agree with the consensus that experienced 4X / strategy gamers should crank the difficulty up), Humankind consistently puts me under pressure — I have yet to win a single game. In my first game, I barely scraped into #2 place on the final turn. In my second game, I came a more distant #3 — and I consider that an accomplishment, given that at one point, I was dead last! This kind of game is all about snowballing — setting up a virtuous cycle of more food, production, and science, which allows more upgrades, which allow more food, production, and science — and it’s notable that the computer knows how to do that. In my second game, one AI player built a gigantic lead by playing as a series of agrarian cultures, amassing a huge population (at one point its cities were size 40+ at a time when mine were in the 20s), and conquering a neighbour early on. It hit the final era well before me, and won by a commanding margin.

Not only is the computer capable of building strong empires, it’s perfectly willing to muster its armies and batter down my gates, just as Civ IV‘s AI did back in the day. My first game was very military-focused — the computer was much more bellicose than I expected, and from the ancient through to the early modern eras, I was almost constantly at war. An early-game rush from the neighbouring player made me fight for my life — the computer cleverly took advantage of my neglecting the military. Subsequently my neighbours hated me for most of the game, until something changed and most of them suddenly wanted to be my friend — my best guess is that when I converted to the dominant religion on the continent, that removed the main source of friction.

The setup for the first battle I (blue) ever fought against a rival empire (yellow), which rushed my capital early in the ancient era. I only had a single professional military unit, so most of my force comprised city militia. Fortunately for me, the canyon channelled the enemy army, and it took a further penalty from having to cross the rivers.

Combat itself is solid and the individual battles are generally interesting to play out. Early on, the rule is “strong melee units in front, archers behind”. New eras introduce new wrinkles — for instance, early gunpowder units can’t move and attack in the same turn, so my Medieval-era Varangians were still extremely useful for flanking and charging enemies even once I started fielding arquebusiers. By the modern era, battleships and bombers can deliver devastating bombardments to support land battles. I really like that cities generate freespawn militia (this is what saved me from that first rush!), and the clever way in which sieges gradually shift the advantage from the defender to the attacker over time, as powerful siege engines go up. Choke points are important, and help a defender against a numerically superior attacker. I do find it unintuitive to read the terrain — it’s often unclear to me which differences in elevation can be traversed by units.

The setup for a later battle – this time, I was on the offensive against the purple city. I put my tough melee fighters, the Varangian Guards, up against the wall, while my trebuchets were further behimd.

So far I have a pretty good grasp on the military aspects of Humankind, to the point where I can consistently beat the computer on the battlefield. In contrast, I’m still learning how Humankind’s economic “engine” works — the key seems to be a combination of territorial expansion; placing districts to take advantage of adjacency bonuses; and savvy use of each culture’s powerful unique buildings. It seems easier to amass science than the other resources — is that actually the case, or is that simply because my playstyle focuses on science? There are other systems I have yet to engage with, such as religion and cultural influence — I don’t know how important or deep they are.

One area that could do with fine-tuning is the final, modern era. The foundations are there and the ideas are interesting – for instance, when pollution caused a penalty to food production on each tile of a large city, I sat up and took notice. When it comes to implementation, the numbers feel as though they still need tweaks: science costs in the final era seem a little low, production costs seem a little (or more than a little) high, and the pace feels a little too brisk, as if the developers overcompensated in a bid to avoid Civilization’s sluggish late game.

Humankind’s space race is less important than Civilization’s – it grants bonus points but doesn’t trigger victory. I won the space race in my second game, but still finished in third place.

One thing I do like is the game’s signature mechanic: the player chooses a new culture for each era, instead of being locked into one for the entire game. The decision to change culture plays into what is required at any given moment. In my first game, I started as the Zhou in the ancient era to take advantage of their science bonus. When it became clear that I’d be locked into brutal wars, I needed the toughest fighters I could find — I chose the Romans, and steamrolled my foes with Rome’s unique units. After that I continued as the Byzantines — with their equally impressive Varangians — in the medieval era, then finished with a string of science cultures: Joseon, industrial-age France, and modern Japan. I love that these options all feel cool and powerful — this is the right way to implement a philosophy of “interesting decisions”.

Each culture in Humankind receives a passive bonus (which carries forward to future eras), a unique city district, and a unique military unit. At first I picked the Byzantines for thematic reasons after earlier playing as the Romans, but I soon became glad for their powerful Varangian Guards.

Edit: I haven’t found bugs to be too bad. So far I’ve encountered 2 crashes (not a big deal, as the game auto-saves each turn) and a couple of cosmetic glitches. The most concerning is that the computer player in a particular slot (purple) reportedly receives automatic influence upon independent minor factions — hopefully Amplitude will address this soon.

A final point is that the art lives up to Amplitude’s typical high standard. Special mention for the art that illustrates each technology in the game — with a quick scroll left and right along the tech tree, I can see Humankind’s (and humankind’s) progress from the pyramids to the space age.

Overall, I’m very positive so far. Humankind already has the “just one more turn” magic backed by solid strategic gameplay, and I expect it will have room to grow via DLC and patches. I look forward to my next game!

I love that the ending narration congratulates you on your achievements.

Revisiting Humankind, three years on

This entry is part 2 of 2 in the series Humankind

Humankind launched three years ago, in August 2021. Since then, I’ve played it many times (on both Game Pass and Steam), and with Civ VII now on the horizon, the time is right for me to revisit my thoughts. I still like Humankind — but I don’t think it ever reached its potential.

What I liked about Humankind

At launch, I was glowing about Humankind. It was a very good “traditional” and warfare-focused 4X, focused on building an empire, investing in food, production & science, raising armies, and fighting other empires. It featured:

  • Engaging warfare and tactical combat — terrain, technology, unit types, and tactics all mattered. Armies marched as stacks, before deploying to fight. Battles took part over several turns (allowing reinforcing stacks to join the fray), and siege battles felt distinct thanks to the presence of city militia and mechanics such as fortifications and siege engines. Warfare evolved through the eras, especially once gunpowder made its debut.
  • Computer players that could give me a challenge — from my very first game, I had to fight for my life against early rushes.
  • Points-based victory conditions that rewarded doing well across multiple dimensions:
    • Being both tall (building plenty of districts) and wide (geographic size).
    • Succeeding both at peace (having a large population, being scientifically advanced, earning lots of gold and influence) and at war (destroying lots of enemies).
    • Doing well consistently through the game — it was often worth lingering in earlier eras to earn points.
  • Novel mechanics such as choosing new cultures over the course of a game, opening with a hunter-gatherer phase in the Neolithic period, building cities region by region, and attaching outposts to expand cities.
  • A great aesthetic, both in terms of art and music.
Fighting off an early rush in Humankind.
The computer players in Humankind really, really like rushing me in the early game. This contributes to the challenge and is one reason why I consider it a military-focused game.

… and what I didn’t

At the same time, Humankind had its flaws:

  • From a design perspective, anything that wasn’t part of that “traditional” 4X core was vestigial — notably religion & culture. Diplomacy was bland and different computer opponents never felt distinct.
  • From an execution perspective, the late game was noticeably less polished than the early game1. Cool ideas suffered in implementation, such as a pollution system that was too punitive in practice.

The problem is that the game never really evolved from there. In an ideal world, expansions would have plugged the holes in its design; but none of its DLC2 did. As for execution, I think it says a lot that later updates simply disabled pollution by default.

By now, I think it’s unlikely this situation will change. The last DLC, a culture pack, came out nearly 12 months ago (back in September 2023), and the cadence of patches has slowed — there have been none since January 2024, although the developers have said they’re working on an update.

Why Old World was, overall, better

Another comparison is Old World, the other major historical 4X besides Civ. Between the two, Old World is the better game:

At the same time, Humankind is the better-looking of the two, and its familiarity made it easier for me to learn at first.

Conclusions

Would I still recommend Humankind? Yes, with the caveat that prospective players should understand its focus — and its limitations. I still think it’s a good 4X game with great ideas (underscored by how many of them are recurring in Civ VII). It does particularly well at warfare — better than any Civ game probably has in decades — and that makes it well-suited for those seeking military challenges in their 4X games. On sale, it’s cheap enough (90% off!) to justify taking a look. What it does not do well is anything outside that core focus.

Would I replay it? Maybe. I thought I was ready to move on, but writing this article reminded me of what I enjoyed. I’ve yet to try the mod scene, and even without mods, sometimes I just feel like 4X comfort food: something that Humankind offers.

Enjoying this site? Subscribe to updates here:

  1. I don’t think it’s a coincidence that the late game never went through the rounds of open beta testing that the early game did.
  2. Humankind received several rounds of DLC that added new playable cultures, plus one expansion, “Together We Rule”, all the way back in November 2022. “Together We Rule” added new diplomacy and world congress mechanics, which I never really understood or enjoyed.