Making games is quick and easy. Unless you want to do it right.
It's been over one and a half year since I decided to delay the launch and rework major parts of the game. In this post I will attempt to summarize the most most significant changes I have been working on and what lies ahead.
What changed in Manor Lords?
Residential plot system
Before that change, whatever I did, I couldn't get the towns to look quite as real as I wanted. This was partially because of the lack of integration between the houses and their surroundings. After studying medieval settlement layouts as well as drone footage I decided to try and adapt the mechanic I was using for organic field placement and apply it to form residential zones.
If you ever heard about medieval "burgage plots" this is what inspired me.
Currently you place the residential plots by placing four points, and the edges try to conform to surrounding roads, creating more organic shapes. In the backyards, if you leave enough space, there is room for "backyard extensions" - vegetable gardens, goat sheds & chicken coops.
Late game towns & Artisan conversions
I went through a lot of iterations of late game towns. I tried procedural generation but ended up using a more standard approach with randomized pre-made buildings.
However, another subject I tackled is that in a late game town, the high level production buildings like Bakeries and Tailors were located on the outskirts - cause they were placed last.
"Artisan conversions" mean that after you upgrade your residential plot to a certain level, you can convert it to a artisan workshop - brewery, blacksmith, tailor or a cobbler. This means that high level jobs remain in the very center of the town. It also "locks" the artisans from how you appoint normal peasants to work, as they become permanently attached to their profession.
Don't get accustomed to the architecture style though, more on that below...
Region sizes
Manor Lords originally started as more or less a classic medieval RTS. In 2018 it had a grid, a fog of war, and the camera was similar to what you can see in Age of Empires series. But what I really wanted was to make the towns look and feel more real, I ditched the grid, the fog, and divided the map into regions. But since the camera was low, the regions were therefore quite small too - I hoped it will be interesting to work with the limited space.
In early 2020, the region sizes were more or less symbolic, reminiscent of games like Northgard
Time and time again during conversations with the community and later on with first closed testers giving feedback it became clear that people expect more realism. I kept making the regions larger, so in effect the amount of regions was getting smaller (we're down from 50 to around 11-12).
There is room to make bigger maps, but that I leave for a later update. Keep in mind however that due to gameplay functionality and technical limitations, it won't ever be 1:1 to real world, so those 4 square kilometers of landscape are meant to represent a bigger area.
Aerial view of one of the starting regions after reworks. WIP.
Freedom of the camera movement & zoom levels
Another very challenging request from the early testers was to free up the camera. The technical problem comes from the fact that you could already zoom in pretty close - so now I had to try and make the game look good from close, medium and far, while all being optimized. It's like art directing 3 games at once.
But you can't treat it like 3 games or it's unoptimized, so it's all about being creative with the complexity of the terrain shader. In a game like this, the terrain covers the whole player's screen (the bigger the area covered by the object, the more expensive it is to render as it has to calculate all the nifty shader tricks per pixel). Additionally, the terrain textures shift throughout the seasons, and can be influenced by the player during play.
Furthermore typically you flip far trees to flat billboards. Freeing up the camera means you can view the trees straight down from high up. To solve this I had used a technique developed by Ryan Brucks (for Fortnite actually) called Octahedral Impostors, but with the dynamic seasons it was a large extra headache.
Those birds loved to glitch out.
Narrowing down the timeframe and region to unify the visual style
When I started the project I wasn't quite sure where I'll end up. I always said that it's "inspired by history" and you can see in the first announcement trailer that it's a big mix of styles and periods. I wanted something similar to Age of Empires, where as you unlock technologies, your soldiers and buildings change their look.
However, this ended up looking very messy to anyone who's even partially interested in history. Even in the 2021 gameplay trailer you can see both knights wearing plate armour and "viking" style mercenaries. It had to be resolved and it cost a lot of time and effort to rework this.
In the end after consulting with the community I decided that end of XIV century and Holy Roman Empire (roughly the Franconia region) will be a fairly good fit so I went with it. It's still "inspired by", but it is definitely less chaotic and anachronistic.
Complete architecture rework
The first aspect of this was to replace the architectural mix of my Anglo-Saxon pit houses, Tudor and French timber framing and overall random influences. Luckily I met Jakob on the Discord, who specializes in medieval architecture of southern Germany and offered to help redesign all the buildings in the game in a accurate fashion.
Nowadays Jakob draws the desgins for me and I do the 3D models
Complete character rework
Parallel to this I moved to characters. I've been very obsessed over optimization of the game and a lot of my character models were rough and very low poly. I definitely did not plan to model individual fingers as I thought nobody will ever see it in a strategy game. I kept thinking about this ever since Mack from the WorthABuy channel mentioned it in his reaction after I posted the announcement trailer.
For this I was lucky to meet Marcos, a AAA character artist who previously had the opportunity to work on characters for Kingdom Come: Deliverance, who was happy to bring the characters to another level.
Community on the history channel on our Manor Lords discord server guided us for historical accuracy, with one knowledgably member: Eol sacrificing his time to draw concept art on how the armour could still be good for gameplay progression but also historically accurate.
This includes faces and hairstyles...
Marcos' work was mostly about clothing and armour. Before the changes I simply used 3D scanned heads bought online. They looked somewhat OK, but there was a limited amount of them and they often looked too young or had modern looking haircuts.
I scrapped what I had here and tried another workflow. I generated head photos with deep learning AI which allowed me to pick and choose faces that would be more fitting. Then I used Character Creator software to convert it into 3d models efficiently - the benefit here is that it has a unified texture topology, so I was able to randomly swap the textures and models to create multitude of unique variations. Then I added hair as a separate mesh and again followed the community guidance on what hairstyles were popular at the time. The effect of all this is less obvious repetition in how peasants look, and if I want I can easily generate more variations.
Recruitment system
When I posted the announcement trailer in July 2020, Manor Lords was a far simpler game. There were no "late game towns" yet. I never expected the player to get to a high enough population for large battles so the only units that were in the game - all that you see in that 2020 trailer - were mercenaries. What would probably shock you today is how they were spawned too:
I had reasons for it, first one was the population limits, but also I wanted the player to have multiple ways to develop economically instead of just going for iron and making a ton of weapons. Wealth being the sole requirement for hiring units helped to achieve that. However I was later convinced to change it completely. Players really wanted to send their peasants to battle.
Therefore I added a Militia rally system and a personal Retinue system.
Retinue editor and it's rework
The first version of the Retinue had two major problems. One is the lack of historical accuracy that I already mentioned, and another one was unification of equipment. Initially I've made the customization system so that you pick a weapon, eg a "sword" or a "war axe" and it would effectively create a unit of "swordsmen" or "axemen". Such unification wasn't really in medieval spirit.
After a few debates I decided to finally go for variety over unity and the game instantly became more authentic looking. This however limited some customization as there was no longer a 'swordsmen' unit with a single helmet type. More variety & randomness, less customization, but more authenticity.
However I left room for future specializations weapon types divided into categories (types still in the works, for now they are planned to be mixed, long polearms, short polearms, pavise infantry, cavalry). Furthermore I implemented [WIP] a system where armour can be upgraded individually per member, giving even more individuality to the soldiers. It's still a WIP though.
Combat changes
If you've seen the battle trailer, you know all the features I've been working on. However it might not be clear to some that none of these features existed in the time when I published the first announcement. If you missed it...
And if you're curious, here's how combat looked before (note: the map is a test map, it won't be present in the early access launch).
The addition of occasional finishers and synchronised mocapped animations really helped, but I made sure to keep it minimalistic and not over the top as I know that many Total War fans dislike that units "wait their turn" as their animations have to be synced.
Further animation reworks
After Marcos worked on the new character models, we had to adjust the animation skeletons - a foundation that you use to animate the characters.
I used my friends mocap system for capturing motion and sadly, because it is more of a budget solution, it created some less-than-perfect bone placements that had to be manually corrected. Unfortunately that meant that a lot of animations broke, cause the length of the arms changed. This had to all be manually fixed.
In the end every animation in the game got recorded again, and most of them had to be further manually altered. It took forever, and I still would love to revisit some of them, but I fear I won't have enough time. Perhaps with the new funds I can find some help - more on that later.
Endless mocap cleanup and tweaks are not my favourite thing to do.
Gameplay. Micromanagement. Farming.
In March 2021 I started gathering a small tester group on the Discord. I started from my close friends, other game developers who were interested in helping out, and Discord moderators who did a lot of work for me managing the community there. Overall it was around 20 people.
Another major feedback of theirs was that the farming system was too repetitive and micro heavy. Basically what I had is a three element system, the Ox Farm for ploughing, a Field for sowing and harvesting, and a Barn for transporting and storing the wheat. This meant that if a player has 10 farms, every in game year he has to micromanage peasants to plough the fields, then sow them, then harvest, and every field was handled separately. That was a lot of clicking.
Here I had to try several systems but what we ended up with is a central "Farm" building. You employ the workers there and they do their thing automatically. You can however set each field's priority and crop type. If they all have the same priority, the peasants will split proportionally. It required a whole AI subsystem of Farms dividing the workforce "intelligently" and due to extra individual AI autonomy it wasn't as easy as it should be.
The testers weren't fond of the Ox mechanics too. There was some frustration due to the limitations of how the ox farm worked. After discussion on the Discord I introduced hand ploughing and plan to reintroduce Ox Ploughs differently soon.
Livestock system
Which brings me to the livestock. In previous builds it was simply bound to the building, you constructed a Sheep Farm and the Sheep just periodically spawned there. I didn't expect people to want realism here, but they did.
I had to invent a way to buy and sell livestock so I introduced travelling livestock traders and livestock markets.
However, livestock traders caused some traffic jams as they squeezed their way through the narrow town streets with their herds. Also, the livestock market was an empty, ugly looking place 99% of the time. So I scrapped the whole system and created a livestock trader building instead - a trader who handles import and export of animals.
I also added an age to sheep, so you can buy lambs if you wish to invest long term. More livestock management functionality still needs to be done.
New trade system
Which brings me to the trade system. Originally I had two trade buildings:
a) A merchant, sometimes called a trading post. The workers imported goods from other regions to your region.
b) A market stall. Merchant visit market stalls to buy stuff, so if you want to export something, you have to build a stall and hope that a merchant will visit you.
This added some trade competitiveness as whoever attracted more merchants had more money. You couldn't just click a 'sell' button to generate money.
The problem was that it made trade between regions very cumbersome. So instead I introduced a trade overview where you can set your export/import limits and the workers handle the rest. The market attractiveness still matters, because using your own peasants for export takes more time and effort than attracting foreign ones.
Water & Soil
One of the things that annoyed me that due to handmade map, when players landed on the same spot, their towns tended to look the same. To combat this I introduced an underground water vein layer. The well has to be placed on top and since the well marks the town center, the towns are always a bit different too.
Lately I've done the same for crop fertility. Here you can see various overlays showing how fertile the land is for different types of crops (emmer wheat, flax, barley and rye - not sure if rye will stay though as it's basically a clone of the wheat resource).
Rotating the crop type improves the fertility and yields, but it has to be semi-automated in the future cause the testers tend to ignore this mechanic.
The market square evolution
The announcement trailer version had individual market stalls. Each stall sold one type of good. This was fine for a couple of stalls, but in longer sessions players tended to create these awful snake market roads that required a ton of micromanagement.
Because you placed stalls individually by the nearest trade route, it resulted in these long market stall trains
What I tried to do instead is utilize the field system again, and create a centralized market zone. It looked cooler and solved the boring micromanagement, but had two problems:
-You had to setup a large empty plaza in the beginning of the game where you only need one or two stalls.
-The market was placed in the outskirts where there is a lot of space.
I really wanted it to be in the town center instead, as it was often in reality. So I ended up combining the well and the market into a one town square building that provided water and slots for stalls.
A "circle" market version with a central well.
What I disliked about it is that every market square looked the same. It was also a bit awkward to place plots around it as I noticed testers struggling to fit them perfectly. So I did yet another take and split it into individual modules.
The core module is a well and stalls are additional modules. You still have a central UI for making decisions, but more control over the look of your town square. Oh and coloured cloth unfolds when you employ sellers!
Wealth propagation
Another thing players did that I did not like is they upgraded every single building in town to the max. In real towns you get nicer town center and less developed outskirts, but why should player do that if he or she has the means to upgrade everything eventually?
Even in 2021 the lack of variety and hierarchy really made larger towns flat and boring
My take on it was a wealth propagation system. When a good is sold on the market, the wealth "propagates" outwards to nearby homes. And you can only upgrade the buildings that have enough wealth stored. However I've been receiving feedback that this system is unclear and confusing ("why can't I upgrade this? I have enough silver"). It did however produce far more organic towns. Maybe another twist on this mechanic is needed that is more clear but achieves the same result?
Wood situation
The ability to zoom in is a frequent curse in places where you wouldn't normally expect it. Wood in strategy games tends to be just that. It's a symbol meant to represent everything from huge construction timber to arrow shaft sticks. It becomes weird if you can zoom and peep on what's happening with it.
In previous ver., I had "wood" and "timber". Wood was a generic resource, used for fuel in the winter and to create timber, which was used for construction.
What annoyed me about it is that it was represented differently depending on where it landed. After chopping the tree it was these small logs that people could actually physically transport. After becoming timber I represented it visually as planks. But of course when the construction animation is happening you see it as huge timber beams. It's not a *big* problem but, I always wondered if I can fix it. The appearance of wood flip flopped constantly between different visuals depending on what it was used for, and that caused it often to magically get longer or shorter.
You do not want too many similar resources cause they clutter the UI and make the game harder to learn. It also requires extra safeguards to prevent a situation where there is no 'firewood' and peasants freeze just because they have a lot of the wrong 'wood' resource. Again not all huge problems, but extra headaches.
What I ended up with for now is "Timber" and "Firewood". Timber is the large logs and must be transported by Oxen. Firewood is the chopped wood and can be transported by people. I thought that 3 wood resources clutter the UI too much so for now these are the two left in the game. I have a hunch people will like planks back however.
Sound design rework
Ambience proved to be more difficult than expected. The map is large and the seasons are changing in the game, the ambience has to adapt and evolve. The player can fly around like a madman, zoom in, zoom out in a split second. If two ambient nodes with the same sound synced by a very rare chance, it produces weird phase cancellation issues, aka sounds terrible and weird.
What I tried is to separate the ambiences into individual bird calls and noises that could be combined to create an ambient flow. However for some reason individual birds never quite sound like a forest. I dabbled with ambisonics 360 audio but the availability of sounds was limited, the equipment too expensive and the Unreal implementation didn't sound as amazing as I expected (maybe it was the samples?).
So I ended up with a 5 point node system that probes what's around the player and adjusts accordingly. It never uses the same sample to avoid any phasing. Together with the incidental individual sounds and noises it's seamless enough and reacts to the player, aka if you rotate the camera, the ambience rotates, which is far superior to what I had before.
Dirt & paths
To boost the organic look of towns I added an eroding paths system. The paths were created wherever peasants walked. However during a long gameplay it did not quite looked as cool as I expected - for example people took many trips to the forest to chop trees and it created these weird patterns.
Turns out if you simply erode the grass wherever peasants walk, after a long time you get weird geometric results like this.
Due to the size of the map the resolution of the path texture had to be high and was never enough. I ended up scrapping this system and instead take more care to place more small paths between buildings in a more controlled fashion.
Rebellions and fires
At the time of the announcement trailer, when Loyalty of the Region dropped to -100, you lost control over it. When you lost all Regions, you lost the game. It wasn't enough fun to lose, and therefore I couldn't amp up the challenge without it being frustrating.
I managed to add a rebellion mechanic where a part of your population would split and form a band of brigands. Those brigands would loot your town, burn some buildings and eventually, after they collect enough loot, leave. This also paved the way for a raiding mechanic in future updates.
Subsequently I added the ability for peasants to run to the nearest well, grab buckets of water and to fight the fires.
Soundtrack rework
For the OST I partnered with a company called Pressure Cooker Studios. They're working with high profile companies like Epic Games or Netflix and were eager to be a part of Manor Lords. We worked on a limited budget and it took a few iterations to find a perfect balance between 'cinematic' and 'authentic'.
We've been going through each track and revising it. While the music in the trailers I am very happy with, in the gameplay I think we had problems with too lighthearted of a mood and in places a lack of authenticity to the sound. We brought new talented musicians who helped us boost the "medievalness" of the OST while keeping some cinematic flavour and not making it sound like your typical medieval fair.
[music comparisons]
Coat of arms and banner creator
In the old system all you had to chose was a color and a symbol. It was "good enough", but a lot of players were interested in something more, so I created a more advanced coat of arms creator.
The unit banners in the game will display your coat, but part of the banner should remain individual per every unit. The unit symbol is their personal insignia helping to identify them in a battle, so I'm cooperating with a few artists to bring a variety of designs.
Evolution of one of the patron saints, St. George.
The symbols don't show up on units anymore however. I know, some of you will dislike it, but after the reworks I was all about individuality of soldiers and I think it is the best, most realistic choice to make them non-uniform.
AI rework
Every single change I made meant that the AI become obsolete. Residential plots? It can't build houses anymore. New market system? It can't trade. New recruitment system? It can't fight. So for the most part I stopped working on the AI altogether all until I know that the core reworks are more or less done. Lately I've been able to teach the AI to use the systems again but I still need a couple of months to make it fully competitive.
Fixes, fixes, fixes...
Keep in mind that for every major feature, there are 10 minor fixes or quality of life improvements. You don't see it in the trailers cause it's not flashy, but that's what makes the game good in the end.
So much more stuff!
I wanted to talk about victory conditions, tech progression, storage, diplo, dynamic sfx cues, AI autonomy - but turns out I hit the maximum word limit. Maybe it's for the best - next time!
New partners in the industry
A while ago I promised another video update that I didn't make in the end. That's because in the midst of character, animation and architecture rework, the game looked like a mess. However there were some big announcements that I missed to share.
1. Manor Lords has received an Epic Megagrant!
It was a no strings attached financial boost from Epic Games. The only requirement is that I am to continue working on the game and utilize Unreal Engine. This was really great news. For the first 2-3 years of the development I still worked at my filmmaking company, I had a lot of flexibility but Patreon and the Megagrant really given me the opportunity to pause my day job and focus on the game + still have some funds for outsourcing artists like Marcos.
2. Manor Lords on GOG
Since we're on Steam, I don't want to advertise it too much but the game is now also available on that one friendly Polish game store. GOG also supported the development of the game, for which I am very thankful.
3. Cooperating with Nvidia
First we cooperated on getting DLSS implemented in Manor Lords. Nvidia was kind enough to provide a newer GPU for me since I was still on my trusty 1070. I also received confirmation that Manor Lords will be available on Geforce Now which is great news for anyone with an older PC but a reliable internet connection.
4. I partnered with Hooded Horse publishing!
I was very lucky to meet a lot of great people from the industry who offered help after I announced the game. I planned to stay independent, but in the end due to certain chain of events I decided to partner with Hooded Horse. It's a publisher dedicated solely to strategy games (you may know them from Old World or Falling Frontier) with relatively short history, but growing rapidly and most importantly sharing the same values as me. I also have retained full freedom. They'll help with translations, marketing and given me more development budget to play with. "Finally!" some of you will surely say.
Now for the big announcement...
Closed alpha applications open!
I'm looking for players who would like to help with the development, test the game, give feedback and participate in discussions.
If you have a lot of free time and are passionate about the project, it would be great to have you in the closed alpha group.
However, if you just want to simply play the game earlier/for free - please wait instead, as you'll only be frustrated because of the unfinished state. Same if you don't want to participate in internal discussions or votes. Ability to record footage is also a plus.
If you'd like to join the Closed Alpha:
1. Join the Manor Lords Discord server https://discord.gg/RRD55RZdB7
2. Find a bot called AlphaBot.
3. Send a private message to the bot with the command !alpha
4. The bot will ask you a couple of questions.
5. Wait for an e-mail confirmation. I plan to pick new testers irregularly so don't worry if you don't receive a message immediately - as always patience is appreciated.
If you're applying and you're a Patreon supporter please make sure to connect your Discord account in the Patreon settings. If you joined Patreon before but aren't there now, or if you're a long time supporter who just followed the game, first of all thank you, secondly please mention it.
Note: If you can't send the message, please check this setting:
When game?
I really appreciate all your patience and I hope you now better understand why it takes longer than expected. I didn't have to rework all this stuff, but I wanted to and I took a risk. I think all the major reworks are behind me, but first I wanted to hear feedback from more testers before I make the early access public, and there are still some buildings left to redo and the AI to improve.
For now I hope you'll excuse me for leaving the release date as: To be announced.
That's it for today, thanks for following and all the help!