This is the 2nd blog in our new series, which discusses developing our updated engine that will power the next title in the Grand Tactician-series. We’re keeping some details, like historical setting, vague on purpose to discuss at a later stage. But, we’ll keep releasing these peeks behind the scenes to let you know what to expect in terms of game play and improvements.

The New Unit System

One of the first things we developed for The Civil War (during 2016-17) was what we call the Unit System. This is how units are created and handled in the battle layer, and it has a big impact on a multitude of other things: visuals, controls, performance, AI, to name a few.

Re-building the Unit System is a key part of improving the battle layer for the upcoming title. The results are very encouraging and we’re excited about the possibilities it brings to the table, both for game play and also for performance and clunkiness (or lack of, this time around) of the battles. This will of course work hand-in-hand with terrain and movement, group-system, combat system(s), AI, and so on.

The Old System

In the old system, we built the units from what we call “company blocks”. These are sub-elements of the units, and platforms, that are moved depending on the unit’s formation and positioning. The 3D models (or as the game originally featured, sprite soldiers!) are attached to each platform.

In The Civil War (1861-1865), or GT1, each platform has 4 soldier models in 2 ranks. This is to show line formations in 2 ranks and marching columns in 4 files. Only 1 model type is supported, and all special situations are “hardcoded” (a word we use to determine something that is non-adjustable by editing the game’s preference files). Such special situations include the officer model and the artillery units’ model arrangement. The different unit formations (line, assault column, column, square) were built by determining how the company blocks are placed side by side or behind one another.

Tracking each platform allows curving and quite OK-looking column movement (“snaking”) for each unit. When resizing the unit in battle, due to changing game settings or due to casualties, the number of platforms per unit changes. In the system, the more platforms there are, the more severe the performance hit, as the company blocks are used for many combat and terrain placement calculations. The end result could be described as “clunky”, and it also limited the performance being rather CPU heavy (like the game is, in general, instead of many modern games that stress the GPU more.)

Image 1: Here you can see how the units are built in the system. The “company blocks” are highlighted in red. Each contains a set number of 3D soldier models (2×4), the officer is added as a separate block, and the 3D unit flag simply floats in the air in the coordinates where the unit is calculated to be stationed in. When giving a movement order, the positioning of each company block is calculated, as seen in the terrain (right).

With the tools we had when developing the old unit system, we had many restrictions. For example, each unit shows only one type of soldier model (except for hardcoded special situations mentioned above), and the number of soldier models in the battle is limited to a few thousand in total due to performance. This means a regiment of 1000 soldiers is normally visualized by way less than a hundred models (in the above image the unit shows some 40). This means a 1:25 scale (one 3D model = 25 real soldiers). Not very “grand”, right?

The Rebuilt System

Just tinkering with the old unit system would have improved the situation in general by a limited degree. For that reason, we scrapped the old system completely and rebuilt the new one using the knowledge we’ve gained and also modern tools that have only become available more recently.

In the updated GT engine, units are built differently. Instead of using set coy blocks like Legos, we determine the composition of whole units, soldier by soldier. Each unit can contain many different soldier types including special ones such as officers, and drummers, each with their own 3D models and animations. The per soldier data also contains info for combat simulation. What this allows, for example, is units engaging in multiple directions depending on the combat situation and using different kinds of firing systems, such as different numbers of ranks firing simultaneous volleys or in succession.

The units now have more flexible formations for different combat situations they can assume depending on conditions around the unit, the situation of the unit itself, and the training level. As an example, a cavalry unit may use different formations for marching, battlefield movement, ranged engagement, skirmishing, and a few different for charging. Together with the more flexible firing mechanisms and ability to fight in multiple directions the battle micro-AI will have fewer difficulties in countering player’s moves on the lowest, i.e. unit, level. (The battle AI is divided into unit-level “micro-AI” that reacts to surrounding conditions and a “macro-AI” that handles moving groups of units and assigning stances.)

From a management point of view, drill is more important than before, as raw troops cannot perform the more complex maneuvers as a cohesive unit. This limits their effectiveness even more compared to GT1.

Image 2: This is a sketch of a Civil War Infantry Regiment in line. With the new unit system, a more detailed unit deployment such as this can be created, including positioning of the flags, officers, NCOs, musicians, and so on.

Sounds rather performance consuming, right? On the contrary: with the new tools and much improved coding magic, instead of showing a maximum of a few thousand soldiers in a battle (GT1 upper limit), we are talking about tens of thousands – and this with better performance than in GT1!

And much of this is moddable.

How this is technically pulled off, shall remain our secret for now, so let’s not dive deeper into details. You may read between the lines we’re quite excited about the opportunities this new unit system brings to the table. But that’s not all! In the next blog we’ll take one step deeper into what it means when we use the new system to simulate ranged and melee combat on the single soldier level – as part of the improved combat model!

Most Respy,
Gen’l. Ilja Varha
Lead Designer – Grand Engineer Corps

Gen’l,

With this log entry, we start to look into the future of Grand Tactician and what new and reworked features the next title will have when compared to The Civil War. Currently we’re rebuilding the battle layer, so we’ll focus on the changes there during this and the upcoming log entries.

The New Battle Layer.

As discussed in our earlier log entry, there are many aspects we want to change in the Grand Tactician’s battle layer. This is anything from the limited number of battlefields, rigid unit system, and visual scale of the action, to the combat system itself (especially lack of proper melee system), performance, AI… As was discussed, it’s better to do this as one big overhaul vs. trying to improve each sub-system within the main system. While doing a re-write, we can also include other ambitions, like improved moddability, support for localization, and modularity that will later make patching and further development much easier.

The first new thing to discuss has to do with the battlefields themselves. In The Civil War (1861-1865), which we will refer to as GT1 from here on, all battlefields are drawn by hand and then integrated into the campaign map using coordinates, radius, and any randomization options. The system was developed as early as 2016-17 and was among the very first things we ever created for Grand Tactician. As each map took me up to two weeks to draw and integrate into the battle layer (to create battlefield functionality for the units), this naturally meant that only a limited number of battlefields are available in GT1 – 30 historical ones and 13 non-historical ones to be exact. It took a similar time for Wasel to draw the beautiful papermaps in parallel.

How the GT1 maps were created is discussed in further detail here in an old log entry from 2019.

Image – This is how GT1’s main campaign map looks with all the historical battlefields.

This approach has its pros, but also quite heavy cons:

1. - As you can see in the above image, most of North America is void of battlefields. As each battlefield is 8×8 miles in scale, the populated area is limited. We added 4 “random map” sets to fill the void. The maps within these sets (mountainous, farmlands, wetlands, plains) are called “random” as one in the set is randomly chosen when a battle is initiated.
2. - Only the most important battlefields have sufficient reference material available. The topography of historical battlefields looks very different today, and proper maps (or battle descriptions also focusing on the topography and terrain features) are hard to come by. Even if drawing a map out of only rough or almost non-existent reference material, doing it by hand still takes a lot of time. When we look at different historical periods from the 17th to 19th Centuries, the amount and quality of reference material for the Civil War battlefields is probably the best, all other conflicts have it mostly much worse.
3. - Most importantly from the gameplay point of view, the campaign to battlefield topography link is missing. This means the cunning maneuvers on the campaign map play no role once battle is joined. As the nearest available battle map is loaded, this could be a hundred miles away from the area where the campaign map depicts the battle taking place, with a completely different topography. The river you were following and hoping to anchor your flank to is missing, and so are the roads, hills, forests…

The 20th Battle of Manassas?

We tackled all 3 main issues by developing a map generator, which creates the battle maps from campaign map data. The data is created from multiple bitmaps that cover terrain height, infrastructure and vegetation. Location data allows adding the names on the map in correct locations.

In GT1, whenever you fight a battle in the Northern part of Virginia, you end up fighting it out in the Manassas (Bull Run) map, and always more or less around the Bull Run -line. The new system allows creating an unlimited number of battlefields that will always represent the campaign map situation accurately. The terrain is generated according to data in that exact location, including all the main terrain features you’d expect to see on the map. Now, army placement on the campaign map plays a much bigger role, introducing one important aspect more to the aspiring Grand Tacticians!

While the introduction of the terrain generator does not remove the possibility to use hand-drawn maps, it also introduces two more pros: scalability and moddability. Depending on the campaign map dataset scale (i.e. how many meters does one pixel in the bitmaps represent), and the size of the “slice” (X pixels vertically/horizontally), the battlefield scale can be easily changed. The data is read from bitmaps and text files, meaning it does not require any dedicated modding tools to create a complete dataset to generate maps.

The Terrain.

We also made some changes in the terrain itself, and the unit to terrain link, to allow more realistic unit behavior in different types of terrain. These aspects are built around terrain types (and properties) and navigation data, that also control units’ pathfinding.

In GT1 the terrain types and navigation data have a quite simple connection. The terrain types are limited to only a few (open ground, fields, forest, swamp, water), while most differences are visual only. Further granularity is added through linear objects like roads, streams, fences… Units can move on all terrain but water (if not frozen), and very steep slopes, but otherwise, there are no limits. Also, how the terrain affects units is always the same: forest slows down skirmishers and cavalry brigades just the same, and does not prevent artillery deployment like in reality.

From a pathfinding point of view, all terrain is the same, so when giving a movement order (player or AI), the units do know to prioritize roads for fast movement, but do not make a difference between an open plain or swamp. This leads to units moving through difficult terrain more often than necessary, getting tired and slowing down, which in turn breaks group (for example division) cohesion, and affects especially the AI.

In the new terrain engine, the number of different terrain types has been expanded. For example, there are forests of different tree density. They are now included in the navigation engine, so each terrain type can have its own navigation characteristics. This is further augmented with unit heaviness, which determines how different types of units are affected by terrain. For example, very light units (such as skirmishers) have no issues moving through a forest, but a large cavalry formation or artillery battery cannot enter a thicket, or try to cross a swamp. This is part of pathfinding, so giving movement orders will take this into account, units preferring easier terrain, and going around obstacles – when allowed to.

Image – Left: the old terrain system. From a pathfinding point of view open terrain, forests, fields are the same, even if the latter ones slow units down more. Right: the new terrain system. Here each terrain type (different colors) has its own pathfinding properties and characteristics for different types of units.

While at it, we also added further detail on how weather affects the terrain. Instead of raining true/false as in GT1, we now track precipitation and terrain wetness separately. The former increases the latter, and when the rain stops, the terrain will slowly start to dry up. During winter, the same is true with snowing and terrain snow coverage, which increases or melts gradually.

Like said, this is the first building block in the re-worked battle layer. In the next log entry we’ll take a look at how the unit system has changed, and what kind of effects this will have on game play, performance, AI and moddability. Until next time!

Most Respy,
Ilja, Lead Designer
The Grand Engineer Corps.

Gen’l,

For some time now, an eerie silence has surrounded the Grand Engineer Corps camp. The troops are wondering, what is happening inside the fog of war? What does this mean to the Grand Tacticians, present and future? Major changes have taken place in the background, with the Engineer Corps going through a reorganization. As the pieces are falling in place, it’s a good time to share some Engineering intel from behind the scenes.

The Greenhorn Volunteers.

Grand Tactician’s story began in 2016, with Oliver (a finance banker turned coder from Austria) and Ilja (a wargame enthusiast in the Finnish Army) deciding to develop a strategy game independently. The idea was to create something unique in the genre, the game taking place in the American Civil War – a conflict of great interest to us. Soon after Peter (a chemical industry worker and mod artist from Germany) joined the team as an artist, and since then our team of three worked hard to make our dream a reality. We started from zero, new game engine, new coding language, and a lot to learn…

The Civil War (1861-1865) has been a true independent “garage project”, being developed outside the working hours as a second job. It started as a very interesting (yet ambitious) hobby project, with no commercial expectations. The development didn’t have a fixed schedule, we worked as much on the game as we could, learning as we went. We have been blessed with many helpful and talented people with enthusiasm toward our project and the will to help, and this helped shape the game into what it is today – not perfect by any means, but something we’re proud of. We managed to create a unique game among its peers. 2000 reviews with > 70% score is high praise!

What we never expected was that so many players would find the game and enjoy it, providing us ideas and constructive criticism from the Early Access of late 2020 continuing all the way to this day. This of course caused immense pressure to deliver. The post-release development has been even more hectic than the time before. At times, there was exhaustion among the ranks, as all the work was still done after returning home from a day (or night) job, work trip or even military exercise. The COVID years and increased tempo at work for some left a mark in the development as well. This was especially true during development of the Whiskey & Lemons DLC. In these circumstances, while frustrating at times, it has been very flattering that our humble game has been compared to titles created by much larger developers with much larger resources at hand.

At the same time we knew we could not continue like this in the future. Demands at work, with family and with the game were taking an increasing toll. We’re not young guys anymore.

The Long Road Ahead.

We always wanted to make Grand Tactician a game series to portray multiple wars of the gunpowder and linear warfare era (17th to 19th Centuries). If for now we disregard the personal issues we had with resources to invest in game development, we also had reached a point in the Grand Tactician -engine where small fixing here and there no longer was achieving sufficient results. Like said, we know the game is far from perfect, and there are the shortfalls here and there we want to improve. These are mostly the ones the community knows very well and have been providing us feedback about (there are also some technical ones, that are not apparent to players). “Clunkiness”, performance, AI, battle balance, player onboarding, information…

Military minded people remember the approach to putting new wine into old bottles… In order to really improve the battle AI, as an example, working on the AI itself alone no longer helps – work is required in other areas as well, like the whole unit system, pathfinding, order of battle, parts of the combat systems. Tinkering with all these systems individually to help the battle AI a bit will cause issues in other areas, that in turn need to be reworked, and the butterfly effect continues. And still we would have the old unit system, pathfinding, etc. that we want to improve anyways, and that are affecting other areas (clunkiness, performance…) So, in some cases we need a new bottle for the new wine.

During the eight years we of course learned a lot, but also the tools available have constantly developed. Especially the battle layer we developed first during 2017 is showing age and no longer facilitates much improvement without a rehaul. So, we decided that in order to move forward we need to rewrite big blocks of the game engine. This of course takes a lot of time – it’s “revolution vs. evolution”, like game developers like to call it. At the same time it allows further changes like increasing modding opportunities and adding localization.

But, back to reality and the resources… If the volunteers are worn out from years of battle, to rally the Corps, veteran regulars are needed.

The Regulars.

Recently, the Grand Engineer Corps has gone through a major reorganization. As things were still in motion and results were foggy, we kept quiet about this, while already re-working hard the first elements of the game engine. At the same time we released multiple patches to fix the remaining major issues reported by players.

Now we’re very happy to be able to say that things have turned around for us in a major way. We have teamed up with a like minded publisher, the right one for us and for our project – a Grand Tactician -series. With this move we have been able to transform from exhausted garage-developers into professionals – volunteers into regulars – resigning our earlier jobs and now working as full-time game developers! To put it mildly, life is now very different and so are our development opportunities.

What exactly we are working on will be disclosed at a much, much later date. What can be said this early on, is that we’re working on taking Grand Tactician forward in a way that earlier was not possible for us. While not much more can be said at this point, we’re planning on sharing a developer log regularly, approximately every 2-3 months, to give you a small glimpse into our upcoming game. While our vision for the next game is already set – and this takes into account the feedback we’ve received –, we do want to give you an opportunity to provide feedback, during development, at a more detailed level. We’ll start with some more technical matters, and – in full transparency – intend to hold back the major details, until there is a better opportunity to share them. We hope you share our ambition of making Grand Tactician into something greater, and that to do that we need to let the marketing machine have something to run on, when the time is right!

While our work continues, players of The Civil War 1861-1865 will be supported by the development team like before. The best way is to drop an e-mail to our message box info(at)grandtactician(dot)com, as we are deep “in the matrix” – there are only so many pairs of hands. We’ll do our best to help you with issues and questions you have. What is also fair to say at this point is that no new features or downloadable content are to be expected for The Civil War. We hope this is understandable, as large parts of the game’s engine are already cut open on our field surgeon’s table.

Having shared this piece of intel, we continue our long march toward what we hope and believe will be a bright future for Grand Tactician. Touch the elbow!

Very respectfully, your obedient servants
Oliver, Ilja, Peter
The Grand Engineer Corps.

Gen’l,

There were many legendary generals in the American Civil War, like “Stonewall” Jackson, “Old Pete” Longstreet, “Grumble” Jones, “Bull Head” Sumner… each with their own unique stories. In the Grand Tactician DLC, it’s your story.

The Way to the Top – or Not.

Whiskey & Lemons will introduce a new career mode. You will start by creating your character, determining your Civil War personality via a questionnaire, and taking command of a regiment.

As a regimental commander, you are but a small gear in the huge military steam engine. You will have very limited influence over the war, and you will receive orders from your AI commanders. According to your actions and performance, you will generate prestige and get promoted to higher places, where your actions will matter more.

As the DLC campaign will be regiment level, on a zoomed in campaign map, divisions will be the smallest units moving on the campaign map. Single brigades can also be used as garrisons to protect important terrain.

Once commanding a division, you will be able to control the unit on the campaign map. With increased prestige you will be able to influence politics or projects, but this is very limited before rising to the top, commanding the armies of the Union or the Confederacy – provided that you will ever make it that far. Or want to – you can play the whole campaign without accepting any promotions, if you so wish.

Your Story.

There are many things that can get in the way. Your previously mentioned personality could be one of them. Personality is determined on six axes: Loyalty, Piety, Social Life, Heroism, Stress Control and Political Views. These may provide positive and/or negative traits to your career.

You may also attract companions that will be part of your staff. For example, if you are known for lack of loyalty, you may be followed by a Committee Investigator, reporting to the military brass, who may end up limiting your freedom of action by giving political guidance you cannot ignore.

There are also other new features. You will be able to choose to buy your own equipment, recruit your own staff and carry out actions that all will influence your, and your nation’s Civil War journey. You may also get injured in combat, with some of your injuries becoming permanent disabilities.

When you start to gain fame, you may end up with a grudge with some of your fellow officers. This may have a serious impact on your performance in the battle-field, like what happened between Bragg and his lieutenants. And getting rid of them may prove difficult, as it will require further prestige to overrule choices made by the President.

It will be very different, if you become known as “Old Tippler” due to your drinking habits, keep falling off the horse during battles and get into drunken fights with your fellow officers, or as “Skedaddle Joe” due to lack of courage and losing the respect of your men.

Everything you do and experience during the campaign, will be recorded into a Biography. When the campaign ends, you may read your whole story, and see how you compared with historical civil war commanders.

The Other Things.

We’re looking to release Whiskey & Lemons during Q2 of 2023. The plan and development has suffered some real-life related delays, mainly caused by a war in Europe, where the development team is from.

But, while on some fronts there have been delays, on others extra progress has been made. We’re constantly working on the game engine even if the DLC content has not proceeded as earlier planned.

Also the band is recording new music to the game. Along with the DLC release, we will also release the new music as a Soundtrack, the same way we did with the game’s original music – meaning buying the soundtrack is not mandatory to hear the music in the game. The Soundtrack II (working title) allows listening to the tunes outside of the game, and the revenues will go to support our great band, Wasel and the Weasels.

With the DLC release, we’ll also release the next major patch, which is focused on fixing reported issues, implementing some player wishes (like different color fire arcs on paper map), and some performance improvements.

Most Respy,
The Grand Tactician -Team.

Gen’l,

Weapon production and procurement system has seen some changes during the development of Grand Tactician: The Civil War (1861-1865), most recently in the 1.08 patch. Let’s take a quick look at how the system works now, vs. previously.

The ability to equip and re-equip units with different weapon types has been one of the unit customization features in the game since the beginning. Originally, since the game’s Early Access release in fall of 2020, the weapon production system was quite heavily abstracted, which many players found unsatisfying, as they had no real direct control over the production.

The Old System.

In short, it used to work like this:
- Industrialization & Diplomacy policies made certain weapon types available, with the better weapon types requiring higher level policy to “unlock”.
- Once the weapon was available, production/import started automatically. Depending on fixed “standardization years” and “equipment modifiers”, the weapons were added to player’s “pool”.
- Standardization year: each weapon had a year, from where onward the weapon type was available. During approximately 2 years the standardization would increase following a curve. Example: weapon standardization year is 1861. In 1861, more weapons of this type became available very slowly, and in around 1863 the production would reach the highest level, as per equipment modifier.
- Equipment modifier: using this value, the different weapon types were balanced according to historical data. Example: Springfield Rifle-Musket, with historical production numbers of more than 1 million units, had a very high modifier, meaning much faster production than say Richmond Rifle, that had historical production of just a bit over 30,000 units.
- Weapons were not tracked as individual pieces, but rather per units to equip.
- Weapon costs were realized at the moment units were equipped, and the costs were depending on number of Weapons & Artillery goods types in markets, which tied the system to economy.
- If a weapon type was commonly used by units, it’s production would speed up a bit.

As said, the system was quite abstract, and limited player’s options. For example, if the player would have wanted to focus the industry to produce repeating rifles, this was not possible, as these weapon types became available very slowly. For this reason in patch 1.06, which focused on economy in general, we changed the system. But there were some bugs, plus a few remnants from the old system, so we wanted to revamp it further in version 1.08 of the game.

Fayetteville Rifle has low standardization, as only 7509 examples of this weapon type have been produced so far.

The New System.

The new system works quite differently if compared to the old one:
- New weapon types are made available via Projects, instead of policies. This adds the possibility to more directly influence which weapon to procure. If the player wants to focus on repeating rifles early on, it is possible, but it requires much higher subsidies than more simple weapon types.
- Using the new Weapons -panel, player can place orders for available weapon types. These can be either produced domestically, or imported from Europe.
- Weapons are tracked by actual number of pieces per weapon type. This also allows tracking of weapons captured in battles, including them in the stock.
- Each weapon is ranked per Production Complexity. The more simple the weapon, the cheaper and faster it is to produce.
- Equipment Factor is used only to balance production times between weapon types, or availability of imported weapons in Europe. For example, it’s used to make a difference between 10-, 20-, and 30-pounder Parrott Rifles, making the larger ones a bit slower to produce, or making sure more commonly available Enfields are faster to import than, say, Dreyses (which historically were not imported).
- Standardization, which affects only domestic production, plays a big role. This value represents the readiness of the nation’s arms industry to produce said weapon type, which in turn depends on the number of weapons of said type in use. If a weapon type is common, it means weapon industry has the required machinery ready to produce further pieces, and also has the means to quickly produce required ammunition and other accessories. When standardization is low, the industry needs to set up the required infrastructure and machinery…
- …which means weapons with high standardization are much quicker and cheaper to produce.
- Weapon costs are realized when an order is placed. The player can place one order per weapon type at a time. Once the weapons are delivered, they can be used to equip units. At this point the costs will cover the required logistics.
- Shortages of needed materials will make production slower and more expensive.
- Weapon imports are affected by blockades. If blockade efficiency is high, less goods get through, meaning orders will take much longer time to be delivered.
- In the Weapons -panel, effects of these mechanics are explained in the offered tooltips.

Example, purchasing 1000 rifles. Left: Union order of Springfield Rifles with standardization of 100 costs $201k and takes 14 days. Middle: CSA order of Richmond Rifles with standardization of 6 costs $293k and takes 62 days. Right: CSA order of Enfield Rifles from Europe costs $344k and takes 17 days. Conclusion: highly standardized, domestically produced weapons are cheaper and quicker to get, but ramping up the industry takes time.

With this new system, weapon production and import (= procurement by the government) are more controllable by the player, and also better balanced. If the player wants to get a lot of weapons quickly, importing is a better option, as long as trade routes are open, but the costs are higher due to the required logistics. If player wants to domestically produce weapons, it’s not possible to start mass production right away, as the industry is not ready immediately. It’s better to place smaller orders first, to increase standardization – later production will be much faster and cheaper than importing, but it takes some time. It also means it’s possible to balance between fast mass production of simple weapon types, such as muzzle-loading rifle-muskets, or more complex types like repeating rifles. And the link to economy is also more apparent: if production facilities suffer shortages on materials, or trade routes are not open, weapon procurement will be directly affected, in the short term as well as in the long term.

I hope this short study explains how weapon production works in the 1.08 version of the game!

Most Respy,

Gen’l. Ilja Varha,
Chief Designer, &c.,
The Grand Tactician -Team