• Total War: Three Kingdoms Review (Part 2)

    Total War: Three Kingdoms Review (Part 2)
    by zoner16

    zoner16's review continues from Part 1


    Another feature that got majorly overhauled was espionage. After the controversial removal of agents in Thrones of Britannia, Three Kingdoms brings back most of their offensive functionality through the espionage system. Spies are now just characters that you inject into other factionís recruitment pools in the hopes that theyíll get recruited to serve the faction while also completing espionage tasks for you. The system has a lot of potential and can be used to pull off some devastating twists, especially late game. However, getting a spy into position is a difficult task, as factions will rarely pick-up low-level characters, requiring costly investments to even get started. Furthermore, your target faction might not have enough turnover to be recruiting at all. This means that spying can end up doing little for much of the game. However, once you have enough spies or good luck, there are some fun options to use.

    Depending on the position of the spy in the target faction, one can do everything from discrediting certain characters to defecting with entire provinces in tow to starting a civil war if one makes it to the right positions and has enough of the Cover and Network resources, which build up over time. These options can get powerful, but it often feels like its controlled by RNG as to when and if your spy can get into the right places. Visibility at least comes with any amount of spying, and this is always useful.

    At the end of the day, the espionage system is an enjoyable add on when it works, but itís a bit too RNG dependent to have much strategy other than casting a wide net with spies and hoping that one eventually worms their way into a useful position.


    Recruitment may be the strangest part of the game. On the surface, itís taken from Thrones of Britanniaís universal recruitment system. Units can be recruited anywhere in friendly territory, start at a fraction of their strength and have to muster for a few turns, and have no infrastructure requirements. However, there are some additional restrictions layered on top of this. The first is level, as a low rank general can only recruit low rank units, with more units coming as time goes on. As mentioned previously, the Wu Xing class system means that most mid tier units can only be recruited by generals who match their element, though everyone can recruit the five basic militia units. Bizarrely though, the highest tier unitsóthe so called ďDragonĒ units as well as imperial guard unitsóbecome available to everyone after the respective reform has been researched or the proper faction rank gained. Additionally, the faction specific units are available to everyone within that faction.

    Once recruited into a generalís retinue, the unit is bound to that general. They cannot be moved to another retinue unless their original general dies and is succeeded. Units no longer lose experience from taking casualties and replenishing, and if the unit is destroyed to the point it must disband, it will automatically start remustering with its exact same experience level after a couple turns. You can now replace units such that the new unit gets the soldiers that the old one had and doesnít have to muster, but for some reason, this doesnít maintain experience level, even if upgrading within the same class (i.e. going from Ji Militia to Ji Infantry).

    This system creates a lot of weird priorities and is riddled with contradictions. Because they have no building or region requirements, units donít really feel like they come from anywhere . Since they donít lose experience and just remuster when destroyed, they donít feel like theyíre alive and thereís no incentive to care about their fate. Youíre encouraged to match units to the right general, but you canít move units between retinues if the generals change or you hire a new general with an existing retinue. Replacing units is supposed to save you time remustering them, but the hit to the unitís effectiveness from the loss of experience is somehow greater than that of the unit being wiped out. Added to the fact that mid-tier units are locked to specific classes while high tier units arenít, this just encourages you to rush top tier units and skip the mid tier entirely while relying on experienced militia units.

    Overall, the system reduces units in campaign to little more than gear for your generals, with no life or presence of their own that you just trade in for the next best one when they become available. It canít make up its mind what it wants you to focus on, so it overloads you with counterintuitive rules that it communicates poorly.

    Army Control

    The only truly ďnewĒ part of army control is the way that retinues play into it. An army is at least one retinue but can be up to three. One of the three generals is chosen as the commanding officer, which will allow them to confer certain buffs that only apply if commanding but has no impact on relationships or satisfaction. This also means that armies can combine and split so long as retinues stay together, allowing for more flexibility compared to the indivisible armies of recent titles. However, it runs into the problem of the retinues themselves being limited by the recruitment system. Having balanced retinues capable of operating independently is difficult to achieve without late game unlocks of high-level units recruitable everywhere. The game is actively encouraging you to combine the right generals to make a balanced army, so balanced individual retinues are discouraged, which also discourages splitting the army. The choice is good, but the class system gets in the way again.

    Otherwise, moving about the campaign map is the same as it has been since Rome 2. You have the Normal, March, Ambush, and Fortify stances, all of which mostly operate like they have in the past. March does not get you ambushed if attacked and still allows you to retreat, but this applies to the enemy as well and will make chasing down weak stacks in your territory very annoying, especially since they cause public order penalties in the regions they are in. Balancing this is the military supply system from Thrones of Britannia, which limits stays in enemy territory for everyone. Supplies drain incredibly fast in rough terrain and during winter, which gives some more weight to planning the timing of attacks. This does do a certain amount to limit the extent to which enemies can run away from you within your territory, but what theyíre capable of is still more than enough to cause major frustration.

    Getting into a fight will prompt the autoresolve, which has been changed so that itís completely deterministic rather than partially random. While it does take away from the gamble of autoresolve and make it a bit too easy to rely upon, given the number of battles you will be faced with, Iíve found it to be more of a mercy that prevents the need to waste my time repeating encounters I know I can win, but wouldnít be interesting in the slightest. Fighting every battle manually to prevent the autoresolve from destroying my army and killing my generals by chance is not something I find enjoyable, especially due to some battle issues mentioned below. The random elements are instead around the post-battle rewards, which now include captured generals once again, on top of loot and regular prisoners. This gives victory an even sweeter aftertaste and encourages difficult fights for to gain good generals, cash, and ancillaries.


    The meat of Total War has always been its real time battles, and most titles in the series are made or broken on their quality. The battles in this newest title take lots of steps, but not all of them are forward, and itís difficult to say if it ends up going anywhere.

    From a command and control level, battles are mostly the same, with the only major difference coming from the segmentation of the retinues. Each stack is three retinues max, with each retinue being one general and six other units max. Battles are locked to 6 retinues per side on the field at a given time. This just replaces the old 20/40 unit limit for armies and battles with a potential 21/42 unit one if each retinue is fully built. While there is a ďcommandingĒ general among the three for each army, if a general dies, only their retinue takes a morale hit. However, the morale boost from aura that each general gives to friendly units is not retinue-specific, so there isnít any bonus to keeping generals with their own retinues in battle.

    Aside from the strange absence of the Rally and Inspire abilities, essentially the entire suite of unit options and abilities is present from previous titles. Even dismounting cavalry makes its return. There are also several unit formations available in the game, though these can be somewhat hard to come by for a couple reasons. First you need at least mid-tier units for most of them, as militia tier and lower units cannot adopt any formation other than Loose Formation. The issue then becomes that you canít use any formations without either a mid-level strategist or a few specific ancillaries, and even then, those only work per army, so reinforcements donít get the bonus. Getting that convergence of mid-tier units and mid-level strategists is a rare occurrence in the first half of the campaign, so actually being able to use them becomes logistical chore. Whatís more irritating is that it is another place where color matching the elements becomes more important than picking the best generals for anything but class. An army with all level 10 generals may have been fighting with distinction for decades, but if none of them are Strategists, none of the men will ever be able to use any formation unless you equip an officer with a rare book that tells him how. This gets worse if you have familiarity with the source material, as most of the Strategist characters are political advisors or army group level commanders, not drillmasters or tacticians, so the binding of unit formations to them is just as nonsensical as it is denying it to everyone else.

    This turns out to be a contributing factor to a feeling of sameyness between the battles. While the roster is respectable and you can run into basically any unit combination under the sun, theyíre mostly going to play out similarly due to redundancy in application and a lack of real versatility in their roles. Missiles winnow the line in the opening, infantry holds the line for most of the battle, and then the cavalry flank to break the line at the end. Some deviations to this formula can arise in non-standard battles like sieges and ambushes, or in the case of unbalanced numbers, but youíll rarely get anything that enables truly new tactics as the game progresses. Each unit type will usually be too terrible at other roles to allow for flexibility. Building an army around infantry shock for example, will usually be pointless because even units whose role is listed as ďAssaultĒ usually will never be able to break through another infantry unit before the cavalry engagement has finished and the victor is decided by whoever won it. Certain formations might help with this, but they end up being too rare in the early game, and by the time they get common, unit types have specialized even harder to make them almost obsolete. There are plenty of viable army builds, but there are very few viable army tactics.

    The aspect that usually ends up mixing up the tactics is the terrain. Elevation still takes its toll on stamina and line of sight, snow and shallow water will slow down units to make them easy targets, and the occasional rock formation or village will block off movement entirely in certain areas but, the real standout is forests. Not only do forests absorb missiles, hide units, and block line of sight, but they can also now be set aflame, dealing health and morale damage to any units caught in the fire. This brings to life the ďFire AttackĒ which is so often referenced in both the history and novel of the Three Kingdoms. Fire makes fighting in them a real gamble. Their protective effects can be used against the defender by setting the treeline alight, but the blaze will affect both sides and may spread beyond the scope of where the igniter was planning. If used poorly, one can find themselves in an inferno of their own making while the intended victim retreats to an untouched area. By contrast, a disappointment compared to other titles would be rivers, which now just count as shallow water and are fordable at almost any point. The few maps that do have bridges donít play like bridge battles from previous games as theyíre not really chokepoints and more just express lanes, albeit rarely used ones.

    Keeping a cohesive formation is less important than keeping momentum on your side. Charges are brutal, especially cavalry ones.

    Sieges are in a strange place. The cities are large and expansive, with good fortifications and plenty of options for layered defense, and representing the infrastructure built in campaign on the battle map finally makes its return. They are noticeably very square and follow a very obvious layout paradigm, but this is historical to how Han cities were built. However, thereís a strange lack of siege equipment present for a siege assault. The only options constructible on the campaign map are rams and sapping points (which just collapse random sections of wall). Additionally, there is the trebuchet unit, and all other units can use grappling hooks to climb walls at a cost of stamina. Historicity problems aside, most of these options end up being boring ways to play. Rams are as straightforward as it gets, sapping doesnít have any interaction after the battle starts, and trebuchets have incredible accuracy and range but low ammo, thereby ending their role almost immediately. The grappling hooks are the only method of siege assault that ends up being tactically interesting, though it often feels cheap due to how long the walls are compared to how small the garrisons are.

    For the most part, the remaining battle types move forward. Ambushes now have an escape point that allows for a player to salvage what would have been a doomed situation. Encampment battles have more space and less useless obstacles to mess up the pathfinding, though for some reason the rubble from burned down walls blocks movement. Minor settlements now have maps with interesting layouts that are themed to the type of resource, so attacking an iron mine plays very different from attacking a horse pasture. It helps break up the monotony of the more formulaic field battles and sieges.

    The only truly new feature in battles are the duels in Romance mode. How they work is that any general not of the Strategist class (who are barred from dueling) can challenge any other general not of the Strategist class to a duel if close enough to each other. If the other general accepts, the two will run towards each other and then be locked in combat until one dies or runs away. All in all, thereís not a lot to say about this feature. Itís a good way to snipe enemy generals if you know you can win since the AI will never run away. However, the debuffs for losing or running away make it mostly not worth it if thereís any chance of your general losing. Itís almost always safer to just tell a general to attack another normally if thatís the case, since you can easily run away without penalty if things get hairy. Thereís also very little interaction aside from remembering to hit abilities when theyíre off cooldown, to which thereís essentially no thought or tactics to. Some duels also just last absurdly long times, to the point where the rest of the battle has ended and the generals are still dueling. I personally find the tactics around trapping and killing general units in Records mode more engaging.

    The end of a typical duel. Probably the most exciting part, but easily missable since you'll usually be too busy fighting the rest of the battle.


    Perhaps the biggest omission in this game is the complete lack of naval combat. The game takes the pre-Vampire Coast Warhammer method of having no navies and every meeting between army transports being only autoresolvable. CAís stated reasoning for this is that most combat during the period took place on land and that the Vampire Coast option of having island battles in place of autoresolve would turn all action on the water into just more land fights, rather than being its own aspect of the game.

    The first part of this does have some merit. There were only a small number of naval actions during the period, but this is mostly because the naval war was won early on and almost never contested. Of more immediate concern to the game is that unlike most other Total War campaign maps, there is barely any water to navigate. Aside from some use of the rivers as highways, youíll rarely have reason to go out onto what small amounts of water there is. On the second point though, the current system does nothing to make river combat any different other than arbitrarily forcing you to autoresolve. The autoresolve odds will be the same whether on land or on water and there is no way to separately influence them. In practice, the water ends up being just a different kind of road with some weird rules. Since most of the campaign doesnít involve any sort of naval interaction and the autoresolve is mostly fair, the lack of naval battles is not the worst thing in the world. However, the inability for the game to represent any sort of naval advantage on a strategic level can be very irritating when you are forced to deal with enemy stacks using the rivers as a shortcut past your land defenses and have no option for fortifying the rivers as was done historically. Merely the ability to get better naval autoresolves through techs or general traits would be welcome.

    Artificial Intelligence

    The AI has always been a major limitation of the Total War experience, ranging from the completely insane to the grudgingly acceptable. Fortunately, the Three Kingdoms AI is on the latter half of the scale. The campaign AI is mostly rational, providing a competent, if predictable opponent. Itís rarely, if ever, optimal or even challenging, but while it wonít have the absolute best building setups or army compositions, itíll have workable ones that make sense. The diplomatic AI is also reasonable, with a couple issues that are mostly just balancing, like how it doesnít seem to take a vassalís overlord into consideration when dealing with a vassal. Battle AI mostly follows basic tactics and will be familiar to anyone who has played the Warhammer games, with a couple new tricks like bending its formation to prevent itself being outflanked and some better targeting for ranged units. Siege AI is the weakest link, as it will still get stuck committing huge portions of its army to a single chokepoint, though it will burn down towers before attacking if it has the capability to do so.

    The AI will bend and shift its formation to prevent itself from being outflanked. It usually doesn't keep reserves, but it will pull units out of combat to chase down enemies loose in its back line.

    Overall the AI is an incremental improvement from previous titles with some of its more egregious problems mitigated.

    User Interface

    The UI is a strange case of good and bad. On one hand, itís quick, responsive, and logically organized. On the other hand, itís trying to convey too much information in not enough space, and several important actions arenít possible from the places where the relevant information is displayed.

    The good is that the UI rarely obfuscates the game itself and is very quick about transitioning between various displays. Scrolling in and out smoothly toggles the level of detail on the campaign map without impacting performance or readability, which makes it easy to identify and move to points of interest. Info cards and tooltips quickly provide extra information without being overwhelming. A simple right click will usually take you to a more detailed display in a second, and other windows are fast and easy to navigate to from related screens. The game uses color very well, having bright and well delineated segments that communicate groups and relations as well as preventing misclicks. Thereís also a wealth of further UI options that can be toggled on or off for battle and campaign.

    The bad is that the UI doesnít seem to understand the connection between information and action. Despite having more characters than ever, thereís no central place to manage them. Viewing all your administrator slots for example, is done in an interface on the court screen where you canít view the details of the commanderies they are administrating. Often, youíll have to close the window where youíre weighing an action in in order to open another window where the information you need to make that action is. Once you finish, you have to open the first window again, and do several more clicks in order to get back to where you were. Diplomacy and character management, easily the two most important aspects of the campaign, both suffer from this. If you want to decide between two generals to lead an army, you have to open one generalís window, open the sub window with the detailed info, scroll down to find out what bonuses they confer to their commanded army, then close that generalís window and do the same sequence with the other one. It leads to some of the gameís strongest features being tiring to use.

    The character management window is easy to use and quickly expands to give more information, but the effects go far beyond just one character, and its not possible to view those without leaving this first.

    Art and Aesthetic

    The game is beautiful. Thereís no denying that. In both Romance and Records mode, the game is teeming with gorgeous landscapes, skyboxes, and all manner of details. Unit models are well detailed and expressive, and though the general models are a step down from their 2D artwork, this is mostly a result of how great the 2D artwork is. There is a certain amount of sameface between units, which isnít to say that the faces are the same, just that the differences are very hard to see without zooming in. The soundtrack is one of the best in the series, with a good mix of bombastic battle tracks, melodic ambient music, and tense transitional themes, all immersive Chinese instrumental pieces. The sound design is mostly solid, with loud gongs and drumbeats to accentuate the playerís actions, while units will chant, cheer, and yell as they move and fight.

    Of less consistent quality are voice acting and animation. There is thankfully an option for Mandarin voiceover, as the English voices sound both inauthentic and exaggerated with the exception of the advisor. A handful of well animated cutscenes and well delivered intro speeches help set the tone very well, but the one liners delivered on loading screens and shouted between characters in battle end up being too cheesy to be dramatic and not cheesy enough to be funny. Unit callouts are much livelier and the responses from characters when selected are great as well. Animation is a mixed bag, with fantastic looking marching, charges (especially cavalry), and duels, but mediocre combat and death animations. This might be helped by the eventual blood DLC, but without it, battles look very pantomime, with a large amount of air-striking and robotic motions compared to the much more energetic and reactive combat of the Warhammer games or even more recent historical titles. This animation style thankfully doesnít bother one much from a distance and has no impact on gameplay, but it does make zooming in less desirable.

    The cutscenes are beautiful and appropriately dramatic, especially the faction intro and victory ones.

    The only aspects of the game that donít come across well at all aesthetically are the projectiles and the unit icons. The former feel like they have little weight and look and sound almost like laser bolts. Their bright trails can be turned off in the options menu, but this causes the same problem that the Warhammer games have where they essentially become invisible. The unit icons meanwhile are incredibly out of place circle replacements for the large unit banners. Theyíre small, easily obscured by each other, bad at conveying information, and really break with the rest of the aesthetic by looking as generic a video game icon as you can get. Unit banners still exist (and are used to great effect aesthetically), but theyíre much smaller and donít communicate unit type or health. These two issues are small and usually get subsumed by the rest of the experience, but theyíre worth mentioning.

    Performance and Stability

    For context, I have a NVIDIA GTX 1060 with 6GB of VRAM, an Intel i5-8400 @ 2.8GHz processor, 16GB of standard RAM, and the game is installed on an SSD.

    The game mostly runs stably and smoothly, though others have reported some campaign stuttering, and multi-stack battles will get choppy if put on fast-forward. I have had only two CTDs in 100 hours that I could not replicate nor discover the cause of, and the only noticeable bugs were issues with stacking general buffs in Romance mode and family tree strangeness after annexing a vassal. Loading screens were noticeable but not intrusive, and usually had something on screen to distract, though the situation report that the advisor gives while loading a save gets repetitive.


    At the end of the day, the game is more than the sum of its parts, and overall, it works. Rational diplomacy and dynamic character management create a fantastically organic campaign experience that still hasnít gotten even close to stale for me after 100 hours. The game rarely feels like its wasting my time, and itís been very easy to lose hours and hours plotting a single operation. For those into roleplaying, the game is very good at communicating narrative without letting the fluff get in the way of the gameplay. There is solid payoff for every success and ways forward through every defeat. The not as stellar parts of the game like province administration and recruitment donít make a nuisance of themselves and drag down other aspects, and while battles can get rather predictable, there are still ways to spice them up without feeling like the game is hobbling you for wanting to have fun.

    Total War: Three Kingdoms is a definite step forward for the franchise overall. While there are a couple dead ends and some design decisions that donít feel like they were completely thought through, most of the game confidently introduces new concepts and ideas and makes them work with the Total War formula and the Three Kingdoms time period. I sincerely hope the series takes the right lessons from this titleís apparent success and carries them forward into their next games.

    If you found this article interesting, you might like to read some of the Eagle Standard's previous articles. A full list of these can be found here.
    Comments 1 Comment
    1. Welsh Dragon's Avatar
      Welsh Dragon -
      As with Part one, I found this to be a really good in depth review of Three Kingdoms various elements. Also glad you mentioned the use of little round icons vs the banners of old. This change in the post Rome 2 titles is a real let down for me, and I find it both generic and immersion breaking.

      One other issue I would add about the asthetics and UI is that the use of a stark white background for menus and loading screens is a real problem for some people, myself included, as this can make the screens painful and tiring to look at. I'm hopeful that either CA or modders will be able to address this in future, perhaps with the option to choose between themes (like the main menu theme choice in Rome 2,) but at present it's a definite negative point about the game for me.

      All the Best, Welsh Dragon.