Design: Movement, travel, and distance

One of the fundamental moment-to-moment gameplay decisions of Civ games is movement. The first game had no pathfinding, and you had to do everything manually. Later games used the increase in CPU power to do automatic pathfinding on your behalf, using the computer. Yet movement decisions still dominated Civ gameplay.

The basic model

Depending on version, the game board is made of hexes or squares. The common name for them is “plots”.

Tier 1 and Tier 2 units move “one plot per turn”.

Tier 3 units move two plots per turn, and Tier 4 units move 3 plots.

Water units move at least 3 plots, and at most 15 or more plots. Air units move up to 8 or 16 plots, but have to finish their turn in a City or Aircraft Carrier.


Most terrains cost 1.5x as many plots as normal to traverse. If you have insufficient remaining moves, but at least 1, you are allowed to move as if you had enough.

For Tier 1 units this has zero effect. For Tier 2 units, it means they often only move as T1 units.

Roads cost 0.33x as many plots as normal to traverse, and often nullify the 1.5x penalty of terrains.

The basic strategic issues

Most units have a vision/sight distance of 1-2 plots. On many terrains, that’s reduced to 1 plot.

In the first Civ, every battle resulted in the death of one of the units, either the attacker or the attacked. Battle was always a high-stakes issue. This hurt player’s fun in many ways, made the game hard – but the game was balanced around it. Features in later sequels (e.g. promotions) made it more fun but ruined balance, and only with Civ5 was the balance fully restored for the new combat design.

Pre-civ5, it was usually fatal to finish a turn with any unit adjacent to an enemy, unless your army contained a high-defence unit. The units you sent out into wilderness to explore/attack/wage war were always high-attack units (otherwise they’d be useless). Civ5 prevents stacking units, so it was essential that they made combat less “all or nothing”.

The problem with movement ending in a bad place was a triple whammy:

  1. Most attacking units are extremely weak and vulnerable on defence. If you can avoid ever defending, they are much more powerful. With careful choice of movement each turn, you could sustain an attacking-to-defending ratio of 90% or more.
  2. Allowing your enemy to decide “whether” to attack places all the strategic power in their hands; units cannot disengage
  3. The terrains you can easily move across (no penalty) have no defence boost (or negative boosts). Being forced to wage a battle on those plots was often suicidal

Other gameplay effects of movement

90% or more of any run-through of a civ-style game is spent exploring the map. You generally don’t know what you’ll meet, and what will be adjacent to you when you end your turn.

Roads – and enemy borders (cultural borders, and unit “zones of control” in some versions) – have massive impact on where you can explore, how quickly, in what directions, and how you much you find.

Roads and valleys (areas of non-slowing terrain), and their opposite: the impassable Mountain and Sea/Lake plots, encourage movement in particular directions.

My changes that affect movement

I’ve designed a unique set of combat rules that are more balanced, and more human-predictable, than the modern Civ games. This decouples movement from combat – I’m more free to do what I want with moves.

This is subtle, but: Civ5 scaled-up the size of plots by a factor of approximately 2-4, and then further scaled them up by forcing a rule of only one unit per tile. De facto, a Civ5 plot is – in game-strategy terms – approximately 2-5 times the size of a Civ/Civ2/Civ3/Civ4 plot. But even in the old games, I was unhappy that the plots were too broad. So I designed my game around approximately 3 plots for each Civ1 plot – approximately 3 plots across, and a total area of 7 plots (hexes) in a circle.

On the implementation side I considered Civ’s system of “moves” to be clunky and a poor technical design. Instead, I created a pool of energy per unit, and different actions drain different amounts of energy. Energy re-fills each turn. This gives me a lot of freedom to do small tweaks to game-balance while I’m playtesting the game – much finer-grained than in traditional Civ games.

How is this turning out, in practice?

Since I had fine-grained energy, I tried playtesting with different designs…

Design 1: You only get what you get

Each attempted action deducts from your energy. If there’s not enough energy to fulfil it, you cannot make that move.

Design 2: Debt, Fair movement

Each action has a cost, and so long as you have more than zero energy, you can perform any single action. If you try to perform 3 actions in a row (say: moving to a plot 3 plots away), the game attempts each action in turn, deducting from energy, until it ends up with 0 or less energy.

Each turn, each unit regenerates it’s maximum amount of energy. If it started negative it will not recharge to its full amount. e.g. 50 energy, spends 20, then 20, then 20 (3 moves), leaving -10 energy. Next turn it regens 50, giving a total (at start of second turn) of only 40 energy.

This is true and fair: movement averaged over turns is exactly correct.

Design 3: Debt, but limited

As for Design 2, except you can only hit a maximum negative of -10 energy.

Design 4: Civ1-Civ5 (mostly)

If you have any energy left, you can complete your move.

Actually, in the pre-Civ5 games this isn’t entirely true – there are cases where very low remaining moves prevent moving to high-cost plots, or where having “less than 1 full move left” prevent an attack. Moving on roads used to cost 0.33 moves, so that moving 2 plots along a road could scupper your chances of attacking, or entering a very expensive move-cost plot.

ALL Designs: elevation impacts movement

In my game, every plot has an elevation. Moving uphill costs slightly more energy, and moving downhill costs slightly less. Combat bonuses are proportional to the actual steepness of the terrain – if my archer stands atop a cliff, it gets a much higher defence bonus than if it’s on a low hill.

In practice, this adds or subtracts between 2% and 10% to the cost of each move.

Playtesting results

I started with Design 1. It made playtesting and debugging easy, but … for units with fewer than 5 or 6 plots of movement it felt a bit clunky. It also felt dull, strategically: most units end up moving the same distance in practice. The player often expected to move a bit further than they actually could. No surprises from AI moving further than planned, but movement felt bogged-down.

On to Design 2: the most accurate and “fair” one. Some Civilization games/mods had a similar system, but it created weird situations, whereby a unit would have e.g. 0.1 moves left, but would act as if it had 1.5 moves left (e.g. moving onto a hill, at 1.5x cost). This would appear to be a “bonus” move, often adding up to 50% to the unit’s quoted movement!

Not a problem for the player – until they counted the possible moves for an AI / enemy player, carefully manouevred to barely avoid it … and then on the enemy player’s turn they got crushed, as if by cheating. Worse, once you get into “debt”, it’s hard to get out of it, and every turn begins with you having fewer than your unit’s quoted allowance. This is potentially more interesting and realistic – but in practice I found it annoying and disappointing. The problem is simply it’s very hard to judge how far you’ll actually travel.

I tried playing for a while with routinely resting before embarking on long journeys, and it didn’t help. I played around with two ideas of Design 3: one that limited your potential debt, and the other that allowed a “rest” to give you a temporary boost to max-energy on the next turn. This second variant I like a lot more, but I haven’t playtested; I suspect it’s going to be the best.

Finally, I switched to Design 4, which most closely follows vanilla Civ. No negative/debt, which makes it unfair in edge-cases, but it’s predictable enough to leave human players not feeling cheated. In particular, the uphill/downhill issues in my game make the fairness of Designs 1 and 4 more obvious: going uphill nearly always costs you a move in Design 1, and going downhill nearly always gains you a move in Design 4.

So, at the moment, I’m sticking with 4, but I want to go back and try 3 with the “rest to get temporary bonus” version. I also like the precedent: it allows the player to strategically boost their units in a small way a few turns before taking a big action – e.g. starting a new wave of attackers. I might have to make it that you have to remain still for two full turns in order to get the movement boost, to ensure it doesn’t become overused.

What to do with Roads and Railways

In the late game, road-based units become insanely far-travelling. Railways in some versions of Civ give you unlimited free movement. This is desperately necessary to overcome the boring micromanagement of large empires, and the irritating barbarians who kite around the map evading your units, forcing you to divert entire armies to capture them quickly, a distraction from the serious World Wars you’re engaged in with enemy civs.

Also: they have to justify the upgrade costs.

It disappoints me how much this obsoletes mid-game units that don’t need to be obsoleted. There are plenty of end-of-the-tech-tree units that would be worth keeping around by the player as strategic alternatives. Civ5’s design says “I can’t be bothered trying to balance all this” and forces all units into a very narrow, boring, limited “everyone gets ALL the techs” valley. C.f. my very brief, half-written post comparing the techtrees in differnet Civ versions, and their effects. Also, more usefully, look at the complete tech-trees for all versions. At some point, I’m going to re-write / add to that half-written post, and look at more detail on Civ1, Civ3, Civ4, and Civ5.

Meanwhile, I’ve decided to add a new unit-type (alongside Melee, Ranged, Recon, etc): Wheeled/Engined. Wheeled/Engined units do not increase their movement in the standard Civilization way: they have the same total movement as medieval-era units, but they get a huge boost when travelling on roads. They are blocked from traversing many terrain types unless there is a road/railway there.

A variant – Offroad – gets a small boost when travelling on roads, but is able to travel on almost all terrain types.

All previous-era units are classed as Pedestrian.

(yes, this isn’t perfect – e.g. I need to rethink the names to cover Horses, and early wheeled-but-not-engined units!)

I’m also allowing you to use any road, whether its in your territory or an enemy’s. It’s absurd that Civilization blocked you from using them (but was required for gameplay reasons – otherwise it was too easy to declare war and conquer an enemy civ all in the same turn, without them getting to even move!)

Impact of Wheeled, Offroad, and Pedestrian

Now we have the situation where Medieval Era (and even earlier!) units still have relevanve in their movement. Suddenly, terrain is restored to having relevance in late-game expansion, and in wars. Your Recon units in late-game are the ones that can charge across enemy lands – but your powerful city-attackers and fighting units need to have land cleared in front of them, or else be funnelled along predictable lines of attack, allowing defending civilization to much more effectively plan their defence.

In turn, this allows me to reduce the nerfing of attacks and conquest in Civ5 (they were too powerful in earlier games, all the way back to Civilization 1’s overpowered Chariots), and restore some of the excitement and danger of wars.


My plan is that as you hit a modern Era, you get automatic roads on every plot in your cultural borders. The roads you manually built get upgraded to Highways, and enemy units can use Highways but not the automatically-generated roads (these are “rural” roads and don’t give them bonuses – but do work for you).

Why build Highways at all then? They make you vulnerable!

I plan to make them a tempting gamble for the owning civ: they give a unique boost to your economy, that makes them highly desirable for non-warfaring civs, and possibly a small extra boost to Wheeled vehicles that makes them slightly desirable to warfaring civs. The player gets to choose whether to keep them for the bonuses but making themselves more vulnerable to attack from weaker civs coveting their land, or blow them up and harm their economy growth.

…but all this still has yet to be implemented and playtested.


  1. The link to “very brief, half-written post comparing the techtrees in differnet Civ versions, and their effects” is giving 404. I also don’t see the post listed in blog archive.

    What do you mean by Civ1 having no pathfinding? AI had to move somehow. Also, there was GoTo command available to the player.

    The energy system in design 2 sound very similar to classic roguelikes and manifests the same problem of unexpected double-move by opponent. It will probably go away if you remove these sub-10% movement modifiers.
    The energy system can also can be used for multi-turn worker commands.

    Reading this post gave me a thought about treating railways similarly to sea transport. A unit wastes a turn to embark or disembark a train. Once on a train, it moves using train’s movement rules. The train unit should not be built like a ship, but rather automatically appear when needed.

    • I may have unpublished the techtrees post, because it was so incomplete, sorry.

      Civ1 had no interactive pathfinding for the player – goto was horrendously underpowered by comparison. But it was designed as keyboard-only game (mouse was an optional extra!) so this is no surprise 😉

    • For opponent moves, I want to add explicit previews of opponent possible moves – something MASSIVELY important that no Civ version has had :(. This should remove any issues with unexpected opponent moves

    • Railways using ad-hoc vehicles is an interesting idea, I like it.

      Also allows for e.g. some ultra-special units could use the railways directly without sacrificing their combat strength, whereas normal units would become highly vulnerable (like most sea transport units in Civ5).

Comments are closed.