CONQUEST Fortify: How Far Armies Move vs Real RISK
One fortify move in CONQUEST crosses your whole connected territory. Printed RISK allows one adjacent step: 3.2 armies reach the front, not 25.5.

Fortify is your last move of the turn, and in CONQUEST it can cross your whole connected territory. You pick one territory, one destination, any number of armies up to all but one, and the move ends your turn. Printed RISK only allows a step to an adjacent territory.
That difference is worth a lot. Aimed at your most threatened border, a connected move brought 25.5 armies on average in our tests. An adjacent-only move brought 3.2. The rest of this page is the working and what to do about it.
CONQUEST is a RISK-style game and is not affiliated with Hasbro, the maker of RISK. The numbers below describe CONQUEST's rules and its Hard bots, not human opponents.
The rule in CONQUEST and in the box
CONQUEST's how-to-play page says fortify is optional: move armies once between two of your connected territories, then your turn ends. The engine checks three things. Both territories are yours, you move between 1 and all but one of the armies on the source, and a path of your own territories links the two. The sea routes drawn as dashed lines on the map count as links.
Hasbro's printed rules are narrower. The RISK rulebook (Parker Brothers rules text, copyright dates 1959 to 1993) says to move "as many armies as you'd like from one (and only one) of your territories into one (and only one) of your adjacent territories" and to leave at least one army behind. It also calls this the "free move." Wikipedia's summary of RISK words it more loosely, as a move to another "connected" territory, so check which version your table plays before you argue about it.
| Printed RISK (Hasbro) | CONQUEST | |
|---|---|---|
| Moves per turn | One | One |
| Source and destination | One territory to one adjacent territory | Any two territories linked by your own territories |
| Armies moved | As many as you like, leave one | Any number up to all but one |
| After the move | The turn is over | The turn is over |
| Skippable | Yes | Yes |
Half your armies are idle when you reach it
By the time you fortify, many armies are standing on territories that touch no enemy. They cannot attack anyone and they cannot defend anything. Fortify is the only way to put them to work.
We measured how many. The script scripts/research/conquest-fortify.mjs plays four-player games between Hard bots on World, Europe and USA, using the game's real engine and bot brains. At the start of every fortify phase it counts armies on interior territories, meaning territories with no enemy neighbor, minus the one that has to stay home. It covers 450 games and 31,320 fortify phases, all up to round 40. Later rounds are bot stalemates with stacks in the thousands, and they are left out.
| Rounds | Fortify phases | Idle share of all armies | Idle armies per turn | Biggest single stack |
|---|---|---|---|---|
| 1 to 5 | 8,800 | 4.4% | 0.8 | 0.7 |
| 6 to 15 | 11,723 | 35.0% | 13.2 | 10.3 |
| 16 to 25 | 5,947 | 56.6% | 53.1 | 45.7 |
| 26 to 40 | 4,850 | 68.6% | 96.5 | 87.7 |
| All | 31,320 | 51.2% | 30.2 | 26.3 |
Two things stand out. Early on, fortify hardly matters: with 4.4% idle in the first five rounds there is almost nothing to move. After round 15 it becomes the main way to shift weight, and by round 26 more than two thirds of a Hard bot's armies are interior.
The second is in the last column. The biggest single stack held 87% of the idle armies. You only get one move, so that figure decides how much the move can do. A board with one big interior stack is easy: one fortify moves it. A board with idle armies spread over six interior territories means one stack goes this turn, and the others wait.

One connected move reaches the border, one adjacent step mostly does not
Idle armies only matter if they can reach the front. To compare the two rules without trusting how the bot picks its target, the script takes your most threatened border territory at each fortify phase. It uses the bot's own measure of threat: the enemy armies next door minus your armies on the territory. Then it asks one question for both rules: what is the most armies a single move could deliver there?
| Map | Phases with a threatened border | Adjacent-only, armies delivered | Connected (CONQUEST), armies delivered | CONQUEST brings 2x or more | Adjacent-only brings 0 |
|---|---|---|---|---|---|
| World | 12,440 | 3.4 | 26.3 | 42.1% | 33.5% |
| Europe | 7,607 | 2.8 | 32.5 | 42.8% | 30.7% |
| USA | 10,922 | 3.2 | 19.6 | 37.9% | 29.0% |
| All maps | 30,969 | 3.2 | 25.5 | 40.8% | 31.2% |
The adjacent-only column is the best single neighbor of that territory, so it is the most that rule could send. In 31% of phases no neighbor had a spare army to send at all. A connected move can still reach a stack further back.
So CONQUEST plays differently from the board game even though both give you one move. Under the printed rule a stack has to already be standing next to the border it will help. In CONQUEST it can be anywhere in your connected land.
How Hard bots use the move
The bot code in shared/bots.js is a useful reference for what a decent fortify looks like. A Hard bot lists its interior territories with more than one army, biggest first, and its borders sorted by threat. It then takes the first of its four biggest interior stacks that connects to a border, starting with the most threatened border, and sends all but one army. If it has no interior stack, it balances two borders instead.
The odds that a bot does this at all depend on its level. The profiles give Easy 40%, Medium 80%, Hard and God Mode 100%, which matches the fortify column in our bot difficulty write-up. In our runs a Hard bot made a move in 61.7% of its fortify phases, which is slightly more than the 58.8% of phases where it had an idle army, since the balancing rule also fires.
Distance is mostly decided by where the threat is. Of 19,310 bot moves, 13.6% moved one step, 26.0% two steps, 23.8% three, 16.0% four and 20.5% five or more. The median move carried 7 armies and the mean was 43, because late-game stacks are large. Under the printed rule only the one-step moves would have been legal.
What to do with the move
These are recommendations that follow from the numbers above, and they are tested against bots, not against people.
- Attack first, fortify last. Fortify ends your turn, so anything you want to attack with has to happen before it. The card for a capture is drawn at the end of the turn either way, so skipping the move costs nothing in cards.
- Move one big stack, not several small ones. Only one source is allowed per turn. Reinforce your border directly, and treat interior armies as a single pile you refill from.
- Send it to the border under the most pressure. The bot's own threat measure is enemy armies next door minus yours. A border with a negative number is safe, and the biggest positive number is where the move should land.
- Keep your territory in one piece. The path has to run through your own territories. A single enemy territory between your stack and your border breaks the chain and makes the move illegal, so do not leave a big stack stranded behind one.
- Expect late-game stacks to move fast. By rounds 26 to 40, 69% of a Hard bot's armies were interior. A bot with a big stack in the back can be at the border the same turn, so a quiet border is not safe if it is connected to a rich interior.
Turn order has its own numbers in the first-player advantage test. The dice odds page has what a stack is worth once it arrives, and the card post has the army income that decides how many interior armies you have to move in the first place. For which region to build that connected base around, see which continent to take first.
How we got the numbers
The script is scripts/research/conquest-fortify.mjs in the blog repository. It imports the real engine (shared/engine.js) and the real bot brains (shared/bots.js), seats four Hard bots on each of World, Europe and USA, and plays 150 games per map with Wild Dice and auto-deal. 440 of the 450 games finished. Before each bot chooses, the script measures idle armies, the biggest interior stack and the threatened border, and then applies the bot's real choice. The bots use random choices, so a rerun moves each figure by a few points. The adjacent-only comparison is a count of what the rule would allow, not a replay of a game played under it.
To try the move yourself, start a bot game at CONQUEST and fortify from your largest back-line territory to your weakest border.
More from the blog.

Manual Draft or Auto-Deal in CONQUEST: What to Claim
Auto-deal gave 0.00 whole regions per player in 2,400 starts. A draft gave 0.72 to 1.37. Claim the cleanest region: +9 points on Europe, none on World.
Oct 27, 2026 · 6 min
CONQUEST Fog of War: How Much of the Map You See
With fog of war on, you see 96% of the map at 2 players but only 62% at 6. What stays visible, what turns into a question mark, and how to play it.
Oct 22, 2026 · 5 min
RISK-Style Browser Games: Which Need No Account?
Five RISK-style games you can play online, checked on their own sites: who asks for an account, who has bots, and how long a 2-player game runs.
Oct 18, 2026 · 5 min