The Terraforming Mars Midgame: Engines, Milestones, and Tempo

By the middle of a Terraforming Mars game, everyone has an engine of some kind. Income is up, blue cards are working, and the tracks are climbing. The opening asked what you could build. The midgame asks a harder question: how long do you have left to use it?

Terraforming is a shared clock

There are 42 steps of terraforming in the game: 19 for temperature, 14 for oxygen, and 9 oceans. The game ends after the generation in which the last one is taken. Every player’s terraforming moves the same clock.

That gives every player a reason to speed it up or slow it down:

• Players with big engines want a long game. Every extra generation pays them again.
• Players ahead on points want a short one. The fewer generations left, the less time others have to catch up.

Watch how many steps the table takes each generation. If 42 steps are being eaten at six or seven a generation, you can estimate the end several generations ahead. That estimate drives every decision below.

When to stop growing the engine

A production step pays once per generation left. With eight generations to go, 1 M€ production is worth about 8 M€. With two generations left, it is worth about 2.

So at some point, production cards stop paying for themselves. Compare a production card’s cost with its value over the remaining generations. When the value drops below the cost, stop buying production and start buying points: TR, tiles, and cards with victory points.

It often pays to make this switch a generation earlier than feels natural.

Milestones are a race for three

A milestone costs 8 M€ and is worth 5 VP. That is one of the best rates in the game. But only 3 of the 5 can be claimed, and each by one player.

The five on the base board:

• Terraformer: TR of at least 35.
• Mayor: at least 3 cities.
• Gardener: at least 3 greeneries.
• Builder: at least 8 building tags in play.
• Planner: at least 16 cards in hand.

Track who is close to each one. If you qualify and someone else is one step away, claim it now. Two actions in one turn can let you qualify and claim before anyone reacts. The rulebook’s own example does it: a player builds a third city, then claims Mayor with the second action.

Fund only what you can hold

Awards work differently. Funding costs 8 M€, then 14, then 20, and only three are funded. At the end, first place in each funded award gets 5 VP and second place gets 2, no matter who paid. Ties are friendly: tied players all score.

That means funding an award you then lose is a gift to someone else. Before you fund, ask: can anyone catch me here?

• Landlord (most tiles) and Banker (highest M€ production) build slowly, so leads tend to hold.
• Scientist (most science tags) can swing with a single card.
• Thermalist (most heat) and Miner (most steel and titanium) count the cubes you hold at the end. Anyone can spend theirs down, or save up, in the last generation.

The first award is cheap. The third, at 20 M€, is rarely worth it unless it decides first place.

Reading the table

Before your turn, look past your own cards:

1. Which parameter is closest to done? Cards needing a low value are about to become unplayable. Cards needing a high value are about to open.
2. Who is close to a milestone? You may need to claim first, or accept that one is gone.
3. Who leads each award? Spot the award that someone is about to fund.
4. Who is the leader? Red-bordered cards can take resources or production from any player. Point them at whoever is winning.

The map fight

Cities and greeneries score each other, whoever owns them. A city scores 1 point for every adjacent greenery, so:

• Your greenery next to a rival’s city gives them a point. Sometimes a placement bonus or a spot next to your own city is worth it anyway. Know the cost.
• Your city next to rival greeneries collects points from their work.
• Cities cannot touch. A city placed in the right spot can also block a rival’s best city site.

Managing your hand

A midgame hand fills up fast. Every card you cannot play before the end is wasted money. Sell the ones that will never fit, using the Sell patents standard project. And when you are waiting to see what others do, a single small action keeps you in the generation without committing.

For the bigger picture of how engines rise and when they stop paying, see our engine building explainer, where Terraforming Mars is the long-form example.