How to Win at Catan: Strategy Tips From 2,000 Bot Games
To win at Catan you need 10 victory points, and the fastest route is cities, a spread of goods and one of the two road or army awards. Claimridge is a separate, independent game of the same genre, so we measured it: 2,000 four-seat bot matches through its real engine, counting what the winners did. These are medium bots, not people. Treat every number below as a pattern, not proof of cause.
The target: 10 points, and where they come from
The classic target is 10 points, and our default is the same ("Target 10 VP"). A village is worth 1, a city 2, Longest Road and Largest Army 2 each, and a Victory Card 1. In the bot matches, the winner's points split like this:
| Source of the winner's points | Share |
|---|---|
| Cities | 56.6% |
| Villages | 16.4% |
| Victory Cards | 10.7% |
| Longest Road | 9.4% |
| Largest Army | 6.9% |
More than half of every winning score was cities. That sets the plan for the rest of this guide.
Ten tips, each with a number behind it
1. Spread your goods in the opening
In the draft, each seat places two villages. The second one pays one good per producing hex it touches, and that is your opening hand. We counted how many different goods the two villages touched. Seats touching 2 goods won 12.6% of the time, 3 goods 19.3%, 4 goods 25.9% and 5 goods 34.6%. An even split at four seats is 25%, so four goods is par and five is a real edge. If you are curious how to choose the spots, read our opening placement guide.
2. Aim at ore and grain
A city costs 2 grain and 3 ore. Seats whose opening touched both ore and grain won 31.7% of the time, against 17.1% for seats that did not. Timber and brick together (the road and village goods) won 26.3% against 23.5%, which is a small gap. Early roads are cheap. Cities are the bottleneck.
3. Build cities, not extra villages
You have 5 village pieces and 4 city pieces. Winners took 5.69 of their points from cities on average, 1.64 from villages and 1.08 from Victory Cards. The median seat waited 40 dice rolls at the table for its first city, and only 83% of seats ever built one. Upgrade your best producers early instead of filling the map with 1-point villages.
4. Hexes marked 6 and 8 matter, total pips do not
Seats whose opening touched no 6 or 8 hex won 13.2% of the time, one such hex 23.0%, two or more 27.6%. Chasing the highest total pip count is a different story: bots almost always land between 18 and 23, and seats at 18 to 20 won 24.0% while 21 to 23 won 25.9%, a difference that is not significant. Take the strong numbers, then fix your goods, and stop adding up dots.
5. Do not pay for a trading post in the opening
Seats with an opening village on a trading post won 22.1% of the time, seats without won 25.5%. Posts give 3:1 or 2:1 against the bank's 4:1, which only pays off when you have a surplus of one good. In the opening you do not. For when trading is worth it, see the trading guide.
6. Seat order is not your excuse
Seat order is a random shuffle, and the draft runs in snake order. Win rates at four seats were 23.5% for the first seat, 25.4% for the second, 24.7% for the third and 26.4% for the fourth. None of that is significant. The first seat rolls first, the last seat places twice in a row, and it evens out.
7. Keep your hand at 7 or fewer
When a 7 is rolled nobody collects, and anyone holding more than the hand limit (7 by default, so 8 or more) discards half, rounded down. A four-seat match averaged 18 sevens, and the bots discarded 103.6 cards per match across the whole table. The dice are not the cause: 216,589 recorded rolls matched theory (chi-square p = 0.71). Spend before you reach 8 and you keep the goods. See the dice odds for how often a 7 really comes.
8. Longest Road is a closer
Longest Road needs a run of at least 5 roads, and ships count. A rival village or city on a corner breaks the run through it. Seats holding it at the end won 63.2% of the time, against 17.7% for seats without it. It is worth +2, so it often decides a close finish. The rules are in the Longest Road rules guide.
9. Largest Army, and counting the deck
Largest Army needs 3 played Gunslingers, and a rival must have strictly more to take it. Seats holding it won 47.6% of the time. Seats holding both awards won 98.1%, but that happened only 54 times in 2,000 matches. Seats holding neither won 9.5%. Buying cards alone was no recipe: winners bought 4.7 cards on average and everyone else 4.8. That is a selection effect, so do not read it as advice to buy none.
Hidden Victory Cards matter more. The deck at 2 to 4 seats has 25 cards: 14 Gunslinger, 5 Victory, 2 Road Master, 2 Supply Wagon and 2 Boom. A Victory Card counts the instant it is drawn, so the visible score can understate a rival. Track how many cards a rival has bought and add a possible hidden point before you let someone sit on 9.
10. The Bandit finds the leader
On a 7 the roller moves the Bandit to a hex and steals a card. Per match, the Bandit landed on a hex touching an eventual winner's building 13.04 times, against 10.37 for other seats. Part of that is simple: a winner has more buildings, so more hexes touch them. Still, expect to be targeted when you lead, and keep your hand small (tip 7).
The habits in one view
| Habit | Won with | Won without |
|---|---|---|
| Ore and grain in the opening | 31.7% | 17.1% |
| Timber and brick in the opening | 26.3% | 23.5% |
| Opening village on a trading post | 22.1% | 25.5% |
| Holding Longest Road at the end | 63.2% | 17.7% |
| 5 goods in the opening (vs 2) | 34.6% | 12.6% |
What bots cannot teach you
Bots do not bluff, do not table-talk and do not refuse a trade out of spite. They also hoard cards and sometimes stall: 13.4% of our four-seat bot matches ended on the dead-board rule, which closes a stuck match. Real trading is negotiation, and the best habit with people is to watch who is ahead and stop feeding them goods. Use these numbers for the opening and the build order, and your table for the rest.
Try it yourself
Open the 2 to 6 player game on the Claimridge page. It is free in the browser with no signup, and a quick room seats you plus two bots. Pick an opening that touches four or five goods, aim at ore and grain, and upgrade to cities as soon as you can.
CATAN® is a registered trademark of CATAN GmbH. Playgridia and Claimridge are not affiliated with, sponsored or endorsed by CATAN GmbH, its licensees or publishers. The name is used only to identify the game readers are searching for and to describe the genre; Claimridge is an independent game with its own rules, names and artwork.
Frequently Asked Questions
How many points do you need to win Catan?
The classic game is won at 10 victory points. In Claimridge the default target is also 10, and in larger 5 to 6 seat rooms the host can pick 8, 10, 12 or 14. Hidden Victory Cards count at once, so a win can arrive the moment one is drawn.
What is the best opening strategy?
Place your two villages so they touch four or five different goods, ideally including ore and grain, and at least one 6 or 8 hex. In our bot matches, seats touching 5 goods won 34.6% of the time, against 12.6% for 2 goods. The second village pays out the opening hand, so plan that one carefully.
Should I build cities or more villages?
In our bot matches, 56.6% of a winner's points came from cities and 16.4% from villages. A city is worth 2 points and costs 2 grain and 3 ore. Upgrade your best producing villages before you spend on new ones.
Does playing first give an advantage?
In our four-seat bot matches the win rates were 23.5%, 25.4%, 24.7% and 26.4% by seat, and the differences were not significant. Seat order is a random shuffle, and the draft runs in snake order, so the last seat places twice in a row.
What happens when a 7 is rolled?
Nobody collects goods that roll. Anyone holding more than 7 cards discards half, rounded down, and the roller moves the Bandit and steals one random card from a player next to it. Spending your hand down to 7 or fewer is the simplest defence.
Do these numbers apply to games against people?
Not directly. They come from 2,000 matches between medium bots, so they show patterns, not proof of cause. People trade and bluff differently, so treat them as a starting point for your opening.