I realize I've been winning Settlers of Catan regularly. The last time we played, I felt like people were getting frustrated that I am the only person who wins, especially when I mentioned that I once read a strategy guide online when we first started playing. So, I've decided to post the strategy that I've found wins most, in the hope that people will want to continue to play SOC when we meet. The strategy described here is general; there will always be exceptions, and the layout of the board will always be more of a determinant than any strategy you study. Take all of my points with a grain of salt.

Initial placement of settlements I've found the order of initial placement of settlements can give one player an advantage. Placing last is slightly better and placing first is a slight disadvantage. If you have to place first, then you'll be guaranteed the best spot on the board, but this is of little consequence as there is typically more than one spot with good production. It also makes it difficult to balance the placement of your second settlement in a way that rounds out the diversity of resources you'll need in the beginning. If you place last, its easy to get a good positional advantage and you can usually find at least one decent spot.

It is helpful, but not vital, to control a 6 or 8 tile (or both). However, if you have settlements on the 6 or 8 tiles, there's more of a chance someone will want to place the robber on that tile, creating more problems for you. Consider all surrounding tiles when placing a settlement. A 6-4-3 placement is frequently inferior to a 5-9-10 placement, even though the 6 tile will statistically be rolled most often.

When possible, don't cluster your settlements around a tile. A robber can easily wipe out all of their production. A clustered formation around a 6 or 8 tile is always a target. Lastly, placing settlements on differently numbered tiles is better than placing on duplicate numbered tiles; it increases the chances of your numbers being rolled.

Expand early, and as often as possible. Forget about Longest Road, Largest Army or development cards early in the game. You may think that each time a settlement is built, the rate of resource gathering increases by a linear amount. In fact, the rate at which you gather resource cards increases exponentially for each settlement. This is why it is so difficult to get ahead in the game if you fall behind early. A player who builds settlements early in the game will almost always win against a player who builds later.

The player who places an initial settlement on a port is immediately at a disadvantage because he has less chance of collecting resources that are critical for early expansion. The only time this might make sense is when you can grab at least two high-production tiles of the same type of resource, the corresponding 2:1 port, and the port settlement also adjoins another tile (don't initially settle on a single tile port). Build your port settlement on a location where you get the benefit of two tiles/resources. Another argument against placing an initial settlement on a port is the fact that its so easy to place on a tile just one or two links away (thereby getting the early benefit of all three adjoining tiles) and then build a settlement out to the port quickly.

If you have a choice to build either a city or a settlement, choose the settlement. You'll expand your empire and control more tiles. An empire made of two cities is usually inferior to that made of four settlements, assuming the settlements are well placed.

Wheat is the most important resource. You'll win more games with wheat than any other resource, especially if paired with ore. Wheat is a resource that everyone needs to build a settlement, city, or to buy a development card. You need at least one good tile (I always shoot for two) of wheat to have a chance at the game. If possible, I always make sure I get at least one ore tile as well, as this will give me a greater chance to get cities early.

Try to place your settlements on tiles that will provide you with a good mix of resources. Diversity is always better than having too much of a single resource. For instance, a spot that provides sheep-sheep-brick will be outplayed more often than an ore-wheat-wood, even if the former has an advantage in numerical statistics. (The exception might be a wheat-wheat-ore or ore-ore-wheat spot, as long as you have other spots that complement this). Also, if you place your settlements last in the initial setup, you'll have the opportunity to build a great hand for yourself. If you've decided to settle on a wood-brick spot, its a good plan to place your last settlement (rather than your first) on the wood-brick so that you'll be able to build a road on your first turn.

Trade only when you're certain it benefits you. Hesitate to perform a trade if your turn isn't coming up for a while. The card you traded for has more chance of being stolen; if this happens, you've enhanced your rival's hand to no benefit of your own. Best is to trade on your turn or with the person just before you. Think of your rivals as 1:1 ports. If you can get a resource you need by trading one of your own, usually this is better than giving up 3 or 4 resources to the bank for that same card. If someone offers to trade with you on their turn, this usually means they are about to get points. Don't trade with them unless you're fairly certain to get some sort of development out of it on your next turn.

Never end your turn with more than four cards in your hand. Best is to end your turn with three or less to lessen the chance of getting caught by the robber with more than 7 in your hand. Spend your cards, especially after you have a few settlements working for you, because there's a good chance someone will roll your numbers anyway. If you have cities or many settlements around a single tile, its easy to see how you can get caught and have to give up half your hand.

Try to have a soldier card ready at all times. This is to get rid of the robber if it lands on one of your best tiles. Don't be afraid to use it. Also, few people seem to try to get Largest Army, so if you have this card, you're already 1/3 of the way to getting the 2 points.

Try to keep track of the resources people are holding. If you roll a 7 and know that a certain player, on the previous turn, just gathered a resource you need, it makes it easy to choose who you will rob.

Brick and wood are poor resources to have late in the game. Most of the roads and settlements have already been built, and thus there is little room to grow. Since it gets more difficult to expand as the board gets crowded, avoid wood and brick late in the game. Sheep is also a poor resource; there are four tiles for it in every game and usually its easy to get someone to trade it to you any time you need it. Ore and wheat are normally always in demand, since players are always wanting to build cities. It will also help you build your own cities. If you can build in such a way as to get good production values for ore or wheat while limiting other players' access, you can either prevent your rivals' growth or you can get them to trade you two or even three resources for one of yours.

You can't count on winning with development cards. You need at least seven points in buildings before you can think about development cards pushing you to ten points, and even that is a stretch. I would only buy them to get points when I have nine points already and the board is too crowded to build more settlements. Largest army is handy to get, however, and if you have the cards in hand, can tilt the game in your favor in a pinch. Over half of the cards in the development card deck are soldier cards, so Largest Army is much more likely to win you points than the University or Market cards. The Monopoly card is rarely useful, as you have to spend three resources to buy the card, and you probably won't get more than that amount in return when you use it. Lastly, don't buy development cards early in the game. Use the resources instead to expand.