Get into the habit of counting other players' victory points every turn, and figuring out how they will probably act to get what they need to win. This may seem obvious, but most people do not do it.
2) Don't be a target
Hide points to near the end if possible (do you really want the longest road with five segments, and being ahead in points that early on?). Or wait to put out that last knight if you can.
1) They're essential.
You will probably need a port to win the game, or a lotta luck. Some strategies are better served with a 3:1 port, while others by a specific resource port. Make sure you know how you will get them. Don't stress too much over ports early on, they are usually more important in the mid- and endgame (unless a resource is particularly rare). But don't get blocked from reaching one, either.
2) Don't compare apples & oranges
Note that many players think that 3:1 ports are inferior to 2:1 ports. This isn't true. Different ports are good for different situations. Think about which kind you need. People who advocate one kind over another often just play with one particular strategy. For example, 2:1 ore and grain ports seem to be particularly popular, as players using the ore-grain or card-builder strategies really need them later on in the game.
You may not want a 2:1 port if you just have one source for that resource, even if it is a good source. It will draw the robber. Conversely, watch for other people in which this happens when considering where to place the robber.
3) Count the cost
A popular strategy is to go for a port on the first turn, which means you are on an intersection with only two hexes that produce, at most. It had better be worth it, because it is going to have to offset the increased production of someone who placed two inland settlements, with a higher total production value. Sometimes ports are just screaming to have a settlement placed on them at the initial setup (for example, an intersection with an ore port on an eight ore hex). Expect the robber to show up at hexes like these.
4) Rare is good
If a particular resource looks like it's going to be extremely rare, it may be worthwhile to start off with a port, especially if you're the last player and can coordinate a good combination of spots. (Isaac Kuo
5) Easy is good
Remember, as ports are on water hexes, it is a lot easier to block them off or reserve them - just build a road to the port area, and sometimes all you need is one more road on the other side to block that particular port off.