We all have different strengths and weaknesses when it comes to world-building. It’s generally best to focus on and develop our areas of strength. For instance, a linguist (say J.R.R. Tolkien) will probably have a world with very well developed languages, but be weaker on the minute aspects of culture and religious practice (for instance, outside of the Silmirillion, it’s hard to figure out what or who the peoples of Middle Earth worship; much less what their rituals of worship look like). On the other hand an anthropologist (say Steven Erikson) will have intricately detailed cultures, but will be more likely to be lax on languages (In 6000+ pages of reading I think the number of indigenous words, outside of names, in Erikson’s books amounts to maybe a sentence and a half).
I point this out to say that geography is one of my weaker areas. I love maps, but often I don’t put a lot of coherent thought into what a world SHOULD look like over what I WANT it to look like. As such I wind up with geographical features that don’t make any sense, like a world with no islands…anywhere…at all…or a world with lakes and oceans made in perfect circles…why…because that’s how I drew it. I’m getting better about this, but mostly because I’ve had a lot of help. I show my maps to people, I particularly love pilots because they see everything from a bird’s eye view and are very familiar with maps, and ask them to point out features that don’t make sense. I actually have Brian to thank for pointing out that one of my mountain ranges (in an earlier version of Avnul) formed a perfect right angle. It was convenient for me at the time…it has since been corrected.
Ultimately what our worlds look like will have an impact on their believability. If, in the maps for my book, the world looks like something that I drew in my spare time then the book itself suffers because it immediately becomes less believable. Don’t just draw a map and then walk away. Keep working on what your world looks like, go back and make sure that the features still work, figure out distances (I had to do this for Among The Neshelim…oh…right…the launch for the book is going to be called Among the Neshelim), and decide how this geography has affected the cultures that live in it. For instance, Japanese culture developed largely in isolation. If the Japanese had not had long periods of isolation then their culture and history would be very different. If you want to have culture in your world similar to the Japanese then you need to take this into account.
Do your people live on a large plain? If so, do they have access to wood? What do they make their fires out of? How do your people get water? Many ancient cities were built on or near rivers so you might want to consider this when placing your cities. Are there any volcanoes in your mountain ranges? If not, why not? Volcanoes aren’t exactly rare in the world. If there are volcanoes, then how does the local populace react to them?
Some basic geographical faux pas from my experience are:
Mountain ranges that don’t follow fault lines
Lakes and inland seas that have a perfectly circular coastline
Continents that have a straight coastline lacking in bays, inlets, fjords, etc
Large areas with no forests without explanation
A complete lack of volcanoes in the world
large, non-desert areas with no obvious water supply (rivers and lakes)
A complete lack of islands
These are a few in my experience to watch out for. However, remember that anything can be explained by the history of your world. For instance, if your world has no islands then this is a problem. However, if your world has no islands because seven hundred years ago an insane wizard cast a spell that made every island in the world sink into the ocean, then it is no longer a problem. You can explain away just about anything, just make sure that your explanation fits within the scope of your world.