So how often should I change my oil?

You can never change your engine oil too frequently. The more you do it, the longer the engine will last. The whole debate about exactly when you change your oil is somewhat of a grey area. Manufacturers tell you every 10,000 miles or so. Your mate with a classic car tells you every 3,000 miles. Ole' Bob with the bad breath who drives a truck tells you he'* never once changed the oil in his car. Fact is, large quantities of water are produced by the normal combustion process and, depending on engine wear, some of it gets into the crank case. If you have a good crank case breathing system it gets removed from there PDQ, but even so, in cold weather a lot of condensation will take place. This is bad enough in itself, since water is not noted for its lubrication qualities in an engine, but even worse, that water dissolves any nitrates formed during the combustion process. If my memory of chemistry serves me right, that leaves you with a mixture of Nitric (HNO3) and Nitrous (HNO2) acid circulating round your engine! So not only do you suffer a high rate of wear at start-up and when the engine is cold, you suffer a high rate of subsequent corrosion during normal running or even when stationary.
The point I'm trying to make is that the optimum time for changing oil ought to be related to a number of factors, of which distance travelled is probably one of the least important in most cases. Here is my selection in rough order of importance:

Number of cold starts (more condensation in a cold engine)
Ambient temperature (how long before warm enough to stop serious condensation)
Effectiveness of crank case scavenging (more of that anon)
State of wear of the engine (piston blow-by multiplies the problem)
Accuracy of carburation during warm-up period (extra gook produced)
Distance travelled (well, lets get that one out of the way)
If you were clever (or anal) enough, you could probably come up with a really clever formula incorporating all those factors. However, I would give 1, 2, and 3 equal top weighting. Items 1 to 3 have to be taken together since a given number of "cold" starts in the Dakar in summer is not the same as an equal number conducted in Fargo in January. The effect in either case will be modified by how much gas gets past the pistons. What we are really after is the severity and duration of the initial condensation period. All other things being equal, that will give you how much condensate will be produced and I would suggest that more than anything else determines when the oil should be dumped.