In the end it depends on what you want to predict from your assumptions that tell you whether it’s okay to assume a ball is round or the roughness matters. Wou

In the end it depends on what you want to predict from your assumptions that tell you whether it’s okay to assume a ball is round or the roughness matters. Would you call a golf ball round? Depends. If you want to calculate how a golf ball travels upon being hit by a golf club, then you’re going to get more predictive results if you consider the dimples in the golf ball. If you only want to calculate the number of balls you can fit in a bag, by all means ignore the roughness. Just as in the flat earth debate: it’s perfectly reasonable to assume flat earth when planning your bike trip, but the assumption breaks down when you want to fly large distances. For me it is all about knowing which models and abstractions are useful in which contexts rather than philosophising what is “real”.