In evaluating a city site, ask yourself three questions:
1) How quickly will it become a net contributor? This could mean it's producing more commerce than it consumes in maintenance, or it's contributing hammers to building units. Or it could mean claiming an important resource, or blocking off land. Usually this depends on the best 2 or 3 tiles. Whether it needs a border pop and how that's to be achieved, and distance costs from the capital, factor in here as well. (The lowest distance maintenance is for cities within 4 tiles, couting diagonals as 1.5 and rounding down -- equivalent to the 60% culture shape. For early non-blocking cities especially, 1 or 2 gold per turn can be a significant factor in favor of tighter packing).
2) What will it look like at size 6/8/10? Generally this is the best N tiles unless that city is running specialists. Use the lower range for early cities, the higher range for later cities.
3) What is its late-game potential? Many people focus way too much on this third question, when it's arguably the least important. It's mainly worth considering for your National Wonder cities. You do want a few really good production cities for the HE and Ironworks, a super cottage city for Oxford, etc. And cities that overlap with one of these uber-cities will have to relinquish some tiles in the late game, and that should be accounted for. But it's only part of one of the three questions.
These three questions leave a lot of room for overlap. Borrowing a food resource greatly helps (1), and can even help (2) if your size 6/8/10 cities are whipping or drafting and need to balance regrowth rates. Overlap of "generic tiles" e.g. grass cottages generally allows you to have more good cities at stage (2), even if they won't all grow into stage (3) monsters.
1) How quickly will it become a net contributor? This could mean it's producing more commerce than it consumes in maintenance, or it's contributing hammers to building units. Or it could mean claiming an important resource, or blocking off land. Usually this depends on the best 2 or 3 tiles. Whether it needs a border pop and how that's to be achieved, and distance costs from the capital, factor in here as well. (The lowest distance maintenance is for cities within 4 tiles, couting diagonals as 1.5 and rounding down -- equivalent to the 60% culture shape. For early non-blocking cities especially, 1 or 2 gold per turn can be a significant factor in favor of tighter packing).
2) What will it look like at size 6/8/10? Generally this is the best N tiles unless that city is running specialists. Use the lower range for early cities, the higher range for later cities.
3) What is its late-game potential? Many people focus way too much on this third question, when it's arguably the least important. It's mainly worth considering for your National Wonder cities. You do want a few really good production cities for the HE and Ironworks, a super cottage city for Oxford, etc. And cities that overlap with one of these uber-cities will have to relinquish some tiles in the late game, and that should be accounted for. But it's only part of one of the three questions.
These three questions leave a lot of room for overlap. Borrowing a food resource greatly helps (1), and can even help (2) if your size 6/8/10 cities are whipping or drafting and need to balance regrowth rates. Overlap of "generic tiles" e.g. grass cottages generally allows you to have more good cities at stage (2), even if they won't all grow into stage (3) monsters.