Paris is its own bubble, with prices that are less tied to changing things like new bus routes or restaurants, and more to the perception of what other people pay for it. The historical aggregates usually serve as a bullshit test (if you sell at 25% above historical prices and don't have a very good reason to do so).
It is a given that the price will be very different depending on the state of the property, the quality of the building and its era (1900-1930 flats command significantly higher prices than 1970-1980), whether it's a narrow alley or a large boulevard, whether it has windows on only one side or both sides, whether there's a window in the kitchen and bathroom, and whether part of the building belongs to the public housing agency (among many others). After a while, you start to get a pretty good intuition of how high a property will sell above or below the average.
Another thing about subjective experience is some properties fit certain lifestyles and will be priced accordingly. A flat in the Bastille area (lots of bars, restaurants and night clubs) appeals to people who would pay for the privilege a bonus that a boring father of two (like me) wouldn't consider. Conversely, having an address in the Quartier Latin makes it easier to enter the top high schools, so tiny one-bedroom flats sell surprisingly high (and are then rented out to university students for a far lower price than the purchase price would suggest).