Calculating Demand

[AirlineTycoon]

About 10 years ago I started writing a airline simulation game for fun. However marriage, babies and a few other personal things got in the way.

Over the past couple of months I've been tapping away and some aspects of this again.

One of the original issues I had and still face is calculating demand between two ports.

In real life there are so many factors that writing logic for all of these will be complex and although I'd like to continuously improve the logic it's impracticable to include everything at this stage.

So my question is how can I start to narrow it down.

I've played with:

Population and Origin - Destination. Have a distance calculation so the further apart they are the lower the demand.

For example, LAX-NYC vs TYO-NYC. Similar sizes in a way but because TYO is further away it's got lower demand.

International v Domestic also factors into it.

But, how to I lower if for Chinese cities for example. Based on population i could fly into also ever cities in China 5 times a day.

Also, some places have small populations. Fiji for example. It's all inbound tourists. Based on locals i'd only fly a ATR-42 in there.

Any thoughts would be good.

[bycrewlgw]

About 10 years ago I started writing a airline simulation game for fun. However marriage, babies and a few other personal things got in the way.

Over the past couple of months I've been tapping away and some aspects of this again.

One of the original issues I had and still face is calculating demand between two ports.

In real life there are so many factors that writing logic for all of these will be complex and although I'd like to continuously improve the logic it's impracticable to include everything at this stage.

So my question is how can I start to narrow it down.

I've played with:

Population and Origin - Destination. Have a distance calculation so the further apart they are the lower the demand.

For example, LAX-NYC vs TYO-NYC. Similar sizes in a way but because TYO is further away it's got lower demand.

International v Domestic also factors into it.

But, how to I lower if for Chinese cities for example. Based on population i could fly into also ever cities in China 5 times a day.

Also, some places have small populations. Fiji for example. It's all inbound tourists. Based on locals i'd only fly a ATR-42 in there.

Any thoughts would be good.

Look at affluence too. You can have cities of infinite population but if they can’t afford to fly then demand will be lower.

[AirlineTycoon]

Look at affluence too. You can have cities of infinite population but if they can’t afford to fly then demand will be lower.

oh yeah, thought of that. what's the best global value to use which will be used globally.

[DaveReidUK]

One of the original issues I had and still face is calculating demand between two ports.

Those who know how to do that accurately and reliably aren't on here answering questions, but out there running hugely profitable airlines.

There aren't many of them. :O

[anothertyke]

Try Googling Air Traffic Demand Models and start from there. You will need to choose what you want to represent in your model eg whether you want to represent tourism, land mode competition or just population, income per cap and distance.


Then there is hubbing.....................

[22/04]

And the level of competition

[Flitefone]

Try Googling Air Traffic Demand Models and start from there. You will need to choose what you want to represent in your model eg whether you want to represent tourism, land mode competition or just population, income per cap and distance.


Then there is hubbing.....................

Airbus and Boeing spend a good deal on working all this out and publish comprehensive data. Google; Boeing Current Market Outlook and, Airbus Global Market Forecast...

FF

[DaveReidUK]

Airbus and Boeing spend a good deal on working all this out and publish comprehensive data. Google; Boeing Current Market Outlook and, Airbus Global Market Forecast...

The OP is looking for demand between individual city/airport pairs.

Last time I looked at a Boeing or Airbus forecast, it didn't go into that level of detail.

[840]

When modelling commuting patterns in a borderless environment, it's the square of the distance that matters rather than the distance itself.

I'm not sure it would apply in aviation though. There are all sorts of other complications like alternative means of travel. Look at Dublin and Newcastle to London as an example. Both airports are a similar distance and have similar catchment, but because of the sea blocking rail and road as means of travel, passenger numbers to Dublin are 10 times higher.

Even on long distance, there are factors like historic cultural ties. E.g. look at routes from Europe to the West Indies where routes from the UK go to the former UK colonies, from France it's their former colonies and overseas departments, from The Netherlands it's Aruba, Bonaire, Curacao, Sint Maarten, while from Germany, which has no real ties, it's a mix and match of all of them.