Calculating Demand

Hey AS Community,

Just a quick question on how demand is calculated: is it based more off of real world demand, or solely on how much capacity travels through 2 airports? I'm planning on expanding out of Honolulu, but I can't seem to figure out if it's better to start only on large airports (LAX, SFO, SEA, etc.) or provide service to lower-demand airports (SJC, SAN, etc.). I know a lot of traffic passes to HNL through both airports, however places like NAN where there is barely any O+D demand between both seems like a waste of time, especially considering how hard it is to employ connections on such a route. Considering both HNL and NAN are of good size (8 bars and 5 bars), and could fill a 320/CS1 on that demand alone I could start the route easily, however if it is real-world based I face a struggle in O+D with a lack of connections.

For USA I think it have per state figure? As in the country of origin and state of origin breakdown of travellers count.

And i think connection traffic also count towarf traffic in game

Here is my understanding:

Demand on any given route is based on a combination of real world capacity (how many seats are offered on a given route) and real world total airport traffic. It does not take into account how many of those passengers are transferring.

The real world capacity part of the equation allows the game to model demand between markets with special relations ships (CDG-YUL being more popular than CDG-YYZ due to the french connection for example). Because it does not take connecting passengers into account it also creates what feels like a falsely high demand between some hub routes.

The total airport traffic (what you see approximated by the green passenger bars) is used to model traffic that is served by connections in real life. This passenger demand is distributed based on distance (more folks going SFO-LAX than SFO-JFK), size (more folks going SFO-LAX than SFO-LGB), and country (more folks going LAX-MIA than LAX-MEX). The home country effect is very strong, you’ll see a lot more passengers on domestic US flights than those going to Canada. It is worth noting that the bars are based on total passengers and distributed throughout the country. This means that GUA-FRS (5 and 3 bars respectively) will be easy to fill since GUA’s 5 bars of passengers aren’t divided between many airports, while SAN-BFL (8 and 3 bars respectively) will struggle because SAN’s 8 bars of passengers are being distributed over many more airports.

HNL can be a challenge because most of the natural demand is to the NW of you (CONUS). You can offer connections to the neighbor isles, but that is hard to do with wave scheduling since the mainland flights are long and the inter island flights are short.

Sooooo, will NAN work? I don’t know, try it! A whole lot of words for a non-commital answer, I know, but you can improve your chances if you offer good connections to SFO and LAX.

Here is my understanding:

Demand on any given route is based on a combination of real world capacity (how many seats are offered on a given route) and real world total airport traffic. It does not take into account how many of those passengers are transferring.

The real world capacity part of the equation allows the game to model demand between markets with special relations ships (CDG-YUL being more popular than CDG-YYZ due to the french connection for example). Because it does not take connecting passengers into account it also creates what feels like a falsely high demand between some hub routes.

The total airport traffic (what you see approximated by the green passenger bars) is used to model traffic that is served by connections in real life. This passenger demand is distributed based on distance (more folks going SFO-LAX than SFO-JFK), size (more folks going SFO-LAX than SFO-LGB), and country (more folks going LAX-MIA than LAX-MEX). The home country effect is very strong, you’ll see a lot more passengers on domestic US flights than those going to Canada. It is worth noting that the bars are based on total passengers and distributed throughout the country. This means that GUA-FRS (5 and 3 bars respectively) will be easy to fill since GUA’s 5 bars of passengers aren’t divided between many airports, while SAN-BFL (8 and 3 bars respectively) will struggle because SAN’s 8 bars of passengers are being distributed over many more airports.

HNL can be a challenge because most of the natural demand is to the NW of you (CONUS). You can offer connections to the neighbor isles, but that is hard to do with wave scheduling since the mainland flights are long and the inter island flights are short.

Sooooo, will NAN work? I don’t know, try it! A whole lot of words for a non-commital answer, I know, but you can improve your chances if you offer good connections to SFO and LAX.

I have to disagree, demand is not really based on specific routes but rather on the overall demand between two countries / regions which is then broken on the respective airports.

For instance when operating flights to La Guardia (LGA) and Regan National (DCA) airports have a sufficient demand for transatlantic flights although in real life they are only used for short distance flights.

Your example with YYZ and YUL might relate to the two provinces Quebec and Ontario who have a difference in demand for flights to mainland France.

as for the initial question, I would research which state in the US has the highest amount of real life connections to hawaii (I guess that will be California). Try to create connections to all airports in that region. Within a region demand shouldn't vary for two similar sized airports so for instance OAK and SMF should have similar demands.

Also note that this demand is also divided among all aiprots in the Hawaii region, so you will have demand to California even on those airports that in real life only have intra state connections.

A great site that is free from the US govt is this one: http://www.transtats.bts.gov/airports.asp

It lists all airports with commercial traffic in the United States, and their top 10 destinations domestically.

If you want specific info for HNL, for example, here it is: http://www.transtats.bts.gov/airports.asp?pn=1&Airport=HNL&Airport_Name=Honolulu,%20HI:%20Honolulu%20International&carrier=FACTS

Off-topic (maybe), but kinda relevant (definitely)...

The link to the excellent site above, which gave transportation statistics for US airports is now dead. If you still want to find them, they are here: http://transtats.bts.gov/airports.asp

Basically the same link, just no "www." before "transtats"

Still a great source of information!

"The link is dead! Long live the link!"

I've operated a NAN-based airline for quite a while on Gatow (now liquidated as I decided to start all over in Australia). NAN-HNL was a moderate success with relatively full flights most likely amounting to about 400 pax a day spread over 3 flights, if I recall correctly, but only if you have lots of connections on either or both ends. I remember making a consideration similar to yours early in the game, and was disappointed by the lack of O&D between both.

For me, what works best in cases like this, is to check ORS for connections that make sense (in terms of geography and total flight time) and use that to make your decisions.

What I did for example, was using NAN as a hub for southern pacific travel - there was nothing to get from French New Caledonia and East Coast Australia, to French Polynesia, for instance. That gave me a LOT of transit in NAN which is in between. Then, having found an IL partner in HNL, I restarted NAN-HNL and could fill a 788. In HNL in your case, just check how hard it is to get to NAN from west-coast USA. If you see plenty of direct flights, then you can probably forget about a successful HNL-NAN route, but if you get much traffic from that region to HNL and then succeed in offering the best connection time and rating onward to NAN, it may work well.