Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Demand
#1
Hi

As I wrote in the latest blog update I am currently looking into having demand between airports being calculated on demand between cities, since people are not travelling for ex. Newark but for New York. I got a response from an user that it is just rubbish and unrealistic. So before I move forward with that idea could you please provide me with some inputs on the subject?
Reply
#2
Although I really like the idea that people are traveling to places instead of airports, I don't think you can make it an algorithm. Ofcourse there are a few factors which you can include in an algorithm. 

For example: 
  • if an airport has a lot of destinations the demand increases. 
  • if an airport has low prices it attracts more holiday travelers.
  • if an airport has more services it attracts more business travelers.
  • the airport size influence the amount of people that will fly from there, so not demand but supply
I think it's way harder to worker with destination to destination demand then airport to airport demand. It both takes a lot of hard data, but the destination demand also requires a lot of factors so that realistic data is created. For example, London City lays on the perfect location (based on coordinates), and it has an ok size for a regional airport with lots of holiday travelers. The problem is that based on all data available you can not decide that this airport is mainly used by business travelers etc. 

Of course you could add a flag or something to this airport, that it mainly handels business or holiday, etc etc. But I think that in the end this would be more data then keeping it at airport to airport data.

These are just my thoughts Smile

Nick
Reply
#3
Just had similar thoughts as nnick. I guess it will be the "easiest" way to keep it the way it was before. If there is an airport with e.g. 30 million pax a year, it is not really simple to find out, to which destinations they all are flying.

In so far, as nnick said, it depends on the offer of destinations. And this might be different on every user. E.g. one has his homebase in TXL and offers 500 destinatons he will get pax for all routes (especially if global advertising is set).

My proposal: keep it as it is. Thats why we have total pax per year in the database, let the users decide, where they are going to fly to. Otherwise the coding gets too complicated in my opinion and I guess most players will have some kind of "free choice", which routes they can / will offer.

Nico Tongue
Reply
#4
Thanks for the input and it looks like I better use the old way of calculating demand :-)
Reply
#5
(04-26-2017, 03:26 AM)pjank42 Wrote: Thanks for the input and it looks like I better use the old way of calculating demand :-)

Cool  Cool
Reply
#6
I have almost 20 academic papers on how to forecast demand for airports and route pairings, but because of uni, grad applications and work i didn't manage to get around to doing summary/info docs on them over the break and with group projects on the horizon I'm unsure of how much time I'll be able to devote to it. I'll try to get something done by the time stuvac comes around.
Reply
#7
(04-27-2017, 04:05 PM)r3dhawk95 Wrote: I have almost 20 academic papers on how to forecast demand for airports and route pairings, but because of uni, grad applications and work i didn't manage to get around to doing summary/info docs on them over the break and with group projects on the horizon I'm unsure of how much time I'll be able to devote to it. I'll try to get something done by the time stuvac comes around.

Do you think it is necessary to program it as teached in University?? I think @pjank has done the right choice to keep it as it was. And to be honest: I do not want to play a game where I can only fly predefined routes! Let us have some kind of free gameplay, P L E A S E ! ! !
Reply
#8
(04-27-2017, 08:40 PM)Therealnick Wrote:
(04-27-2017, 04:05 PM)r3dhawk95 Wrote: I have almost 20 academic papers on how to forecast demand for airports and route pairings, but because of uni, grad applications and work i didn't manage to get around to doing summary/info docs on them over the break and with group projects on the horizon I'm unsure of how much time I'll be able to devote to it. I'll try to get something done by the time stuvac comes around.

Do you think it is necessary to program it as teached in University?? I think @pjank has done the right choice to keep it as it was. And to be honest: I do not want to play a game where I can only fly predefined routes! Let us have some kind of free gameplay, P L E A S E ! ! !

it's only to get an idea of how to relatively accurately forecast demand between the cities. And I'm not sure where you get this idea of pre-defined routes from. Yes there are essential routes that will always exist because of the significant traffic, but there are numerous routes that in the last 5-10 years never existed and no one had predicted would exist and yet they are now being operated due to changing demand forecasts
Reply
#9
@r3dhawk95: I will still need the data because I need it for calculating demand between destinations where we don't have point to point demand specified :-)
Reply
#10
That's why I fly from London to Germany for!
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)