AirwaySim

Hello Everyone,

Forgive me if I'm missing something, but I believe the passenger demand at EGGP (Liverpool, UK) is much lower than it is in RL.  For example, the most popular route from Liverpool is to Belfast International.  In 2009 397,052 passengers flew this route in RL, while in 2007 in AWS 300,000 Passengers flew this route.  Another example is the second most popular route from Liverpool, Liverpool - Malaga where in 2009 297,157 Passengers flew this route while in AWS only 168,000 passengers want to go on this route. 

If I'm missing something please let me know, it just seems like a mistake.

Thank You,
Jordan

I forgot to mention, this is where I found my statistics.
http://en.wikipedia.org/wiki/Liverpool_John_Lennon_Airport

We cannot model single routes according to real life, and that is not the intention even actually. As it would be impossible.

So what do you model?  I'm just curious.

Quote from: J. Oates on April 17, 2011, 09:34:51 AM
So what do you model?  I'm just curious.

Calvin Klein jeans...

sorry, I couldn't resist.

Quote from: swiftus27 on April 17, 2011, 10:23:25 AM
Calvin Klein jeans...

sorry, I couldn't resist.

CLASSIC!  lol good one mate ;)

It's some sort of formula based on real life data. However it's not the intention to model real life 1:1 in AWS and many people (including me) hope, one time AirwaySim will be completely different from real life, starting at every airport with the same requirements and not be forcen to use unknown cities in the middle of a nowhere, just because in real life a big carrier operates it's base there.  :)

Quote from: Curse on April 17, 2011, 10:28:33 AM
It's some sort of formula based on real life data. However it's not the intention to model real life 1:1 in AWS and many people (including me) hope, one time AirwaySim will be completely different from real life, starting at every airport with the same requirements and not be forcen to use unknown cities in the middle of a nowhere, just because in real life a big carrier operates it's base there.  :)

go read about city based demand.  that will help. 

I understand that, but it is a simulation based on RL.  All the aircraft are based on RL, the Airports are based on RL.  I think the Passenger number should be based on RL.  Or at least be able to increase the amount of passenger numbers by increased marketing.  If it wasn't based on RL why would people base themselves at Heathrow/Atlanta etc if the passenger numbers weren't like the ones in RL.  I know it seems I'm being a bit picky but I am actually losing out on a few hundred thousand passengers a year and I think that its not exactly realistic.  Don't get me wrong I love the game and I'll continue to play it :p

Jordan

Quote from: swiftus27 on April 17, 2011, 10:30:01 AM
go read about city based demand.  that will help.  

I don't believe it until it is a live feature  ;D

@ J. Oates

That's the fact. Why people in this game base 1950 in Atlanta? Because Atlanta is now, 2011, an incredible big airport and a very important US hub which generates endless demand of passengers.
But what exactly was Atlanta in 1950? Compared to New York? Detroit?

I dream of a game one can found an airline 1950 in Portland, Oregon, and if you build up a very big airline, this should be in this specific ingame world in the year 2011 a big airport like Atlanta is in real life, while in the same specific game Atlanta is just a normal big airport.

If the current system is followed, we will see forever the same thing; the same big city names, the same limited amount of slots. If there were in the 50s an airline at Frankfurt/Main, Germany, that had double or third the size of Lufthansa then, we had a second and third runway in the same decade and 2011 we would not have a big controversy about building a third one.

Why is Heathrow so extremely popular in AWS? It has the by far biggest international demand. Pure luck Heathrow is now so well known.

Sebastian,

It's not just about having a big airline or not. Passengers should be willing to fly to a certain city in the first place, which is based on so many variables that it is almost impossible to code. Swiftus' example -- city based demand -- is just one of the many things which determines the growth of an airline and, mostly as a result, the demand position of the airport (intercontinental, domestic etc). There is a reason why airports have developed the way they did in the past - after all, the variables that determine passenger demand can obviously only be put into practice in the real world.

I think you base airport growth too much on the airline, which is again just one of the many variables which determine the development of the airport.

To further illustrate my point:

Quote from: Curse on April 17, 2011, 10:47:50 AM
If there were in the 50s an airline at Frankfurt/Main, Germany, that had double or third the size of Lufthansa then, we had a second and third runway in the same decade and 2011 we would not have a big controversy about building a third one.

Take a step back and think about why there actually wasn't an airline three times the size of Lufthansa at the time. Think about why a second or a third runway was not needed back then. It's certainly not because Lufthansa was poorly managed - they're still around which means they must have been doing something right. I mean, if they had the chance to realize more significant growth back then, they obviously would have. But they didn't because of those aforementioned variables composing the demand, which is the point I'm trying make.

both good points, however Portland, Oregon couldn't possibly be as big as the big names it just doesn't have the population and its out of the way in terms of holiday destinations, because the actual way passenger demand should be determined is the amount of people in the area, size of the airport, financial status of the airport city and neighboring cities etc.... City based Demand would be the only I can see that Passenger demand can be truly modelled unless Sami did route modelling... because the actual passenger demand in some areas is completely unrealistic.

In my example the demand is EXTREMELY low compared to RL.  and whether or not the passenger demand is modelled by a formula or not it is incredibly unrealistic...

City Based Demand seems like a very good idea :)

There are some airports that are big because they have the demand - no problem.

But why is Heathrow this big? There was not alternative in the beginning of aviation to travel with aircraft through the British Empire except from Heathrow. It's not the fact it's just a cool airport.

I named Atlanta as second example. Atlanta has the advantage to be a central place of the US and a big city, like Chicago. This makes it an optimal hub.

How would Atlanta demand be without hub for Delta and Eastern?

The problem is - not in every game is an airline that's like Delta or Eastern. Many base at Atlanta because it has such a big demand, but the demand isn't because everybody in Georgia swims in money and needs to travel every day around the world. It's just the fact people from all over the USA and all over the world travel there. You want to fly from Frankfurt to Los Angeles? If you choose Delta, you'll see Atlanta Int. Airport.


In a game like AWS you should be able to make every airport of a bigger city to something like Atlanta. The hub of your airline. A big hub. That growths - or growths not if you want to have a smaller airline.

The player should define the game and not make best out of data that was achieved in real life. I can understand why sami has build in the current model and it is ok, but it's boring to base always at the same airports.


And the fact demand can be "created" everybody can see in the real world if he looks to Dubai (Emirates Hub), Atlanta (Delta) or even Frankfurt (which is a much smaller area than Berlin or Ruhrgebiet, but Lufthansa hub for decades). I see there's not alternative to London in the UK, Paris in France or Tokyo in Japan. But take also a look to China. There are some good alternatives to Bejing as main hub for an China based airline. But in AWS it's not possible, because nearly all international demand is based at Bejing.


Edit:
@ J. Oates

Portland could be a good hub. You can get people from Seattle etc. there and if you build a strong international network with good prices, why one shouldn't fly from Tokyo to Portland and from Portland to Chicago?

By the way, Portland was chosen totally random. Insert any other airport instead, I named already Detroit in the 50s which was a much more important city than Los Angeles or Atlanta in my perception.

You keep forgetting that there are reasons behind those hubs being so large. In fact, you almost assume that airlines like Delta and Eastern have the wrong strategies. After all, why would an airline like Eastern still compete with Delta at Atlanta, if plenty of other hubs would provide them with the same opportunities? We can't define all those reasons and variables just like that, but they're definitely there. And again, they just cannot be coded. And if they could, we would turn out with an airport system that is at least similar to, if not the same as the system we have now simply due to the way logistics are distributed throughout the world.

The reasons are simple. Sometimes it's because of the real demand needed (New York airports), sometimes it's due to space (not a big problem in many US cities like Atlanta, Dallas-Ft. Worth) or just because they are extremely cheap to fly to (Ryan Air airports). Their are some other points: Personal preferrence of CEO/owner (was important earlier) or place on the map (what makes Atlanta and Chicago better than Los Angeles for example).

I don't know what exactly what the benefits of Atlanta were. Due to the fact it wasn't this big industrial city etc. it might be a mix of cheap possibilites and the strategical position in the US and North America.

Why airlines often open hubs at established airports is also verly clear. No new infrastructure must be build and you can get some pax from the other airline.

I don't say Delta or Eastern choose a bad strategy. I only say that, if we were in the 40s or 50s or even 60s we could decide ourselves what factor we prefer most. Will we build a big airline out of our homecity? Indianapolis for example?
Or will we choose an airport in a not-to-small city with a good location? Kansas City could do well.


Coding seems not a problem for me. Sami already has the city based demand. What must be coded is some model that allows increase and decrease of this demand in combination with of transfer between aircraft (1x or 2x) at least within a specific airline. So a route SanFrancisco-KansasCity-Boston could preferred to SanFrancisco-Boston or SanFrancisco-Atlanta-Boston.
Also there must be the possiblity to increase airport size. If x% of slots have gone, the city must decide to build a new runway. This shouldn't be based on city demand. See airports with extremely big amount of slots in smaller cities. Also there could be a hard coded limit for some existing airports that just can't be extended, this limit could also change over time when cities grow and airports that were some miles away are now in the middle of a city.

So, I hope you see my point. I don't want play Earth Aviation 2.0. I did this in the last games and even in Early Days the Atlanta-supermode worked well. I will make my own choices. If this would be real life and my airline had such an impressive hub at Atlanta, why people still don't want to fly there from Paris, Roma, Milano and so on?

I do see your point how this increases the fun and challenge of the game, but I just think it is hard to code and implement. :)

For example, the reason that Schiphol is now such a big hub, is because local logistics make this possible (the entire Dutch train system comes together at Schiphol, for instance). Let's take Groningen Eelde airport - it would never be able to support the same logistic system as Schiphol unless substantial government expenses are put into place. And if there would be any increase in traffic, it would be minimal (also because of the low population density here). Not to mention certain legal requirements as noise and runway limits (e.g. a fully loaded B737 cannot depart from Groningen Eelde). Therefore, it depends heavily on local situations whether an airport is able to actually grow substantially or not. Of course, in some situations it is definitely possible but I think there are too many exceptions to really make a change. Again, Lufthansa chose Frankfurt as their hub for a reason that's presumably quite well thought about.

I think it eventually comes down to where you start (in Dutch - the chicken and the egg :laugh:). If there were no land, noise etc restrictions, no logistical systems in place already and so on, changes are more easily made. But in 2010, you can't just throw 100 aircraft in an airport like Groningen and expect the airport to grow along. :)

Why the whole railway system comes together at Schiphol, Nick?

Image it's the year 1950. And KLM decides not to base at Schiphol, KLM decides to base somewhere else. Where do you think in ourdays all the railways will go to?  Wink

"Again, Lufthansa chose Frankfurt as their hub for a reason that's presumably quite well thought about."

They had to, because they weren't allowed to serve Berlin or out of Berlin due to the second world war. Before that, Lufthansa's Berlin hub was bigger than Frankfurt. What will happen when Berlin-Brandenburg-International opens, nobody knows... it depends if one of the two big German airlines will move there... or not. Exactly the same situation I want to make in AWS for all airports.

Groningen might be a bad example. But if there will be all flights located, everybody is forced to get there from Amsterdam, Rotterdam etc. I'm sure one can't make a very big international and domestic hub out of very small airports, but Ryan Air shows how smaller airports far away from big cities could create demand. Or who wanted to fly to Frankfurt-Hahn or London-Stansted 1985?

Noise restriction etc. should be just country wide. You are absolutely right it would be too extremious to make this for each airport and airport build levels extra.

QuoteBut in 2010, you can't just throw 100 aircraft in an airport like Groningen and expect the airport to grow along.

This might be correct. But on the one hand, I don't know employee situation there, so maybe it could benefit the region. On the other hand, like I said, my idea is more for scenarios that will run from 1950 to 2030... not for Air Travel Boom scenarios  :)

Quote from: Curse on April 17, 2011, 01:37:28 PM
Why the whole railway system comes together at Schiphol, Nick?

Maybe I exaggerated a bit, but if I want to travel from Groningen to Amsterdam I will have to travel via Schiphol. It's mostly because of the construction of a huge train station underneath Schiphol because Schiphol is such an important economic motor of the Netherlands (thus many commuters and workers travel there from throughout the Netherlands). And it is at the center the most important economic area of the Netherlands - the Randstad.

Quote from: Curse on April 17, 2011, 01:37:28 PM
On the other hand, like I said, my idea is more for scenarios that will run from 1950 to 2030... not for Air Travel Boom scenarios  :)

Ah. In that case, it might be possible to make a change in hubs but weren't logistical systems in the US quite established by that time?

Such system can change. See JFK -> Newark, Heathrow -> Gatwick or the complete build up of Boeing at Everett. Other examples avoiding aviation are Nokia who left Germany a few years ago and built up a new and big factory with all their suppliers in Romania.

The biggest example is, still said, Dubai. 30 years ago it was the same desert jerkwater town it is now, completely out of any industry or anything else that's worth something (except oil), but they have established a very big airline and building an incredible big airport in the middle of nowhere.

Again, to close this part: I see the importance of Schiphol. It's like in France... an airport in Marseille can never get as big as one in Paris, because of the range between both cities and the fact Paris is 50x bigger, including the center of everything in france.

Yup, but the reason such an airline in Dubai was needed was the complete shift of a rich oil state to a rich tourism/service state (0% tax, anyone? ;)).

Also, what you're referring to is organizational change which is completely different. Nokia and Boeing strived for efficiency within their companies, and sometimes quite rigorous change is needed to achieve this.

Newark and Gatwick were just additions to the logistical system that was already in place.

Anyhow, let's just wait and see if and how Sami wants to implement such changes to the current system. :)

Quote from: Quinoky on April 17, 2011, 12:44:22 PMWe can't define all those reasons and variables just like that, but they're definitely there. And again, they just cannot be coded.  And if they could, we would turn out with an airport system that is at least similar to, if not the same as the system we have now simply due to the way logistics are distributed throughout the world.

I'm not taking sides, I'm just curious - could you elaborate on this?  What exactly are these variables?  
I have a feeling that the random political element (which is present in real life, but often illogical) plays a large part in this discussion.  If this is the case, many of today's hubs may, in part, be products of politically-motivated decisions that are not necessarily in line with the logical, hard numbers that effectively drive player decisions in AWS.

As far as fixed variables are concerned, local population may be one, but then again... (and this is just an example), Atlanta is only the 40th most populous US city at 420,003 (2010 US Census) while Memphis is #20 at 646,889.  Yet somehow, KMEM only boards just 5.5 million passengers annually while its home city has a 154% larger population than that of KATL, which boards 42.2 million pax yearly (FAA CY2009 data (http://www.faa.gov/airports/planning_capacity/passenger_allcargo_stats/passenger/media/cy09_cs_enplanements.pdf)).

Again, you narrow it down to just city based demand. Here, you assume that airport growth is merely based on the people in the area that are willing to fly. Let's put it this way: Amsterdam Schiphol is one of the largest hubs on the planet, yet the entire population of the country of the Netherlands consists of about 16.5 million people. That's less than most large cities on earth. So there must be a different explanation for its current size, which is not because some huge airline uses it as its HQ (KLM isn't exactly the biggest airline out there), nor because it serves a large population.

It remains clear, therefore, that many other factors are involved here, most of which we're probably not able to define.

I still see your point, Nick. And you are right, today in the year 2011 space is extremely limited in Europe and the existing airports are well established. So AWS represents the situation in (which year is sami's data mostly? 2005?) this current era very well.

Again, all I written is only useful in games that will last long enough to establish an airport by a player himself or some airlines at a specific airport. It doesn't make sense to use such a model if the game world lasts 2000-2020 and the creation of another runway would start ~2008 and finished ~2011. This might be helpful for already big airports to make sure there are still slots available, but it isn't something one can count on if he wants to establish a fully new base.

In the 50s and 60s aviation was very small and many completely new airports were built since than and passengers use them. So in the long time scenarios, which sami said they will come (and I'm focussing mostly on that because they seem endless more interesting than scenarios without fleet replacement), you can form an airport. You have time to and passengers in the early days are willing to follow you, because it isn't established this way.

For sure we can count in all the things you said - but only some of them were already a point in the 50s and 60s. We often forget the world is more than the space limited Europe with it's small countrys and the well established US. I named already China, but there are also countrys like India or maybe even the Soviet Union where big chances are possible to make even ourdays. 

Obviously theres more to it than the population of the city.  But more back to my point... On AWS information page under What is Airwaysim it states:

"The simulation includes over 2500 airports around the world and over 350 different aircraft models - all of them with accurate details and specifications."

All of them with accurate details and specifications, that is not entirely correct as we have already established, I know it looks like I'm being a bit of a jerk, but I want my couple hundred thousands passengers lol  :P

I read this as it is based on the last part of the sentence - the aircraft thing.  ;)

Quote from: Quinoky on April 18, 2011, 05:59:27 AM
Again, you narrow it down to just city based demand. Here, you assume that airport growth is merely based on the people in the area that are willing to fly.

No, if you will re-read my post, I use the case of KATL/KMEM specifically to underline the fact that population is not necessarily the main driver behind hub selection by an airline.

Quote from: Quinoky on April 18, 2011, 05:59:27 AMIt remains clear, therefore, that many other factors are involved here, most of which we're probably not able to define.

Again, I ask - what are these intangible factors?  I am genuinely interested, but you have yet to expound on this statement.  The only thing that comes to mind are political motives, which are not present in AWS, and thus would mean that hub distribution in the game would probably not be the same as in real life (which you assert in previous posts).

Quote from: J. Oates on April 18, 2011, 08:17:23 AMAll of them with accurate details and specifications, that is not entirely correct as we have already established, I know it looks like I'm being a bit of a jerk, but I want my couple hundred thousands passengers lol  :P

I understand your frustration, but perhaps the precision & accuracy promised by that statement refers more to the technical specifications of the airports (runway lengths, average seasonal weather, distances from other airports, etc.) and not necessarily the statistical representations of passenger/cargo demand.

Quote from: headphase on April 18, 2011, 09:38:08 AM
No, if you will re-read my post, I use the case of KATL/KMEM specifically to underline the fact that population is not necessarily the main driver behind hub selection by an airline.

I see, I interpreted it in the wrong way, sorry.

Quote from: headphase on April 18, 2011, 09:38:08 AM
Again, I ask - what are these intangible factors?  I am genuinely interested, but you have yet to expound on this statement.  The only thing that comes to mind are political motives, which are not present in AWS, and thus would mean that hub distribution in the game would probably not be the same as in real life (which you assert in previous posts).

Primarily the infinite amount of local exceptions. I named Groningen Eelde airport before - logistic systems, legal issues (noise, runway limits etc.) and population density, combined with political issues as you pointed out. Not to mention economic value of certain established airports, by which I mean that certain airports serve different economic aspects of regions, countries or even continents. There probably are many more factors such as employment rates, tax benefits, economic stability, political stability, maybe even geographical issues (Osaka Kansai comes to mind) and forces of nature (Sendai airport, for example).. Again, there are just so many variables that it is impossible to determine them all and take them into account at local levels.

Quote from: J. Oates on April 18, 2011, 08:17:23 AM
All of them with accurate details and specifications, that is not entirely correct as we have already established, I know it looks like I'm being a bit of a jerk, but I want my couple hundred thousands passengers lol

Fine. There are 2500 airports (or more). Meaning that there are at least 6 million different route combinations. Now please find the 'true and 100% accurate demans' for all of these route-pairs.... Get what I'm after? :P

Once more: We are not going to even try to model all of the routes after real life. Within certain accuracy, yes. But not to the last straw.

Quote from: sami on April 18, 2011, 11:00:54 AM
Fine. There are 2500 airports (or more). Meaning that there is at least 6 million different route combinations. Now please find the 'true and 100% accurate demans' for all of these route-pairs....

Get what I'm after? :P

Once more: We are not going to even try to model the routes after real life. Within certain accuracy, yes. But not to the last straw.

LOL yes I understand that but when you say within certain accuracy LPL is within almost no accuracy... I've not gone through every single route but I'd say its around 300,000 Passengers short a year on some routes.

http://www.liverpoolairport.com/assets/_files/documents/sep_08/peel__1221147943_Summer_2008_Timetable_-_low_re.pdf

If you have a look at that link, it'd be impossible to serve those destinations with the AWS passenger demand as it is... Also have a look at the statistics on this wikipedia link: http://en.wikipedia.org/wiki/Liverpool_Airport

Each route is about 100,000 passengers missing in AWS... Im not sure whether this is a general problem with every airport or just LPL...I  understand you can't model each route absolutely perfect... but come on this is quite a big difference between RL and AWS

btw. Checked, and Liverpool-Malaga demand (ref. first post of this topic) is based on the very real and actual demand figures of that route (UK CAA). As well as Belfast. (or at least that is what's been marked to database)

The base year for all data is 2006 and it is scaled between years (future and past). So if actual demand in 2009 is something different, then it's because the economy system cannot predict the future and cannot predict and model the changes on economy status of individual airports or routes; and the travel index and factors are the same globally for all routes and airports.

Anyway, the current demand system will not be changed at all. Time will be focused on the research of the new system.

Quote from: Quinoky on April 18, 2011, 10:08:58 AM
I see, I interpreted it in the wrong way, sorry.

No worries :)

Quote from: Quinoky on April 18, 2011, 10:08:58 AMAgain, there are just so many variables that it is impossible to determine them all and take them into account at local levels.

All good points; but I still submit that in some cases an airport truly is only massive primarily because of its airline hub status (thanks to all the connecting traffic that flows through it).  I suppose the big limitation we must recognize is that AWS doesn't understand connecting pax; route demand is only assumed to be O&D.  In this simulation environment it will never be possible to build hubs from the ground up until connections are somehow modeled within the game engine.

Quote from: CurseIn a game like AWS you should be able to make every airport of a bigger city to something like Atlanta. The hub of your airline. A big hub. That growths - or growths not if you want to have a smaller airline.

Hopefully one day it will look like this  :laugh:

Quote from: headphase on April 19, 2011, 07:34:50 AM
All good points; but I still submit that in some cases an airport truly is only massive primarily because of its airline hub status (thanks to all the connecting traffic that flows through it). 

Certainly, but like I said, if you want a model that differs from the real world which permits airports to grow to sizes like Atlanta, there are too many local exceptions to take into account. An airport can't just grow because the airline grows, it is bound to an incredible amount of variables. :)

Quote from: headphase on April 19, 2011, 07:34:50 AM
I suppose the big limitation we must recognize is that AWS doesn't understand connecting pax; route demand is only assumed to be O&D.  In this simulation environment it will never be possible to build hubs from the ground up until connections are somehow modeled within the game engine.

Which is one of those variables, indeed. :)

All those variables could be ignored in a game like AWS.

Would the goverment of Groningen reject the offer the build the biggest international airport in the 50s? I don't know. But this is a game, so it doesn't care.

It's de facto not possible to model such things, because there is no data. There was never the idea to make every small airport a big one, so all these points can and must be ignored. But, again, Ryan Air shows us how even very small airports in the middle of nowhere can increase in demand and importance, even in 2011.

Ryan Air have a base in LPL.... increase passenger demand  plleeeaasee  :'(

Ugh stupid CAA don't know how to make correct statistics, wikipedia is SOOO much more truthful ;)

http://www.dallas-lovefield.com/pdf/masterplan/Chap2-Forecasts.pdf