Hub preference

English Boards General
1

Hello,

    I would just like to ask for your opinion. I currently have an airline based in Athens. Recently, my overall traffic in and out of my hub airport (Athens) has fallen dramatically, plus the fact that I can't get slots at suitable times anymore. I figured out that it may be so because of the harsh competition. So, I thought that if I move my operations to a non central airport (that does not have many direct connections) I may be able to secure more traffic, higher profit and the slots I need. Howerver, that airport will no longer be in the centre of the country (Greece) and thus MAY make domestic connections worse. What is your opinion on the subject?

2

Choosing a hub is one of the most important parts of starting an airline, if not the most important. You are right in that a central location does help, but that doesn't mean coastal areas or hubs around the edge of a country shouldn't be considered. One thing you need to do is decide what you want your airline to be early on. Is it a regional airline? Does it fly internationally? Will it have interlining partners inside the country or outside, both, or neither?

I will always say the key to running a successful airline for more than a few months is making sure you have connections. Usually people run the connections through their hub... that's why it's important in choosing a hub. You should use an airport that has the ability to make good connections the way you want them to be, and also have enough general traffic to help get things going. Greece (like most all European countries) is small. Flying domestically will get you some passengers if you use a small enough aircraft (20-60 seats total), but the growth rate is small. If you're wanting to be larger, you will need international flights that time up well for maximum ORS rating. Usually the most efficient way to do this is to use one aircraft.

I'll explain... but keep in mind I don't know the travel times between airports, so if I'm wrong on the specifics, don't worry about it. If I were to start an airline in Athens, I would probably get aircraft like the Embraer 195s, 737s, and maybe some CRJs. For any of these, you should be able to fly from places like Rome, Munich, or other central European cities into Athens. Now that you have the plane in Athens, you should schedule it to go somewhere east. Think of it like a flight from Rome, through Athens, to.. Beirut or somewhere. Istanbul, maybe. Go the opposite direction. If you can line up the slots correctly, you'll have passengers booking three flights. Rome, via Athens, to Beirut. Rome to Athens. Athens to Beirut. That's three possible chances to fill up that flight, and it really does make all the difference.

I run a rather successful airline out of a 7-bar airport in the USA. 43% of my passengers are transfer passengers through my hub. I'd actually rather have that closer to 50-55%. I feel that's about the butter zone.

3

i try to answer the original question in a harsh way:

you have issues with competition and your loads and you try to solve these issues by moving your airline to a place where you have the same competition* and a smaller demand for traffic?

Be patient. Slots will become available in the future. Someone is closing down his airline everyday.

Try to optimize your offer. As long as your flights don't show up at the first positions in the ORS, you didn't reached the best settings

Fight for every passenger.

* passengers from AMS, MAN, CPH and ARN don't mind whether they connect through ATH or SKG to CHQ, CFU and other airports

4

Guys really thank you for your answers.

i try to answer the original question in a harsh way:

you have issues with competition and your loads and you try to solve these issues by moving your airline to a place where you have the same competition* and a smaller demand for traffic?

Be patient. Slots will become available in the future. Someone is closing down his airline everyday.

Try to optimize your offer. As long as your flights don't show up at the first positions in the ORS, you didn't reached the best settings

Fight for every passenger.

* passengers from AMS, MAN, CPH and ARN don't mind whether they connect through ATH or SKG to CHQ, CFU and other airports

Actually, the airport I am considering switching to does not have many connections to Europe or Africa or Asia   (or even direct connections within Greece) meaning that I can generate a lot p2p traffic and there I can choose whatever slot I want. Its traffic concerning passengers is actually just one bar less than Athens. Does this make any difference?

5

... you have issues with competition and your loads and you try to solve these issues by moving your airline to a place where you have the same competition* and a smaller demand for traffic? ...

Hi,

I was thinking exactly the same.

@Filippos

You have a hub and slots in Athens. In most countries, the capital is the best place for a hub  :-)

Ask yourself why your seat loads go down. Check the ORS and see how your flights compare with other flights. If necessary, improve your ratings by better service or lower prices. Check the weekly statistics and see how your competitors are doing. If everybody is suffering, make sure you can survive longer than your neighbour. If he fails before you, you wil get his passengers and his slots.

And Caithes is right about connections. Flights from Europe should connect with flights to the Middle East or Asia. I also think relatively few passengers want to fly from Korfu to Kreta. Many "domestic" passengers are tourists who fly via Athens. Think about these connections too.

Jan

6

And Caithes is right about connections. Flights from Europe should connect with flights to the Middle East or Asia. I also think relatively few passengers want to fly from Korfu to Kreta. Many "domestic" passengers are tourists who fly via Athens. Think about these connections too.

Jan you are right about that IRL, but in AS I don{t think that necessarily holds true. it is my understanding that AS generates traffic based on the combination of these factors:

1) Total traffic for the airport

2) Number of departures (and arrivals) per day

3) Distribution of domestic vs international traffic

4) ...and "maybe" distribution of short/medium haul vs long haul traffic

5) for some airports, where strong preference of trunk routes is known (e.g. A-B specific traffic or country X to country Y) that may be calculated as well.

Based on this, a flight between two 6-bar airports should generate roughly the same traffic. Meaning: Flights between ATH and Korfu would have preference over ATH to XYZ international with same bars as Korfu.

6) It is my understanding that AS does not know origin (or destination) on IRL connecting flights. That means DFW to PHX will have strong traffic because both are connecting hubs, even if O/D between DFW and PHX may not be that huge.

Another example: I run an airline in Brazil and I tried a route that IRL is operated daily by a 757, from a deep secondary airport (6 bars). In my case I was the "ONLY" flight in ORS at all and I generated exactly 2 pax per day. This also confirms points 2) and 6) as I listed above.

Getting back to the OP's case, airport stats lists departures (and number of pax) as domestic and international. That means that even if most ATH to Korfu are connecting FROM international, on that leg they would be listed as Domestic in airport stats, and that information would be extrapolated into AS pax demand for domestic routes (in General, because again, AS does not apparently differentiate between specific destinations). So all domestic pax from ATH (even if many/most connecting from INTL) would be distributed in domestic demand calculation from ATH to domestic airports in the order of domestic bar-importance.

This does not mean, though, that international trafic to airports in Greece beyond ATH terminates with demand to ATH. That is described by my point 5) above, if approximate pax numbers from Germany into Greece are known, those pax would be distributed in a cross matrix among all German airports on one side and all Greece airports on the other side, again based on airports bar-importance. So a DE 8-bar to GR 7-bar will have more pax than DE-8 to GR-5, and DE-5 to GR-7 will have more pax than DE-6 to GR-1. This demand calculation, in my opinion, is different from domestic demand calculation for Greece (or any other country for that matter) I am even starting to believe that the "domestic leg" of "international-connecting-into-domestic" flight is given double demand, once as domestic O/D traffic and one as international connection-via-matrix traffic. Maybe not exactly double pax, but certainly more than RL proportion.

7

@adfilipposgr

DO NOT abandon ATH base, start experimenting with your secondary airport and see how it goes.

First, get ATH in order. See how you stand in ORS and make changes if necessary. Service and prices, and maybe even seats. Once your ATH loads stabilize, start experimenting with that second airport. First, put in one plane and start flying domestic high-bar airports, O/D traffic, and see how it goes.

8

Hey guys, really thank you for your answers!

@adfilipposgr

DO NOT abandon ATH base, start experimenting with your secondary airport and see how it goes.

First, get ATH in order. See how you stand in ORS and make changes if necessary. Service and prices, and maybe even seats. Once your ATH loads stabilize, start experimenting with that second airport. First, put in one plane and start flying domestic high-bar airports, O/D traffic, and see how it goes.

I am trying to do exactly this thing and I can say for sure that my second airport generates A LOT of demand, thought I have only tried domestic.

So if I understand well what everybody says, having a hub in A is no different than having a hub in B, as long as I have sufficient connections.

Also would like to ask how can I gradually completely reschedule all my flight in the same airport (without changing hub) in order to get the maximum connection. Please check out my airline A-Via in the Tempelhof server.

9

Rescheduling in ATH will be difficult because most of the slots there are taken, 92-99% average slot, unless you make the connection wave in 17-21.00 hrs window, but that is only one wave.

10

Hi ...,

I am not sure what you mean when you say that the game generates passenger demand based on

1) Total traffic for the airport

2) Number of departures (and arrivals) per day

3) Distribution of domestic vs international traffic

4) ...and "maybe" distribution of short/medium haul vs long haul traffic

5) for some airports, where strong preference of trunk routes is known (e.g. A-B specific traffic or country X to country Y) that may be calculated as well.

I know AS doesn't use origin to destination data. They use airport to airport data. That is indeed why there is too much passenger demand between real life hubs. With this in mind, there would be no (holiday) transfer passengers in Athens. However, the game also has passengers who are distributed over a region.

The big question is: how is passenger demand calculated  :-)

We know passenger demand is calculated per airport, and the game status page doesn't show any "regional" passenger demand calculations. So I assume Sascha divides part of the passengers over several aiports within a region.

If (for example) 100 real passengers fly from Athens to Hamburg, the game may have 50 passengers from ATH to HAM, 10 passengers from HER to HAM, 8 passengers from SKG to HAM, 6 from RHO to HAM, and so on. In my opinion, AS specifies the origin/destination of every passenger.

Apart from the AGEX, passenger demand is static. I don't think there are virtual passengers from the region Greece who choose between flying from HER to HAM and flying from SKG to HAM. They are appointed to one airport and they will fly if there is an eligible route.

I also don't think the number of departures has an influence on passenger demand. However, flying from HER to ATH may increase bookings on the ATH-HAM flight. Just like starting a daily flight from ATH to HAM may increase bookings on your HER-ATH flight. These are the 10 passengers that I mentioned before.

...

Based on this, a flight between two 6-bar airports should generate roughly the same traffic. Meaning: Flights between ATH and Korfu would have preference over ATH to XYZ international with same bars as Korfu.

...

Again, I am not sure I understand you correctly. In my opinion, there is no preference involved. Either there is passenger demand on a route, or there is no passenger demand. But I agree that in general there will be more passenger demand between a 6 bar domestic airport and the capital, than between a 6 bar foreign airport and (in this case) Athens.

...

Another example: I run an airline in Brazil and I tried a route that IRL is operated daily by a 757, from a deep secondary airport (6 bars). In my case I was the "ONLY" flight in ORS at all and I generated exactly 2 pax per day.

...

Again, I don't think you generate passengers. And if there are no bookings, there is no demand for that route.

AS uses real life data. But not every real flight is reflected in the game. And some routes may be newer than the data used by the game. I have also tried routes that are flown in real life, but had no demand in the game. As a general rule however, if there are several daily flights for real, you can expect passenger demand in the game.

Allow me to end with a disclaimer. All of the above is guessed. I only prefer my view above yours because mine is... simpler  ;-)

Jan

11

So, to put things together, will changing my hub make any difference, assuming that I keep all my existing connections ?

12

Hi,

it will make a difference...

Passengers who fly from Beirut to Hamburg don't care if they fly via Athens or via Heraklion (if the total travel time is about the same). These passengers would "move" to your new hub.

Passengers who fly from Beirut to Athens, and passengers who fly from Athens to Hamburg will still fly to/from Athens. You would loose these passengers.

However, perhaps there are also passengers from Beirut who want to fly to Heraklion, and passengers who want to fly from Heraklion to Hamburg. You would gain these passengers  ;-)

I only gave an example, but I hope you get my point. For some routes it won't make a difference, for many routes it will make a difference. And if you move your hub to an airport with less passenger demand, I guess on the whole you will loose passengers.

Jan

13

@Jan - I wrote my assumptions (about demand) based on airport pax data and number of arrivals/departures as a basis for AS pax demand calculation. Your assumption goes a step further, assuming that besides what I assume there are also actual route pairs taken into consideration.

E.g. I assume airport ABC has 1.000.000 pax and 22 daily aircraft movements. You assume the same but also assume there are 8 aircraft movements related to ABC-DEF route, 4 movements related to ABC-GHI route, 8 movements related to ABC-JKL route and 4 movements related to ABC-MNO route. Meaning... your assumption goes a bit further than mine.

All I wrote above, is based on NOT knowing the exact demand pairs, and that demand pairs are spread out evenly (6 bar A is equal to 6 bar B, etc.). You assume that the demand pairs ARE known exactly. Basically, that is the only difference in our assumptions, everything else remains the same.

P.S. Was it ever mentioned in the forums, by AS team, that demand pairs copy real life, or only that airport data (number of pax and distribution domestic/international) are calculated? Was it ever discussed a bit more deeply on the forums in the past?

14

So, to put things together, will changing my hub make any difference, assuming that I keep all my existing connections ?

Well, as Jan said, I think on the overall you will loose passengers, UNLESS most of your actual passengers are CONNECTING pax.

You can find stats by going into offices tab, selecting ATHENS, and there you will see % pf pax that are direct and % that are connecting.

If 80% or so are connecting it could make sense to move the hub. If, on the other hand, 80 % are direct pax, by moving hub you will lose those pax.

You may gain some direct pax for your new hub, but it will not be so many as for Athens.

15

@Jan - I wrote my assumptions (about demand) based on airport pax data and number of arrivals/departures as a basis for AS pax demand calculation. Your assumption goes a step further, assuming that besides what I assume there are also actual route pairs taken into consideration.

E.g. I assume airport ABC has 1.000.000 pax and 22 daily aircraft movements. You assume the same but also assume there are 8 aircraft movements related to ABC-DEF route, 4 movements related to ABC-GHI route, 8 movements related to ABC-JKL route and 4 movements related to ABC-MNO route. Meaning... your assumption goes a bit further than mine.

All I wrote above, is based on NOT knowing the exact demand pairs, and that demand pairs are spread out evenly (6 bar A is equal to 6 bar B, etc.). You assume that the demand pairs ARE known exactly. Basically, that is the only difference in our assumptions, everything else remains the same.

P.S. Was it ever mentioned in the forums, by AS team, that demand pairs copy real life, or only that airport data (number of pax and distribution domestic/international) are calculated? Was it ever discussed a bit more deeply on the forums in the past?

Hi,

I am afraid I was not clear enough in explaining my thoughts.

I don't think that aircraft movements - direct flights or pairs of flights - have an influence on passenger demand. I think every airport has a fixed list of destinations, and a fixed number of passengers for every destination. Some of these destinations & passengers are based on actual airport-to-airport traffic. Some of these destinations & passengers are based on actual traffic in a nearby airport (regional demand).

I gave an imaginary example. If 100 passengers fly from ATH to HAM, the AS-team may decide to change this and put some of these passengers in Heraklion and Rhodos (as regional passenger demand). In real life Rhodos airport may have 20 destinations. In the game Rhodos may have 30 destinations, of which 10 destinations have smaller numbers of passengers. These added destinations & passengers are based on real life traffic in nearby airports.

It is not important if we operate one or ten daily flights between RHO and ATH, and between ATH and HAM.

If there are 10 passengers in Rhodos who want to fly to Hamburg (as part of regional passenger demand) they will look for eligible routes to Hamburg. Could be a domestic flight to Athens and a connecting flight to Hamburg. But it could also be a flight to Manchester and a connecting flight to Hamburg.

So in my opinion, our flights do not generate passenger demand. Passenger demand is in the AS data base, and the ORS looks for possible routes to get there. I guess that even small airports who in real life only connect to their capital, in the game have small numbers of (added "regional") passengers who fly to various destinations.

A new airline with a few domestic flights, will be able to fill one Dash per day to such a small airport. But if that airline grows and offers the right connections, he may be able to fill three Dashes per day to that same small airport. Did the airline generate passenger demand ? No, the passengers were there. But without a possible route to their destination, they simply stayed at home  ;-)

As for numbers... airport A with 6 bars will probably have (more or less) the same number of passengers as airport B with also 6 bars. But that does not mean there will be demand between these two airports. Passengers do not fly because we offer a flight. Passengers book a seat if our flight brings them where they want to be.

AS passengers don't say "I am from the Bavaria/Bayern region and I want to fly to one of the Greek islands". Martin and/or Sascha have added passengers to precise airports and have given them exact destinations.

Passenger demand has been discussed many times. But mostly by players like you and me. People who try to figure it out and guess  :-)

If I remember correctly, the only "official" explanation is that passenger demand is based on real life airport to airport traffic, that the data are not always perfectly up to date, and that there is also regional passenger demand. Oh, and passenger demand is equal in both directions.

Cheers,

Jan

16

Jan,

Maybe you misunderstood me, ... I never suggested that AS airlines/connections "generates" traffic ... that would be too good to be true

If LCCs were supported (which officially was stated they are not) than they would be able to generate traffic. Because there is no traffic generation, they are not supported (yet). But with certain statements made by AS we might expect "passengers with a mind" or pax demand based on price elasticity.

What I actually wanted to say with my post is that I thought the demand was generated by airport type and location combination. I never meant to suggest that AS airlines can generate demand (and if somebody understood it that way that was not what I implied so please do not take such idea into consideration).

What I meant was something like that:

Airport X... has 10 million pax per year in real life.

80% domestic, 20% international.

Now we know AS demand is "more" than IRL demand in static pax numbers.

So let's say this airport X has 20 million pax per year demand in-game, and same domestic/international split.

What I suggested was that in such vase 80%@20MM=16MM would fly domestic, and 4MM would fly international.

Those 16MM domestic would be distributed among domestic airports based on bar-size (8 bar would get proportionally more pax than 7 bar; 5 bar would get proportionally more than  bar, etc.)

In my supposition, this is how those 16MM domestic pax would be distributed. Same would go for international, if we know that 50$ of int'l fly to Canada, for example, those 2MM would be proportionally distributed among Canada airports base don bar-size.

When I said your supposition went a step further, I meant that you suppose that if we know that (e.g. above Canada int'l pax example) 90% of pax IRL travel to Toronto (because e.g. IRL Airport X to YYZ is a trunk route), that such route would have 90% demand of X-to-Canada in AS (or approximately same).

So the only difference between your and my supposition si that you put a specific route pair into equation, while I put airport size.

So let's say airport A is IAD.

In my supposition, Toronto would get the highest proportion of pax, followed by Vancouver and Quebec. E.g. Toronto could fill 5 flights, and Vancouver and Quebec could fill 2 flights each

In your supposition, Toronto would get most of the pax, because InRL 90% of flights are via YYZ, so in AS Toronto would fill 7 flights Vacouver and Quebec only 1,

That is the only difference in our suppositions, everything else is same.

17

Thanks for explaining again. And yes, for most parts we think along the same lines.

I agree that in-game passenger demand seems to be higher than real life demand.

And sure, domestic traffic is probably based on airport size.

As for traffic out of (for example) IAD... your assumption could well be correct.

I only guess that real life traffic between IAD and YYZ is reflected in the game, with the addition of regional traffic. But even then, we don't know if Toronto region means airports in Ottawa or all Canadian airports. Nor do we know the percentage of passengers that flies the "real" route and the percentage that flies to airports in the region.

Isn't this a fun game ?

:P

Jan