AirwaySim
Online Airline Management Simulation
Login
Username
Password
 
or login using:
 
My Account
Username:
E-mail:
Edit account
» Achievements
» Logout
Game Credits
Credit balance: 0 Cr
Buy credits
» Credit history
» Credits FAQ

Author Topic: Long-term playability brainstorming  (Read 8672 times)

Ansettaddict123

  • Former member
Re: Long-term playability brainstorming
« Reply #40 on: July 13, 2010, 11:34:24 AM »
+1!!

I think its a great idea:)

Offline DarkVenegance

  • Members
  • Posts: 19
Re: Long-term playability brainstorming
« Reply #41 on: July 13, 2010, 11:46:03 AM »
Transfers would definitely be good, they open up a new dimension of playing by allowing everyone to grow to a big airline regardless of whether they are based at an airport that's just big because a real world airline is hubbing there. And it opens up the game up for code sharing and other real alliance advantages.

I do realise that implementing this is difficult, but I think it is possible without changing big parts of the game engine. Basically, there is demand for lots of A-B relations and a network of flights that connects these destinations, associated with various costs for each flight (duration, price ...) and between each flight (connection time, same airline ...). This is something called a Multi-commodity flow problem. Now it is possible to approximately solve this problem for example once a week and map the result to the demand between each airport pair. This way it is possible to have demand based on connections without having to model the actual connection each and every time.

Online JumboShrimp

  • Members
  • Posts: 7916
Re: Long-term playability brainstorming
« Reply #42 on: July 13, 2010, 02:16:27 PM »
Transfers would definitely be good, they open up a new dimension of playing by allowing everyone to grow to a big airline regardless of whether they are based at an airport that's just big because a real world airline is hubbing there. And it opens up the game up for code sharing and other real alliance advantages.

Yup.  All the airports would be more or less equal, and an active player starting in not so big airport with many routes could make the airport very busy with a lot of traffic going through.  The only limitation would be the starting number of slots.  Airports could grow dynamically based on percentage of slots sold, rather than at constant pace.

I do realise that implementing this is difficult, but I think it is possible without changing big parts of the game engine. Basically, there is demand for lots of A-B relations and a network of flights that connects these destinations, associated with various costs for each flight (duration, price ...) and between each flight (connection time, same airline ...). This is something called a Multi-commodity flow problem. Now it is possible to approximately solve this problem for example once a week and map the result to the demand between each airport pair. This way it is possible to have demand based on connections without having to model the actual connection each and every time.

This could be really simple.  Each airport would have a numeric value assigned to it based on, say population surrounding the airport * GDP per person in the country.  Demand between any AB pairs could be determined by a function:
#1 - of these numerical values for each airport
#2 - distance (Very short distance would be 0, then sloping up up to certain distance, then sloping down)
#3 - country of the pairs of airports.  The factor could be, say 1.0 for the same country, 0.75 for EU / NAFTA, 0.5 for unrelated countries  There could even be a Warsaw Pact grouping in place up to 1989, which would then be disbanded.

The recalculation of inherent demand between each pair could take place even less frequently than once per week you mentioned.  It could be just once per year.  Maybe the system could have a population growth figures and GDP growth figures for the countries.  So, say on January 1st, population growth and GDP growth figures would adjust numerical value of each airport (component #1 above of the formula above).

All of this would would not only make the simulation more realistic, but also the game to be more true multi-player.


Filippo

  • Former member
Re: Long-term playability brainstorming
« Reply #43 on: July 13, 2010, 03:02:47 PM »
I understand that the one aircraft a month rule has been introduced to make the game less boring.

Well, I remember how playing V1.1 was without this rule and I can say that it was much more interesting :P. Personally, I am getting bored of receiving just one aircraft. Apart from the fact that this feature is not even realistic, I understand that we are trying to stop boredom, not trying to create it!

For the future, I would say that updating this aircraft delivery system to a realistic standard (4 small eg. A320, 737 a/c a month, 2 big eg. A340, 777 a/c a month - depending on how many you ordered maybe, or you could even choose how many per month to have as real airlines do) and having connecting traffic are the top priorities.

Offline DarkVenegance

  • Members
  • Posts: 19
Re: Long-term playability brainstorming
« Reply #44 on: July 13, 2010, 06:26:24 PM »
This could be really simple.  Each airport would have a numeric value assigned to it based on, say population surrounding the airport * GDP per person in the country.  Demand between any AB pairs could be determined by a function:
...

Yes, but that has nothing to do with connections. What I described is a way to map the A-B demand to the various flown routes, e.g. A-C-B, A-D-B, etc. The pure demand A-B is already calculated some way and does grow. No idea if sami is using a formular, database or whatever for that, though.

Online JumboShrimp

  • Members
  • Posts: 7916
Re: Long-term playability brainstorming
« Reply #45 on: July 13, 2010, 09:49:21 PM »
Yes, but that has nothing to do with connections. What I described is a way to map the A-B demand to the various flown routes, e.g. A-C-B, A-D-B, etc. The pure demand A-B is already calculated some way and does grow. No idea if sami is using a formular, database or whatever for that, though.

But the model for demand of all A-B connections in place in the game is "wrong".

Well, they may very well be a good snapshot of reality, but that's wrong way to go about it.  Today's snapshot is a reflection of how major airlines built their hubs.  Let me give you an example:  Suppose Delta decided to build their hub in Philadelphia rather than Atlanta, and flew a ton of flights from Philadelphia to just about every other place on the earth.  Which airport do you thing would be the world's bussiest?  Atlanta or Philadelphia?

If you look at local demand, Georgians flying from Atlanta, they are only a tiny fraction of people who go through Atlanta.  Philadelphia is a larger metropolitan area, with larger local demand compared to Atlanta.  Whether Atlanta or Philadelphia (or any other airport) becomes the world busiest airport should depend on where airlines (in the game) form their hubs and how many passengers they fly through these hubs.

Let me say it differently:  Suppose a person from Savannah, GA wants to go to Sacramento, CA:

A person flying from from Savannah Georgia to, say Sacramento, CA would take a Delta flight from Savannah to Atlanta, transfer there to another Delta flight from Atlanga, GA to Sacramento, CA (SMF).  If Delta hub was in Philadelphia, the same person would fly through Philadelphia instead.

Now here is why the demand model is wrong (using ATB demand), extreme example of how wrong it is:
SMF-ATL: 2000
SMF-PHL: 310

The way it should be modeled is (approximately):
SMF-ATL: 120
SMF-PHL: 150

reflecting local demand.  Now if DarkVengence Air builds a huge hub in Philadelphia, is flying route pairs to every other airport in the North East and PHL, DarkVengence Air can create a demand of 2000 between SMF and PHL, but it will be a sum of 150 local demand + 1850 demand from connecting flights.

Curse

  • Former member
Re: Long-term playability brainstorming
« Reply #46 on: July 13, 2010, 10:03:55 PM »
Very nice idea!

In addition to that, airlines should buy/build new terminals/runways and get slots from this, because the slot situation ingame is like the slot situation in rl, like DarkVenegance said in the example with Delta and Atlanta/Philadelphia.

Online JumboShrimp

  • Members
  • Posts: 7916
Re: Long-term playability brainstorming
« Reply #47 on: July 13, 2010, 10:24:42 PM »
Very nice idea!

In addition to that, airlines should buy/build new terminals/runways and get slots from this, because the slot situation ingame is like the slot situation in rl, like DarkVenegance said in the example with Delta and Atlanta/Philadelphia.

Terminals would be a nice extra, something along the lines of lease space from the airport vs. buy/build your own terminal.  Building own terminal would cost upfront money, but long term operating expense would be less than leasing space...

But terminals do not equal to slots.  Slots correspond landing / takeoff capacity basically runways.

With slots, when out of slots, player could be prompted to pay to expand the capacity, by say $50 mil (inflation adjusted) to get additional 15 slots / hr, with the player paying for the expansion would be automatically allocated, say half of those slots for a period of time and the other half would be generally available...

Sigma

  • Former member
Re: Long-term playability brainstorming
« Reply #48 on: July 13, 2010, 10:26:00 PM »
JumboShrimp, the exact system that you describe (with proper demand between routes, not recreating the hub-spoke system of reality that we have today) is something that sami already has in the works, and it's been discussed quite a bit (just not lately).  It's quite a bit more complicated than you describe, but the basic premise is exactly the same.

Filippo

  • Former member
Re: Long-term playability brainstorming
« Reply #49 on: July 13, 2010, 11:42:29 PM »
By the way, i was trying to convince a friend to play AWS, but when i told him about the slow delivery dates he changed his mind! Hahahaha

Online JumboShrimp

  • Members
  • Posts: 7916
Re: Long-term playability brainstorming
« Reply #50 on: July 14, 2010, 12:14:19 AM »
JumboShrimp, the exact system that you describe (with proper demand between routes, not recreating the hub-spoke system of reality that we have today) is something that sami already has in the works, and it's been discussed quite a bit (just not lately).  It's quite a bit more complicated than you describe, but the basic premise is exactly the same.

Excellent.  I am looking forward to it.  I would definitely be interested in alpha / beta testing when it gets to that stage.

Offline DarkVenegance

  • Members
  • Posts: 19
Re: Long-term playability brainstorming
« Reply #51 on: July 14, 2010, 01:39:24 PM »
But the model for demand of all A-B connections in place in the game is "wrong".
Of course, it is just a different problem. I guess there's some database with route demand for the simulation and as such the demand could be changed right now to different values without changing the actual code. For that, your formula probably makes sense, but might need to be extended to handle multiple airports in one city.

Distributing that demand is the hard part, because it requires writing new code. I just sketched a way to do it that allows to reuse as much existing code as possible (like the LF calculations for each flight etc.).

To make hubbing really work airport slot growth also needs to be changed, but I don't know which way is best for that (big one-time cost for expansion, high slot cost, automatic growth or whatever).
« Last Edit: July 14, 2010, 01:42:50 PM by DarkVenegance »

Filippo

  • Former member
Re: Long-term playability brainstorming
« Reply #52 on: July 14, 2010, 01:51:28 PM »
I had this idea:

what if you have to pay for your extra delivery slots? Say you order 12 737-700s. You could receive them in 12 months for no extra charge (this sounds very easujet  ). But, you want them in 3 months. So you buy the extra delivery slot from the manufacturer or even other airlines (give a maximum of 4 a month  for small a/c and 2 a month for big a/c to stop airlines buying entire production lines!). The cost of the slot would be determined by the number of aircraft ordered of that production line.

This would allow everybody to be happy as you could receive aircraft in the way you want (fast or slow), it could be an investment as a production slot would rise or decrease in value to add a new dimension to the game and it could help small airlines to grow faster either by getting planes faster or getting good cash because of their valuable production slots!

What do you think?

Filippo

  • Former member
Re: Long-term playability brainstorming
« Reply #53 on: July 14, 2010, 05:38:02 PM »
There's a nice thread on subsidiary lines here in the requests forum

Offline Sami

  • Administrator
  • Members
  • Posts: 16786
    • AirwaySim - Are you the next Richard Branson?
Re: Long-term playability brainstorming
« Reply #54 on: July 14, 2010, 09:54:01 PM »
But the model for demand of all A-B connections in place in the game is "wrong".

Yes, the model is based loosely on real hub systems and reflects what you'd see point-to-point with present hubs. That is "wrong" indeed and plan is to make the regional demand system. It's of course a two sided thing as regional demand and flight connections sort of go together, so I'd have to think if these could be implemented simultaneously (but that will cause quite a bit of "uuuufff" since it would be a huge update). But have to think this with a pint someday...  :P


By the way, i was trying to convince a friend to play AWS, but when i told him about the slow delivery dates he changed his mind! Hahahaha

Sorry but please don't lobby this on every possible thread. This is unrelated to this topic too.

Online JumboShrimp

  • Members
  • Posts: 7916
Re: Long-term playability brainstorming
« Reply #55 on: July 14, 2010, 11:16:22 PM »
Yes, the model is based loosely on real hub systems and reflects what you'd see point-to-point with present hubs. That is "wrong" indeed and plan is to make the regional demand system. It's of course a two sided thing as regional demand and flight connections sort of go together, so I'd have to think if these could be implemented simultaneously (but that will cause quite a bit of "uuuufff" since it would be a huge update). But have to think this with a pint someday...  :P

Yes, I think the 2 (regional demand and flight connections) would have to be implemented simultaneously.  Changing demand to regional only - meaning demand between any A-B pair to reflect only demand of the city / region surrounding the airport - that probably could be done with no code changes.  It could probably be done by updating your set-up databases to the new figures.  But a slightly better system, which I described above might need some programming, but it would be a more elegant solution.  Let me copy it here:

------------------------------------------------------
This could be really simple.  Each airport would have a numeric value assigned to it based on, say population surrounding the airport * GDP per person in the country.  Demand between any AB pairs could be determined by a function:
#1 - of these numerical values for each airport
#2 - distance (Very short distance would be 0, then sloping up up to certain distance, then sloping down)
#3 - country of the pairs of airports.  The factor could be, say 1.0 for the same country, 0.75 for EU / NAFTA, 0.5 for unrelated countries  There could even be a Warsaw Pact grouping in place up to 1989, which would then be disbanded.
#4 - vacation destination - would go into #1.  So #1 would be local demand based on local population + a vacation demand component.
-------------------------------------------------------

To implement this, the only values that would have to be obtained and populate into setup databases:
1 - population of the region surrounding the airport (a lot of figures, but the only really time consuming thing)
2 - population growth of the country (good enough proxy for population of the cities).  2 snapshots for population figures would be needed to obtain this for every country.  But only 1 value per country
3 - GDP per person.  1 per country
4 - GDP growth for the country, again only 1 figure per country.  To add a twist to this, during the actual running of the game, this could be non-linear to simulate economic boom / bust cycles.

None of these demand figures would ever be zero.  Many may be tiny, but none zero.  For example, as is, in the game, demand between LHR and EIS (Tortola, British Virgin Islands) is zero.  It does reflect the fact that there are no flights between these airports, but there are definitely quite a few British subjects that go back and forth between those 2 airports.  If someone is operating LHR - SJU (San Juan, Puerto Rico) route and SJU - EIS route, he should get some (most) of this traffic.

From these 4 set of figures above, any A-B demand can be calculated for any timeframe of any game scenarios.  Updating this once per year during the running of the game should be sufficient.  This figure would be (more or less) static, updated once per year.

Then there would be the actual demand for the player for a route.  This would be a combination of
1. static component (regional demand between A-B pair)
2. dynamic component - connecting traffic based on the existing set of routes the player is operating that may add traffic to A-B (for example X-A, Y-A, B-M, B-N) plus connections from other airlines the player is willing to share connecting traffic with.  For this dynamic demand, every player should see a different figure. (I would make connecting between Player 1 flight and Player 2 flight conditional on whether these 2 players are willing to share connecting traffic.  This will make the game more exciting and competitive).

The realized demand (actual tickets purcased) should have enhanced elasticity (greatly enhaced).  Time to go from A to B, number of connections needed to get there, price, seating quality etc should greatly influence on whether 5 of 25 potential passengers (demand) or 23 out of 25 passengers purchase the tickets.

If someone needs to go to their child's wedding, that person will pay any price, endure any hardship to get there.  OTOH, going on a vacation should greatly depend on low price.  It should also be important for player competition.  If Player 1 is flying A-C-B and Player 2 is flying A-D-B, the player offering better value should get a lot more of the A-B traffic.

One more thing on the demand:  Let's say we come up with demand figure of 300 for an A-B pair.  What should this figure represent?  It could represent:
- cap - there can never be more people flying even if the ticket price is $1.  So the realized demand would always be a lot less
- average - demand based on default price and average level of accommodation.  If the airline has a direct flight in a decent a/c and has 30% discount, this 300 demand could turn into 400
« Last Edit: July 14, 2010, 11:24:09 PM by JumboShrimp »

Offline Minto Typhoon

  • Members
  • Posts: 1035
Re: Long-term playability brainstorming
« Reply #56 on: July 15, 2010, 07:22:59 AM »

Don't forget that demand can be stimulated from close to zero.  Ryanair's fare structure allows them to significantly stimulate demand on a city pair from a low base by basically building a new reason for travel - 'its cheap' - if an airline can be profitable with fares 50% below suggested, they should be able to fill 100 seats on a 10 demand route.

Online JumboShrimp

  • Members
  • Posts: 7916
Re: Long-term playability brainstorming
« Reply #57 on: July 15, 2010, 07:57:26 AM »
Don't forget that demand can be stimulated from close to zero.  Ryanair's fare structure allows them to significantly stimulate demand on a city pair from a low base by basically building a new reason for travel - 'its cheap' - if an airline can be profitable with fares 50% below suggested, they should be able to fill 100 seats on a 10 demand route.

Yeah, exactly.  That's why I brought up a concept of demand being "Average" - average aircraft, default prices.  At lower prices new demand can be created.

Online MuzhikRB

  • Members
  • Posts: 902
Re: Long-term playability brainstorming
« Reply #58 on: January 04, 2017, 01:51:17 PM »
Hello community. Sorry for long post.
Cannot start new topic so I searched for something related to.

I want to propose and discuss the idea of «basic» stock system.

Purpose:

1. Initiate stocks system in AWS.
2. Slow down growth of Multi-Billion companies.
3. Add random seeds in gameplay at medium and later stages of GWs. Increase unpredictedness.

Note: «Basic» system means there will be NO stock markets and NO possibility to buy out other companies.

System definition:

Currently new player starts with the following balance sheet (downside):
Shareholders' equity   
Common stocks   
Retained earnings / accumulated deficit

Some digits initially written in Common Stocks. Example: In GW4 my company received 2 000 009 USD. (started in 1970).
This is our initial capital.
After it it doesnt increased and all profit accumulated in «Retained earnings / accumulated deficit».

My proposal is not to invent something new but start to use this numbers for purpose.

I define 3 life stages for every company:

Stage 1. Initial growth.

Stage length 5 years. From day 1 till end of Calendar Year 5. If company started in 1970 (any day) - first stage will end at 31 December 1974.
During this stage nothing will be changed from current game mechanic.

Stage 2. Investment returning.

Stage length - 5 calendar years Or when company reached A bank rating.
It means if after stage 1 company already have A credit rating it will skip Stage 2 and goes directly to stage 3.
If company starts Stage 2 and during year 2 reached A bank rating then it will go to stage 3 after calendar year 2 is finished.

During this stage company start to pay dividends to its initial owners only.
Dividends is paid every Q.
Dividends amount (DA) is calculated by formula:

DA = Last Q profit * (Common stocks/Shareholders Equity) * Dividend Payout Ratio (DPR) (will be described later)

Stage 3. Public company.

Stage length 10 calendar years from finishing Stage 2.

At this stage company will slowly transferred to public company.
How? Based on the following formula:
Last Q profit + 2,5% of Retained earnings will be added to Common Stocks string.
Retained Earnings will be decreased correspondingly by 2,5%.
And this is will be done every Q during Stage 3

After it company will pay out dividend by mentioned above formula (in Stage 2).
Stage 3 purpose is to slowly increase dividend pay outs level for the company not making it like one time shock, so player can adjust to it.
Finally Share Holders equity will be 100% in Common Stocks.

DPR definition.
the most interesting thing is DPR value.
I propose the following:
DPR is individual for every company and in public access.
DPR is defined each year (January 1st) for the upcoming year.
DPR formula = random seed (5%-30%) + last year company Profit Margin (%) = DPR %.
Why to add company Margin ? to level down different profit tax influence.
So company in UAE (0%) will not have advantage before USA (30%).

What it will provide to GW.
1. It will remove extra money from the companies. it is simply additional tax.
2. because DPR is individual it will add another level to competition factor between companies.
3. because DPR is changed randomly every year it will decrease stability level in GWs therefore add more interest to long-term strategy.
4. Based on this we can start to develop «advanced» stock market features in the future.
« Last Edit: January 04, 2017, 01:54:29 PM by MuzhikRB »

Offline Springbal

  • Members
  • Posts: 241
Re: Long-term playability brainstorming
« Reply #59 on: January 12, 2017, 04:10:09 PM »
Some points may have been mentionned, but my lack of concentration sometimes prevents me keeping it all in mind.

About Aircrafts:
- make new seatings possible like Brussels Airlines/Lufthansa update (I tought due to new seating design there was 4cm more legroom)
- Charter flights: you can organise flights for eg a touroperator, sportsleague, big congress, sport or culturalevent
- Adaptive scheduele: having an AC that is not doing regular flights but is in reserve for the charter market

About Airports:
- Allow lounges and bars (eg lounge for C and F, and a bar for Economy)

About Alliances:
- Allow group purchases (e.g. Company A buys 1 A330-300, Company B buys 2 A330-200 and 1 A330-300, Company C buys 1 A330-200 and Company D buys 2 A330-200 abd 2 A330-300. Few discounts, but when They are part of an alliance the alliance (or one player) can place the order of 4 A330-300 and 5 A330-200 and have more chance of a discount) --> might be already but I havn't played enough in alliances
- Code Sharing
- Alliance marketing

About flights:
- connective flights, eventualy with your alliance partners
- inflight service

I know not all these things are possible but it is a brainstorm he :)

 

 

WARNING! This website is not compatible with the old version of Internet Explorer you are using.

If you are using the latest version please turn OFF the compatibility mode.