Online Airline Management Simulation
or login using:
My Account
Edit account
» Achievements
» Logout
Game Credits
Credit balance: 0 Cr
Buy credits
» Credit history
Main Menu

Weekly based demand

Started by Shubinine, December 07, 2010, 07:21:37 PM


I have searched if this has been brought up before but I couldn't find anything so if it has alredy been I'm sorry...

I was thinking of the idea that this way we can fly (let's say) only once a week to an airport and still get the plane filled up...
I still think that the graphic of dayly demand is good as there are lods of routes where the demand is too big for.. you know.. only once a week, but what if you make a separate graphic, in which the demand for the whole week appears, as I'm thinking that, aswell as it happens in real, if you really need to get to a city and there's only 1 airline making a 1/week flight to your destination, you gotta take that flight on the day it happens rather than not going at all...

If your airline operates a 737-NG only fleet, you can't fly to an airport where the avg. daily demand is 40 passengers, but if the week total is 280-something you can easily send a 737 one day and a 738 on another, as the total seats you send is ~ the weekly total.

So in this case, both the company and the passengers are happy...


This is a good idea; however not entirely in this proposed form.

I completely agree with its implementation; however i think the demand (no matter how small) needs to be spread out over at least 3 services a week.
Otherwise some airlines could run a single service a week and still get great loads; from a marketing and RI perspective i think this is most valid (ie: most people in reality wouldn't bother shifting their travel plans if only one service a week is offered. Ooooh! another idea! what about with the more services offered throughout the week, the higher percentage of passengers come across from days without a service?)

But yes, it is true that in the real world passengers come across from days that services are not operated, onto days that a service is operated. However not all passengers do; maybe a fixed percentage of say 70% would theoretically move their travel over onto a day with a service (that increases as more services in the given week are added)


The idea is quite good, and something I've often thought of myself.
Although, I think certain criteria needs to be applied to avoid some of the problems already mentioned.

Longhaul pax for instance, are less sensitive on frequencies. The same with holliday travelers.


Something like this would make small airports viable.  As for pax demand for more frequency, this is already modeled in the system.


You might want to direct your attention to this thread from just a few days ago where the same thing was requested (as it has been numerous times):,27059.0.html


I disagree, this makes the established carriers with big jets far too competative on smaller routes (assuming usable runways). As a smaller airline run by a new player a few years into the game, I depend on smaller demand routes to have any decent route offerings.

The only thing I might support would be a one day roll-over, i.e. if no pax are flown one day they carry to the secong - but no further.  Then you'd have to fly 3 or 4 times a week, but it may make the very small demand routes possible. 


Please discuss the 'weekly demand' subject ..

In other words, if we have a route that is not flown every day, how should the passengers behave. I would say; simply so that certain percentage of the demand is willing to fly on another day if there are no flights on that particular day at all. So let's say 50% of monday's people could move to tuesday if they have a flight then.

Wouldn't see the need to shift more than 1 day back/forth since if we look at real world usually most destinations are served at least 3x weekly, less than that is rare (well 2x weekly still to some leisure destinations is common..).

This would help operating on small routes and on thin longhaulers.


I know several flights out of Tier2/Tier3 airports here in Germany depart just twice or thrice a week - so maybe extend the waiting time to 48 hours instead of 24 hours? And 75% of Economy demand but just 15% of business demand, while first class doesn't wait at all?


I guess the simplest way to do achieve it, without maintaining any sort of accumulation is to just look at the previous day flights, figuring out the number of pax not flown the day before, and a fraction of those pax (not flown the day before) would be added to present day demand.

The combined demand of "today" would not be stored anywhere, the history would not have to be kept for thousands of route combinations.  It would just be one day look back.

So if we start with no pending pax, 80 pax daily demand, 100 pax a/c every other day:
Day 1: New demand (80) + 50% of previous day (0) - pax flown (0: no flights) = unserved demand from today's demand (80)
Day 2: New demand (80) + 50% of previous day unserved demand (40) - pax flown from todays demand (80) - pax flown from previous day (20: aircraft full) = unserved demand from today's demand (0)

This way, there is no need to store any more data than the system already has, all it would do is look back 1 day.


It would be a major overhaul of the way of thinking lines, though. Real life example : WIZZAIR flys a A320 between Beauvais-Tillé and Wrocław twice a week. That's a low-cost flight. It's nearly full during holidays, and rather full off-season, so probably 250/300 pax per week. In 2014. That's 40 pax/day, roughly

In game terms, even a CRJ100 would be overdimensioned. Most people flying this plane are bound to take it anyways, as the only alternative is 17 hours of bus. In GW3, it currently has 14 passengers per day - 1999. But even with such a low demand, a CRJ100 twice a week could fdo the trick.

I completely agree with the idea of separating classes, though. First class flyers would never wait. Business really not much. Economic have no choice, and will push back their flight for a few days.

Could be an accumulation. Let's stay on this line. Demand is, from monday to sunday :
15-13-13-14-15-12-12 - that's unflyable.
Monday ; 15, no flight
Tuesday ; (15*0.75)12 + 13 = 25, no flight
Wednesday ;  (25*0.75)18 + 13 = 31, flight
Thursday ; 14
Friday ; (14*0.75)10 + 15 = 25
Saturday ; (25*0.75)18 + 12 = 30
Sunday ; (30*0.75)23 + 12 = 35, flight

That's flights with 31 & 35 of potential, Meat for a small Embraer, and the line is now playable. And somewhat realist. you go down from 94 to 66 potential pax, but at least you make some money.
count 25% instead of 75% for Business, 10% for First class, and you've got something interesting.

For simplicity's sake, it would go back until the previous day with ANY flight, not counting excessive passengers. For example, even if my plane is a 30-seater, it would still begin at zero on mondays and thursdays, on my example.


I think flying only a couple times a week is a lot more common than you think sami.  Allegiant in the US flies nearly all of their routes only 2x or 3x/week and fly with 90%+ load factors systemwide.

In a true economic sense, demand always exists, but it is quantities demanded that change in response to pricing and supply.  I'm a dude and if pantyhose goes on sale, I don't run out and buy it because it is on sale because my demand is zero.  However, if you're a woman your pantyhose demand is always there, but your quantities demanded fluctuate based on price and supply.

Likewise in AWS the demand from Taipei to Hong Kong is ~8000 pax/day.  If there are no airlines flying the route, the demand is still ~8000 pax/day, they just take the ferry or otherwise to make it happen or if there are no alternatives they stay home, but they still have demand to travel the route.  Since flying 2x/week is much more common than you think and pax will often shift their travel plans by a few days to accommodate travel schedules, you could simply add on top of the existing demand graphs/structure a value called "shifted demand".

The daily demand would stay fixed at 100%, but if the route is not flown daily, the adjacent days would shift their demand.  The scale would be 100% (day flown), 50% (+/- 1 day), 25% (+/- 2 days).  This would allow a player flying 1x/week to capture 100% of the demand on that day, 50% 1 day in each direction, and 25% 2 days in each direction.  So if demand were:


Then a player flying the route only on Wed would look like this:

total demand302010020304040
shifted demand-10-20+60-20-1000

So if you flew a route 3x/week, you could virtually capture 100% of the weekly demand:

total demand208008008020
shifted demand-20+40-40+40-40+40-20

This scheme would also mean that if you fly a route 7x/week and you go in for a B-check and don't fly the route on any day of the week, 100% of the unserved demand will shift to the adjacent days, which I think we would all like :)  Of course, this would only happen to routes that were underserved on a given day, so not every route would see this when B-checks happen.


I'm not sure to have entirely understood the LemonButt's preamble  :)

But I agree with the conclusions. +50% from adjacent days without flight and 25% from +/- 2 days.

I also propose to don't considered the day as with or without flight but as "exceeding demand". So all the exceeding demand (because there are no flights or too small plane) will be shifted to adjacent days. I'm not sure if it worth but could be nice to have.


I completely agree with this, people rather have frequency but they will change or plan around flight times when there is no alternative they are willing to take (other airline has bad CI or there are no other airlines). This could for example allow for up gauges of lesser demand routes but at lower frequency.

I would add however that for long haul routes this dynamic be different then for short haul routes. Just have different brackets, thus having lower penalties for very long haul routes in such a way that they could even fly 1x Weekly (eg: CDG:PPT), the penalty would then be over 4 days perhaps?   


Splitting demand into discrete timeframes, such that pax will take that timeframe if offered, but take one up to a week away if their preferred one isn't, will also help with forcing people to put more effort into scheduling, rather than cluster routes as tightly as the current restrictions allow. Because once the programming is there for Tuesday's pax to take a Monday or Thursday flight, the programming is also there for evening peak, 5-8pm pax to take that if offered, and only take a 10am flight if their preferred time isn't there. Smaller routes & airports suddenly become much more interesting, particularly smaller LH airports, e.g. Jomo Kenyatta, and bigger routes & airports become a bit less mindless, scheduling takes a bit more effort. People also have more options for 7-day schedules, too. Instead of only in groups of 7, could do a group of 3 LH planes, overall flying 3 times/week to 5 different destinations. I've made a long post about it somewhere in the past, will have to try and find it...




Six years later. Was this ever seriously considered?