![]() |
|
Demand formula - Printable Version +- Airline Enterprise - Forums (https://www.airlineenterprise.com/forums) +-- Forum: Airline Enterprise (https://www.airlineenterprise.com/forums/forumdisplay.php?fid=1) +--- Forum: Game Updates (https://www.airlineenterprise.com/forums/forumdisplay.php?fid=2) +--- Thread: Demand formula (/showthread.php?tid=80) |
Demand formula - Unknown98 - 03-01-2017 ![]() As you can see by my so elegantly edited screenshot, I've implemented (albeit a rather simple) demand formula. So now flights can actually have passengers, revenues and costs associated with them. So how does it work? (warning - long read) 1. Airports are "ranked" by their passenger numbers, much like they were in CA, except in this game it's based off real-world data. As it stands ATL is ranked #1. 2. The ranks are converted into percentages based on total # of airports. So for example, ATL would be 100%. An airport with half as many passengers as ATL would be 50%. Then we have what I call "factors". These are things that affect the final number of passengers each flight will receive. The first and most important is the "rank factor". 3. Rank Factor takes into account the respective ranks of the origin and destination airports. The formula is [(rank1+rank2)/2]*min(rank1,rank2) So wtf does that mean? It takes the average of the two ranks, and then multiplies that average by the smaller of the two ranks. What that does, is put emphasis on the smaller rank, so that you cannot simply fly from a major airport into a tiny airport and still get hundreds of passengers. It minimizes the skew that the larger airport puts on the formula. I believe this in itself is already better than what CA provided us 4. Base passengers. This is rather simple, and subject to change. This is the "base" amount of passengers that your route starts out with. It is then modified (passengers added & subtracted) by the other various "factors" that I will discuss below. The formula is (500*rank factor) That can be translated to 500 passengers, multiplied by the rank factor. The rank factor will always be < 100% if you do the math. Even if you fly from airport ranked #1 to airport ranked #2, the rank factor will be somewhere around 99.7% - give or take. That would lead to ~ 499 "base" passengers. 5. Reputation factor. This will take into account your airline's reputation. Any reputation under 100% will cause you to lose passengers, but the closer you are to 100% reputation the less passengers you lose. The sensitivity for this factor is 0.65. The formula for pax lost from reputation is [(base pax*reputation)-base pax]*sensitivity Keep in mind reputation is a percentage, so for example a reputation of 88.46% would be represented in decimal form as 0.8846. 6. Ticket price factor. This will take into account your ticket price compared to 1; the average ticket price for that route among competing airlines, and 2; the "base" ticket price the game calculates. This will prevent cooperation among airlines to raise the average price to an unreasonable amount. The sensitivity for this factor is 0.65. For now, this factor does NOT account for average ticket prices among the competition (as there is no way to accurately gauge competition without active players). The "base" ticket price is calculated by the game as follows: $50+(distance*$0.11) That's distance in miles. It then ranks your ticket price compared to the base ticket price like so: price rank = base price/your price The price factor is then calculated: price factor = price rank^1.5 The formula for pax gained/lost from ticket price is: [(base pax*price factor)-base pax]*sensitivity The following factors are not yet implemented into the formula in-game, but will be eventually. All of these factors will modify the passengers previously calculated in the "base passengers" step. Because I am still working on implementing these factors I do not have formulas for them yet. 7. Competition factor. This will take into account competition on the route. The more airlines that have flown the route recently, the less passengers will be available to your airline. 8. Time of day factor. This will take into account local departure/arrival times. If you are taking off or landing in the middle of the night, you will get less passengers than if you operate during "normal" hours. This won't be a huge factor, because obviously there is a demand for flights throughout the entire day. But if you don't plan well for a long-haul flight for example, and you depart at 3am and arrive at 2am, then you can expect a negative impact. 9. IFS factor. If you have a better in-flight service package, you will attract passengers. A poor IFS package (or lack of one) will deter passengers. 10. IFE factor. If you have good in-flight entertainment options, it will attract passengers. A poor IFE will deter passengers. 11. Random factor. This adds a random +/- 5% to the total amount of passengers at the end of the calculation. For example if the game calculates a total of 100 passengers after all the factors are accounted for, you could actually end up with anywhere from 95-105 passengers. Of course, the final number of passengers you can carry is ultimately limited by the aircraft you have assigned to fly the route. This brings up an interesting strategy element. Each route will have a certain amount of passengers willing to fly it - it's up to you to utilize the correct aircraft for the respective demand. If demand is only 85 passengers, it would not be economical to use a large aircraft. Another thing that I have yet to account for is passenger classes. For now, all the passenger demand is allocated to economy class. Eventually I want to create a formula that spreads out demand over all five classes. If you managed to read all of that, great job! You get a cookie. Obviously, all of this is still a work in progress and subject to change. It is the main thing that I want to test during public beta here sometime soon. RE: We're getting somewhere ;) [Demand formula] - skifans - 03-01-2017 I have to turn over my cookie. Looks great and we done!
RE: We're getting somewhere ;) [Demand formula] - Carde - 03-01-2017 Lovely to see some work here and I bet before it is all said and done this formula will be a lot more complicated Guess the real pudding will be in the testing to see how it plays out. RE: We're getting somewhere ;) [Demand formula] - POTKC - 03-02-2017 This seems...excellent, and more amazing than anything I would ever be able to create. Question: Will we be able to see the (approximate) demand, before the 'time of day' and 'random' factors are applied to it, before leasing slots and flying a route? RE: We're getting somewhere ;) [Demand formula] - Frenchie - 03-02-2017 Good news indeed! A comprehensive formula, we must wait and see how it actually turns out! RE: We're getting somewhere ;) [Demand formula] - Unknown98 - 03-02-2017 (03-02-2017, 03:52 AM)POTKC Wrote: This seems...excellent, and more amazing than anything I would ever be able to create. Question: Will we be able to see the (approximate) demand, before the 'time of day' and 'random' factors are applied to it, before leasing slots and flying a route? Yeah, I'll create a route research page where you can select origin and destination and get some info about each route. RE: We're getting somewhere ;) [Demand formula] - Unknown98 - 03-02-2017 Ticket pricing and reputation have now been factored into the formula. I'll update the main post with exact formulas, but basically a higher ticket price deters customers and a lower ticket price attracts customers. Reputation is also simple, the higher your reputation, the more passengers are willing to fly with you. RE: We're getting somewhere ;) [Demand formula] - Unknown98 - 03-10-2017 I added a distance factor into the formula as well that aims to decrease demand for cities extremely close together. It may need to be altered a bit, but we can sort that out later.
RE: We're getting somewhere ;) [Demand formula] - Carde - 03-10-2017 Aw, but my Air Bus venture I used to run in CA would not work now between those two cities
RE: We're getting somewhere ;) [Demand formula] - Unknown98 - 03-10-2017 Precisely
|