AirwaySim

Could find a related topic for this one.

Looking for advise and tips for growing your airline in big game worlds, 300 to 400 plus players.
I started a airline and am doing ok with 30+ aircraft. But now most of the routes are full for my headquarters airport, I want to open a new base but looking at route planning, almost every airport worth moving to have either a lot of airlines already there or most routes already full. It seams to me that unless you open an airline at the begging of a game world start, its hard to grow when you start mid or late game.   

Hi there,

That is not true. And I'll tell you why. You always need to find a niche within the game and use that niche for your profit. For example, last game world 2 was opened in the game year 1960, but I joined a bit later (1965 if I recall correctly). Obviously all "good" bases were gone and the demand was fully supplied pretty much everywhere. I joined the game in Chania (a really small airport in Greece). I did fine for the first 2-3 years, made sure to "own" the place and then found out that a Greek competitor in Rhodes and Corfu was struggling. I immediately expanded at the bigger airport (which was Rhodes) and took over the market share. Eventually my competitor went bankrupt and I expanded also to Corfu. Rhodes and Corfu were quite large airports to play with until EU-wide bases could be opened. I expanded where I could and also expanded into Zakytnhos (not the biggest airport, but demand was good enough).

In 1998 the OpenSky agreement allowed me to open bases in the European Union. I was looking for struggling airlines and found a "victim" in Vienna (Austria). 2-3 game years later he was gone and I had another big base all for my own. The same strategy was applied elsewhere (e.g Prague). I don't quite recall which other bases I had, but I think I also went into Hamburg later on and another few smaller places here and there. Again, my HQ was Chania! Nobody ever (me included) would have thought I'd become one of the bigger airlines in Europe. But it's possible if you have a plan.

In current GW2 I joined late (again) and was "forced" to go into Stuttgart. It's not exactly a bad place to start, but certainly not the biggest HQ either. And now? I killed all competitors, expanded to Cologne, Hamburg and Düsseldorf, have around 500 million passengers a week and have nothing to fear. My score is currently a bit low as I fired quite a lot of people recently, but that'll recover soon. Again, started in a rather small place, fought my way up, owned the place(s).

Nothing is impossible, just give it time! Make a plan, find a niche, look out for opportunities. You're based in Denver. There should always be room for improvement. E.g "outfrequency" competitors if that needs to be done or focus on smaller demand niches...even cargo. I am sure there is some room for expansion. If not, the US has lots of potential. I am sure you'll find a place with weak airlines, unfulfilled demand etc...

EDIT: Also, the "airport list" doesn't end on page 2. I have just had a brief look what I would do. Look at Louisville, Tulsa, Dayton etc. 230 demand and only 2 daily flights. Open a third daily flight there and you could easily fill a plane with 70-80 seats. And what about Greater Rochester, Syracuse or Long Island? 250 demand, only 114 covered. You could easily fly two daily 80-seaters on each of those routes! As I said...the airport list doesn't end on page 2. I would be happy to have so much opportunities for expansion! ;)

Wow, learn something new every day, thanks!

I meant to put "its hard to grow when you start mid or late game." with a question mark at the end, because I wasn't sure, but again thanks for answering that.  ;D ;D ;D

Andre is completely right.

Situation in smaller countries can be complicated, but in the EU, USA, Japan or China, you can always find a way to expand.

Starting mid-game can sometimes be complicated if you start relatively early, as prod lines can be filled already and the UM completely depleted. But that's only for a few years: after that, things get better and better, and bankruptcies begin to happen.

Last GW#4 (1970-2030) I joined in 1988 and within 10 years I was the 3rd biggest airline in China, HQed in Xian. And I came close to the scores of the first 2 that started on day 1. Sure, they held Beijing tight, and out of the majors airports I was able to set a foothold "only" in Guangzhou and Shanghai (still, small footholds), but the other 2nd row bases were still large enough for my needs. Grew past 600 planes without a problem.

The thing is to know when to move and where. Don't go to a said place because you want to be there (emotional reasons), but because there is an opportunity: like Andre said, someone that's just gone BK / about to go BK, mainly. Or, even if the guy there is strong, if he is flying a different sector than yours (props vs classic jets; MH vs LH, etc.), so you won't be direct competitors. Or if he's not well settled yet.

Doesn't mean you can't move to those bases on your wish-list, but do so only when there is that opportunity. Look for the demand from all potential airports (sometimes people fly to only a handful destinations), on various sectors (are props covered? are workhorses jets covered?), look at the health of airlines based there, etc.

So for example sometimes MH is completely locked, but SH is not, and that's a nice way to set a foot in the door, even if you have to wait for years for the guy to BK. Might even never happen, but that won't prevent you from having a profitable base. A lot of airports have enough room for 2-3 players. But not all of them.

I moved into Atlanta late, when the two big boys had the place sewn up and all routes 150% supplied with A320s/737s. Now I make loads of money.

The way I did it was use smaller planes, 50 seaters, against their 180 seaters, and with double their frequency. All the airwaysim computer does is look at how many passengers fly the route that day, say, 2000, divide by the number of flights that day, say 10, so all flights get 200 passengers. If I have smaller planes and double the frequency then my LF% are 90% and the bigger planes will be 50% or so, even when the route is 150% or 200% supplied. I suppose there is logic to allocate, say, 80% of passengers to the 6am-9pm prime slot, and the other 20% to the rest of the day but I don't think the logic goes much beyond what I've described.

The only downside of the above is slot costs at Atlanta and other large airports are really expensive, and small planes carry a small amount of light cargo and zero standard cargo so that lucrative cashcow is denied me. I wish there was a small regional jet that carried some standard cargo...

Hope this helps, Simon

Quote from: groundbum2 on February 14, 2018, 11:19:15 PM
All the airwaysim computer does is look at how many passengers fly the route that day, say, 2000, divide by the number of flights that day, say 10, so all flights get 200 passengers. If I have smaller planes and double the frequency then my LF% are 90% and the bigger planes will be 50% or so, even when the route is 150% or 200% supplied. I suppose there is logic to allocate, say, 80% of passengers to the 6am-9pm prime slot, and the other 20% to the rest of the day but I don't think the logic goes much beyond what I've described.

Sadly, you are 100% correct.

Allocating passengers by counting the flights, and (nearly) ignoring everything else should at some point be fixed, the code responsible for this allocation should be rewritten from scratch, instead of applying bandaids to it.

Quote from: JumboShrimp on February 15, 2018, 03:16:14 AM
Sadly, you are 100% correct.

Allocating passengers by counting the flights, and (nearly) ignoring everything else should at some point be fixed, the code responsible for this allocation should be rewritten from scratch, instead of applying bandaids to it.

Time for AI! Since your average toaster now has AI built in, time for airwaysim to have it! 8-)

Simon

Quote from: groundbum2 on February 14, 2018, 11:19:15 PM
All the airwaysim computer does is look at how many passengers fly the route that day, say, 2000, divide by the number of flights that day, say 10, so all flights get 200 passengers.

...that conception of yours on how the background systems work could not be more wrong. It does not work on such a simple system.

Quote from: Sami on February 15, 2018, 08:45:35 AM
...that conception of yours on how the background systems work could not be more wrong. It does not work on such a simple system.

Possible. At the same time, his abstraction of the system, how leaked and wrong as it may be, is more than enough to build scheduling strategies that dominate the game. I personally assume what he said is true, at least enough for my needs, and my performances in some Game worlds, let me think that it's a theory accurate enough to be very successful(even if I saw a subtle difference while RI is low in current regional game, which I included in my strategy).

Quote from: Sami on February 15, 2018, 08:45:35 AM
...that conception of yours on how the background systems work could not be more wrong. It does not work on such a simple system.

I'm glad to hear an impassioned defence of the system. Obviously, those in the know understand the phrasing used is a rank generalisation, and the system is a complicated one.

The problem is, however wrong it is in theory, in practice following many years of playing, one realises it's a good basis for thinking about demand allocation and how to best serve t. An alliance colleague describes the effects of many of the variables that go into the system calculations as "a rounding error". What he means by that is that in isolation, each seemingly important variable has such a small effect, as to be fundamentally unimportant, and further to that, is easily negated the other way by the next variable.

What the game has seen is a race to the bottom, and the most successful aircraft on the game are the smallest ones that can go the furthest distance, and usually frequency wins out.  The too small warning came in to deal with the 320 and the 757 spamming and killing MH/LH, but it didn't really solve the root cause of what's happening - small and often still wins out, that means certain aircraft always have an advantage. But as well as nerfing the smaller planes spamming the 3000-3500nm market, it ended up nerfing the higher end as well, with the 707 and DC-8 going to small in 1980. That shouldn't happen, because the economics of running such planes  in this time period are marginal anyway. And if you look at the current GW2, the whole gw is full of airlines struggling financially because of this. Nerfing the 757 you are protecting spamming of cheap to run planes from frequency spamming. Nerfing the DC-8 and 707  we are killing airlines for no reason that haven't yet switched (probably because the production rates are so low in 2, the 350 players can't source enough planes to switich!

Back to the small and often - the 767 is the most economical small plane. It's small, it's cheap, it's plentiful,  it never goes too small and it flies over 6000nm. If you have a 400 demand route, one player flies a single 747, and another flies two 767-200s - however the system is meant to work, with certainty I can say the person flying two 767s has lower operating costs, and in all probability, more than 60% of the passengers. Neither will make much money on such a  route, but the 767 flyer will be bankrupting the 747 flyer if too much of that happens across the base.

In my opinion, the DC-8 and 707 should be removed from the too small nerfing for reasons stated above (and to allow the super 70s to be a realistic proporsition), but once the 787/A350/330NEO/A350 are in operation, the 767 (which with 80s technology remains viable to 2036), should be giving too small warnings after 2015.

Then there is the Embraer E190 series - there is a too small warning at 1600nm within a country, and 1200-1300nm outside. That needs to come down as it's overpowered.

Anyway back to the original point, reading the comment you commented on raises two fundamental points; it's factually wrong for sure, but the system does err towards working as if it's based on fact. Knowing how difficult like this to program, I don't know if I have any answers or better ways of creating a system, because each would be flawed and inherently biased towards favouring one kind of approach. But small tweaks, and incentives towards larger aircraft, or away from the same old work horses (767, 320 etc) might leaven the bread somewhat. Just some rambling thoughts.

Before playing this game I never thought that Caravelles would lose for Beechcrafts and "inexistent" Accountants when it comes to pax preference... It was after that that I understood that passengers in AWS prefer choosing more flights a day on a Short or other "cheatcraft"  than 1 comfortable flight on a DC-9, which is way bizarre!

The two main things that influence ticket sales is: Price and Schedule. First and Business class pay more attention to schedule while Economy class passengers
are more price sensitive. So, it would be nice if AWS would allocate F & C class more to the airline with the higher frequency, while Y class to the airline with lower
fare (ignoring the frequency to some extent). Maybe that's already the case but not very prominent.

What I haven't tried myself yet and would be interested in knowing: Did anyone successfully run a low cost airline out of a major hub by using only high density seating at considerably lower fares?

Cheers!

Quote from: MikeS on February 15, 2018, 02:33:29 PM
(.../...)
What I haven't tried myself yet and would be interested in knowing: Did anyone successfully run a low cost airline out of a major hub by using only high density seating at considerably lower fares?

I've seen opponents try this method against me. None of them survived long.

Quote from: Sami on February 15, 2018, 08:45:35 AM
...that conception of yours on how the background systems work could not be more wrong. It does not work on such a simple system.

The problem where the starting point of the allocation is.  The main unit a flight.  As in 1 flight, 2 flight, N flights.

While the excesses were curbed in the corner cases (small aircraft flying very long distances), the center of distribution, where there are most flights (0nm - 1500nm) is unaffected.

A route with 150 demand flown by
- Airline A with 75 pax aircraft
- Airline B with 150 pax aircraft

The system will alocate 75 pax to each.  That is where the center of gravity is.  All the variables that the system has have such a miniscule effect that they just can't materially change the distribution to be 75 pax to each flight.

The result is
- Airine A has 100% LF, half the costs, and is printing money
- Airline B has 50% LF, double costs, and is losing monehy.

The system strongly favors flying smaller aircraft.

What needs to happen is that this center of gravity needs to move from 1 flight = 1 flight to 1 seat = 1 seat, which would result in:
- Airline A with 50 pax, 67% LF
- Airline B with 100 pax, 67% LF

In this case, the system is neutral, as far as size of aircraft.

From this starting point, where the seat is the main unit of distribution, other variable can have their proper effect (price, seating quality, flight duration etc.)

Quote from: Sami on February 15, 2018, 08:45:35 AM
...that conception of yours on how the background systems work could not be more wrong. It does not work on such a simple system.

I'm sure it is, Sami. The problem is that what we all observe is a system that behave like if "look at how many passengers fly the route that day, and divide by the number of flights" regardless the underline algorithm.

Nothing else works. Indeed you have to add the too small penalty, that is, in my opinion, and artificial patch to fix a not well working algorithm in assigning passenger.
The too small penalty doesn't have any counterpart in real life. People look for price, length of the journey, the departure or arrival time in some case, and to the service offered by the company. The majority of people doesn't know, nor cares, about the type of aircraft and its size.

So if someone ask me how to beat a well established player on a base, I'll answer to take smaller planes and send more flights per route than the competitor. And he will win.

Quote from: Luperco on February 15, 2018, 03:35:52 PM
I'm sure it is, Sami. The problem is that what we all observe is a system that behave like if "look at how many passengers fly the route that day, and divide by the number of flights" regardless the underline algorithm.

Quote from: wilian.souza2 on February 15, 2018, 01:20:12 PM
Before playing this game I never thought that Caravelles would lose for Beechcrafts and "inexistent" Accountants when it comes to pax preference... It was after that that I understood that passengers in AWS prefer choosing more flights a day on a Short or other "cheatcraft"  than 1 comfortable flight on a DC-9, which is way bizarre!

Sorry, that is simply not correct. Here's a test with some raw data for you:

Quote
Details applicable to all cases:
- Year 1970.
- Two airlines in game, both with:
   - Company Image 100.
   - Route Image 100.
   - Based at the same airport.
   - Salaries at the standard level (= same for both).
- Both are flying the same one route, with:
   - Route pricing the same.
   - Same seat quality (HD Y).
   - Different a/c types, other uses a modern jet for the era and other a small prop (also modern for the era).
   - Schedules done with at least 1 hr spacing between flights and no night flights.
   - All other variables the same.
- No delays, cancellations, or maintenances in effect (= pure flight data calculation).
- Route distance 235nm (within full payload range of both a/c).
- All in all, everything "normal" and route and setup is also "standard".


Case 1:
- Airline A: Supplies 160 seats daily on TWO flights (dep 0600 and 1500), jet (Caravelle VI-R), 80 HD seats per flight. One aircraft in use.
- Airline B: Supplies 160 seats daily on EIGHT flights (evenly spaced between 0500 and 2000), small prop (DHC-6-300), 20 HD seats per flight. Three aircraft in use.

- Route demand 450pax/day, total supply is 340 seats (route "underserved").

=> Route sales result: Both airlines sell 100% of their seats. 160 and 160 seats sold for both, market share will be 50/50%. (all flights full, nothing spectacular here, since there is still demand to be served)

- Economy comparison: Both airlines earn $16320 ticket revenue. Direct route expenses for airline A are $3114 while airline B has $2304 (pax fees are the same of course, but B has lower landing, handling and fuel fees but larger navigation fees; small aircraft are very much cheaper to operate).  => So airline A's direct route operational result was $13206 and airline B's $14016.

- However airline B has paid some $60 000 more for slots (but this is very airport specific and not comparable; but later in the game and at large airports the difference towards fewer operations per day will be VERY significant).

- With this scheduling (1 large a/c vs 3 small in use) airline A is paying $40000 is staff expenses and airline B is paying $90000 (more aircraft in use, and more operations per day). However since both airlines are new/small this comparison is not fully accurate, but in principle more operations per day need more operational staff.

==> Overall in this case it is more economical to run the larger jet twice daily, than small a/c eight times a day. (however you'd need to account the a/c ownership/lease costs and maintenance costs too, but that's beyond this comparison)



Case 2:
- Same airlines, same equipment, same number of daily flights and same schedules as in #1.

- Route demand is 250pax/day, and supply remains at 340 seats (route "overserved").

=> Airline A sells 142 seats, airline B sells 148 seats. Market share would be 49/51% (but number of flights per day is 20/80%). So here you can clearly see that route sales are not allocated to whom just flies the most!

- Economy comparison: Airline A earns $11376 and airline B $12905 after direct expenses (B had the small aircraft, hence lower ops expenses). Slot cost and staff cost (etc) not taken into account.

==> Airline B with six more flights per day has a very minor advantage but it is offset by the other indirect operational costs. So again it's smarter to fly the bigger aircraft in this case, but you are not broke if you wish to run a small airline either.



Case 3:
- Again the same route, and all the same details, but let's lower the demand even more.

- Route demand is 150pax/day and both airlines serve a total of 340 seats a day again (route "very overserved"). Both airlines are slightly overserving the demand (160 supply from both and demand is 150 per day). This is a very normal case, as all airlines would of course like to fill the available demand with their supply.

=> Airline A sells 86 seats, airline B sells 86 seats too. Market share 50/50% again (but number of flights per day is 20/80%). Again the number of flights per day makes no difference here.

- Economy comparison: Airline A earns $5939 and airline B $6985 after direct expenses. Again slot cost and staff cost (etc) not taken into account.

==> No noticeable difference in who gets the passengers. Like earlier, airline B has a few more bucks in the bank after the direct expenses since small a/c is a tad cheaper to operate (but again needs more slots and staff etc.).



Case 4:
- Same routes, airlines and details. But airline B is supplying only three flights.

- Route demand 150pax/day, airline A supplies 160 seats per day (two flights) and airline B supplies 60 seats (three flights); total supply 220pax/day.

==> Airline A sells 114 seats, airline B sells 60 seats. (market share 66/34%)

- If you look at load factors alone, you might think, "wow" airline B has 100% LF, but that is merely because the aircraft is small. The system does not calculate the sales based on the load factor, but instead the specifications of the flights.

- Economy comparison: Airline A earns $8680 and airline B $5256 after all direct expenses.

==> You can see that it's relatively easy to get some market share with small aircraft on a route with larger airlines established. But you cannot use the frequency to your advantage really. You'd get the similar results for using a lower frequency and a bigger plane...

As you see, you cannot "win" by frequency spamming. In no case of these does the system allocate any significant higher number of passengers to the player flying the route a lot more times than the opposition. And to point out in this case the frequency has no benefit either as the route is so small, and no penalty either as player B flies it only 8 times a day. On larger routes (2000 pax a day or similar) you would get a small bonus for flying let's say 10 times a day in comparison to 2 daily flights.

Since this route is so short (<300nm) the slower prop has no noticeable disadvantage over the more comfortable and faster jet. At larger distances the prop would start to lose (chose this short route as it was the easiest to set up for demonstration). But of course since the prop is slower, you need more aircraft, more staff etc. in order to fullfill the whole demand here.

The system is built so that also the small airlines can survive (but their life will be hard at large airports due to the slot costs). So basically if you think that you will set up shop at LHR and spam 19-seaters to all routes in order to gain the market share, then you're screwed. Yes, you will get your share of the sales eventually but it will be much more costly overall.

But the statement that number of seats sold are allocated in the order of who flies the most flights per day is just incorrect, it depends on so many other things too.


Most of you seem to be too fixed on some notices and warnings the system gives to help you, and to some other things you might have heard on how the system works (but such data is probably never disclosed?)... Hate to say but perhaps it could be best to hide all those helpers etc. since it seems to lead to misconceptions. For example the "too small" notification - it does not automatically mean that flying that route with that a/c type is a complete failure (it's not a "on/off" type trigger).


/edit, typo

The system has another "special case" to combat a corner case scenario, where the system severely penalizes high number of flights to a destinations (in relation to demand of the route.

It seems that the cases 1 - 3 are really this special scenario, where airline B is penalized for high number of flights, and only case 4 is a pure one, where no special rules or penalties apply.

As I said elsewhere the system is a lot better with these 2 out of 3 bandaids (than it was before):
1. "too small"l penalty
2. penalty for too many flights
3. tech stop penalty is a mixed bag

But in the vast space where the pure allocation formula operates, where none of these 3 bandaids apply, the allocation is just per flight, and it results in a small planes being full, bigger airplanes being half empty.

Quote
But the statement that number of seats sold are allocated in the order of who flies the most flights per day is just incorrect, it depends on so many other things too.

Better way to look at is to ask, as Luperco mentioned:

Q: How many passengers will be allocated to a flight?
A: Total demand / number of flights.

Case 1:
450 / 10 = 45 per flight
Airine B 20 seater full
(even though it may be in penalty phase of 8 flights to 450 demand)

Case 2:
250 / 10 = 25  pax per flight
Airline B gets 148 / 8 = 18.5 (nearly full, nearly all of 20 seats sold
(Airline B is slightly more into the penalty phase of 8 flights to 250 demand)

Case 3
150 / 10 = 15 pax per flight
Airline B gets 86 / 8 = 11
(Airline B is most likely very deep in the penalty phase, for flying 8 flights to 150 pax demand)

Case 4   <- this is the only normal scenario where no penalties apply
150 / 5 = 30 pax per flight
(Airline B gets 60, all 20 seats are full.  System tries to give it 30, so a lot of room to spare for Airline B.  They probably don't even have to clean the lavatories between the flights, because the system will always (no matter what) allocate full plane.  As far as penalties, supplying 2 flights to 150 demand has no penalties, so this really how the system works across AWS, perhaps in of 80% of all passenger allocations.

As long as the allocation starts with the formula of:

Allocated pax = Total demand / number of flights.

this issue problem will persist.

It should be noted Sami, that many of us are aware of mechanisms (we don't know what they are of course), which mean in the scenarios you pose, one airline won't monopolize the routes. Indeed, the first flight is always the most important, and when I play the US, I often operate a single flight policy knowing that lf's will be sweet across the board, costs low and yields high.

The issues raised really pose problems higher up the food chain. Frequency spamming is probably the wrong phrase for it, because operating 2 or 3 flights isn't spamming. But the reality is this - buying 2 767-200s is cheaper than 1 747-400. Operating 2 767-200s on a 3500nm will cost less (fuel, staff etc combined) than 1 747-400. And in most cases, all things being equal, the two 767-200s will pick up more passengers than the single 747-400.  I can't speak for anyone else, but this is what I mean by small and frequent over size,  not talking about irrelevant oversupplying on small regional routes - I apologise for not making that clear.

My question is reasonably simple; what is the benefit to buying a 747-400. Now I do understand there has to be a benefit to frequency, because pax like frequency (If by using the word frequency I'm mistunderstanding the system, again sorry, but the point is about because the two 767s carry more passengers, it's hard to nail down a better word). But there is an argument it's too primitive. And the planes people are interested in buying, long haul are very  narrow. Basically, everyone wants B767s and A330s, because the 777, the 747 and the MD11 are considered too big, and they bankrupt airlines. This isn't conjecture, it's a fact of gameplay

In real life, the 777 IS the most popular widebody on the market, and the 747-400 was the most successful variant of the series. They don't really work in game, except for a few select airports. This is a shame I think, and hopefully cargo will address this (although I don't think so, as the same game play rules apply for cargo, and the 767 and 330f also look hard to beat).

In real life, the 777 is not too big to be economically run. In real life, the 747 has had its day a viable pax plane because of the 777 really, but it had 40 years as a lynchpin of the skies. In game, they're very difficult to make work anywhere. It's a shame they're too big. Perhaps using the word frequency muddies the water, but if there's any competition, you have to go small and more frequent (or not).

EDIT: Actually - forget the word frequency. Small beats large. Frequency is a by product of the realisation of players that large is bad. So...

If a 400 demand route has a 767 and a 747 on it, a perfect system would allocate demand in a way that wasn't 50/50, but favoured more passengers on the larger jet (but overall meant neither airline loses out too much/costs are covered more relatively to the metal being flown). This does actually happen IRL, through traveller preference for an A380, and of course airlines offering more cheaper seats on bigger jets to fill them. Frankfurt to Toronto is a route I take a lot for example, and Air Canada flies 777. Lufthansa flies 747. Lufthansa flights are always full. Air Canada flights are a mixed bag.

surely in real life the bigger plane has a cost advantage? The whole point of an A380 is it's one set of airport slots, 3 flight deck crew, one load of avgas so surely lower price tickets can be offered, compared to smaller planes that cost more per revenue seat mile to run? I would think this would be modelled in AWS?

So in the scenario's above, the one 747 carrying 450 pax should be profitable at a lower ticket price than 2x767 250 paxs? So the clever player will get the Jumbo, offer lower ticket prices, and harvest the price sensitive Y passengers? Just like the real world.

What breaks the above is the lease/purchase cost of bigger planes is sky high compared to smaller planes. I can buy the 50 seat Embrear for $12M in GW3 whilst 180 seat A320s are $85M, so cost per seat is twice as much for the bigger plane.

Other things AWS takes into account when allocating seats,
relative CI
relative RI
age of plane
attractiveness of plane (eg Russian vs Western)
seat style (high density etc)
ticket prices
direct vs 1 vs 2 stop
departure time

but all the above is tiny swings 2-3% compared to frequency.

hefty penalties apply when an airline has 2 flights within 60 minutes of each other or the too small plane penalty

Simon

can I also say I'm happy with the system as it is now, it reflects real life pretty well within the constraints of what a 3GL computer program can do. And makes for an enjoyable game where big and little players can co-exist.

Simon

Quote from: yearofthecactus on February 15, 2018, 06:09:40 PM
And in most cases, all things being equal, the two 767-200s will pick up more passengers than the single 747-400. 

My question is reasonably simple; what is the benefit to buying a 747-400.

In real life, the 777 IS the most popular widebody on the market, and the 747-400 was the most successful variant of the series.

I think these are really interesting and valuable points/questions. 

I think the problem is that people are conflating the reality of the world too much with the game.  Consider the following: within this game we are all flying what are, for all practical purposes, point to point airlines (as opposed to hub and spoke).  In the real world, point to point airlines want to fly the smallest airplanes that make sense for their route structure.  Southwest, Porter, Ryan, etc., all opt for (relatively speaking) smaller airplanes with higher route frequency.  Consider also newer "disruptive" operators in the LH market like Norwegian--they are operating a lot of point-to-point flights and are doing so using ships like the 787 that are certainly a step down from the 777/747/A380s--its the nature of point-to-point flying.

So, why do the A380 and the 747 not work, except in certain circumstances?  Well, here the real world is actually an excellent analog.  A380 and 747 work well for cities (New York, London, Paris) where--despite being hub-and-spoke hubs, the cities themselves draw HUGE numbers of passengers and which send out huge numbers of passengers.  They also work in the Middle Eastern hubs.  The problem for us in AWS is that the game does not model the hub-and-spoke system of the Middle Eastern 3.  Dubai doesn't receive that many travelers for whom the Middle East is their ultimate destination, but it is a great stopover when flying between North/South America and India.

I would also point to history.  I am an American and am most familiar with American aviation history, so that is what I will use as my example.  Before deregulation of the aviation industry, when point-to-point flying was very common for the big domestic operators, smaller aircraft used to prevail.  The ending of regulation allowed for the advent of a handful of mega hubs.  It was really only then that the largest aircraft really started to make great financial sense.  Just look at the order history for the 747.  83 orders in 1966, the launch year, then 43 or less until 1977.  Then, 50, 60, 70, or more orders became the norm with the peak in 1990 (120+ orders) at a time when hub consolidation and industry consolidation made the very large aircraft make tremendous sense.

So, how can we "fix" AWS?  Well, short of having a functional system that will allow for passenger transfers in hub cities OR dramatically reducing the number of airlines, you can't.  There are presently 4 truly large airlines in the United States (6, if you include UPS and FedEx).  How many large airlines are there in a typical AWS game in the US in the 2000s? 

Ultimately, you can have reality or you can have a fun game that allows many people to play and enjoy at least a modicum of success.  You can't have both.

I mentioned this in another thread:

The difference between two 767s vs one 747 is now cargo. A 747 (400) @ 5000nm will take much more cargo then 2 767s. Thus the total gross of the 747 will outweigh the gross of 66/33 pax split.

This is totally observable in MT3 were we can see 777s out gross smaller wide-bodies that would of been considered a mistake to go up against.

True, for the past 9 years, this was not the case, but AWS has started changing and alot of our understanding will be challenged over the coming months. At the end of this all, our overall thinking is going to be noticeably different from 6 months ago.


Talentz

Quote from: Talentz on February 16, 2018, 03:28:06 AM
A 747 (400) @ 5000nm will take much more cargo then 2 767s.

If you have demand that is!  ::)

Quote from: Andre090904 on February 16, 2018, 04:43:56 AM
If you have demand that is!  ::)

True that... but then again, why use a 744 on a route were you couldn't make the use of its full potential?  :P


Talentz

Well, maybe we are all wrong from the beginning, and the real issue is that lowering prices does not bring the rewards it should? ' cause honestly, price wars don't work well. You can have a minor advantage against your opponent by lowering prices, in terms of market share, but the loss in revenue more than offsets this. That's why I always apply an insanely high pricing strategy. Which, at least for Y seats, should not work THAT well.

Quote from: groundbum2 on February 14, 2018, 11:19:15 PM
I moved into Atlanta late, when the two big boys had the place sewn up and all routes 150% supplied with A320s/737s. Now I make loads of money.

The way I did it was use smaller planes, 50 seaters, against their 180 seaters, and with double their frequency. All the airwaysim computer does is look at how many passengers fly the route that day, say, 2000, divide by the number of flights that day, say 10, so all flights get 200 passengers. If I have smaller planes and double the frequency then my LF% are 90% and the bigger planes will be 50% or so, even when the route is 150% or 200% supplied. I suppose there is logic to allocate, say, 80% of passengers to the 6am-9pm prime slot, and the other 20% to the rest of the day but I don't think the logic goes much beyond what I've described.

The only downside of the above is slot costs at Atlanta and other large airports are really expensive, and small planes carry a small amount of light cargo and zero standard cargo so that lucrative cashcow is denied me. I wish there was a small regional jet that carried some standard cargo...

Hope this helps, Simon

Quote from: groundbum2 on February 15, 2018, 06:43:49 PM
surely in real life the bigger plane has a cost advantage? The whole point of an A380 is it's one set of airport slots, 3 flight deck crew, one load of avgas so surely lower price tickets can be offered, compared to smaller planes that cost more per revenue seat mile to run? I would think this would be modelled in AWS?

What breaks the above is the lease/purchase cost of bigger planes is sky high compared to smaller planes. I can buy the 50 seat Embrear for $12M in GW3 whilst 180 seat A320s are $85M, so cost per seat is twice as much for the bigger plane.

Simon

You're not factoring in fuel, staff, landing fees, etc. The initial cost of the plane by itself is much less, but you would need to buy 2 additional sets of slots to make up for the passenger volume. In Atlanta that's going to set you back $2M per slot set. 3 E145s will also consume 50% more fuel than a 737NG or A320, require far more staff and maybe most importantly they count 3x towards your outside HQ operation.

I would point out that if you had invested the same resources into an empty airport like Milwaukee you'd probably be doing much better. The cost differential between operating a level 2 vs level 4 base alone would be enough to buy a E-145 every 2 weeks.

Quote from: yearofthecactus on February 15, 2018, 06:09:40 PM
The issues raised really pose problems higher up the food chain. Frequency spamming is probably the wrong phrase for it, because operating 2 or 3 flights isn't spamming. But the reality is this - buying 2 767-200s is cheaper than 1 747-400. Operating 2 767-200s on a 3500nm will cost less (fuel, staff etc combined) than 1 747-400. And in most cases, all things being equal, the two 767-200s will pick up more passengers than the single 747-400.  I can't speak for anyone else, but this is what I mean by small and frequent over size,  not talking about irrelevant oversupplying on small regional routes - I apologise for not making that clear.

My question is reasonably simple; what is the benefit to buying a 747-400. Now I do understand there has to be a benefit to frequency, because pax like frequency (If by using the word frequency I'm mistunderstanding the system, again sorry, but the point is about because the two 767s carry more passengers, it's hard to nail down a better word). But there is an argument it's too primitive. And the planes people are interested in buying, long haul are very  narrow. Basically, everyone wants B767s and A330s, because the 777, the 747 and the MD11 are considered too big, and they bankrupt airlines. This isn't conjecture, it's a fact of gameplay

In real life, the 777 IS the most popular widebody on the market, and the 747-400 was the most successful variant of the series. They don't really work in game, except for a few select airports. This is a shame I think, and hopefully cargo will address this (although I don't think so, as the same game play rules apply for cargo, and the 767 and 330f also look hard to beat).

In real life, the 777 is not too big to be economically run. In real life, the 747 has had its day a viable pax plane because of the 777 really, but it had 40 years as a lynchpin of the skies. In game, they're very difficult to make work anywhere. It's a shame they're too big. Perhaps using the word frequency muddies the water, but if there's any competition, you have to go small and more frequent (or not).

The economy of the game is far larger and more generous than in real life. Because of that, it will always favor the smallest plausible aircraft for a given route. Imagine what the current industry would look like if the economy supported 15 US carriers instead of the 4 main ones in real life. Combine this with LH production lines that operate 3-4x faster than they do in real life and the outcome in game is inevitable.

Yeh, that may be so.

I don't really have any answers to the questions I raise in all honesty. It's a difficult one and even more difficult to write a code for, and then please anyone.

I am steadfast in my belief that the too small warning on the DC-8 and 707 in the 1980s, (however small to start) is unjustified, because they aren't too small aircraft and they don't provide a big advantage over going with the newer jets. Too small was designed and enforced to solve a different problem, and with experience of the gameworlds, the early 1980s cull where airlines go bankrupt en mass, as is beginning to happen in 2 right now, is disappointing and can be reversed, without upsetting anyway.

As for using bigger planes that have more competitiveness, maybe it's a problem that can't be solved. It is just frustrating as a managing member of 2 alliances, that I'm constantly advising people to either go 767 or A330, and sometimes get the response... can't I use 777, I'm bored of that routine. A world where the most popular and successful wide body airline in the world is actually usable/not a fast track to bankruptcy, I dunno that just appeals to me is all - call me a maverick!

But if such a system is impossible without harming the other end of the market, or is too difficult to write, well I guess that's fair enough.

Quote from: yearofthecactus on February 16, 2018, 02:02:22 PM
As for using bigger planes that have more competitiveness, maybe it's a problem that can't be solved. It is just frustrating as a managing member of 2 alliances, that I'm constantly advising people to either go 767 or A330, and sometimes get the response... can't I use 777, I'm bored of that routine. A world where the most popular and successful wide body airline in the world is actually usable/not a fast track to bankruptcy, I dunno that just appeals to me is all - call me a maverick!

But if such a system is impossible without harming the other end of the market, or is too difficult to write, well I guess that's fair enough.

It can surely be solved.  All it needs is that the starting point of passenger (and cargo allocation) is:

Allocation = Demand / Capacity (in pax or ko of cargo)

instead of

Allocation = Demand / Number of Flights

IDK - I think setting allocation to equal the demand/capacity has huge unintended (or maybe intended) consequences. For starters...

1. It would incentive players to a plane on a route that is equal to 200% of demand. Since it doesn't matter how many flights go to a destination, just how many seats you serve, then you would get 747's flying from DC to NYC all the time. Certainly not very realistic.

2. It further penalizes small aircraft operators. If you have a 100 seat route and you put a 50 seat plane on it, and your friend puts a 767 on it, you will likely get a 40% load factor, which given the staffing and other fixed costs, would be challenging to the viability of those airlines.

I much prefer a more dynamic pricing mechanism that allows airlines who choose to fly larger planes with presumably a lower cost per seat be able to offer a lower price and get a lower margin per seat, but potentially greater overall profits because more passengers move to that airline. The challenge now is that the pricing model isn't very sensitive so there's really no benefit to lowering prices.

Quote from: JumboShrimp on February 16, 2018, 03:32:32 PM
It can surely be solved.  All it needs is that the starting point of passenger (and cargo allocation) is:

Allocation = Demand / Capacity (in pax or ko of cargo)

instead of

Allocation = Demand / Number of Flights

That would fix the current "problem," but would create a new one that would make this game far LESS like the real world.  I live in a spoke city in the United States that probably sends 300-400 passengers per day each to large hubs like Chicago, Charlotte, Atlanta, and Houston with smaller loads to other hubs.  My city is served by RJs and the occasional A319/737 and some large props.  In AWS, this is EXACTLY what you would expect for such a city.  Under your proposal, rather than flying 3x 737-500/A318/etc., it would make sense to fly 1x 757-300.  Rather than having a race to the bottom for smaller ships, this would create a race to the top for larger aircraft.

What you propose makes some sense for International LH where some (but certainly not all) passengers prefer to fly on larger ships.  For domestic and SH, people seem much more interested in frequency and availability of connecting flights.

Quote from: dmoose42 on February 16, 2018, 04:03:36 PM
IDK - I think setting allocation to equal the demand/capacity has huge unintended (or maybe intended) consequences. For starters...

1. It would incentive players to a plane on a route that is equal to 200% of demand. Since it doesn't matter how many flights go to a destination, just how many seats you serve, then you would get 747's flying from DC to NYC all the time. Certainly not very realistic.

You don't want to fly with 50% LF now, why would you want to fly with 50% LF under this hypothetical new scenario?

The difference is that both the small and big plane would have 50% LF.  So this scenario is neutral
Currently only big aircraft has 50% LF, smaller aircraft has 100% LF.  Current allocation is lopsidedly favoring smaller aircraft.

Quote from: dmoose42 on February 16, 2018, 04:03:36 PM
2. It further penalizes small aircraft operators. If you have a 100 seat route and you put a 50 seat plane on it, and your friend puts a 767 on it, you will likely get a 40% load factor, which given the staffing and other fixed costs, would be challenging to the viability of those airlines.

I think you need much better examples.  This one, again, makes absolutely no sense.  767 makes absolutely no sense on a 100 pax route.

As far as cost of staffing, staffing a Very Large aircraft is far higher than staffing medium aircraft.  VL aircraft allocate a lot more staff per position, and in case of pilots, a more expensive pilot.  767 cost is probably 6x cost of Medium aircraft it terms of pilots.

Quote from: dmoose42 on February 16, 2018, 04:03:36 PM
I much prefer a more dynamic pricing mechanism that allows airlines who choose to fly larger planes with presumably a lower cost per seat be able to offer a lower price and get a lower margin per seat, but potentially greater overall profits because more passengers move to that airline. The challenge now is that the pricing model isn't very sensitive so there's really no benefit to lowering prices.

Sami just has not been able to make any other variable matter, at all, including the pricing.  Everything except frequency is noise.  Nothing makes any difference.  10% discount?  Nothing.  Better seating?  Nothing.  Faster aircraft?  Nothing

There is nothing you can do to attract more passengers other than flying 2 flights instead of 1.

Quote from: chwatuva on February 16, 2018, 04:13:17 PM
That would fix the current "problem," but would create a new one that would make this game far LESS like the real world.  I live in a spoke city in the United States that probably sends 300-400 passengers per day each to large hubs like Chicago, Charlotte, Atlanta, and Houston with smaller loads to other hubs.  My city is served by RJs and the occasional A319/737 and some large props.  In AWS, this is EXACTLY what you would expect for such a city.  Under your proposal, rather than flying 3x 737-500/A318/etc., it would make sense to fly 1x 757-300.  Rather than having a race to the bottom for smaller ships, this would create a race to the top for larger aircraft.

What you propose makes some sense for International LH where some (but certainly not all) passengers prefer to fly on larger ships.  For domestic and SH, people seem much more interested in frequency and availability of connecting flights.

I imagine it would be possible to give a specific bonus to higher frequency under the capacity allocation, rather than AWS having only one variable that matters - frequency

After read all your comments, including the Sami answer, I don't see this system so bad.

It is natural that the passengers are equally divided per number of aircrafts as a base.
Provided, of course, that there are other factors that bias the people toward one flight rather than another.

In AWS there are such factors. For example the starting and arrival time, or the ticket prices, or the company/route image.

The  problem in my opinion is that some factors doesn't have the right weight. For example the ticket prices or the confort abroad (seats configuration).

Furthermore there are factors that are not considered at all.

The more important is the length of the trip. Instead of put an artificial penalty to the route with tech stop for example, in case of competition on a route, the shorter flight time should bring more passengers. This, if well balanced, would solve the prop vs jet problem and the tech stop itself.

Another factor could be the design age of the aircraft used on the route.

Everything should be balanced so that players can choose different strategies to success. Actually, always in my opinion, they are rather limited.

Speaking about the "too short penalty" the problem is not in the penalty itself. But in the lack of information.

I cannot tell if my flight is suffering from the penalty or not. And cannot search for all the flights with the penalty and, eventually, fix them moving on another plane (or accept it).

I think a small message on the route information page (those with sold tickets and expenses) and a filter on the route list page, would help on that matter.

Quote from: Luperco on February 16, 2018, 05:27:12 PM
It is natural that the passengers are equally divided per number of aircrafts as a base.

Is it?  Why would 1 passenger get on a 750nm Turbo Prop flight, in HD seating if there is a standard seat jet for the same price?

In AWS, the pax will cram into those HD seats, up to full 100% LF, get blood clots, all while leaving half of the jet seats empty.  And at the same price (since price makes pretty much no difference).

Quote from: Luperco on February 16, 2018, 05:27:12 PM
Provided, of course, that there are other factors that bias the people toward one flight rather than another.

With frequency being the untamed and untamable beast, all those other factors don't really matter.

Quote from: Luperco on February 16, 2018, 05:27:12 PM
The more important is the length of the trip. Instead of put an artificial penalty to the route with tech stop for example, in case of competition on a route, the shorter flight time should bring more passengers. This, if well balanced, would solve the prop vs jet problem and the tech stop itself.

Yes, very true.  But we get the tech stop penalty (on top of the too small penalty) because otherwise, the system would give the small, tech stop flight 100% LF, while leaving non-stop wide-body half empty

So we get 2 medicines to fight the symptoms rather than looking at the cause of the disease (frequency).

I posted yesterday in the ongoing "too small" thread in "feature request". Was about the "too small", but also about "frequency". And I can't say I agree on your vision, JS. Got the feeling your focusing on only one idea while forgetting the rest of the picture, if I may.
It was initially written for this thread but then moved on, I was uncertain about duplicating it.
But here it is.

--------

Quote from: JumboShrimp on February 15, 2018, 03:24:53 PM
The problem where the starting point of the allocation is.  The main unit a flight.  As in 1 flight, 2 flight, N flights.

While the excesses were curbed in the corner cases (small aircraft flying very long distances), the center of distribution, where there are most flights (0nm - 1500nm) is unaffected.

A route with 150 demand flown by
- Airline A with 75 pax aircraft
- Airline B with 150 pax aircraft

The system will alocate 75 pax to each.  That is where the center of gravity is.  All the variables that the system has have such a miniscule effect that they just can't materially change the distribution to be 75 pax to each flight.

The result is
- Airine A has 100% LF, half the costs, and is printing money
- Airline B has 50% LF, double costs, and is losing monehy.

The system strongly favors flying smaller aircraft.

What needs to happen is that this center of gravity needs to move from 1 flight = 1 flight to 1 seat = 1 seat, which would result in:
- Airline A with 50 pax, 67% LF
- Airline B with 100 pax, 67% LF

In this case, the system is neutral, as far as size of aircraft.

From this starting point, where the seat is the main unit of distribution, other variable can have their proper effect (price, seating quality, flight duration etc.)

While I agree on the analysis, I don't agree on the proposal.

Taking that example, if airline A leaves at 0700 and airline B at 1700, why should one have more pax than the other?
And if they leave at the same time, again, why, if not for pricing or service (CI/RI), age of the bird, etc.?

As Cactus said, the "too small" nerf is not working properly, and another system has to be found but spreading the LF evenly is not a solution.
Caricature: a CRJ and a 747 on the same route. So the CRJ gets 8 pax because the 747 took 95?
Cactus' post: https://www.airwaysim.com/forum/index.php/topic,75679.msg444075.html#msg444075 (https://www.airwaysim.com/forum/index.php/topic,75679.msg444075.html#msg444075)

Some airlines are flying small birds for good reasons (the majority of their routes are thin). Should they be globally penalized because of this necessity? Obviously no.

However, if you're flying a smaller bird for good reasons and suddenly fly a thick route, it's obvious that you're not flying your basic target market (or with the wrong bird). Then a penalty could apply.

So what I can say is that theoretically the "too small" warning was a good idea. But (and again), as Cactus said, it's sometimes applying in an improper way.

There needs to be a relationship between the plane and the route: is that plane made for that kind of route, yes or no? Then a penalty would apply.
But there needs to be some margins too, and if both aircrafts are considered perfectly suited for that route, then there's no reason (apart the above-mentioned ones) to allocate more pax to one or the other.

Currently, we're limited to flying 200% of the demand. Don't you think that's a bit too much when economically the best thing would be 90% if flown all alone? Blocking seat on 50% of our routes would become painful, but 130 or 150% would already be effective without being a pain.

Going further and applying common sense: was the 757 intended for LHR-JFK? No. Should I get penalized if I fly this route with a 757 20 times a day? Sure. Should I get penalized if I fly it once a day? Maybe not, because I'm "stealing" so few pax compared to the overall demand that is doesn't matter much.

So what I would find even better is a relationship between the frequency and the size. On your example, a 150 demand route is equally made for a 75 pax or 150 pax plane. No reason the 150 pax plane gets more pax. However spamming CRJs on a CDG-FRA is not ok, and would trigger a penalty if flown more than once (even if the distance is short).

What would be penalized is not the size relatively to the length and the thickness, but rather the size of the aircraft related to the frequency related to the thickness of the route.
And should I get a "too small" penalty if I'm all alone ? No, because I'm the only solution (just like for tech-stops).
(and if you're flying a 150 pax route with a 150 pax bird and somebody comes in with a 75 pax plane, it's normal that you cry, as it would happen the same exact way IRL).

You guys are scaring the OP who is new to the game and looking for some basic advice. If there are proposed changes to be made to the game, I suspect they will be far more effective to be debated in the Feature Request forum rather than this thread...

At The_Ape...

Just briefly, I want to say my suggestion of wish for a more proportionate way of making bigger aircraft like the 777, and the 747 more viable would involve making sure apples are matched with apples. The CRJ is a medium size plane. Medium size planes are best used up to 1000 miles only. You shouldn't be flying a 777 on such a route, and if you do, you get rightly penalised under the current system. That's because the of the current way of allocating demand, and this is the prime example of how it does work. So lets be honest, I'm not saying it's a wrecked system for everything. I'm purely compentmartmentalising small vs small, medium vs medium, large vs large and v large vs v large, and my hope is making these match ups more equal.

Anyway, anything between 1000nm and 2000nm is best for a large plane (2000-2500nm being the extra useful range of a 757). Anything above 2000nm you need a wide body or you get small warnings.

How to you solve this in a new system? Nerf very large aircraft under 2000nm would be one. Then nerf sub-500-600nm routes for large aircraft (the point at which props lose their competitive edge IRL, although not making those props unusable upto 900-1000nm so there is a crossover and choice to be made. A so called "too big warning". Passengers don't want it to take 1 hour to board a 300 mile flight from Montreal to Toronto, take longer to take off, land and actually be in the air. A prop ore regional jet to 300nm is the way to fly 300nm, quick to load, quick to maintain/clean/turnaround and quick up and down).

Lets be honest, such a system is unlikely to come to fruition, and Sami doesn't see it as an important part of his vision and that's fine, because this is his baby and his vision has created a great game/simulation that is getting better. But it is still important to put these proposals forward and at least have dialogue as to how such things might work. And being aprehensive and offering counter arguments is fundamental to that debate.

My personal longing is just to make more use of the available planes, and not be limited by a race to the bottom, being limited to one or two choices, and having game mechanics life physical nerfings having unintended consequences, such as the 757 range nerf that affects far more than its intended remit for example. And any suggestions that rectify the various alledged problems would need equally robust testing, consultation and understanding of their potential consequences.

@ Cactus

the 747/CRJ was just a caricature. Only used to highlight the flaws of Shrimp's thought (in my opinion).
On that 150 pax route he's speaking about, if it's 900 nm, both a 737 and a CRJ would be suitable planes according to real life practices. 737 until monopoly, but serving this route well would mean 2 flights per day and here comes the CRJ. So there's no reason the CRJ should get penalized in this case.

Hence, a "small" plane (not the category) would get penalized only if used to spam, but not on its target market like Shrimp's example would suggest. If he thinks he should get 67% LF on this route, I think it's neither realistic nor desirable. Putting a 737/A320 on such a route is exposing ourselves to concurrence, both IRL and in the game.

Because small vs small, medium vs medium, etc. ok, that's fine, but at some point these aircrafts' target market meet each other and this has to be dealt with too.

(sorry for the OP)

Edit: actually I think we globally agree. I'm just a little more conscious of the use of small planes now that I play in a country where distances are huge and demand thin (plenty of routes, but all very thin). So I try to see how things could apply in a context that's a bit different from JFK, LHR or NRT.

Quote from: schro on February 16, 2018, 06:29:15 PM
You guys are scaring the OP who is new to the game and looking for some basic advice. If there are proposed changes to be made to the game, I suspect they will be far more effective to be debated in the Feature Request forum rather than this thread...

Well, at least he will now the road to success in AWS is FREQUENCY!  ;D

Quote from: wilian.souza2 on February 16, 2018, 09:03:40 PM
Well, at least he will now the road to success in AWS is FREQUENCY!  ;D

But it's not....

If you have a 3000 demand route, supply is 6300, and  3 airlines are all flying 300 pax per flight 7 time daily on the same type of plane and you are one of those airlines.

You reduce your frequency from 7 to 6 flights per day, your supply falls from 2100 to 1800, and overall supply from 6300 to 6000.

Who wins? Or should I say, whose costs are lower overall, and likely to make the most money for their outgoings, or in tough times, make the least loss? The answer, as far as I understand it is the airline flying 6 times a day. All things being equal, the 6 flights will improve their load factors more than the 7 flights will. I know David will talk about overall trip cost being the key factor, but they are linked. Anyway, this alone defeats the idea of frequency being king. So I do agree with Sami that we talk too much about it.

I reiterate my apology with regards to bringing frequency into the equation. The problem is, we (and me for sure) use the wrong word to describe a different issue.

Quote from: Tha_Ape on February 16, 2018, 06:19:45 PM
Taking that example, if airline A leaves at 0700 and airline B at 1700, why should one have more pax than the other?
And if they leave at the same time, again, why, if not for pricing or service (CI/RI), age of the bird, etc.?

This is a very good example.  And because it is so good, I will answer it in 4 ways:  ;)

1. Your proposition lead to extinction of 747, 777, A380 from AWS.  A330 and 767 were the next ones to go, as more players took the frequency rape even further.  So if your proposition that a flight is a flight can completely upend VL aircraft flying, perhaps there is more to this that is not being captured from the Real World that AWS allocation of passengers proportionately by # of flights is not recognizing.

2. When United flies 772 between JFK and LAX  in RL, they fully anticipate that 200 of the 350 seats would be empty, because why would there ever be more passengers flying on this flight than on all the other 150 seaters?

3. Proportional allocation of passengers by number of flights is the basic building block of frequency rape strategy.  The rest of the frequency rape strategy rests on this basic building block.  All it takes is slightly re-shaping this basic building block, and the frequency is defeated universally

When LH became unplayable with LH aircraft, Sami applied several bandaids (too small, tech stop, # of flights) that masked some of the symptoms in LH but it was by no means fixed universally.  It got fixed because there was a lot of squealing.

Playing AWS by frequency rape is extremely effective, but ultimately the most boring way to play.  Which is why this weapon is not being wielded by those most capable of wielding it.  Just imagine the amount of squealing if top 20-30 players employed this strategy and BKd every player flying A320, 737, MD-80/90...

The way you destroy this building block of the frequency rape by not subtracting every passenger from oversupplied route from the biggest aircraft, you subtract the passengers more proportionately.  And suddenly, the system is aircraft size agnostic, instead of always favoring small aircraft over bigger aircraft.  Suddenly, all aircraft, including A380, 747, 777 are safe to fly.

4. There are RL situations that reward flight being flown at a particular time.  And maybe 2nd flight at another time and even 3rd flight at yet another time.  Maybe AM, PM.  Perhaps for better passenger connections.  AWS does not have any of this built in.  There is no granularity of demand by hour.  It is by day, that's the smallest granule.  Passenger connections are not implemented.

So when player has flights taking off at 0500, 0600, 0700, it is not for anything useful related to RL or AWS.  It is purely for frequency rape.  It is purely to get 3x passengers vs. flying 1 flight with bigger aircraft.

When AWS implements these features (demand granular by hours, passenger connectivity), there will be real need for more flight.
But now, players are not serving a need, they are exploiting the algorithm that fills up small planes and empties out big planes.

Quote from: Tha_Ape on February 16, 2018, 06:19:45 PM
As Cactus said, the "too small" nerf is not working properly, and another system has to be found but spreading the LF evenly is not a solution.
Caricature: a CRJ and a 747 on the same route. So the CRJ gets 8 pax because the 747 took 95?
Cactus' post: https://www.airwaysim.com/forum/index.php/topic,75679.msg444075.html#msg444075 (https://www.airwaysim.com/forum/index.php/topic,75679.msg444075.html#msg444075)

There is zero overlap between 747 and CRJ, because there is no overlap on the routes.  It is always between one size smaller.  767 can BK 747.  CRJ, E-Jet can BK A321

Quote from: Tha_Ape on February 16, 2018, 06:19:45 PM
Some airlines are flying small birds for good reasons (the majority of their routes are thin). Should they be globally penalized because of this necessity? Obviously no.

Not at all penalized.  They would just not get an unfair advantage against bigger aircraft.  Removing unfair advantage does not equal penalized.

Smaller aircraft would continue flying as they did before.  The only difference is that the small aircraft operators would also feel the pain of the route being oversupplied.

Small aircraft operator can theoretically keep adding fights to an oversupplied route and keep getting 100% LF, be completely unaffected by the route being oversupplied, putting all the pain on large aircraft operator.

So who, again is penalized and who is advantaged here?

Quote from: Tha_Ape on February 16, 2018, 06:19:45 PM
Going further and applying common sense: was the 757 intended for LHR-JFK? No. Should I get penalized if I fly this route with a 757 20 times a day? Sure. Should I get penalized if I fly it once a day? Maybe not, because I'm "stealing" so few pax compared to the overall demand that is doesn't matter much.

This was already tried and it destroyed LH.  And that was just shortly before players deployed the real terminator of LH: A321 / 739ER.  Those 2 just nuked everything.

"Too small" penalty is a bandaid that fixed / covered up many symptoms.  The real problem is that 757 suffered zero consequences of oversupplied route.  All of the pain of the oversupplied route was allocated by the systems exclusively to larger aircraft, such as 747, 777, A380, making them go extinct in AWS.

Quote from: Tha_Ape on February 16, 2018, 06:19:45 PM
So what I would find even better is a relationship between the frequency and the size. On your example, a 150 demand route is equally made for a 75 pax or 150 pax plane.

So when the route becomes oversupplied, you allocate 100% of the pain to the larger aircraft.  That is the very basis of frequency rape strategy.

I hope you understand and agree with this point.

Quote from: Tha_Ape on February 16, 2018, 06:19:45 PM
What would be penalized is not the size relatively to the length and the thickness, but rather the size of the aircraft related to the frequency related to the thickness of the route.
And should I get a "too small" penalty if I'm all alone ? No, because I'm the only solution (just like for tech-stops).
(and if you're flying a 150 pax route with a 150 pax bird and somebody comes in with a 75 pax plane, it's normal that you cry, as it would happen the same exact way IRL).

Can you explain?

Started this thread not to long ago, just looking for a couple tips..........damn, you guys got deep. Love it!  :D

lol, uncle Owl is right, were going off the deep end here  :laugh:


Taletz

So, it seems that the various ideas/arguments would invariably serve to favor either the largest or smallest aircraft within a size group--race to the bottom or race to the top.  Not sure how workable this would be, but I have been kicking around the following in my head for a while:

Suppose each route had an ideal airplane which would service it, ideal in terms of size (default capacity).  If you fly the ideal aircraft people will naturally flock to your airline/aircraft.  As your aircraft diverges more and more from the ideal standard (larger or smaller), fewer and fewer people would naturally want to fly (most people don't want to fly 747 between London and Paris, after all).  Distribution of passengers would then theoretically be even amongst all the aircraft equal or greater in size to the ideal, but of course divergence from the ideal would reduce the actual allocation for sub-optimal types and re-allocate those displaced passengers toward types more ideal/optimal.  Of course, the divergence factor would have to be expressed as a ratio/percentage which would be carried to aircraft smaller than the optimal and applied relative to the aircraft's maximum capacity (so, if the ideal aircraft size is 200 and you are flying a ship with 150, you might have a 20% divergence factor, such that the most the aircraft could carry if demand is more than fully met and the 150 ship opposed would be 120) otherwise you are right back to the race to the bottom.

Commuter routes would naturally place a premium on smaller (relative to demand) aircraft, allowing for more flights per day.  In marked contrast, LH and VLH would naturally place a premium on larger (relative to demand) aircraft, expecting fewer flights per day.  Tech stopping would, of course, also lead to introduction of a divergence from ideal factor.

Such a system would serve to make a lot of long, thin routes viable that really aren't viable now by returning the rangiest 737s and A32x's to the game.  You would no longer need a 767 to fly 120 passengers per day between Kansas City and Madrid.  At the same time, the ideal aircraft for Chicago-Madrid would likely be something more like an A330, so flying that rangy 737 would hit you with a substantial penalty.  747s and A380s would be more viable than at present flying between Chicago and Madrid and still would certainly be more viable than the rangy 737, but, because they are not ideal, would not fly full or close to it.

I suppose this is, in a way, a refinement of the too small penalty with addition of a comprehensive too large penalty.  It would also apply throughout the game, not just on longer routes and would vary in what aircraft would be considered too large/small relative to range and passenger demand.

As Jumboshrimp pointed out, the effectiveness of tools we have to differentiate our flight offerings are very modest. The strongest RL tool, price, is very underrated here.
But to be honest, if it were made stronger, then big cash rich airlines could torture any opponent that tries to take a piece of the cake. Balance is key and really hard to  implement to perfection.
Frequency should be rewarded as it totally reflects RL, even on simple point to point traffic.
As players figure out the mechanics, the game becomes more predictable and to a degree boring as we repeat the same strategies.
What the game mechanics don't really help replicate, is the ability to play very different airline strategies:
- Leisure/holiday airline (low frequency+low price+ 2class HD seating)
- Low cost airline (high frequency, low price, 1class HD seating), 
- Mainline carrier (three class service, high frequency, high price & cost)
That being said, the game is already very mature, well balanced and new features (cargo, CBD) being added all the time.

A new approach could be to use a weekly demand concept instead of the current daily demand one. So demand from A to B could be 700 per week instead of 100 per day.
This way leisure airlines could fly twice weekly to a sunny island and fully meet demand. Africa would get into the game as well as many other low demand areas.

...it's getting late and I'm veering off topic... this has been one of the more interesting threads ... night

And there is also a matter of opposition. This is a multiplayer game, after all, and opposition is tough to predict. Which means if you are alone on the Beauvais-Prestwick line, for example(a likely scenario, it's not common for both airports to be played) where demand is around 300, well, a 777 flight is the less costly option.

Until an opponent appears and spams you with 4 daily A148 flights.

Thise game is a dynamic game, and each situation is different because of that. All the scenario shown are nice, but they are just possible examples. And depending on opponent's strategy, frequency can be the best choice...or not. if opposition is non existant, capacity is the way to go, as it reduces cost per seat. Things become complex when there is opposition, and sometimes, frequency wins. Sometimes not.

WOW! Talking about "rape" now? ???

Well, just a side note before I start answering: I haven't been here for long (just a year), so if things have been tried in the past, I much probably just don't know about them. Sorry if I repeat old ideas/mechanism, then.

I won't reply directly to the 1st part of Shrimp's last post as I think he misunderstood my proposal. I'm not trying to make the smaller planes (on a said route) king of the game, not at all. Only that it shouldn't get penalized in some situations.

And a second side note: I think you're (Shrimp) focusing always more on frequency-frenquency-frequency. Be careful, it becomes frenzy ;)

Explain my proposal a tad more (the last point you ask me about) could probably clarify things. But first I need to explain the background of my thinking.

--------

You're complaining about airlines spamming smaller planes and getting an unfair advantage, right?

PART ONE - CASE STUDY

Let's check certain situations first without their consequences in AWS, but just relatively to RL.

Examples

Example one
Let's take a relatively short and thin route like AMS-BOD (498 nm, 140 demand in GW#2)

Is one airline using a 70 seater targeting the right market compared to the plane? Yes, absolutely. And one using a 120 seater? Economically, offering one flight a day with no competition is the best solution, but as a matter of fact the 120 seater is ok but not perfect, as it's just choosing 1) to be vulnerable to concurrence, and 2) offering a poor service (because of only one flight a day - yes, frequency IS - in certain cases - a service, and not only because of connecting flights).

Example two
Same distance, but with a 650 demand.

The airline using a 70 seater is now out of its market, even though the distance is relatively short, because the demand is much thicker.
So the 120 seater is perfect here, while the 70 seater is ok but not perfect.
However, would a 70 seater get less pax IRL? Not sure, because it's still a short flight.

Example 3
650 demand and 1500 nm

Now the 70 seater is clearly out of its market, while the 120 seater is in its kingdom.

Application to current AWS

Example one
The 120 seater is penalized because it is not the best plane for the route (system works perfectly).

Quote from: JumboShrimp on February 17, 2018, 01:02:18 AM
So when the route becomes oversupplied, you allocate 100% of the pain to the larger aircraft.

Yes. Because the 120 seater is OK but not perfect. And its user couldn't ignore it in the first place.

Quote from: JumboShrimp on February 17, 2018, 01:02:18 AM
That is the very basis of frequency rape strategy.

No, because at a 1 flight vs 1 flight, you just can't talk of spamming or "frequency rape strategy". This is just wrong.

Example two
The 70 seater don't get any penalty because it's still a short hop. You could complain but at the same time one could argue the opposite and he wouldn't be completely wrong. Borderline case.

Example three
Currently the "too small" penalty applies to the 70 seater.

Conclusion
I'm not trying to privilege the 70 seater (or the "smaller plane" in general), but just to give him what is due to him. He has a target market on which he should reign an this is plain normal. If you're flying bigger on that same market, it's normal that you get a lower LF, because you're flying the "wrong" bird.
While one could always complain, the system is relatively balanced.

And I think you should define more precisely what you call spamming.

More examples

Now, let's consider cases in which spamming really occurs.
Example four
SVO-LED in GW#2 is 323 nm and has ~ 1350 demand, 2600 supply.
Currently flown by Metroliners, F.27, Viscounts, F.28, Tu-134, Il-18, Caravelles and 732. None gets the "too small" penalty because of the short length of the route.
Should the smaller ones get a little penalty? Probably, but also depending on the size:
- a large prop or RJ on that route is not completely unusual but not frequent.
- a Metroliner is completely out of its sector.
And who gets penalized? Everybody, because the route is so oversupplied that everybody suffers. The Metroliner user has sure a high LF, but the cost of slots was probably so high that it compensates.
And for that reason it would be better for everyone if we all flew 100/120 seaters. But the race to the bottom occurs, yes, definitely.

Example five
SVO-AMS in GW#2, 1157 nm, ~ 550 demand, 1630 supply.
The bulk of the supply are 732 and DC-9, but there are also 2 Tu-134 and 2 Caravelles, total 19 flights per day.
It's so heavily oversupplied that the effect of the smaller Tu-134 or Caravelle is quite marginal. Moreover, they represent "only" 22% of the flights and around 17-18% of the supply.
Again, the bulk of the problem is not the smaller birds but the heavy, global oversupply.

Example six
Lets consider the same SVO-AMS (same demand, same supply) but with much more props/RJs.
In this case, yes, the DC-9 / 737 users would suffer quite a bit.

Conclusion
Oversupply (with the "right" bird) is basically more a problem than smaller birds, even though in some cases a limitation would be welcome.
I hope those 6 examples helped us have some info about frequent situations.

PART TWO - THINKING

I think we all agree that:
- each plane category has its own target market and shouldn't be penalized by flying it
- categories / target markets have to meet at some point
However we don't agree about the repartition of pax when 2 types of planes meet on a single route:
- you state that they should have the same LF
- I think there is no reason if they both suit the route that one get more pax than the other, at least until it's not spamming.

Why do I think that?

To start with, a little reductio ad absurdum...
The 737 MAX-7 has 126 seats and the A321 NEO has 200, they're both in the same category, still the 737 has 32% less seats. On a CDG-LHR I'll win with my Boeing against your Airbus.
Same with 767-200ER vs 767-400ER, or 747 SP vs 747-300, that are within the same family. This is a never-ending story.

However let's focus on the idea of target market: there is always a part of the market where your plane will rule, and another part where it won't be so good. Where two categories of planes meet, there are usually drawbacks for both categories and this is normal:
- larger will have lower operational costs but less pax
- smaller will have more pax but more operational costs (+ slots)

The heart of the problem (apart from global oversupply) lies where RJs are used extensively against 737/A320, or 767 vs 777/A350, etc.
And I agree that there needs to be some kind of limitation, but only when spamming occurs. When the use is actually extensive. Not where this is marginal.

PART THREE - PROPOSALS

One single flight is not spamming. Thus I don't think one flight should be penalized.
But more flights, yes. And the thicker the route, the more flights with "smaller" plane there are (ie, getting always more in the definition of spamming), the bigger the penalty should be.

If I fly once with a CRJ on that perfect for 737/A320 route only once, the penalty won't apply as I'm causing no harm and not trying to siphon the demand. If I fly 10 times, yes, I'll be penalized.

But "10 times" is just an example. And penalty could apply right at the 2nd bird or 3rd, depending on the width and length of the route, and also the type of route (for LH, you wont have 10 daily flights from CDG to JNB, at least not in 1981). More flights, bigger penalty.

So this is actually a factor of:
type of plane (more or less in line with the route) + length of the route + width of the route + number of flights
About "type of plane", max seating of the variant would be more accurate, though. Some almost double between the smallest and largest variant and their intended use is not the same.

The calculation would then be a bit like for the actual "too small", but inserting the frequency factor. So what is penalized is not the use but the abuse, the excess.

Effects

1°) Regional airlines don't get penalized for flying to some large cities as long as they don't spam.
2°) Same for LH with thinner birds. If you use a smaller bird because your demand is thin you don't want to get penalized on your few thick routes (that's what is currently happening to me).
3°) People would be unable to use the CRJ or E-Jet or else to spam between main cities (and proportionally on different markets).
4°) Larger birds are safe as no one will be able to really contest their ruling without having penalties (spamming 767* in LHR in 2020 would be very costly if flown to large cities, but having a 767 on LHR-Guatemala City or LHR-Ahmadabad will not be a problem).

* if Cactus proposal to have a dynamic sizing of aircraft is approved, or if it is the max seating that is taken into account.

----------

I'm sorry this post is so long, but as it didn't seem that you got my idea, I thought a detailed explanation was welcome.
I'm not asking you to agree (even though I'd like it, as I think it would solve "your" problem without hitting the smaller bird just because he's smaller), but if you could understand what I mean it could be already great.

Thanks.

Quote from: Tha_Ape on February 17, 2018, 01:27:52 PM
WOW! Talking about "rape" now? ???

It's a gamer term where someone is utterly destroyed by a player or by some strategy.

there's also a factor we're not taking into account.

Small planes are good for frequency spamming, but as we see in GW3 small planes carry next to no cargo, my E175s carry 350kg light cargo max, no standard. Whereas the 737s/320s carry loadsa light and standard cargo and make loads of money. So, unintentionally, this is a tilt to favour large planes over small planes as cargo is a mini gold mine.

I still like the game as is and wouldn't want to see it changed.

Simon

Quote from: groundbum2 on February 17, 2018, 10:32:27 PM
there's also a factor we're not taking into account.

Small planes are good for frequency spamming, but as we see in GW3 small planes carry next to no cargo, my E175s carry 350kg light cargo max, no standard. Whereas the 737s/320s carry loadsa light and standard cargo and make loads of money.

Medium aircraft can carry standard  cargo if they are dedicated freighters, and you can use large aircraft to frequency spam busy routes, too. Remember too that cargo is sensitive to flight frequency, so the problem will still persist