How does Passenger Loadout work?

[Agrajag]

Guys,

I'm in a Dash-8, which I think has 50 seats. My first flight I had something like 35 or 36 passengers, then 36, 37, 38 and now a couple flights at 39.

I'm getting the feeling that the number of passengers is directly tied to my company reputation. It seems to be very much in the ballpark as to how
this is determined.

If that's accurate, does everyone find that to be the way to go?

In the best of all worlds I would have liked to get realistic passenger loads that are dynamically generated based on a few factors. My company
reputation should be one of those. For example:

1) Why shouldn't the hour of the flight matter? Obviously I'd be more likely to sell more seats at busy times of the day instead of for a 2 hour
flight at 3am.

2) How about considering the date? Today is July 3rd (4th as I post this) which is a very busy time for flying. Shouldn't I be pretty booked unless
I'm just radically overpriced or have a really bad rating?

3) A lot has been in the news about how so many flights are booked solid (overbooked in fact) and I'd love to see that play a role here. Imagine how
hard it would be to please passengers who are on your average service, expensive flight only because they had no other choice! Consumers who go into a
process with a positive attitude are a lot easier to impress than those who going into one with a negative attitude.

4) It'd be rather intriguing to consider the locations flown. If I'm flying an "A" route, I'd likely have a backlog of orders. If I'm flying from the
middle of nowhere to the middle of nowhere why is my plane full of passengers?

In the end, what I'm getting at is the desire to see some dynamic seating. It'd be interesting to try and make a flight work where only half the seats
are sold, or I'd have to consider not flying that route any longer or changing an option to not track such a variable if I want to fly such routes.

Thoughts?

[Airliner]

FSP doesn't go that far in depth when it comes to exact numbers of passengers. I haven't tried it yet but, I would guess that if I tried to
load up a B744 in KACK (if it could even get in there!) on a flight to KHYA, it would give the same load as if I were going KJFK-KLAX.

The load is generated based on your company's reputation yes but, also on where you fly. For example, if you fly out of BOS all the time,
you will see a steady increase in loads, whereas if you fly out of BOS once, then PHL, then CLT, they won't increase that dramatically. I
think the idea is repeat service - the more you fly a route, the more PR you get. But as far as specific numbers regarding demand, that's
not modeled as it would involve tons of data entry on location population, daily enplanements, etc.

This is one thing I'm addressing to a small degree in my virtual world thing {plug! } in that, the cities you fly between, compared to the
aircraft you're in will help in determining the loads. For example, LAX and SFO are major cities, both category 5's on my scale. If you flew
a category 5 aircraft on such a route, you would get an average load. If you flew something like an E135, you would overbook. but, if you
also flew the route a number of times a day, this would then spread out that load a bit, though you would still be filling up a lot, so you'd
probably want to go up a class to a B737 or A320 or something to get more pax in there. On a route like the aforementioned ACK-HYA,
you have to Cat 1 airports. If you tried the E135 there you would have something like this:

(1 + 1) / 2 = 1
1 / 3 = .33 or 33%. Not a good match here. You'd want something like a KingAir, C402 or something. This, of course, will be added in
with many other factors such as time of day, # of flights on this route, reputation, safety, etc.

In terms of full-scale modeling of pax, that could get ugly. Here's a quick example - you fly the US Dash 8, right? Sounds like Piedmont,
which would mean operating mostly out of CLT. Now, suppose you fly a route, CLT-DCA. If you had some basis for generating a number
of pax by the airport sizes or the specific route, there are still too many variables left out. What time of day is it? Okay, that's easy. How
many other flights on this route are there? Hmm... How many pax are coming through CLT on their way to DCA? HMMM... What other
airlines are flying this route? AGH!!!! Again, a simple model would be to have some kind of value for airport size for each airport but even
that means a TON of work just to collect the information.

When I was working on a pure airline simulation (not connected to FS9), I got way into deep detail with pax generation and buried myself
in calculations. There are just way too many variables to take into account without a full schedule planner, as well as connection with
other airline's schedules. Let's look at your CLT-DCA route again. Someone else may be flying CLT-ATL-DCA (um, DL or FL
maybe? ). This does involve a stop but, they may also offer a lower fare, which could lure pax away from your nonstop. When I worked
at KMHT (that was before KBOS), I actually saw one of the most bizarre routings (trying to remember the exact route). It started: KMHT-
KLGA. But then it got really weird. If memory serves, it was KLGA-KDFW-KLIT, so we ended up with (IATA style): MHT-LGA-DFW-LIT.
WHY?! The woman simply told me she got a really good deal on the fare. It makes sense, as it was American Eagle to LGA, then AA the
rest of the way, DFW being an AA hub. But, she was going past her destination on her way to the connecting airport! If someone else
was running a flight MHT-LIT, this route would actually be considered in competition with that 3 legger!

Here's a link to what was looking to be a great simulation of day to day airline operations:
http://www.intrinsicalgorithm.com/ATM/index.php

Unfortunately, the guy took ill and it looks like the project's dead. Even now, some of the site's links are down (though I have some great
screenshots around here somewhere of nearly every single airline simulation ever made or conceived).

Okay, enough.

[Agrajag]

Thanks. My view is a simple one based on an old saying:

Even the longest journey begins with a single step out the door.

What I'm getting at is that I think I'd prefer even rudimentary shifts in the design to what is there now regarding passenger loads. Seeing it at
pretty much the same number every time and bumping based on reputation, pulls me out of the fantasy FSP is creating.

Even if the number were randomized lightly (35 today, 38 tomorrow, 31 the next day) I'd think that would feel more realistic. That'd be a first step.

The next step might be to factor in time of day. Busy times bump the randomizer up X% while slow times bump it down X%. The step after that might be
to identify all the big airports. Those lists are out there. Flying out of LAX will bump the randomizer up another X% while flying out of Ithica would
bump it down X%. All this could be done with just factoring in the big airports (all Cat 5 in your lingo).

The step after that might be to factor in both sides. Where you fly out of bumps the number as does where you fly into. Thus, the number will be
bigger if you fly out of LAX and into SFO than if you fly out of JFK and into ITH (Ithica, NY).

If you wanted to, later you could start to develop Cat 4 levels and go from there.

Just for clarity sake, I don't just fly Piedmont. There are three US Airways carriers flying Dash's out of KPHL alone. There used to be a fourth,
Allegheny, which is who I started emulating initially but they were absorbed into Piedmont. US Airways Dash-8's fly out of a lot of places around the
country so I pretty much just fly as "US Air XXXX" wherever I go. Generally I'm "US Air 4027". I do want to limit realism in one area. I want to be
able to fly pretty much anywhere with the plane I fly even if it doesn't actually fly there. That's sort of forced on you by VATSIM. I'm not going to
change planes, paints and call signs just because the best controlled area doesn't match what I'm in now.

[Airliner]

I hear ya. A simple +/- random generator that would adjust the final total to vary from flight to flight would be a quick fix.

In terms of what the pax generator is based on, I believe it's a flat percentage - someone correct me if I'm wrong. This means that if the
parameters come out to 75%, then whatever your aircraft can hold, it takes 75% of that - be it a B744 or a JS31. That would be my
sticking point, cause you couldn't realistically fly a jetstream JFK-ITH and fill it to 75% one flight, while taking the same flight the next day
in a B744 and STILL get 75% of that one. The most realistic would deal with the city pair itself, not the aircraft's payload size and that
might require a huge rework of the code in this area. Just trying to think through it.

About US, I remember Allegheny as they used to fly Dash 8's on the MHT-LGA run but I didn't realize that express carriers other than
Piedmont still flew them.

Colgan - SF34, B190
Chautauqua - E145
Republic - E170
Piedmont - DH8A/B
PSA - CRJ2 (are they still flying?)
Air Midwest - B190
Trans States - E145

Who else flies them?

[Agrajag]

I BELIEVE your assumption is correct and that flat number is what I'm suggesting is a bit too dull.

You could be right on the DASH-8's. The last time I looked others were flying them but then again, I hadn't looked in quite some time so I might have
to start looking for a new plane. Egads..... That's a scary thought.

I show the following as still flying them

Air Creebec "Cree"
Air Canada Regional Inc "Jazz"
Mesa Airlines "Air Shuttle" (I used to use "Air Shuttle" myself but then two controller complained about where I was flying so I dropped it)
Horizon Airlines, Inc. "Horizon Air"
Era Helicopters Inc. "Erah" (Helicopters?)
Labrador Airways Ltd. "Lab Air"
Voyageur Airways Ltd. "Voyageur"
Air Norterra Inc "Empress"
Caribbean Sun Airlines, Inc. "Caribbean Sun"
Airline Of The Marshall Islands "Marshall Islands"
Air Inuit Ltd. "Air Inuit"

None of these are in Philly so I guess it's down to them now there.

I love my turbo-prop and the PSS Dash-8 is the right fit for me. It has a decent number of passenger seats. It has a nice start-up sequence without
requiring a doctoral thesis to pull it off. It had decent support around the skies. I thought I'd move to the ATR but that's extremely complicated,
isn't being flown all that much, etc.

Thought about the ERJ's but I don't like the payware versions and they're jets so I just don't get the same fun out of it for some reason. I also own
PMDG's 1900C and D but neither fly using FMC's and I love flying SID and STAR's. Hmmm.

[jboweruk]

It's a flat percentage based on 2 variables:

1/ company reputation
2/ how many times you fly the route, more often = more pax

I know it's not the best but it works without bogging everything down. FSP is meant to add atmosphere not perfection, other than your
own perfection of course.