(comments) AirwaySim v.1.3 news, previews & info

[LemonButt]

The extremes aren't supposed to happen because they are extremes.  Think of it as a price channel, not a prediction of what will happen.  Statistically, it is a similar concept to Bollinger Bands except forward looking.

[Curse]

Just take a look at the screenshots and compare it with the result. Nobody expects the fuel price to exactly hit the high or low line, however, it should orient way more and the predictions of the staff should not change each week massively! Both screenshots weren't taken too far apart from each other and the staff predicts totally different results.

The predictions were also totally useless - and I don't need useless predictions.

It's simply not the useful feature sami has announced.

[LemonButt]

I was curious so I whipped out Photoshop and did the best I could to overlay the graphs.  It is important to note in your screenshots that the graph scale changes constantly, so you cannot compare the 3 graphs until you put them on the same scale.  Thus, the attached graph uses eyeballed min/max values to overlay the lines.  It appears the predictions are probably more accurate than we both thought.  The first prediction resulted in the fuel price riding the low scenario, the second the high scenario, with the third TBD.

(edit: the colors are terrible because of the stupid 200kb upload limit so I saved in as gif)

[Curse]

Your graphic actually shows that

1) it's not a longterm (that is, in my idea, several month up to several years in this context) as sami announced in the feature announcement
2) adds nothing actually useful to the game (because staff changes their mind on a weekly basis from "fuel will skyrock and we will all BK" to "fuel will be basically free and we will earn money for burning it")


Conclusion: The system must be changed to be longer lasting accurate and actually predicting if the fuel price will increase or decrease the next 24 month. I'm ok with very very few failures in this (wrong predictions), however, those must be then visible for all airlines so nobody knows the truth.

[LemonButt]

Conclusion: The system must be changed to be longer lasting accurate and actually predicting if the fuel price will increase or decrease the next 24 month. I'm ok with very very few failures in this (wrong predictions), however, those must be then visible for all airlines so nobody knows the truth.

You can only hedge for 12 months, so who cares what it will be in 24?
"Very very few failures" is ridiculous.  The brightest minds on Wall Street can't even tell you where interest rates will be in 3 months with an 80%+ confidence rate, much less fuel prices in 24.

Your accuracy is only 80% based on staff morale etc.  That's not even close to statistical significance to say fuel prices will do x.  The system doesn't have to be changed at all--you're just using it wrong.  It goes back to the fact that the purpose of hedging is not to beat the market, but to mitigate risk and remove uncertainty.  If you want a guarantee, hedge your fuel for 12 months and you'll know exactly what you're paying.

[JumboShrimp]

I am going with LemonButt here.  I think it is a useful guide.  It gives 2 useful pieces of information:
- short term, 1 week, is it more likely to go up or down (just by looking at the slopes of the lines)
- long term, 1 year, is the trend up or down?  (take the end points, find the mid point between them, and see if it is above or below the current price.

combining these 2 pieces info, it may not necessarily tell you when to hedge, but it gives you some idea when not to hedge

[Curse]

Then the whole new system is useless. Why would I need people that tell me fuel can go up, down or stay the same - all at the same time? That's basically what I know myself.

And no, it doesn't tell you a trend. That's why I posted three images and you can see it on LemonButts graphic, too. It stayed on the same level even predictions were "fuel will skyrock and we will all BK" to "fuel will be basically free and we will earn money for burning it")


At the moment it appears to be totally random in what is displayed and what actually is going to happen.


Edit:
Another example: End of March 1976 all displayed a HUGE increase. However, since then the fuel stayed the same level, had a very small increase over time and now dropped nearly to the level it was end of March - and that all in less than two month.

Also, there is no greater accuracy than ~82,9%. Staff morale is 100% and sami said that's around the maximum possible. Please read also sami's announcement again before you answer:
https://www.airwaysim.com/forum/index.php/topic,26356.msg311295.html#msg311295

Especially what he described in "example 1" turned out not to work SEVERAL times in real live GameWorld#4 environment.


Edit 2:
This is also the feature feedback thread. It's really not helpful if you defend a feature that is useless and maybe not working as intended. Again, nobody needs a fuel prediction that tells you the same things you, your neighbour or even your great-grandfather could have told you ("price may increase, decrease or stay the same.").

[JumboShrimp]

Posting here if since we don't have the Feature Request forum open:

Along the lines with the projections, it would be nice to have a CI projection.  Based on the level of spending, where the CI will likely end up (I guess taking average delays into considerations, number of airports).

Also, it would be nice to have the amount needed to be spent on marketing to achieve certain level of desired CI.

[Andre]

Sami,  you mentioned that the new base structure might be rolled out to ongoing games as well. Could you please do this? Game world 1 and 2 have very few players now, and people feel that they can't do much more. There are so many opportunities in these game worlds that can't be used because of the base and aircraft limits. Please?

[LemonButt]

Posting here if since we don't have the Feature Request forum open:

Along the lines with the projections, it would be nice to have a CI projection.  Based on the level of spending, where the CI will likely end up (I guess taking average delays into considerations, number of airports).

Also, it would be nice to have the amount needed to be spent on marketing to achieve certain level of desired CI.

I would take this a step further and have a "Strategic Plan" page where, as the CEO, you can set your plan for CI targets, LF targets, fleet planning, profitability, etc. and then you can compare where you're at to where you want to be.

[Sami]

I do have plans for the new bases in other games too but it requires some test runs to make sure the existing airlines do not suffer from that, (have been hugely busy lately)...

[LotusAirways]

"Strategic Plan" page where, as the CEO, you can set your plan for CI targets, LF targets, fleet planning, profitability, etc. and then you can compare where you're at to where you want to be.

LB, as usual, another good idea from you.

And where would this be set-up? And should it be public info? Or maybe the player as a choice of making it public? And perhaps if it is public info, then credits can be earned by the airlines that meet those (long-term) objectives? Also, maybe some of the info that is not public can be made public on a random basis -it happens in real life, industrial spying, breach of information, etc.

As the present Dashboard page is rather "weak", maybe this is the right place to set-up the "Strategic Plan"? And then link it to Airline Information page so others can see it? 

Some ideas for goals vs. results:
- Image
- Load
- Profit margin (before tax please!)
- Airports served
- Countries served (new!)
- Fleet size
- Value

But then, it would be nice to compare it to peers. So maybe four columns: GOAL -- RESULT -- DIFFERENCE -- COMPETITOR -- DIFFERENCE TO COMPETITOR

Sami, maybe something for the next long north European winter...
LA

[Curse]

I do have plans for the new bases in other games too but it requires some test runs to make sure the existing airlines do not suffer from that, (have been hugely busy lately)...

Is there some kind of release date for GW#4? Just so we could prepare for things.

600 aircraft and 9 bases limit or do you want to change it?

[Curse]

Hello sami,

after you included COMAC C919 in one of the last updates and you said you'll include for example 777-8X and 777-9X, are you going to include the A330-800Neo and A330-900Neo, too?

They would be absolutely vital for GameWorld#3. If you need data for them I guess people would help you to make it possible!

[Sami]

Yes, if there is enough data, haven't looked

[ARASKA]

Will the Irkut MC-21 family be added?

[Andre]

I think it's awesome that more prototype models are being added. It would be great if the various MD-95/B717 planned models were added too, to make that family a more viable option. We have the nice opportunity to change aviation history a little bit in this game.

[mcfc2014]

I would like the MC-21/YAK 242 to be added as well.

Looks like it could be a winner.

[dmoose42]

Sami, given the new 'range' of prices that new aircraft can have that is more demand based - is it useful to display what the 'baseline' price would be without that adjustment?  Just so players can know what kind of deal they are getting.

[Aoitsuki]

a little thought(gw2)... the new dynamic plane system will hurt the airline that is behind in current game world. as they didn't have the cash to order before the implement of this system will makes it harder for them to grab decent aircraft. on the other hand the current order of popular aircraft would give everyone high 5 as they have already placed so much aircraft in the "discounted" price, and they can certainly order more since now the lessor player cannot order as much as they can.