Air Grid visuals: follow up

Some of you missed a nuance of coder speak in the last post and clearly its picking up quite a few comments so I don’t want this to get missed in the noise.

The visuals I posted yesterday aren’t a mockup – that means that in no-way do they attempt to reflect the presentation you might see in game. I’m not going to be working on that, I merely borrowed a suggestion in an earlier comment and decided to “draw that” – using data taken from the live game.

My intent is to show you the spread of data. If all you did was click on the small animation, you probably missed the point entirely. Try the large animation (link here) and sit and watch it for 3 minutes. This is assuming that the system, when designed, covered the entire map. Don’t ask me if the system will cover the entire map, because I’m not designing the system and I haven’t yet seen the design for the system.

But, if it did, I think that a 20x20km grid system would provide such a density of information (over the entire map, but quite dense at regular map viewing scales) that complicating it with rules to let people hide from radar is just going to hurt you guys who want to sit on the airfield for 40 minutes forming up and preparing waypoints etc.

mapanistill.jpgLook at the imaginary visual again and really look at it. Don’t just glance at it and try to integrate the squares and colors with the underlying map. I didn’t just randomly color some squares over a map, I took data from the game, so the colors actually mean something. Orange = light enemy presence, Red = Heavy. Now, look again and this time, place yourself on either Wiltz (upper right) or Etain (bottom right) airfields… And remember – color = enemy aircraft. Starting to see something?

As Zheriz accurately observed, the information is spatially dense (2 bits per 3200km3) and yet it is remarkably sparse taken across any significant viewing area. Additionally, it is also spread out over time – 30 or 60 seconds per update? At 400mph you cover 1/4 of a grid in 30 seconds, 1/2 in 60, so you traverse an entire grid every 2 updates.

Some of you are positing the concern that limiting “Radar” to enemy factories/airspace will be more realistic and fairer, that opening it to the whole map will put you at a disadvantage.

I’m going to posit that you’ve listened to just a little too much forum whining and gotten suckered into some real bad math/logic.

If we model “Radar” and do it realisticly, only the enemy sees it. You see “Radar” for your airspace, but right now you are a buff formation heading to the enemy “Radar” zone. You see nothing until you get shot down by the enemy who is being guided by the game to your position.

How is that easier on you than a system that both sides see?

With “Radar” the enemy will see some representation of this

eview.jpg

while you and your escorts are looking at this:

fview1.jpg

instead of this:

gview.jpg

(I leave it as an exercise to Allied pilots to translate the pictures to the Axis factories because I’m too lazy to do two versions)

/report enemy bombers cheating, no way their escorts could have found us unless they are spying with a second account.

30 Comments

BTW – The original animation represents 112 (+/-3) aircraft on one side over 1/4 of the frontline during a non-peak play time.

“I’m going to posit that you’ve listened to just a little too much forum whining and gotten suckered into some real bad math/logic.”

Mr Smith, welcome to Mastermind, and good luck. Your specialty subject is “the bleedin’ obvious”, and your time starts… now!

An interesting thing would be an animation of a complete bomber flight from base to target (and back).

If there is one last comment I’d make about this, it is this:

As we look at the sample animated grid, which is using current pilot flight data, we have to remember that we are looking at pilots flying around the map who have no air grid. They are flying mostly random patterns in the sky looking for something to shoot down or strafe. The result is a large amorphous blob of data.

Once Air Grid is added, pilots will be flying less randomly since they’ll have direction from Air Grid. The Air Grid ‘picture’ of flying activity is going to look very different than it does now.

That change in behavior is hands down the hardest thing to predict. KFS1’s design is sound: it’s simple, fast and reasonable. The actual effect on pilot behavior and hence game play is something we can only guess at. I’d say lets get Air Grid into the next Beta or Intermission super-pronto and see what happens.

My only quip remains the resolution, I think it needs to be higher –10-15km per side — for it to have the desired effect of improving air combat.

That’s all I have to say about Air Grid. I’ll go be quiet somewhere now. :)

really, if you’re flying in over an enemy factory you could at least try to look out the windows on the plane to see if you can get visual on any enemies?

now, if I see a big red blob working it’s way from an allied rear AF I’ll be all over it and if it turns out to be a raid I’ll be able to attack it more than once before they hit the facotires and I’ll probably be able to shave off one or two escorts/bombers every time. now imagine I’m not alone…

if you’ve been in on any of the across the map raids where we trigger lots of ews giving the enemy lots of time you’d know it’s a difficult task that requires lots of escorts. say at least 20-30 escorts for every 10 bombers and there’s still gonna be heavy losses. most regular raids don’t have that kind of escorts so it’ll be really hard to get through.

and if the interceptors want to be really sneaky they can just hide in the clutter around the frontlines and the raid wouldn’t have been any wiser than if there were no radar.

it should be hard to make a successful raid, but not too hard. I’m not trying to be an ass here, I’m just pointing our things that I think are flaws that might hurt the air war rather than help it.

KFS1,

A little humor here:

“but right now you are a buff formation heading to the enemy “Radar” zone.” We don’t have a BUFF, so you’ll have to call our formations LUFF. :-)

The idea appears sound to me. I don’t agree with increasing the resolution of the reported area with the first implementation. It may be needed long term, but I’d start similar to what you show now and improve upon it if you find it’s needed.

With the current Air Templates and rules in the Air Forces, if you report only on bombers, it is going to be fairly easy to determine a Strategic Bombing run leaving from rear airfields. That’s how I’d use the data in any case. Currently we can’t resupply Bombers from Division to Brigades so anyone paying attention would know either one of two things were happening. 1. Bomber formation heading to a front line attack or 2. Strategic Bombing Formation headed to the Factories. So, BARCAP at or near the front will be up waiting on the formations. Just my 2 cents.

I’m not after trying to make the new reporting system useless. We all know something has to be done. My question again though is, will this map wide reporting include areas of the map that are not reported on now? Will Skirting be reported on? I am also curious as to how you will report on off map flights (Skirters). In the past, we’ve been told that you could not report on aircraft flying over the ocean. How now, will that be done?

So, in a short sentence, I’m happy to see us getting PHUCKED (Primary Heading Under Continuous Kilometer Enhanced Detection) in game. I’ve asked the wife what she thought about me getting PHUCKED and she says “not interested”. So, getting PHUCKED in game will alleviate some of the pent up frustrations of real life….ROFLMAO. After all, isn’t that one of the reasons we play this game.

The above paragraph is intended as a humorous addition to the discussion and should in no way be construed to represent my personal feelings about the new Aircraft Reporting System.

In the past, we’ve been told that you could not report on aircraft flying over the ocean. How now, will that be done?

You have to understand how EWS used to work, mechanically.

Your bomber would get near a CP marker, and would send a message to the hosts to turn on EWS in that CP. Since you can’t have a CP marker on the ocean, you couldn’t have EWS over the ocean.

The new system does away with that entirely – instead of being client-triggered based on getting near an object, it will be a new host-based process that keeps track of every plane’s reported vectors and amalgamates them in to the air activity grid.

It’s very much apples and oranges on the mechanics of the two systems.

The grid seems to big for me… because our actual EWS marks, gives almost the same detail.

A 10Km x 10Km grid it’s may be the best option for airwar encouragement. Because our visual range in game is about 4Km… and we speak of a 3D environment… that makes you hardly spot a plane that is 5Km below of you (something that in real life is possible, but it isn’t in the game).

A grid of 10×10 at least gives you a chance to locate the air force in the 2D space… more accurately, leaving the 3rd factor (height) as unknown.

Zheriz wrote:
As we look at the sample animated grid, which is using current pilot flight data, we have to remember that we are looking at pilots flying around the map who have no air grid. They are flying mostly random patterns in the sky looking for something to shoot down or strafe. The result is a large amorphous blob of data.

That’s a very valid point, but with 1 update per 30 or 60 seconds, I’d be surprized if it seriously impacted the data over the front line. The data is far too imprecise.

Several people have suggested more precise data :) Try to imagine 112 points of data instead of 40-something squares. For a start, each one goes from 2 bits *per square* to 48 bits per aircraft. That goes from ~80bits to 5,376 bits. OUCH. Buhbye 32kbps players.

Others have suggested something precise but still aggregated (“phucked”) – that’s going to require some pretty intelligent development work and a lot more data points – still – and in busy areas its going to be a real cluster.

We’ve talked about 16km or 12km (actually 16,384m x 16,384 – anyone guess why?)

Kizmet:
if you report only on bombers

I do not believe that is the intention. We’ve talked about whether it would be possible to depict bombers and fighters separately. Look at the still in the post, and imagine each square divided diagnoally. You might quickly be able to pick out a flight of bombers but … you’d also have 2x the amount of information making it harder to read.

Would like to thank you all for joining this as a discussion, I’m very pleased to have come thru this with my respect for bomber pilots in-tact!

– Oliver

Well, 16,384m is a binary number: 2^14. Divisible by 8 and multiples of 8. So it probably has something to do with byte or memory or matrix efficiency. No idea. Not a code monkey. =P

It’s a grouping of your cells, or a cluster of the cells. just like you can only have so many hulks per cell. Each cell knows what is in it, easier reporting.

The hosts and clients work in meters(*). So, using a power-of-2 meter size (16.384km) means that it becomes really easy to perform binary (fast) calculations on positions etc.


HostVehicle::calculateGridPosition()
{
gridX = myX >> 14 ;
gridY = myY >> 14 ;
}

is significantly cheaper than


gridX = (myX / 16000) ;
gridY = (myY / 16000) ;

* A fact that apparently eluded the original guys designing the terrain. Our terrain is based on cells of 800x800m. Someone thought “8 hundred, that’s a power of 2”. It’s not. 512m or 1024m would have been a power of 2, 800m is just a round number, which doesn’t mean its binary-convenient.

Is the 800M one of those items, that you ask yourself Why? Almost daily?

It was, for about 2 years, and then one morning a voice answered because “because they hired warm bodies“. One of the monkeys was clearly trained enough to know the number needed to be a power of 2, but just not actually intelligent enough to understand it had to be a power of 2 and not a power of 2 multiplied by something else.

Has anyone brought up the idea of only allowing the ML’s to see the grid?

I think this would probably make it more of a teamwork feature and a bit more realistic, as I’d suspect that IRL the flight leader would get the contact info and direct his flight accordingly.

teh?

drkpatch is officially voted off the island for that idea! and for using ‘teh’.

:)

Ooh…. now THAT is an intersting idea!

Furthermore What if it were availiable only to an ML on an “INTECEPT” mission, and the number of said missions allowable was limited?

This would really “elevate” the interception game by forcing proper flight tactics and communications. If that makes interceptions even more lethal (it should), then CRS scales back on the interceptor squadron ToE and/or the number of missions allowable.

Trout

——-
“Has anyone brought up the idea of only allowing the ML’s to see the grid?

I think this would probably make it more of a teamwork feature and a bit more realistic, as I’d suspect that IRL the flight leader would get the contact info and direct his flight accordingly.

teh?”

I thought the 800 meter octet side was directly a result of the DEM data being provided on a 400 meter grid.

In the same way that (per past Rat comments, anyway) the next-generation terrain will be based on DEM data on a 90 meter grid.

Particularly in locales with rapidly and irregularly changing elevation, I’d think it’d be difficult and unreliable to convert 400 meter data to a 256 or 512 meter game-grid, or 90 meter data to 64 or 128.

sniff,

if you know the bombers are coming, the escorts know you’re coming.

While its an interesting idea, there’s the potential negative of fighter jocks insisting on posting their own missions in order to get ‘dar and then ignoring the peons who join their missions hoping for some co-ordination :(

ahhhh good point!

Maybe more feasible after mission system revamp? You guys ever consider that there are way too many missions in general anyways? Like maybe each air brigade should have a mission?

We’ve been talking about creating “Company”s – which would be sort of ‘meta missions’, and limiting how many company’s each bde can have. Inactive companys would be greyed out, but you could still use them to spawn in as defenders, but you would have to take lead to assign an origin/target/etc.

Within each company players could group together inside the company to form a “detail”. The company leader could set specific details that he feels need completing.

There’d be bonus score for completing the task assigned to a detail by the mission leader for both leader and takers – e.g. “defend north deport”. If you go to the north depot on that detail and shoot ei from within 50m of the building, you get an extra bit of score per kill.

That’s how I’d like to see it done, but its probably not what we’ll see for 1.28.

Yeah, you’ve mentioned the “details” concept before and I like it.

Its apparently a given that one player “ordering” another player in any way is a non-starter, but requesting certain tasks be done in a visually represented/verified way would be great, especially if there was side bonuses for completing them.

I think that’d be great to have from the top down:

HC sets an objective, OIC takes on objective, OIC sets a number of objectives (such as air support, tank column, cap S CP, defend S CP, AAA, etc), MLs choose from objectives , ML sets objective (defend area, cover armour, ), grunt chooses objective offered up by ML.

Sounds great.

An excellent approach to integrate someday with three-tier KOTH ranking, points-consumption to spawn high value equipment, and highly encouraged mission-group proximity…I hope.

Yet another reason to update the mission system. Correct me if I’m wrong, but is this not the only game system that has remained unchanged in 6 years?

Strategic bombing missions really need a robust mission system to unlock all the gameplay value there.

Trout

————–
KFS said:

“While its an interesting idea, there’s the potential negative of fighter jocks insisting on posting their own missions in order to get ‘dar and then ignoring the peons who join their missions hoping for some co-ordination :(“

Responding to:

“If we model “Radar” and do it realisticly, only the enemy sees it. You see “Radar” for your airspace, but right now you are a buff formation heading to the enemy “Radar” zone. You see nothing until you get shot down by the enemy who is being guided by the game to your position.

How is that easier on you than a system that both sides see?”

Strat Bombing should not be easy. It sure as hell wasn’t easy by any stretch of the imagination back in the day so why make it easy now?

Unfortunately, though, you’re right. A Realistic Radar-like system would make it far too hard for the Bombers. Not only would the Defenders have the abilty to ‘track’ an inbound bomber group, but, much more significantly, the lack of dispersed targets makes it laughably obvious as to where the Target is. It’ll then almost be child’s play to put up a flight of defending Interceptors between the Attacking Bombers and the clumped together Factories. No real means for the Attackers to use any sort of Diversionary methods and strategies in order to put as many bombs on target as possible. Even the Compromise system, someone will figure it out how to track bombers…with the same result.

Don’t introduce Radar without giving Bombers a much larger array of targets to hit.

Well that’s cool, because we’re not introducing radar. In the meantime, we need to kill skirting and — according to exit polls — help pilots find the air fight and get them off the deck.

With plenty of pilots continuing to participate in the ground war, bombing bridges, attacking FBs, etc, etc, there should be plenty of noise in the center of the grid to act as cover for a lot of flight operations.

“Unfortunately, though, you’re right. A Realistic Radar-like system would make it far too hard for the Bombers. Not only would the Defenders have the abilty to ‘track’ an inbound bomber group, but, much more significantly, the lack of dispersed targets makes it laughably obvious as to where the Target is. It’ll then almost be child’s play to put up a flight of defending Interceptors between the Attacking Bombers and the clumped together Factories.”

In Aces High, radar sems to work fairly well. Of course they also have AI drone wingman bombers – giving the bombers the numbers they need to get a reasonable amount of bombs on target even if they get bounced on the way.

WTF? Why to hell attacking force should see absolutely ANYTHING on their freaking gps map over ENEMY territory?

Seriously?

Well, sorry to make you work for your comprehension, wizardd, but the whole point of this and the related posts was to explain exactly that.

Leave a Reply

Name and email address are required. Your email address will not be published.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

You may use these HTML tags and attributes:

<a href="" title="" rel=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <pre> <q cite=""> <s> <strike> <strong> 

%d bloggers like this: