Ekelund Ranging

[SeaQueen]

I've been experimenting with a spreadsheet that uses the Ekelund ranging formula and time correction give me an estimate of target range. Unfortunately, there's a big problem. Because DW only returns bearing measurments to the nearest degree, it introduces a significant amount of error into the formula, which already depends upon idealizations to begin with, so it frequently gives answers that are radically wrong. It's like... "no way... he can't be 145yds away at the turn, I'd have hit him!"

I was wondering if anyone else had experimented with this idea, and if they had any techniques for minimizing errors, or if they felt the technique was just unsalvagable?

[Henson]

Wow, I hadn't thought of that problem. A way around that is to maximize bearing rate on each leg (which we hope for already), maximize the change in bearing rate on each leg (which we should do already), and use longer legs to get a more accurate average bearing rate over time. The longer you're on a leg, the more likely you are to get a 'good' bearing rate (although longer intervals of data also decrease the accuracy of your RHek. There has to be a balance...grr.)

One question though...Are you calculating T* (time of the range)? That could be your problem too. Maybe he was 145 yds 25 minutes ago and you're trying to apply it for time now.

[SeaQueen]

Originally Posted by Henson

Wow, I hadn't thought of that problem. A way around that is to maximize bearing rate on each leg (which we hope for already), maximize the change in bearing rate on each leg (which we should do already), and use longer legs to get a more accurate average bearing rate over time. The longer you're on a leg, the more likely you are to get a 'good' bearing rate (although longer intervals of data also decrease the accuracy of your RHek. There has to be a balance...grr.)

One question though...Are you calculating T* (time of the range)? That could be your problem too. Maybe he was 145 yds 25 minutes ago and you're trying to apply it for time now.

Yeah... I get both the range at the turn and the best range time. I guess that's my other problem, I've been trying to do it like in the text books, and only use a few bearings. To do it very quickly. I guess I need to work more slowly.

[Linton]

SQ,could you post on here or pm me the formulae to have a look at please?You obviously take a very scientific approach to your game play-how log does it take you to input all the data and can you quantify how your calculations improve your solutions compared to the game generated ones?Finally I use 1 degree bearing error at 60 nm=2000yards,obviously you wouldn't be shooting from that far away so bearing error in minutes/seconds of a degree at maximum torpedo range should fall well within the capability of the seeker head.

[SeaQueen]

Originally Posted by Linton

SQ,could you post on here or pm me the formulae to have a look at please?

The formulas I use are mostly in all in the books I recommended already, except for the stuff I work out myself. If you understand the methodology in the books, though, doing your own calculations becomes easier and you can get creative. The Ekelund ranging formula is detailed in Wagner. It's a little bit subtle, honestly. It's well worth the read, though.

You obviously take a very scientific approach to your game play-

Regarding my approach, ASW is part science. That's what makes it interesting. It's the science of decision making in the face of uncertainty. That's particularly the case in the scenarios I make. I like them to be open ended enough so that there's not always a clear course of action, just better and worse ones that might or might not be obvious. The ASW game, to me, is fun because it's part science, part educated guesswork, part puzzle solving... There's a lot of room for shades of "maybe" in ASW and that's why I like it.

how log does it take you to input all the data and can you quantify how your calculations improve your solutions compared to the game generated ones?

In general not long. I believe in simple calculations. Anything complex that I need to think about, I work out ahead of time, so that way, when I want to apply a particular tactic, I have something "canned" so that I just need to plug in my numbers and go.

Like my torpedo preset calculator - If I want to shoot a salvo of torpedoes in a fan against a target with a certain maximum speed, all I have to do is plug in two or three numbers. Then I just need to read the bearings off the screen and plug them into my torpedoes. DONE. It's not a whole lot slower, just more rational.

That being said, the more steps you stick in any given process the more things can go wrong. I HAVE goofed things up before, like read the wrong number off my screen and shot my torpedoes in the wrong direction, and other goof-ball things. I'll chalk that up to learning, though. :-)

Ekelund ranging may proove to be a counter example to my usual philosophy, though. In an effort to minimize error, I need to take in data a lot more slowly.

Finally I use 1 degree bearing error at 60 nm=2000yards,obviously you wouldn't be shooting from that far away so bearing error in minutes/seconds of a degree at maximum torpedo range should fall well within the capability of the seeker head.

In terms of what direction I shoot my torpedoes that's not a problem. However, Ekelund ranging (and all bearings only ranging schemes) is very sensitive to small errors in bearing. The way to minimize the error is to drive very long legs.

Sometimes, actually, things I come up with don't improve my performance. That was why I posted about Ekelund ranging. I was trying to duplicate a what I found in a book, but couldn't. Ekelund ranging (which only ever yields an approximate range) in that case was coming up with answers that were so wrong that I couldn't explain it.

In the test case, I created, I was better off just using DEMON and building my solution that way.

Now... here's where I think it could be lethal, or at least helpful:

Slow, quiet, contacts -- you don't get DEMON until you're REALLY close, so you have no speed. Therefore, you have a very difficult time accurately estimating his range. I needed a way to more accurately estimate range based on bearings only. HOPEFULLY that will confer some tactical advantage to me. I hope it will be really helpful with accurately shooting ASW missiles.

Quantifying improved results, it's difficult to do that. This is a problem with real life experiments as well as video games. Some of it I just take on faith. I think the biggest payoff is not necessarily in having a better plan than anyone else, but have A PLAN at all that has some logic to it. :-)

I win some, I lose some. As long as the scenario distance scale is realistic, I'd like to think I usually have a pretty good advantage. I still hate these little "9 platforms are in a 10x10 box, have at it!" scenarios. They're lame-o.

[Bellman]

SQ: I promise not to open up 'old fronts' but you disparage the MP scene at your risk !
All are not engaged in 'Dooming' !

As you do not seem to take part in MP activity you grant yourself the indulgence of postulating 'theories'
and techniques which you can enjoy spreadsheet on lap within SP with the ever ready facility of 'PAUSE'
Linton: ''How long does it take you to input all the data ?''

Let the theorists and SP progress 'SAVERS' expose themselves to the challenges of 'PRACTICAL'
realtime competition - let us see a demonstration of theoretical skills in 'ACTION' or even have sight of any claimed 'system.'

My capitals do not shout but merely emphasise the fatal flaw in SP - the ability to enjoy extensive thinking-time
or to revisit situations and problems.

We await your new scenarios, with great interest, and hope, in view of your criticisms of those for MP,
that you will show the way forward ! (With less than 2 days playing time if at all possible )

[SeaQueen]

Originally Posted by Bellman

SQ: I promise not to open up 'old fronts' but you disparage the MP scene at your risk !

As you do not seem to take part in MP activity you

I do play MP games on gamespy. It's fun if you just want to shoot some torpedoes but if you want to play around with different search tactics, or asset allocations it's no good. I suspect that the virtual fleets might be a little better in a certain sense, but I don't have the time to commit to them. So far I have yet to see a scenario that really just nailed it for me. I've seen a lot of people try hard, and some come closer.

grant yourself the indulgence of postulating 'theories'
and techniques which you can enjoy spreadsheet on lap within SP

Absolutely. Part of the fun of wargaming is being able to do precisely that. It's what they do at the Naval War College. "What if?" is probably the driver of most wargames, one of the "What if's?" that people can indulge in is tactics. We do this in professional wargames for the Pentagon too. The term they use is "CONOPS development" but really it's just deciding what tactics they'll use in the game. Tactics, really, are the one thing you CAN experiment with and actually learn a lot about real world naval operations because the system performance numbers we use, while at their best are very well researched and hopefully close, ultimately are made up.

If you can't "what if?" tactics then you might as well not play wargames.

with the ever ready facility of 'PAUSE'

Actually, I specifically didn't want to do that. That's why I do everything ahead of time and keep it canned. All I need to do is type in a few numbers and read the answer off the screen. Everything should be quick and easy.

This actually mirrors what I've seen with real world operations. There exists a whole family of computerized mission planners and other tactical decision aids to be used for making recommendations.

The fact that one has the time to do this also reflects reality. Have you ever heard the joke that ASW means "Awfully Slow Warfare?" If there's anything ASW ISN'T, it's fast-paced. It's one thing I like about naval combat simulators, actually, is that I can read a book or do something else while I'm playing. Flight simulators aren't quite so slow paced. :-)

My capitals do not shout but merely emphasise the fatal flaw in SP - the ability to enjoy extensive thinking-time
or to revisit situations and problems.

I don't think that's a flaw. It's actually exactly what it's good for. You can do that in MP too, sometimes a lot more shamelessly. There's one MP scenario I've played, where I'm convinced that the only way for blue to win, is to have played it before. Without unnatural precognition it's a total sucker punch.

One of the things about wargaming, both professional and commercially available, is that it allows one to revisit things. And also, in a realistic ASW scenario, you should have a lot of think time. It should be possible to search for 16 hours and find NOTHING. It's completely realistic for platforms to go into the game, not find each other, and decide to knock it off.

Planning is important. Thinking things through is important. ASW by it's nature is extrodinarily cerebral. In this game there's a lot of options. The name of the game is, "what would you do?"

(With less than 2 days playing time if at all possible )

I've been working on another one. I'm trying to decide whether to make it MP or SP. It should be difficult, though, because it's a datum search with a fair amount of time late against a target in transit. The SSN in the scenario really only ever has 1 chance to catch the target and if he doesn't the whole thing is over. Because of that, he'll have to rely on the MPA to cue him, otherwise there's only about a 25% chance that he'll catch 'em. I originally made this one because I wanted to ask myself "okay... what if I had an MPA... what if I didn't because the weather was bad.... what do I do?" Obviously, you don't do as well. Regardless, I want to put them out as a pair. Good weather / bad weather.

This one is another long one, though. It takes more than 6 hours to play largely because there's a lot of transiting involved. I'm not sure how MP friendly that might be.

[Bellman]

SQ: I find it sad that your MP experience is gained in Gamespy. Sad because I think that that lower quality experience
has coloured your judgement of MP and MP scenarios in particular. Sad that you feel you cannot commit to one
dive a month which is often the only formal minimum commitment.

But mainly I'm very sad that a great fleet like GNSF is deprived your many skills. Finaly I'm sad that
you obviously dont appreciate the standards and mores of a fine bunch who try to uphold high standards and
sad that you cannot see the opportunity to measure yourself against them, stand tall, and make a
substantial contribution to the Multiplayer scene.

[SeaQueen]

Originally Posted by Bellman

Sad that you feel you cannot commit to one
dive a month which is often the only formal minimum commitment.

I wouldn't feel good about just doing one a month. Maybe in the summer, but during the rest of the year I'm way too busy.

But mainly I'm very sad that a great fleet like GNSF is deprived your many skills. Finaly I'm sad that you obviously dont appreciate the standards and mores of a fine bunch who try to uphold high standards and sad that you cannot see the opportunity to measure yourself against them, stand tall, and make a
substantial contribution to the Multiplayer scene.

I don't know anything about them. I can only speek from experience and so far MP has been mostly depressing. I really can't judge how the fleets do things, because I'm not in them.

Btw, what has any of this got to do with Ekelund ranging?

[Bellman]

Right on ! Just walking in some range finding shots on you SQ.....You know you're worth it !

But I will retire and hope to see you in the water - spreadsheets an' all.

[MaHuJa]

There is *one* place where you get the actual sensor bearing rate - the ffg towed array. The rest you'll have to average by time, as said earlier here.

That said, I somehow fail to see how this would not be redundant to the job done at the tma board...

[SeaQueen]

Originally Posted by MaHuJa

There is *one* place where you get the actual sensor bearing rate - the ffg towed array. The rest you'll have to average by time, as said earlier here.

That said, I somehow fail to see how this would not be redundant to the job done at the tma board...

I don't want the derivative of the bearing with respect to time. I actually WANT the time averaged bearing rate. That's how Ekelund ranging works. It's a way of estimating things based on it. They also calculate the bearing rate and display it on the navigation screen for all platforms. That being said, because I don't like how it calculates it, sometimes it's easier to just do it myself. It's usually more correct.

Sometimes trying to do these things graphically is more confusing and less accurate than just plugging the numbers in and getting the answer. Then, I use the numbers I calculate at the sonar station to inform the picture I paint at the TMA station. It's all about having more information.

One of the things I actually like about the FFG is that it gives you a lot of useful numbers that otherwise I'd have to calculate, like CPA distance.

[Henson]

Originally Posted by SeaQueen

Sometimes trying to do these things graphically is more confusing and less accurate than just plugging the numbers in and getting the answer. Then, I use the numbers I calculate at the sonar station to inform the picture I paint at the TMA station. It's all about having more information.

DING! We have a winner. A 'good' FT (known as TMA crewman in the game) does the basic math in his head first and then plugs the numbers into the computer to see how it fits. You need to be able to look at a sonar BB display and immediately see a direction of motion to be truly effective at self TMA.

The TMA problem is just a best estimate of relative motion, and bearing rate is a HUGE factor in every calculation you could possibly run within that framework. You may not realize it, but those nice lines you see in the TMA station display are nothing but a graphical representation of bearing rate, and using that is actually a lot less exact than using the numbers. By moving that course/speed strip in the window and watching the dots you are just letting the computer do that math for you. Some people have a good feel for moving the strip around, and some people like to do the math. I prefer doing both, because if they both come up with the same answer you're probably in the ballpark. If they don't agree, then either you screwed up, or you need more (better) data...and that's a good thing to know all on its own.

Using the Ekelund range formula and the basic range formula (Range=relative speed across the LOS/Bearing rate) along with a speed computed from the Demon display can give you a dead nuts course. All you need is a tool used by pilots to calculate crosswinds and the ability to count. It seems like a pretty long process when it's explained, but after you do it twice you have it, and it takes about 30 seconds from then on. I won't post it today, but maybe it'll come out later...who knows when I might feel ambitious?

The only thing missing in DW TMA is range rate tracking. THAT would spice things up a bit, but it's just not a part of the game (yet). Maybe someday...

[MaHuJa]

Originally Posted by SeaQueen

They also calculate the bearing rate and display it on the navigation screen for all platforms.

That number, in my experience, corresponds only to the solution entered at the tma board. If the solution is wrong, that number is wrong.

That allows the ffg to compare the solution bearing rate with the sensor bearing rate - though the player needs to remember that they are measured from slightly different points - at and behind the ship.

The number shown on the navmap is most useful for seeing which side you'll pass an object at - presuming the solution is correct, and remains so until that happens.

[SeaQueen]

Originally Posted by MaHuJa

That number, in my experience, corresponds only to the solution entered at the tma board. If the solution is wrong, that number is wrong.

True. That's why I do it on my own, and have a check. :-)

That allows the ffg to compare the solution bearing rate with the sensor bearing rate - though the player needs to remember that they are measured from slightly different points - at and behind the ship.

It's the same idea as what I do, only I do it by hand.

When I start playing with my Ekelund ranging, I don't even look at the TMA screen, I do it entirely from the spherical array display. Then I go back to the TMA screen and decide what to make of it.