Speed number--figuring out ships speed

[Zosimus]

Originally Posted by 3catcircus

You're joking, right? Chasing it for 32 mins? One can easily calculate speed. Take a range/bearing and plot it. Do it again 3 minutes and 15 seconds later. Measure the distance it has traveled and divide by 100. That's its speed in knots. You can also calculate based on start/stop of the stopwatch as its bow and stern cross the wire, knowing ship length.

No, I'm not joking.

Oh, and I think the number you're looking for 3 minutes 14.4 seconds. That is exactly 3.24 minutes. In fact, if you multiply that number by 10, you'll get the 32.4 minutes that I mentioned above.

But no, I don't measure ship speed over 3m15s because it isn't accurate. You see, even if you have map contacts on, it will surely take you at least a half a second to click the x on the screen and then turn the stopwatch on. Even then, you'll notice that that x is not exactly in the middle of that box. And the problem only gets worse if you zoom in or out. Suddenly the x isn't even near the box.

In fact, a lot of times you'll watch the ship for 6m29s only to find that it appears to have covered 1.3 km. Really? The ship speed is 6.5 knots? I find that hard to believe. Maybe we should watch it out to 9m43s to be certain whether it's speed is closer to 6 or to 7. I suppose you propose to do all that underwater, scope up while the ship is moving away from you. Brilliant.

Or what if you're hunting by hydrophone? How exactly do you determine the ship length from 25m under water at 14 km away? Do you use your psychic abilities?

Or are you one of the types who fires a few torpedoes off and, if they miss, you just reload your last save and try again?

[3catcircus]

Originally Posted by Zosimus

No, I'm not joking.

Oh, and I think the number you're looking for 3 minutes 14.4 seconds. That is exactly 3.24 minutes. In fact, if you multiply that number by 10, you'll get the 32.4 minutes that I mentioned above.

But no, I don't measure ship speed over 3m15s because it isn't accurate. You see, even if you have map contacts on, it will surely take you at least a half a second to click the x on the screen and then turn the stopwatch on. Even then, you'll notice that that x is not exactly in the middle of that box. And the problem only gets worse if you zoom in or out. Suddenly the x isn't even near the box.

In fact, a lot of times you'll watch the ship for 6m29s only to find that it appears to have covered 1.3 km. Really? The ship speed is 6.5 knots? I find that hard to believe. Maybe we should watch it out to 9m43s to be certain whether it's speed is closer to 6 or to 7. I suppose you propose to do all that underwater, scope up while the ship is moving away from you. Brilliant.

Or what if you're hunting by hydrophone? How exactly do you determine the ship length from 25m under water at 14 km away? Do you use your psychic abilities?

Or are you one of the types who fires a few torpedoes off and, if they miss, you just reload your last save and try again?

I'll assume you meant no offense since tone can't be implied from your text.

I've done real time/bearing and geoplots as part of a Section Tracking Party in a former life. For purposes of the game, the pause key is the equivalent to that Section Tracking Party made up of multiple people. The minimal amount of tine necessary to switch to the nav map and hit pause is actually a lot less than people doing a real plot. How *I* do it in the game with map contacts off is get a range, note the bearing, go to nav map, pause, plot a mark, go back, hit the stopwatch, and wait. Get the new range/bearing at 3:15, go back, pause plot, get the course and speef from the plot, and do it again. Do that over the span of the 10 minutes spent on your final leg and you've got an adequate solution. I don't bother with the plot if I'm doing an overhaul maneuver once I've established his course.

[Leoz]

I use just about all methods mentioned depending on the situation. I use the final stopwatch check usually before a submerged attack.

Back when playing NYGM, I had the *perception* from observing a variety of attacks that target ships would change speed more often and my previous estimations could be off. Again: perception as I don't know the game programming efforts.

Have never really observed that much with GWX.

[BigWalleye]

Some observations:

1. We are at an advantage over our RL counterparts. We know that AI ships only move at speeds in integer knots. Check in the campaign.xxx files. Or turn on the God's eye view so you can get exact data. You'll see what I mean.

2. Knowing the target's speed to better than the nearest integer is not necessary for a successful firing solution. The target is BIG. At 1000 meters, optimum firing range as recommended by KM doctrine, a 140 meter long ship will subtend 8 degrees. The lead angle for a 90 degree shot is 14.5 degrees for a target moving at 10 knots. So if you aim for the center of the target, you can have a speed error of 27% at 1000 meters and still hit. At 2000 meters, quite a long shot, the allowable error drops to 14%. That's 8.5 kts, instead of 10kts. It is not difficult to distinguish between 8.5 and 10 kts.

3. Submarine commanders were warriors, not engineers and certainly not accountants. If you read the first-person accounts, they didn't attempt to determine target parameters with great accuracy, just enough to get the torpedo on the target with a good probability of success. Better to fire two fish, hit, and retire than to try for the perfect shot, get detected, and have to abort the attack. I don't recall ever having read of an RL sub skipper who regarded the attack as a math problem. That's for staff officers.

Remember that the torpedo is a big, powerful weapon. Any hit will do a lot of damage. A heart shot is not needed.

[Zosimus]

Well, there are lots of ways to skin a cat. For example, let's suppose that you don't know how to use a three-bearing method to determine the ship's course. No problem. Here's an easy and infallible way to determine the ship's course, range, and speed.

Assuming that the ship is either moving away or constant distance, simply steer directly toward the ship on the surface. It won't be long till you will be directly astern of the ship and following the ship on exact same course as the ship.

Then you can easily pace the ship to determine its exact speed.

Let's assume, for example, that the ship is going 22º at 7 knots. Just locate the ship and draw a circle around that ship with a radius of either 7 or 14 km. The ship will be on the 7 km circle in 32.4 minutes and the 14 km circle in 64.8 minutes. So as long as the ship doesn't change course (ships do that sometimes) you'll know exactly where that ship will be at a specific future time.

Then conduct an end around attack. I'm assuming pretty much anyone can do that. Since you know the target's course, speed, and location at a set future time, you can easily move out to 10 km distance (assuming normal, imperfect weather) and race around the target at full speed without being seen. Move in submerged, set up for the attack, and hit it with a two-torpedo salvo at 1 km distance.

All of this can be done without identifying the ship, knowing its draft, or knowing its beam. In fact, the entire operation can be conducted periscope down till the last second when you go up just long enough to take the shot and send it to the bottom.

[Leoz]

Originally Posted by BigWalleye

Some observations:

1. We are at an advantage over our RL counterparts. We know that AI ships only move at speeds in integer knots. Check in the campaign.xxx files. Or turn on the God's eye view so you can get exact data. You'll see what I mean.

2. Knowing the target's speed to better than the nearest integer is not necessary for a successful firing solution. The target is BIG. At 1000 meters, optimum firing range as recommended by KM doctrine, a 140 meter long ship will subtend 8 degrees. The lead angle for a 90 degree shot is 14.5 degrees for a target moving at 10 knots. So if you aim for the center of the target, you can have a speed error of 27% at 1000 meters and still hit. At 2000 meters, quite a long shot, the allowable error drops to 14%. That's 8.5 kts, instead of 10kts. It is not difficult to distinguish between 8.5 and 10 kts.

3. Submarine commanders were warriors, not engineers and certainly not accountants. If you read the first-person accounts, they didn't attempt to determine target parameters with great accuracy, just enough to get the torpedo on the target with a good probability of success. Better to fire two fish, hit, and retire than to try for the perfect shot, get detected, and have to abort the attack. I don't recall ever having read of an RL sub skipper who regarded the attack as a math problem. That's for staff officers.

Remember that the torpedo is a big, powerful weapon. Any hit will do a lot of damage. A heart shot is not needed.

Great read! Thanks.

[BigWalleye]

BTW, LGN1 has made a easy-to-use chart for finding speed by the wire-crossing method. You can download it here: http://www.2shared.com/file/qIEMFogN/lgn1fixedwire.html
Sorry, no pic.

[bstanko6]

Whatever method you use... be consistent. It will fare better with every attack you make.

[Zosimus]

Originally Posted by Leoz

Great read! Thanks.

Originally Posted by BigWalleye

Some observations:

1. We are at an advantage over our RL counterparts. We know that AI ships only move at speeds in integer knots. Check in the campaign.xxx files. Or turn on the God's eye view so you can get exact data. You'll see what I mean.

2. Knowing the target's speed to better than the nearest integer is not necessary for a successful firing solution. The target is BIG. At 1000 meters, optimum firing range as recommended by KM doctrine, a 140 meter long ship will subtend 8 degrees. The lead angle for a 90 degree shot is 14.5 degrees for a target moving at 10 knots. So if you aim for the center of the target, you can have a speed error of 27% at 1000 meters and still hit. At 2000 meters, quite a long shot, the allowable error drops to 14%. That's 8.5 kts, instead of 10kts. It is not difficult to distinguish between 8.5 and 10 kts.

3. Submarine commanders were warriors, not engineers and certainly not accountants. If you read the first-person accounts, they didn't attempt to determine target parameters with great accuracy, just enough to get the torpedo on the target with a good probability of success. Better to fire two fish, hit, and retire than to try for the perfect shot, get detected, and have to abort the attack. I don't recall ever having read of an RL sub skipper who regarded the attack as a math problem. That's for staff officers.

Remember that the torpedo is a big, powerful weapon. Any hit will do a lot of damage. A heart shot is not needed.

Your argument relies on a lot of bad assumptions.

Let's run with your example. The target covers 8º and we are firing not a center shot but a salvo shot (that's what KM recommended, isn't it?) with a 4º salvo angle that we can control when we switch to salvo mode.

For ease of understanding, we will label the boat thus:

3 2 1 0 1 2 3 (0 is dead center).

So your torpedoes are aimed at -2 and +2

Suddenly your 27% permissible error rate goes out the window. The acceptable error rate must surely fall to 13% or less.

Additionally, you are assuming that the ship is going 10 knots. Most of the convoys I attack are going about 7 or 8 knots. If we re-calculate with 7 knots, and 13% error that means +/- 0.91 knots. That means that if you think the ship is going 7 but it's really going 8, then one of your torpedoes is going to miss.

Furthermore, there's another possible error involved. The speed setting is analog. You cannot digitally enter 7.0 knots. You are moving the dial and it looks like it's 7, but maybe you actually set for 6.9 and you didn't know. Or maybe you tried to set the AOB for 90º at the moment of the shot, but because of inaccuracies, it's actually set to 88º

Finally, perhaps you're right that your average u-boat commander didn't care about the exact math involved. However, if we look at the list of the top 50 u-boat commanders at https://en.wikipedia.org/wiki/List_o...oat_commanders and arbitrarily pick 26 and 27 (about the middle of the most successful pack) we see that they are sinking about 3 ships per patrol and #29 drops to fewer than 2 ships per patrol.

So, basically, your average u-boat captain sucked. And if you do things the way he did, you'll suck too.

[BigWalleye]

Originally Posted by Zosimus

Your argument relies on a lot of bad assumptions.

Let's run with your example. The target covers 8º and we are firing not a center shot but a salvo shot (that's what KM recommended, isn't it?) with a 4º salvo angle that we can control when we switch to salvo mode.

For ease of understanding, we will label the boat thus:

3 2 1 0 1 2 3 (0 is dead center).

So your torpedoes are aimed at -2 and +2

Suddenly your 27% permissible error rate goes out the window. The acceptable error rate must surely fall to 13% or less.

Additionally, you are assuming that the ship is going 10 knots. Most of the convoys I attack are going about 7 or 8 knots. If we re-calculate with 7 knots, and 13% error that means +/- 0.91 knots. That means that if you think the ship is going 7 but it's really going 8, then one of your torpedoes is going to miss.

Let's go back and read the Submarine Commander's Handbook, Kriegsmarine publication number 1643, edition of 1943.

172.) If the range is over 1,000 m, or if there is uncertainty as regards the aiming data (high

speed of the enemy, several torpedoes (2, 3, or 4) should be released on the "fan" pattern. The

idea is to make sure of

one hit. It is better to score only one hit than to miss the target with each of several consecutive shots.

The target should therefore be covered by aiming at the boundaries a the area of dispersion on

the target; i.e., the shots should be spread by the width of the dispersion area in relation to one

shot aimed on the basis of the estimated data (if 2 or 4 shots are fired, in relation to an

imaginary middle shot).

So, according to the KM manual, the purpose of the salvo is not to hit with all torpedoes, but to ensure that at least one hits the target. By firing the 4-degree spread, our tolerable speed error increases from 27% to a whopping 45%.

Finally, perhaps you're right that your average u-boat commander didn't care about the exact math involved. However, if we look at the list of the top 50 u-boat commanders at https://en.wikipedia.org/wiki/List_o...oat_commanders and arbitrarily pick 26 and 27 (about the middle of the most successful pack) we see that they are sinking about 3 ships per patrol and #29 drops to fewer than 2 ships per patrol.

So, basically, your average u-boat captain sucked. And if you do things the way he did, you'll suck too.

So where is Kretschmer's treatise on the mathematics of the submerged approach? Where did Topp earn his PhD in Toppology? And what is your basis for claiming that what you do is any closer to Kretschmer's technique than to Rosenstiel's? Evidence, please.

Let's look at some of these commanders who you say "sucked." Because the names are familiar to anyone who knows the history of the U-boat war. Jenisch (26) sank 17 vessels in 6 patrols. Zapp (27) sank 16 in 5 patrols. Endrass (23) sank a "mere" 22 vessesl in 10 patrols. These were brave, seasoned, experienced naval commanders. Do not denigrate their performance just because you and I can achieve higher scores in a game which is a trivialization of the environment they fought in.

Now, you spent good money for your copy of the game, and you can play it any way you want. You can turn it into your high school science project in trigonometry if you wish. Or you can play at 27% realism (sic) and sink the entire British Navy every patrol. Whatever floats your (U-) boat.

But if I can use historically attested tactics and methods in this little game, and get results as good as Hardegen (rank 24, 22 ships in 5 patrols for 115, 656 tons), then I am having fun. If Hardegen sucked, and if Endrass and Jenisch and Zapp sucked, then well, I guess I suck too. It could be worse.

YMMV

Who said that SH3 was a simulation ?

No. SH3 is a game based on the German submarines of WWII, but it's far from a real simulation. Very far !

[Zosimus]

Originally Posted by BigWalleye

Let's go back and read the Submarine Commander's Handbook, Kriegsmarine publication number 1643, edition of 1943.

172.) If the range is over 1,000 m, or if there is uncertainty as regards the aiming data (high

speed of the enemy, several torpedoes (2, 3, or 4) should be released on the "fan" pattern. The

idea is to make sure of

one hit. It is better to score only one hit than to miss the target with each of several consecutive shots.

The target should therefore be covered by aiming at the boundaries a the area of dispersion on

the target; i.e., the shots should be spread by the width of the dispersion area in relation to one

shot aimed on the basis of the estimated data (if 2 or 4 shots are fired, in relation to an

imaginary middle shot).

So, according to the KM manual, the purpose of the salvo is not to hit with all torpedoes, but to ensure that at least one hits the target. By firing the 4-degree spread, our tolerable speed error increases from 27% to a whopping 45%.

So where is Kretschmer's treatise on the mathematics of the submerged approach? Where did Topp earn his PhD in Toppology? And what is your basis for claiming that what you do is any closer to Kretschmer's technique than to Rosenstiel's? Evidence, please.

Let's look at some of these commanders who you say "sucked." Because the names are familiar to anyone who knows the history of the U-boat war. Jenisch (26) sank 17 vessels in 6 patrols. Zapp (27) sank 16 in 5 patrols. Endrass (23) sank a "mere" 22 vessesl in 10 patrols. These were brave, seasoned, experienced naval commanders. Do not denigrate their performance just because you and I can achieve higher scores in a game which is a trivialization of the environment they fought in.

Now, you spent good money for your copy of the game, and you can play it any way you want. You can turn it into your high school science project in trigonometry if you wish. Or you can play at 27% realism (sic) and sink the entire British Navy every patrol. Whatever floats your (U-) boat.

But if I can use historically attested tactics and methods in this little game, and get results as good as Hardegen (rank 24, 22 ships in 5 patrols for 115, 656 tons), then I am having fun. If Hardegen sucked, and if Endrass and Jenisch and Zapp sucked, then well, I guess I suck too. It could be worse.

YMMV

Well, I think you need to go back and read the post again and/or take some remedial English lessons.

As I said, the numbers of U-boats ranged into the thousands. Assuming that there were 500 in operation at one time, each with 500 captains, your average captain is number 250 on the list.

We do not have stats for the top 500 u-boat captains. What we do have stats for is the top 50, and even in the top 30 we have some captains who return from patrols with an average of 3 merchants sunk per patrol. Some of these are big name people—names that students of u-boats would recognize. So if the top of the top are returning with 3 under their belt, a good number of mediocre captains are returning patrol after patrol with 0 kills.

By way of comparison, I returned from my latest patrol with 11 merchant kills—10 by torpedo and one by deck gun. Every merchant ship kill was with a two-torpedo salvo striking fore and aft. So yes, this is different from Kretschmer's slogan of one torpedo one ship. On the other hand, he enjoyed an advantage that I do not. In real life, ships hit by torpedoes fall out of formation and can be picked up later with the deck gun. The ships that I hit generally do not fall out of formation—only hits to the aft of the ship bring the ship out of formation. So your 45% error is just a bunch of bullcrud. If you hit a ship in the fore area and the rear shot misses, bounces, or prematures, the ship will pump out the water and continue on as though nothing ever happened.

Plus, in most cases, my crew cannot man the deck gun. Rain or high winds make manning the deck gun impossible, and I can easily experience 30 days straight of bad weather without so much as a 15 minute break to deliver a two-shot coup de grace to a stationary ship. We all know that SH3 weather is broken.

What do I attribute my success to? Patience and perfectionism. While others on here take 3m15s to figure out the course and speed of a ship, I generally take 65 minutes—more than an hour to make certain that alles in Ordnung. I calculate the exact angle to steer when I'm behind the convoy, one that takes me out to a safe distance quickly without falling farther behind the convoy. I submerge every 32 minutes to ensure that the convoy is still on course. I even calculate the exact angle the ship should be at when I'm leading the pack and planning to dive. Using that angle, I calculate the approach angle to use to put myself in optimum position in front of the convoy.

And I do it all with a compass, a ruler, and a protractor. I don't use any trigonometry at all. I just draw a few lines, sketch a couple of circles, draw a similar triangle or two, and measure the angles I need. There's no "high school science project in trigonometry." Just good, old fashioned human ingenuity—the same thing that let the Egyptians build the pyramids with nothing more than a 3,4,5 right triangle in their hand.

[Leoz]

Sorry I started the thread. My fault.

[Fanuboat]

Ok so someone said the 3min15 method was not accurate.

Ok ok... I subbed to this guy a time ago and well.
Just look this video and see your self how inacurate this method is.

And plz relax everyone, SH III is amazing but... Its just a game....................