Isn't the TMA a little too perfect?

[xXNightEagleXx]

It seems to me that the TMA is just simply too perfect once you reach high amount of intel. It reacts too much in real time with extreme precision. You can see the aiming point moving with excessive accuracy whenever the target changes course and/or speed. If i'm not mistaken TMA requires an amount of data collected in a lapse of time to elaborate correctly any changes and even more if there are multiple and abrupt changes .


I guess that rather than using directly the actor vectors (eg. velocity), which is probably what devs are doing, they should use others method that relies on known data history to elaborate the aiming circle. That would still be a simplification but closer to TMA behavior which is more liked.

[ChaosphereIX]

doesnt really matter in an era of homing torpedoes and wire guided weapons

I never use the blue circle

[Steiger]

Blue circle really only matters with the Mk.14, and if the target is maneuvering then that precludes the use of Mk.14 anyhow.

[Shadriss]

You'd be amazed at how quickly and accurately a good Sonar division can provide a shooting solution without even using the fire control computer. Granted, we'd never shoot a straight-run torpedo on them. The point I make though is that even in real time, it's possible to be very accurate even when the contact is turning and shifting speeds... at least enough to accurately place a seeking weapon.

It's not too perfect at all - in fact my problem is the opposite. It's not accurate enough when your solution percentages go DOWN. We knew where he was a moment ago, and now the solution is 20 Kyds further out suddenly? Ships don't move like that, and the solutions should reflect that even when the accuracy plummets somewhat.

[PL_Harpoon]

Originally Posted by Shadriss

It's not too perfect at all - in fact my problem is the opposite. It's not accurate enough when your solution percentages go DOWN. We knew where he was a moment ago, and now the solution is 20 Kyds further out suddenly? Ships don't move like that, and the solutions should reflect that even when the accuracy plummets somewhat.

I agree with that. In my opinion the time to first firing solution is about right but if after having a 95% solution the target starts to move around you you may not have a proper course or speed but the range should be pretty accurate for a while.

You may get that feeling from playing Dangerous Waters, but the problem with TMA in that game is that you get bearing updates every 2 minutes (IIRC). A modern sub at flank speed can do a 360 turn in that time, so obviously plotting a new solution might be a bit hard. If somehow you could increase the frequency to a couple of seconds (as I suspect real firing computer can do) it would be much easier.

[Destex]

Originally Posted by Shadriss

It's not too perfect at all - in fact my problem is the opposite. It's not accurate enough when your solution percentages go DOWN. We knew where he was a moment ago, and now the solution is 20 Kyds further out suddenly? Ships don't move like that, and the solutions should reflect that even when the accuracy plummets somewhat.

Exactly. I'll also add that ownship maneuvering should accelerate the solution more than it is now. If you have a good bearing measurement to the target (and you do if you track it by any Sonar array), two legs and you should get a reasonable solution.

[shipkiller1]

Originally Posted by Shadriss

It's not too perfect at all - in fact my problem is the opposite. It's not accurate enough when your solution percentages go DOWN. We knew where he was a moment ago, and now the solution is 20 Kyds further out suddenly? Ships don't move like that, and the solutions should reflect that even when the accuracy plummets somewhat.

THIS ^

I just sent a private message to the devs on this...
Solution generation is NOT accurate enough.. but it does make for an interesting game...

[Shadriss]

Originally Posted by Destex

Exactly. I'll also add that ownship maneuvering should accelerate the solution more than it is now. If you have a good bearing measurement to the target (and you do if you track it by any Sonar array), two legs and you should get a reasonable solution.

Truth. Though depending on the exact geometry and range, not always a given.

Originally Posted by shipkiller1

THIS ^

I just sent a private message to the devs on this...
Solution generation is NOT accurate enough.. but it does make for an interesting game...

Thought that might resonate with you, Senior.

Approach Officer : "Skipper! She TURNED! Her range just went from 8 Kyds to 30 Kyds!"

CO: "You are disqualified for life. Somebody get me someone competent up here please?"

Can't argue about the interesting game aspect though... I have to wonder how much of what they did was for 'gameplay' purposes, and how much because they didn't or don't understand how it works at the basic levels?

[Capt Jack Harkness]

I wouldn't say it was for gameplay reasons or a lack of understanding necessarily. Could be it was the quickest way to code it.

[PL_Harpoon]

Originally Posted by Capt Jack Harkness

I wouldn't say it was for gameplay reasons or a lack of understanding necessarily. Could be it was the quickest way to code it.

You could be right.
Thankfully there's a quick fix for that.
When you loose SOL on a contact that had a good solution the new range can't be further from the previous one than distance that target could travel.

[Capt Jack Harkness]

Sure. It adds another layer but shouldn't be a big deal, the game is so light on hardware as it is.

[shipkiller1]

Passive TMA can be a hard thing to do, but it is nothing but a geometry problem. Sometimes with very little data.

When you first pick up a contact, the first thing that is determined is the bearing rate and direction of relative motion. If tactically feasible, you change course across the Line of Site (LOS) and you try to drive the bearing rate. The first one or two OS maneuvers nails down the range to a very good 'ball park' figure, to something that will be 'tuned' over time. The next maneuvers nail down the Target Course (Ct) and Target Speed (Dmht). Sonar can help a great deal if they can pickup some specific target parameters. On a surface contact, this is almost always the case. Dmht is easy with a surface contact. Historical operational data helps keep it in the ball park if you do not have the aural clues...

Solution accuracy is determined after each course change. Does the expected incoming bearing match the solution? but more important, does the bearing rate match?? If not, then you have to adjust your solution for a better fit. This is why on a quiet contact, you may take a couple of hours to get a firing solution and put the boat into the proper firing position.

The problem with bearing data accuracy depends on the frequency of the incoming energy. The higher the frequency, the smaller the beam-width (more accurate), conversely, the lower the freq, the larger the beam-width so there is lots of bearing inaccuracies. This has a lot to do with sonar system design and we will not get into this at all.

The 1.05b update incorporated your submerged contacts counter firing... This is more like real life... so, you have to think about your firing position with respect to evading incoming fire. You may have to evade and re-position OS to press the engagement.

I personally like not having to do any hardcore TMA like you did in the 688i game. That game was just not realistic to do. Oh, it was somewhat realistic but 688i was essentially boring... you spent a long time doing TMA and all the other things that had to be performed. In real life, you had 30 people doing all those things but in 688i, it was all you... You got rushed, you missed stuff.

I think Cold Waters is much more enjoyable..

[stormrider_sp]

Originally Posted by xXNightEagleXx

It seems to me that the TMA is just simply too perfect once you reach high amount of intel. It reacts too much in real time with extreme precision. You can see the aiming point moving with excessive accuracy whenever the target changes course and/or speed. If i'm not mistaken TMA requires an amount of data collected in a lapse of time to elaborate correctly any changes and even more if there are multiple and abrupt changes .


I guess that rather than using directly the actor vectors (eg. velocity), which is probably what devs are doing, they should use others method that relies on known data history to elaborate the aiming circle. That would still be a simplification but closer to TMA behavior which is more liked.

Not only data collected, but accurate TMA also require the receptor vessel to manoeuvre several times during the collection process in order to receive the signal from different bearings as stated in Fast Attack's manual:

Change your course by at least 60 degrees and attempt to cross the
track’s bearing. When determining firing solutions, you should
change course every 10-12 minutes.
After changing course, your sensors are in a new location and, thus,
detect the track from a different bearing. An accurate firing solution is not
affected by your change in course, as your sensors detect the track from
the anticipated bearing. An inaccurate firing solution, however, causes all
new dots to be out of alignment.

It's in special during these situations that one notice the amount of simplifications and shortcuts that made its way into this game in order to shorten the overall development time.

[stormrider_sp]

Originally Posted by shipkiller1

Passive TMA can be a hard thing to do, but it is nothing but a geometry problem. Sometimes with very little data.

When you first pick up a contact, the first thing that is determined is the bearing rate and direction of relative motion. If tactically feasible, you change course across the Line of Site (LOS) and you try to drive the bearing rate. The first one or two OS maneuvers nails down the range to a very good 'ball park' figure, to something that will be 'tuned' over time. The next maneuvers nail down the Target Course (Ct) and Target Speed (Dmht). Sonar can help a great deal if they can pickup some specific target parameters. On a surface contact, this is almost always the case. Dmht is easy with a surface contact. Historical operational data helps keep it in the ball park if you do not have the aural clues...

Solution accuracy is determined after each course change. Does the expected incoming bearing match the solution? but more important, does the bearing rate match?? If not, then you have to adjust your solution for a better fit. This is why on a quiet contact, you may take a couple of hours to get a firing solution and put the boat into the proper firing position.

The problem with bearing data accuracy depends on the frequency of the incoming energy. The higher the frequency, the smaller the beam-width (more accurate), conversely, the lower the freq, the larger the beam-width so there is lots of bearing inaccuracies. This has a lot to do with sonar system design and we will not get into this at all.

The 1.05b update incorporated your submerged contacts counter firing... This is more like real life... so, you have to think about your firing position with respect to evading incoming fire. You may have to evade and re-position OS to press the engagement.

I personally like not having to do any hardcore TMA like you did in the 688i game. That game was just not realistic to do. Oh, it was somewhat realistic but 688i was essentially boring... you spent a long time doing TMA and all the other things that had to be performed. In real life, you had 30 people doing all those things but in 688i, it was all you... You got rushed, you missed stuff.

I think Cold Waters is much more enjoyable..

As far as I remember, one could automate some of the stations in 688i and still get along well with it. For me, personally, I enjoyed a lot more doing this trigonometry problems, handling the sonar station and torpedo setup than driving the boat in 3d with WASDQZ and firing and controlling torpedoes visually.

[shipkiller1]

Change your course by at least 60 degrees and attempt to cross the
track’s bearing. When determining firing solutions, you should
change course every 10-12 minutes.
After changing course, your sensors are in a new location and, thus,
detect the track from a different bearing. An accurate firing solution is not
affected by your change in course, as your sensors detect the track from
the anticipated bearing. An inaccurate firing solution, however, causes all
new dots to be out of alignment.

Shadriss, I hope you got as big a laugh on this proverbial BULL**** has I did... What a load of crap...