Accuracy

[Lempereur1]

So I've heard. We're looking into that. Sorry for the inconvenience.

This may be due to the resetting of everyones Sliders to "Default" for MP.

Right now this is the only option.

We are considering an option to let the host's Slider settings be used for everyone.

 

[kotori87]

This has been a fascinating discussion, based on one short story I thought I should share with the forum. I must say that I am totally and completely disgruntled at the performance of my gun-layers and other fire control guys. The twenty or thirty broadsides it took before they started hitting is completely unacceptable, and these men have had their weekend passes revoked until they start hitting quicker. As for the idiot who notified my fire-control team they were hitting the wrong target, he was, um, "fixed", without anesthetic, and his "pride" and "joy" are both currently pickling in a jar on my desk.

In my other hobby, R/C Model Warship Combat, I have earned a reputation as a deadly accurate gunner. Where others hit 5% of the time, I hit 8% of the time. Where others hit 25% of the time, I hit 50% of the time. No matter what the range, no matter what the situation, ships are always holier after I've left them then before they met me. So whenever I play Jutland, I have absolutely no doubt in my mind that, if I were the one aiming the guns, I would be hitting more often than the guys in my fire-control teams. But in Jutland, I'm not a gunner. I'm an admiral, and all I can do is tell the ships what way to sail and what ships to shoot. And from what I've heard, the accuracy rates in Jutland are mostly about right, so instead of grumbling on the forum about the strange shooting habits of my lead battlecruiser, I put my best effort towards success, laugh at the silly stuff, and punish the fool who told my gunners they were hitting the wrong target. After all, if it isn't fun, then what's the point of playing?

By the way, did I tell you guys about the time my fleet bombarded the british with their own escorts? After all, what better to destroy the enemy than by shooting their own destroyers at them?

 

[rgreat]

This may be due to the resetting of everyones Sliders to "Default" for MP.

Right now this is the only option.

We are considering an option to let the host's Slider settings be used for everyone.

These "Default" settings are much more deadly compared to SP "Default".
At least it was in 1.005.

Check it out please, unless it was fixed in latest patches.

 

[Blutarski]

You really are assuming the ship KNOWS where the target is, and/or what it's course and speed really are. That's the only way you can derive a 45-knot mistake in speed--you have to measure that from the target's actual position. But the problem is, you have no idea where the firing ship THINKS the target is.

Suppose there's a 300m difference between the actual future position and the where firing ship THINKS that will be (which is really not very far off with this tech, at least at long range). Now you've only missed by 200m, based on what the firing ship THINKS is correct, not 500m, even though the shells land 500m away from the target. And 200m is about the same as the broadside length of a capital ship. IOW, not all that bad a shot. But because your only frame of reference is the actual position of the target when the shells arrive, you think it's a very bad shot.

..... I'm not sure what you mean when you say "You really are assuming the ship KNOWS where the target is, and/or what it's course and speed really are." If you mean to say that I have assumed that the firing ship would know exactly the target ship's range, inclination, and speed, the answer is no - I do not assume that. The best quality data that the firing ship could have would be a relative bearing reading to the target. What I do assume is that the firing ship sees the target ship and, through observation, has made and entered into its FC calculations some estimates as to its range, inclination, and speed. These data inputs are, of course, all subject to some degree of error, but calculations using these values will produce predicted ranges, deflections, and associated rates to open fire. Spotting of subsequent salvoes would then detect estimation errors and/or subsequent changes in enemy course/speed.


Taking my original gunnery example:

[a] range 14,000 yards with a 20 second time of shell flight.

opposing ships in battle lines on parallel courses.

[c] interval between ships is 500 yards.

[d] salvo lands correct for range but straddles the next ship ahead of the intended target in line.

We know the salvo is more or less correct for range and range rate because it has straddled the ship next ahead of the intended target. We know that the inclination has been correctly estimated because the salvo has landed upon the intended track of the target ship. We are left to understand what could have caused the salvo to land 500 yards off in deflection. Assuming that the target ship is a dreadnought in battle-line, the target speed would typically have been on the order of 15-19 knots. Let's say that the true target ship speed is 18 knots.

[1] The salvo is fired.

[2] During the 20 second time of flight of the salvo, the 18 knot target ship advances 200 yards.

[3] The salvo, however, still lands 500 yards ahead of the intended target.


Barring the sort of FC targeting mixup which we both agree could occur, the only other way that a deflection error of 500 yards can have occurred in this example WHEN THE SALVO IS CORRECT FOR RANGE (caps for emphasis; not for rhetoric), is if the target ship speed estimate had been entered as 63 knots. If you can suggest any other causes, I'd be interested to know.

 

[HMSWarspite]

..... I'm not sure what you mean when you say "You really are assuming the ship KNOWS where the target is, and/or what it's course and speed really are." If you mean to say that I have assumed that the firing ship would know exactly the target ship's range, inclination, and speed, the answer is no - I do not assume that. The best quality data that the firing ship could have would be a relative bearing reading to the target. What I do assume is that the firing ship sees the target ship and, through observation, has made and entered into its FC calculations some estimates as to its range, inclination, and speed. These data inputs are, of course, all subject to some degree of error, but calculations using these values will produce predicted ranges, deflections, and associated rates to open fire. Spotting of subsequent salvoes would then detect estimation errors and/or subsequent changes in enemy course/speed.


Taking my original gunnery example:

[a] range 14,000 yards with a 20 second time of shell flight.

opposing ships in battle lines on parallel courses.

[c] interval between ships is 500 yards.

[d] salvo lands correct for range but straddles the next ship ahead of the intended target in line.

We know the salvo is more or less correct for range and range rate because it has straddled the ship next ahead of the intended target.

Sorry - no you don't. All you know is the salvo landed at approximately the correct range. The Ship may have estimated a range that is too long, with a range rate that is closing. Or vice versa. You don't know...

We know that the inclination has been correctly estimated because the salvo has landed upon the intended track of the target ship.

No again. The inclination may be completely wrong, and producing the range rate errors above. All you know is that the combination of range, bearing, inclination (hence range rate and bearing rate if these have not been over ridden) put the shells 500m ahead of the ship. A ship can persist with an incorrect inclination for quite a long time if the spotted fall of shot keeps producing corrections in the same sense (al down or all up) and the Director crew do not pick up on it). If the range rate is +100, and the spotter keeps putting down 100 corrections on, the inclination will not necessarily be spotted as wrong (if the crew isn't on the ball)

We are left to understand what could have caused the salvo to land 500 yards off in deflection. Assuming that the target ship is a dreadnought in battle-line, the target speed would typically have been on the order of 15-19 knots. Let's say that the true target ship speed is 18 knots.

[1] The salvo is fired.

[2] During the 20 second time of flight of the salvo, the 18 knot target ship advances 200 yards.

[3] The salvo, however, still lands 500 yards ahead of the intended target.


Barring the sort of FC targeting mixup which we both agree could occur, the only other way that a deflection error of 500 yards can have occurred in this example WHEN THE SALVO IS CORRECT FOR RANGE (caps for emphasis; not for rhetoric), is if the target ship speed estimate had been entered as 63 knots. If you can suggest any other causes, I'd be interested to know.

You are assuming the only source of deflection errors is target speed estimate. As I have shown above this is not the case for the director anyway. But in addition, firing on the incorrect point of the roll can do it (wave motion can move the muzzles in any sense - usually actually a corkscrew motion which moves the muzzles in ovals, or actually Lissajous like figures (http://pagesperso-orange.fr/olivier.granier/meca/simul/lisajou/simul.html))

 

[RAMjb]

You are assuming the only source of deflection errors is target speed estimate. As I have shown above this is not the case for the director anyway.

Actually his example is pretty good. The highest bearing change for a given speed is to be expected if the target is in a 90º course relative to it's bearing as seen from the firing ship. As such in a perfect parallel course, a 500 m hit just in front of the enemy means a grossly speed over-estimating no matter what. If you infer that the director somehow "thought" the target was moving in a relative course other than in a perfect paralell course, then the speed adjustement to compensate the "wrong" course would've had to be HIGHER than the actual estimate the firing ship did in the first place. Not sure if I explain it clear enough, but hope so. Dunno if code commands work here but will try to put it in diagrams

X(splash hit).............................<----(actual ship)
                                                  |           
                                                  |
                                                  |
                                                  |
                                              <---- (firing ship)

Hit as described by Blutarsky, requiring (if range/course estimations are right on the enemy ship, an overestimation of speeds in the range of 20 knots)

                                                     /
                                                    /
                                                   / 
X(Splash hit)...................            |_ (target ship, wrong course and/or range estimation))
                                                  |
                                                  |
                                                  | 
                                                  |
                                             <----(firing ship)

What you describe as "possible explanation" of the gross bearing errors we were witnessing in jutland.
Here the overestimation of the speed of the target is even more, given that the bearing rate of change of the target is much less than in the first example (as the target's speed vector is angled towards us). Meaning, the director, to place a shot there, is estimating an speed MUCH HIGHER than in the previous instance, which was rediculous to start with...


Said that, I played a bit with the new 1.007 version and seems the huge bearing mistakes have been quite corrected now. Ships firing on target can fire off to one or the other side of the target, but not at the big distances we saw pre-patch. That doesn't mean that % have increased or that all shots are done towards the right bearings (it seems the % is similar, its just that the ships now don't go ridiculously off the bearing mark anymore). Misses come mostly for wrong range estimations than from wrong bearings (which also happen, of course, just not as wild as before), which also makes quite some sense. As I say, I've tested it just in a couple battles, but the gunnery in game seems to be perfect for me now

 

[Blutarski]

Sorry - no you don't. All you know is the salvo landed at approximately the correct range. The Ship may have estimated a range that is too long, with a range rate that is closing. Or vice versa. You don't know...

..... With all due respect, Warspite, I had carefully stipulated the prevailing gunnery conditions to reflect what I understood to be the original poster's experience.

No again. The inclination may be completely wrong, and producing the range rate errors above. All you know is that the combination of range, bearing, inclination (hence range rate and bearing rate if these have not been over ridden) put the shells 500m ahead of the ship. A ship can persist with an incorrect inclination for quite a long time if the spotted fall of shot keeps producing corrections in the same sense (al down or all up) and the Director crew do not pick up on it). If the range rate is +100, and the spotter keeps putting down 100 corrections on, the inclination will not necessarily be spotted as wrong (if the crew isn't on the ball)

..... You're absolutely correct that an incorrect inclination estimate can persist for a considerable period of time, because the error can only be detected and correct after the range has been approximated by bracketing or straddling the target. Before that moment, it is extremely difficult if not impossible to verify the inclination estimate by range-finder data only. But the only way that that an error in range rate or estimation could put a salvo 500 yards AHEAD of a target ship is if the target inclination was near 90 degrees. And that was not the tactical situation I laid out.

As regards errors in inclination estimates, they averaged less than 10-15 degrees. Trigonometry shows the effect of inclination error on deflection.

Assume the target bears broadside to the firing ship and is following an approximately parallel course:

> a 10 degree inclination error would produce a deflection error of < 2 pct;

> a 15 degree inclination error would produce a deflection error of < 4 pct;

> a 20 degree inclination error would produce a deflection error of 6 pct.

A 20 degree error in inclination estimate on an 18 knot target ship would effectively represent a +/- 1 knot deflection error, which is only about 11 yards. IMO error in the initial estimate of target ship speed was a far greater contributor to deflection error. Consider this for whatever it's worth: the post-Jutland RN 1916 Spotting Rules stipulated that initial double salvoes searching for deflection were to be spread laterally by only 2/3ds of target ship length.

You are assuming the only source of deflection errors is target speed estimate. As I have shown above this is not the case for the director anyway. But in addition, firing on the incorrect point of the roll can do it (wave motion can move the muzzles in any sense - usually actually a corkscrew motion which moves the muzzles in ovals,

..... No I do not. If fact, I agree with your points.

If the director is trained abeam, roll and heave will produce a vertical displacement of the cross-hairs while pitch will produce a lateral displacement. This typically produces a tall and narrow oscillation pattern

If the director is trained on a fine fore or aft bearing, roll will produce lateral displacement of the cross-hairs while pitch and heave will produce a vertical dsiplacement. This condition typically produces a short and wide oscillation pattern.

If the director is trained on bow or quarter bearing intermediate between the two above extremes, the oscillation pattern will display features midway between the two.

The precise nature of any such oscillation pattern would depend upon the severity of the sea state and the heading of the firing ship in relation to the wave front.

The worst case was firing at targets fine on bow or stern. Cross roll effects and trunnion tilt played havoc with deflection under those particular conditions - a problem not really solved until well after WW1. It was extremely difficult to place a narrow shell pattern upon a narrow aspect target - especially with the target zig-zagging or deflection chasing. A good example of this is the experience of INVINCIBLE and INFLEXIBLE at the Falklands: neither ship scored any hits at all on Von Spee's cruisers during the stern chase period of the action.

But:

[1] Out-to-out roll periods for capital ships were typically 14-15 seconds [keep in mind that a 5 degree roll was considered pretty dramatic] and director layers and trainers were well drilled to deal with roll and pitch effects. 500 yard lateral errors were IMO highly unlikely under any conditions.

[2] We are discussing Jutland - a battle fought in smooth seas.

- - -

Just to put matters in perspective, I'm not on any crusade here. And I am not trying to present myself as any sort of genius. I'm just trying to contribute to a forum related to a period of naval history in which I have had a very great and enduring interest over about 40 years. If my posts are going to upset people to the degree that they appear to be, I'll just hang back and lurk in future.

BTW, Warspite, I note that you live in Bristol. Say hello to David Manley for me.

 

[HMSWarspite]

BTW, Warspite, I note that you live in Bristol. Say hello to David Manley for me.

You know the Bristol has a population of 1/2 million, with another 1/4 mil plus in the greater built up area, right

 

[Blutarski]

You know the Bristol has a population of 1/2 million, with another 1/4 mil plus in the greater built up area, right

..... I was guessing that you might know David. He's a very active naval wargamer. A really nice guy as well.


Rgds / B

 

[HMSWarspite]

No, sorry. I have not done face to face wargaming much since Uni (22 years ago). Mind you was CinC RN for a WW1 naval campaign

 

[MUTbKA]

When we talking about accuracy and various effects from the director - we assume, that any director was a huge step toward increasing in accuracy.

If we see that director can produce a stable misses a way larger than simple eyesight, without any correction for speed an so on, or something like that of the pre-dreadnought era can provide - that's the point that make us feel that something is wrong.

If such a results were in reality - there must be strict order not to aim with director, but use anything, even looking through gun muzzles (for example), to provide a better accuracy.

 

[smithcorp]

Hang on, you're now dealing with ranges of more than 10,000 yards!

 

[rgreat]

In fact we are dealing with ranges of 1000-2000 m...

Check out light ships at these ranges. They suck at aiming compared to DG!
And even heavy ships shoot worse at point blank if compared to RJW...

I want close range DG accuracy back! :bite:

 

[Xavier]

In fact we are dealing with ranges of 1000-2000 m...

Check out light ships at these ranges. They suck at aiming compared to DG!
And even heavy ships shoot worse at point blank if compared to RJW...

I want close range DG accuracy back! :bite:

since version 1.007, short range gun accuracy seems perfect, in my opinion.

 

[rgreat]

It got better, thats for sure.

But still quite not enough.

In DG you got about 50% misses at point blank range. In Jutland you miss like 90% of the time (for CL's).
Thats a big difference.

 

[thewood]

I am not seeing the 90% misses at 1000 to 2000 yards. Can you post a save of a scenario where you are seeing 90% misses. I'ld like to compare to my testing.

 

[RAMjb]

heck, yes, gunnery seems quite fine now. I'm just out of the "old ships test" user scenario, and I've smashed the cr4p out of the british line at 8000-9000 yards while crossing their T (before they could come into 4000 yard range without serious consequences), and being hit only 6-7 times by their main guns in the meantime (so seems finally that the gunnery advantage of a force firing broadsides into a directly incoming line really matters).

Been playing quite some tests with 1.007. Gunnery is much, much improved now over what it was before...

So...SES...THANK YOU!!!!...now it's time to work on the IA and the destroyers' issues with firing torpedoes, and Jutland will be 100 over 100

 

[rgreat]

I am not seeing the 90% misses at 1000 to 2000 yards. Can you post a save of a scenario where you are seeing 90% misses. I'ld like to compare to my testing.

Done more testing.

Yes, seems i was wrong.
While hit pattern can be little weird sometimes, average close range hit rates are ok in 1.007.

Yay!

I never did any special test in later version of Jutland, so my expirience was based on standard campaign battles.

And seems like i was confused by 6'' guns not producing any significant explosion compared to DG.
Only small 'puffs' of smoke at the hit place...
These are kinda hard to notice. Just like if there was not hits at all...

P.S. SES, Please Beef Up graphic hit effects for 15-10-8 cm guns!!
My graphics card will surely sustain it... :smoke:

 

[Firestorm]

Even in 1.007 the gun accuracy at short ranges still needs fine tuning. In one encounter at night with a Full Moon and calm seas, I had 4 German CL's that where steaming on a parallel course with a Brit DD at less than 800 m and didn't hit the ship untill about 5 min. after starting to fire! Heck I almost had a torpedo hit the DD before any of the guns got close to a hit on the sucker. The German CL's should almost instant death to DD's at that range.

 

[rgreat]

Well, at least CL can hit another CL in ideal situation while moving same direction/speed. (As i tested)

Btw, secondary guns on BB's and BC's are easily shredding DD's and CL's to the peaces...
German CL guns (same 15cm) are devinetely not so effective... (in campaign at least).

Thats one of the reasons i told about 90% misses by CL's in 1.007...

Maybe it will be wise to specially test CL vs DD while maneuvering...