Accuracy

[kotori87]

I was just playing a battlecruiser action, and watching the shell-cam. On the message list, I saw:
BC Lutzow firing 30.5cm L/50 at BC Queen Mary
and the shell-cam took off. The massive 30.5cm shells arched across the battlefield, coming closer and closer to the British battlecruiser line, until finally striking true. On HMS Lion.

After a few seconds of chuckling to myself that I'd be better off targeting the ship NEXT to the one I want to hit, I saw again:
BC Lutzow firing 30.5cm L/50 at BC queen Mary
and the shell-cam took off, again. This time, the flight looked a little different, as the shells that the camera wasn't following turned a little wide and landed well clear of the target area. However, the shells that went straight landed, once more, on HMS Lion. By this point, I seriously wrote down in my tactics book to target the ship next to my intended victim, and then Lutzow fired a third time.

This time, Lutzow finally put two shells into the bow of Queen Mary, starting fires and adding to the destruction that Derfflinger had already wreaked upon her. The rest of the battle was delightfully destructive, ending in the complete annihilation of the British BC squadron at minimal cost to my own forces, thanks to their timely collision with the entire HSF. I just thought I ought to share this amusing story with everyone before I start my next battle.

 

[bill44]

I've had that happen hit Princess Royal while tageting Qeen Mary or Lion, I really can't see the gunnery officer being court marshalled for sinking the "wrong BC" :laugh:

 

[Haida]

Just be happy you hit something besides water. :laugh:

 

[MajorDamage]

Perhaps it's a feature? There were plenty of examples historically of ships selecting wrong targets because of smoke & general confusion.

 

[HMSWarspite]

Were the targets broadside on or coming at you? Happens a lot if they are in line charging you... Harder to get range.

 

[babyblue]

Were the targets broadside on or coming at you? Happens a lot if they are in line charging you... Harder to get range.

To me it happens most often when they are charging at you diagonally. When they are still really far away, I usually target the second ship in line, so it has a chance to hit both the ship in front and behind it. And I've been doing that since DG as well btw.

 

[Slider6]

To me it happens most often when they are charging at you diagonally. When they are still really far away, I usually target the second ship in line, so it has a chance to hit both the ship in front and behind it. And I've been doing that since DG as well btw.

Ditto, I do same. A hit is a hit. I'll take it

 

[Blutarski]

New here, but a long-time buff of this period in naval history. I'm very hopeful about SES/Jutland. With respect to gunfire wandering in deflection ...

Assume that a salvo fired at a given target lands upon an adjacent ship.

If:

[1] Average interval between targeted capital ships in line is approx 500 yards.
[2] Average ToF at battle range is approx 25 seconds for a heavy caliber gun.
[2] Target inclination with respect to LoF is 60 to 90 degrees.

Then:

A target ship travelling at 18 knots would advance a distance of 200 yards. A target ship travelling at 27 knots would advance a distance of 300 yards. For a salvo to pitch or land 500 yards ahead of the intended target under the above mentioned conditions of fire implies that the speed of the target had been under-estimated by a margin of 45 knots. This is clearly far outside the spectrum of historical target speed estimation errors [a mean of +/- 2 knots error].

Therefore, the only events this phenomenon could reflect would be -

In a non-director controlled gun battery:
A mix-up between the spotting officer and his turrets as to the actual target. It was possible in difficult visibility conditions for turret trainers in non-director controlled main batteries to be physically unable to see a target ship which was visible from the spotting top. At long ranges a gun trainer could easily mis-identify the target ship in spite of target bearing data provided by the follow-the-pointer gear. A training error of < 2 degrees at 16,000 yards would put the turret onto an adjacent ship.

In a director controlled gun battery:
A misunderstanding between the aloft spotting officer and the director with respect to proper target. British warships circa-Jutland had no means other than telephone or voice tube communication to ensure that the spotting officer aloft in the FC top was observing the same target as the director was in fact targeting. If close attention was not being paid to fall of shot timers [not always perfect in themselves] and both targets were under fire, it was possible that the spotting officer could be fooled for a period of time.

On a related note, TIGER [IIRC] at Jutland had the pointing mechanism of one of her turrets knocked out of alignment by a hit and spent a fair period of time firing 15 or 16 degrees out of proper train with respect to the rest of the main battery.


I've not experimented enough within the game to come to any conclusion as to whether such events are over-represented in game play. However, generally speaking, range and range rate were FAR more difficult to establish than deflection. Gross deflection errors should not be commonplace.


Blutarski

 

[PepsiCan]

There's more. There is the impact of the shock wave in the case of heavy guns closely stacked together. And there are also random errors when the shell leaves the tube.

Even the laser guided gun on a modern battle tank will not hit the target at exactly the same spot twice. This has to do with minute wear in the barrel as well as minute margins of error in the manufacturing of the shell (not to mention the impact of air density and clarity on the laser). Also, weather conditions have an impact. Air is not equally dense everywhere due to tiny temperature differences, which can minutely influence the trajectory of a shell and can differ between shots. And now extrapolate that effect for a distance of 1500 meters to 20,000 meters.

With battleships, you need to add a third discrepancy: the shock wave of the gun. No way, can you fire all guns at exactly the same time. In fact all guns are fired separately. And when one shell leaves the gun slightly earlier than the other shells, at least some of the other shells will be impacted by the shock wave of the first shell.

Last but not least, there is the fact that no ship, no matter how large, is inmobile at sea. They all move. A slightly larger wave or bad timing and the gun is off by a quarter degree and there ya go!

In short, when you fire a gun over such distances, there will be a random error. The whole idea of salvo firing is that all the random errors equal out and thereby gives you an idea of when you are on target.

 

[rgreat]

These errors have a very small effect for a such large shells/guns and such a heavy platform as a large ship is.

If all these errors were severe then different shells in salvo will spread alot. Which they do not, unless there are malfunctions in turrets.

Btw: Even 0.25 degree of error is only give you 65 meters of miss at 15km.

 

[Blutarski]

Pepsi,

I agree with what you say as to sources of gun error. A 4 shot salvo of 13.5in British would produce a pattern of about 300 yards in length and about 50-60 yards in width at typical battle ranges and the errors of the individual shots within that pattern can be accounted for by the factors you mention. Maximum random deviation in deflection, barring a wild shot, could not be more than +/- 80 yards or so. A 500 yard error in deflection is of a completely different order of magnitude. If such a large deflection error is a case of an individual shell, then IMO it could only be classifiable as a wild shot, which is fairly rare. If such a large deflection error involves an entire salvo, the error would have to have originated within the fire process control, as no error of the guns themselves or any realistic meteorological conditions could produce such a large pointing error.

 

[HMSWarspite]

Fair enough points. my gut feel is deflection errors are a little over done. But why wont there be errors in the director? The hit rate was OK at WW1 battle ranges (haven't check it recently). We are only worried about the salvo scatter pattern aren't we?

 

[Bullethead]

New here, but a long-time buff of this period in naval history. I'm very hopeful about SES/Jutland. With respect to gunfire wandering in deflection ...

Welcome aboard! New guy buys the drinks :toast:

Assume that a salvo fired at a given target lands upon an adjacent ship.

If:

[1] Average interval between targeted capital ships in line is approx 500 yards.
[2] Average ToF at battle range is approx 25 seconds for a heavy caliber gun.
[2] Target inclination with respect to LoF is 60 to 90 degrees.

Then:

A target ship travelling at 18 knots would advance a distance of 200 yards. A target ship travelling at 27 knots would advance a distance of 300 yards. For a salvo to pitch or land 500 yards ahead of the intended target under the above mentioned conditions of fire implies that the speed of the target had been under-estimated by a margin of 45 knots. This is clearly far outside the spectrum of historical target speed estimation errors [a mean of +/- 2 knots error.

You are assuming that the ship is actually firing AT the target ship. This is NOT the case. The firing ship does not KNOW where the target really is, but is operating on an estimated target position. Furthermore, it has an estimated target course and speed, which gives an estimated future position of the target, which is what the the firing ship is actually aiming at.

Due to numerous sources of error in the system, where the target actually is, where the firing ship THINKS it is, the target's course and speed, and what the firing ships THINKS are the target's course and speed, NEVER correspond exactly, and at times are very far apart. As a result, it's quite possible for ships to be badly off in deflection for a while, even in a 1-on-1 situation. In a multi-target situation, there's also the chance that the director, rangefinders, bearing-takers, and spotters aren't all looking at the same target, which can really hose up the system.

I've not experimented enough within the game to come to any conclusion as to whether such events are over-represented in game play. However, generally speaking, range and range rate were FAR more difficult to establish than deflection. Gross deflection errors should not be commonplace.

I agree that getting the guns on line for deflection was easier than getting the range, but neither was a trivial job. In the game, the vast majority of misses are in range, not deflection, which reflects this. Still, being badly off in deflection did happen, and this sometimes shows up in the game.

The game system takes a huge number of variables into account. Therefore, the results of any small set of individual cases can show a large degree of variation in results. This is realistic, although it doesn't sit well with folks accustomed to much more deterministic wargames. However, with a big enough sample size, you'll see that on average the system produces very realistic results.

What I find interesting in this whole thing is that people aren't complaining about the hit rate at long range. IOW, this whole thing boils down to people complaining that school of fish A dies instead of school of fish B. .

 

[Paladin851]

Perhaps I view things differently ......How the game represents that missed salvo doesn't confront me much. I tend to look at the overall accuracy during the course of the battle and I believe SES (in single player !!!!!) has it pretty darn close. I have not however recorded the battles and counted every hit and shell splash,mine is strictly an estimate. As rgreat has mentioned above Multiplayer is an entirely different kettle of fish ( both school A and B ) with drastically higher accuracy levels.

 

[Bullethead]

Multiplayer is an entirely different kettle of fish ( both school A and B ) with drastically higher accuracy levels.

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

 

[smithcorp]

I have been reading Death in the Grey Wastes and there's some interesting comments about accuracy in the opening salvoes of the Run to the South. On one of the BC's (Lion?), the rangefinders were seriously overestimating the range to their target (by a huge factor) and so all the salvoes were well over and gunnery officers were probably making corrections from other ships' fall of shot.

It apparently took 9 minutes (!) for the error to be corrected and for the first shots to start to be ranged correctly (if my recollection is right).

I think it's important that the game simulate these kinds of errors and incompetence correctly - it may be inconvenient, but it's definitely correct for the period.

 

[jdkbph]

IOW, this whole thing boils down to people complaining that school of fish A dies instead of school of fish B. .

This kind of thing, as it relates to the perceived veracity of historical wargames, has fascinated me for years. It's not just a statistical exercise... I think this comes down to more of a suspension of disbelief thing. At least it does for me... in general. I have no opinion on this particular complaint.

What I mean by that is the player wants to be able to reconcile the results achieved in the game with what he/she knows or believes to be true. When that doesn't happen, either something is wrong with the game, or the player's perception is faulty. No matter how statistically trivial the error, it's a big deal if it makes the player feel uneasy about the results.

Therefore, if my premise is correct, it really would matter whether School A or School B bites the dust.

JD

 

[Zakalwe]

Well,

from the "Battle of Jutland", Monograph No. 1, Naval War College, Newport published 1920 (and somewhere here as link as ebook), page 51.

"The destroyer attacks carried out by both British and German squadrons showed that, against an enemy who is developing a rapid and effective gunfire, destroyers should not approach within 7,000 yds gun range of the enemy heavy ships. If however, enemy gunfire has been crushed or reduced in volume, the attack should be pressed home until a position of 5,000 yds...."

Ok, this is just a single source. But this is what I would expect. Now back to my CLs "Wiesbaden" and "Elbing", which tried to wipe the floor with these French TBS and British DDs off the Flanders coast. I at times was at 2,500 metres and less in bright daylight, and the hit rate was VERY low.

I think at ranges around 8,000 m and above the hit ratio is okay, but around 5,000 metres and below I would expect things being very dangerous for anyone swimming around modern Cls other than fish school a and b. And extremly dangerous at ranges below 3,000 metres.

That said, this is what I would expect and what is somewhat proven from beforementioned source. But I`m willing to learn.

`night

Z.

 

[Blutarski]

Welcome aboard! New guy buys the drinks :toast:

..... If you show up at Historicon, I will definitely buy you a beer.

You are assuming that the ship is actually firing AT the target ship. This is NOT the case. The firing ship does not KNOW where the target really is, but is operating on an estimated target position. Furthermore, it has an estimated target course and speed, which gives an estimated future position of the target, which is what the the firing ship is actually aiming at.

..... BH, I am by no means working on the assumption you describe above. I am working on the basis that the guns are pointed to the estimated future position of the target ship after passage of time equal to the time of flight of the shells at the estimated future range.

Due to numerous sources of error in the system, where the target actually is, where the firing ship THINKS it is, the target's course and speed, and what the firing ships THINKS are the target's course and speed, NEVER correspond exactly, and at times are very far apart.

..... Target course and speed were indeed imprecise, being largely determined by observer estimates; there were no instruments available in 1916 which could produce precision measurement of either value. But the case that I outlined in my original post, i.e. - a salvo at about 14,000 yards with a 20 second time of flight aimed at target A travelling broadside to the line of fire, but landing upon target B 500 yards ahead - suggests that the speed of target A had to have been entered as a value of true target speed + 45 knots [i.e. 65 or 70 knots]. I'd have to check my references but I don't think that it was even possible to enter such a huge deflection setting on a director sight. In any case, a target speed estimate in excess of 28 or 29 knots would have been unimaginable.

As a result, it's quite possible for ships to be badly off in deflection for a while, even in a 1-on-1 situation.

I agree that getting the guns on line for deflection was easier than getting the range, but neither was a trivial job. In the game, the vast majority of misses are in range, not deflection, which reflects this. Still, being badly off in deflection did happen, and this sometimes shows up in the game.

..... I agree that deflection was not always a "gimme", but I think we'll have to agree to disagree on the extent of the possible error.

In a multi-target situation, there's also the chance that the director, rangefinders, bearing-takers, and spotters aren't all looking at the same target, which can really hose up the system.

..... On this point I agree.

The game system takes a huge number of variables into account. Therefore, the results of any small set of individual cases can show a large degree of variation in results. This is realistic, although it doesn't sit well with folks accustomed to much more deterministic wargames. However, with a big enough sample size, you'll see that on average the system produces very realistic results.

What I find interesting in this whole thing is that people aren't complaining about the hit rate at long range. IOW, this whole thing boils down to people complaining that school of fish A dies instead of school of fish B. .

..... I think that one of the issues in play here is that printed wargame rules always permit a gamer to "look under the hood" in a manner of speaking. He can analyse the game mechanics, weigh the modifier values for various conditions and situations, and ultimately develop a godlike understanding of the game. PC games typically keep such data in a digital "black box" unaccessible to the gamer, who then is forced to develop estimates and assumptions about the workings of that box based upon his experience of the game.

 

[Bullethead]

Therefore, if my premise is correct, it really would matter whether School A or School B bites the dust.

I agree that most people seem to have preconceptions that are at variance with what the game's doing. But I submit that those folks who are doing this complaining haven't dug into the available data enough. They have gut feelings, bolstered by sweeping and widely repeated statements of "conventional wisdom", but neither is backed up by the available data.

Still, I do find it humorous that people are complaining, not about missing in and of itself, but by how far they miss by. That's an absolute first in my experience both as customer and developer .

Ok, this is just a single source. But this is what I would expect. Now back to my CLs "Wiesbaden" and "Elbing", which tried to wipe the floor with these French TBS and British DDs off the Flanders coast. I at times was at 2,500 metres and less in bright daylight, and the hit rate was VERY low.

Go dig into your Campbell. His main focus is capital ships, which of course mostly engaged at long ranges. However, he provides good (although untabulated--you have to do that yourself) data on hit rates at short ranges, from the CL and DD actions, which were mostly what engaged at these short ranges.

The hit rate at 2-3km at Jutland was on the order of 10%. A lot of this was in the dusk or at night, of course, but the ships involved were pretty well illuminated, and all the factors that should make hitting relatively easy at short ranges were in play.

Now, in the game, if you put ships under ideal conditions, 2500m apart, parallel courses and same speed, you'll get 30-50% hits, sometimes more. But if you add combat conditions of rather unparallel courses, significantly different speeds, possible smoke, bad weather, and the potential for human error in high-stress situations, and the hit rate falls down to near the historical level, but are still rather better than what happened in real life at Jutland.

BH, I am by no means working on the assumption you describe above. I am working on the basis that the guns are pointed to the estimated future position of the target ship after passage of time equal to the time of flight of the shells at the estimated future range.

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 think that one of the issues in play here is that printed wargame rules always permit a gamer to "look under the hood" in a manner of speaking. He can analyse the game mechanics, weigh the modifier values for various conditions and situations, and ultimately develop a godlike understanding of the game.

I agree completely. Plus even the most complex boardgame can't handle the number of variables as a computer sim. Thus, even with all the advanced rules, boardgames are much more deterministic than computer sims.