"Zig-zagging and chasing salvos" Topic

I know that a line of battleships would steer in a general direction while all or some of them (especially the ones under fire) would be zig-zagging or be chasing salvos.

My question is to what degree or course change and rate would they use?

I think battleship main guns rotated at 2-3 degrees per second so if a course rate change was greater than that they could not keep on target.

Changing over 45 degrees would most likely take them out of the line and may be out of the arc of fire of some turrets.

So does anyone have any hard data or historical accounts of how shos attempted to avoid salvos while still maintaining the line?

Thanks,
Wolfhag

Maybe you're familiar with the phrase "chasing shell splashes". The theory was that if the firing ship failed to achieve a straddle/bracket/hits, said ship's gunnery officer would order range corrections; the target ship, by "chasing" the old salvo, would thus avoid the new one.

The physics of it mean that a course change didn't have to be drastic. Firstly, a "good" salvo had to be tight--less than 40m, I believe, from one side to another. That means a range-change of 40m will get you out of a straddle/bracket. Next, battleships usually took about 30 seconds to fire a salvo, meaning you have that much time to move those 40m. A ship making 20 knots covers about 300 metres in a straight line in 30 seconds. Do the math, and it means that a ship altering course by about 7.7 degrees will change the range by those necessary 40m. A 10-degree change will give you room to spare.

You can see how even a minor course change would result in a significant change in range between salvoes. Once you factor in the shells' flight time, and the smoke, and the changing ballistic properties of the guns as they heated up and their linings wore away… Well, the miracle is that one battleship ever managed to hit another one at all!

Another interesting topic for discussion, Wolfie.

British WW1 practice allowed for evasive deviations of up to +/- two points (22.5 degrees) from the formation's base course. This could be accomplished without reducing the speed of the formation (formation speed was typically set at two knote less than the maximum speed of the slowest ship), nor would it greatly impact the gunnery of the evading ship. A deviation of +/- one point was in fact deemed sufficient for "fast" ships (say, 25+ knots). Such evasive maneuvering would typically commence when the ship had been straddled or bracketed. Even simple weaving would continually alter the range rate, which frustrated plotting; salvo-chasing would, in addition, frustrate salvo spotting.

Assuming a broadside to broadside engagement, a two point course alteration by a ship making 25 knots would change the range by about 250-300 yards in one minute or 100-150 yards in 30 seconds – ample to hinder opponent's spotting at typical engagement ranges. At close ranges, where time of flight was short, it was not so effective .

A note on salvo patterns -
The standard British expectation for the length of a 4 or 5 shot salvo at 16,000 yards engagement range (as stated in the 1916 Spotting Rules) was -

15in gun – 200 yards
13.5in gun – 300 yards
12in guns – 400 yards

For a four gun salvo, the effective "fifty percent" zone (the length within which fifty pct of the shots would be statistically expected to fall) would be about one-third the above-stated lengths. In the case of a 13.5in salvo, very roughly speaking, every fifty yards of deviation from the MPI of the salvo would reduce the likelihood of a hit by about fifty pct.

Final note –
Such evasive tactics remained effective even in a stern chase, the only difference being that they caused deflection errors rather than range errors.

Hope this helps.

B

One final final note -
Such evasive maneuvering would almost certainly have been carried out under "easy helm" (say, 15deg rudder). Any greater degree of helm would rapidly scrub off too much speed.

OK, one other item -
In this pre-RPC era, ships could not accurately fire while in the process of physically turning, so the price paid for such evasive maneuver would be some reduction in rate of fire.

B

Thanks guys, that's the detail I was looking for but could not find it in my notes.

So here are my thoughts on a gunnery model:
After the ranging shots, you start firing salvos. You never know the exact range but your range keeping error will be +/- X,000 yards. Constant maneuvering or slight course changes of 1-2 points should increase that range keeping error. By knowing the targets speed, course change and the time of flight will serve to increase the range keeping error by a known amount.

Example: Your range keeping error may be +/- 1500 yards but if you straddle with a 400-yard long salvo sheath you now know the range keeping error is +/- 400 yards. However, if the target changes course by two points moving at 18 knots and a time of flight for the next salvo is 15 seconds that should add about 100 yards to the error making the error +/- 500 yards. Is that correct?

A range keeping error of +/- 1000 yards is an error area of 2000 yards that a 400-yard salvo sheath will theoretically (most likely?) land within. If I'm correct that should give a 40% chance to straddle.

Knowing the angle of descent of the round fired (it's available) and the height of the target you'll know the danger space behind the target. Adding the beam of the target you get the total danger/hitting space. Correct?

So a ship with a 30-yard beam and danger space behind it of 50 yards would be a hitting area of 80 yards. If straddled by a 400-yard sheaf of 8 rounds that puts a round an average of every 50 yards in the length of the salvo. Would that translate to one hit with a 60% chance of a second hit?

Knowing the range is +/- 400 yards means a 50% chance of straddling on future 400-yard salvos assuming no course change by the target or shooter. Shortening the salvo sheath to 200 yards (if possible) would give a 25% chance to straddle but would put one round every 25 yards for 3 hits with a 20% chance for a fourth hit. If the target took evasive action that increased the range keeping error by 100 yards (making it +/- 500 yards or a total of 1000 yards) the 200-yard salvo pattern would have a 20% chance to straddle.

With a gunnery model like this, the player has a risk-reward decision to make just like a gunnery officer did. Increase the salvo sheath for a better chance to straddle with less or no hits or shorten it with a lesser chance of a straddle but more hits.

If turrets can rotate at 2 degrees per second they should be able to follow the target if the turn rate is not more than 2 degrees. Turning 2 points at 2 degrees per second takes 11-12 seconds plus the advance for the rudder to start the ship turning which is about 500 yards(?).

Question: Forcing a target to take evasive action of over 2 degrees per second would mean lifting fire until the turn is stabilized?

So am I on the right track?

Wolfhag

"After the ranging shots, you start firing salvos."
>>>>> Forget "ranging shots". Single ranging shots were only (TTBOMK) employed at Dogger Bank and, even then, simply to determine whether the target ship could be physically reached by the guns of LION, whose range-finders were useless at such distance. As soon as the reach was confirmed by the second individual shot, LION immediately commenced ranging by salvo (initially two-shot salvos from turrets A & B due to arc of fire restrictions).
- – -
"You never know the exact range but your range keeping error will be +/- X,000 yards. Constant maneuvering or slight course changes of 1-2 points should increase that range keeping error. By knowing the targets speed, course change and the time of flight will serve to increase the range keeping error by a known amount.
Example: Your range keeping error may be +/- 1500 yards but if you straddle with a 400-yard long salvo sheath you now know the range keeping error is +/- 400 yards. However, if the target changes course by two points moving at 18 knots and a time of flight for the next salvo is 15 seconds that should add about 100 yards to the error making the error +/- 500 yards. Is that correct?
A range keeping error of +/- 1000 yards is an error area of 2000 yards that a 400-yard salvo sheath will theoretically (most likely?) land within. If I'm correct that should give a 40% chance to straddle."
>>>>> Target plotting was a terribly complicated and difficult exercise. I honestly do not know how to establish a simple formula or model.
Establishing proper deflection (where shooter, fall of shot and target are all aligned) was relatively easy. But then it was necessary to establish some idea of the range. Your initial range-finder reading would give a range estimate for that particular moment in time. It might be accurate to within +/- 300 yards if your people were really on their game or +/- 2000 yards if they were not; it was difficult to confirm the relative accuracy of the initial range estimate until the target had been bracketed by successive salvoes.
You then needed to establish a range rate (the rate at which the range was changing over time – X00 yds per minute for example) and set the range clock accordingly. Theoretically this could be easily computed by the FC computer by inputting target speed and inclination versus own ship speed and course. Unfortunately, no means of accurately measuring target inclination existed and the range clock would be set by educated guess. Furthermore, due to geometrical issues, range rate was almost always changing in real time, even if the speeds and courses of both ships remained unchanged; this required constant updating of the range clock to account for "the change in the rate of change of range". The only means of verifying the plot was when the salvo spotting produced a straddle. Complicating that step was that it was quite possible for a misjudged target inclination to couple with a misjudged target speed to produce a situation where an erroneous plot might momentarily cross the real plot to yield a single straddle opportunity, yet be otherwise totally in error. Consequently, plotting was only a crude start.
Actual hitting required straddling and that could only be achieved via visual spotting. An inherent problem with visual spotting was that, at 10,000+ yards, the fall of a salvo could only be sensed as short or over; it was impossible to judge the error distance. Hence, the "bracketing" method was developed, for example – start short; fire salvos in succession with ranges increases of 400 yards until target is crossed and a salvo fall over the target; halve the interval to 200 yards and fire salvos at successively reduced ranges until target re-crossed to a short salvo. Continue process back and forth until a straddle has been achieved, then go to rapid fire and hope that the range rate was accurate enough to provide several straddles before the plot diverged from the actual course of the target.
All this is to say that, except for the most unique and rare situation where no component of relative motion between firer and target existed (perfectly parallel courses at identical speeds, for example), spotting onto target was a complicated and vexing exercise.
- – -
"Knowing the angle of descent of the round fired (it's available) and the height of the target you'll know the danger space behind the target. Adding the beam of the target you get the total danger/hitting space. Correct?"
>>>>> Essentially correct, assuming that the straddle was more or less amidships and the line of fire perpendicular to the center line of the target. If else, certain geek adjustments might apply in a minor way: a straddle more forward would involve a side height one deck higher plus superstructure and a beam slightly narrower; a straddle aft would often involve a height one deck level lower as well as a narrower beam. Also a line of fire not exactly perpendicular to the target center line would slightly increase the target width/beam (@ 30deg from perpendicular, the effective target beam would increase by 15 percent).
- – -
"So a ship with a 30-yard beam and danger space behind it of 50 yards would be a hitting area of 80 yards."
>>>>> 60-70 yards = close enough for government work, assuming a capital ship target and a typical 15deg angle of fall at WW1 battle ranges.
- – -
"If straddled by a 400-yard sheaf of 8 rounds that puts a round an average of every 50 yards in the length of the salvo. Would that translate to one hit with a 60% chance of a second hit?"
>>>>> All the guns participating in a given salvo are customarily aimed at the same specific spot in the ocean, with the fall of individual shots over N salvo iterations statistically distributed about the MPI as follows – 2/7/16/25(MPI)25/16/7/2 (each percentile distribution band equal in linear dimension). Let's assume for the sake of convenience that the effective target beam + danger space = the dimension of one percentile distribution band. If the salvo MPI is perfectly situated exactly at the waterline of the target ship, each shot in the salvo theoretically has a 25pct chance of hitting. In the long run (N salvo iterations) you can expect to score 2 hits on average in such a case; on the other hand (if my math is correct) there is still a 10 pct chance og scoring no hits. If the salvo MPI is 50 yards short but still straddles, each shot has a 16 pct chance of hitting. You could once again over the long run expect to score at least one hit, but there is actually about a 25 pct of scoring no hits.
When talking about the car more likely case of a four shot salvo, a straddle with the target covered by the 25 pct distribution band would be expected to score one hit on average, but there would also be a 30 pct chance of scoring no hits; a short straddle involving the 16 pct distribution band would represent a 50 pct chance of scoring no hits.
[Note: Evidence suggests that the BCF experimented with artificially spreading the individual shots of a salvo after Dogger Bank and may have employed the method at Jutland. The post-Jutland 1916 Spotting Rules deemed the method a failure and it was officially discouraged (my guess is that it probably had to do with difficulty in accurately judging salvo MPI). For reference, only director-equipped ships were capable of effecting such a manipulation.]
- – -
"Knowing the range is +/- 400 yards means a 50% chance of straddling on future 400-yard salvos assuming no course change by the target or shooter. Shortening the salvo sheath to 200 yards (if possible) would give a 25% chance to straddle but would put one round every 25 yards for 3 hits with a 20% chance for a fourth hit. If the target took evasive action that increased the range keeping error by 100 yards (making it +/- 500 yards or a total of 1000 yards) the 200-yard salvo pattern would have a 20% chance to straddle.
With a gunnery model like this, the player has a risk-reward decision to make just like a gunnery officer did. Increase the salvo sheath for a better chance to straddle with less or no hits or shorten it with a lesser chance of a straddle but more hits."
>>>>> Once again, leave the notion of adjusting the salvo sheath aside; it was not done "on the fly" and was unicorn-rare in practice. Salvo patterns were a function of basic gun ballistics and scientific calibration of the guns, the gun mounts & foundations, the sights and the ammunition (both propellant and projectile); even the projectile ramming and seating in the gun breech had an influence. The Germans were very studious in regard to calibration and their salvo patterns were almost invariably tighter than those of the British; German salvoes were consequently slightly less likely to straddle, but likely to produce more hits per straddle.
- – -
"If turrets can rotate at 2 degrees per second they should be able to follow the target if the turn rate is not more than 2 degrees. Turning 2 points at 2 degrees per second takes 11-12 seconds plus the advance for the rudder to start the ship turning which is about 500 yards(?)."
>>>>> My studies suggest strongly that, apart from rare and isolated cases (for example, WARSPITE's turrets firing in local control as she circled out of control at Jutland), the fire control systems of capital ships of the WW1 (pre RPC) era did not permit accurate fire while physically in the course of a turn, but only when upon a straight and steady heading.
- – -
"Question: Forcing a target to take evasive action of over 2 degrees per second would mean lifting fire until the turn is stabilized?"
>>>>> I don't think so in principle, but it depends upon the length of time a game turn might represent. If a turn represents two or three minutes or more, I's suggest just penalizing the rate of fire.

Wolfie,

Let me share a gaming exercise I undertook many years ago that produced some useful insights into this naval gunnery issue.
Take the cars out of your garage (or kick Buffy off the tennis court).
Get two 1:2400 scale battleship models and a 25ft tape measure.
Use a ground scale of 1-inch = 100 yards.
Use a time scale of one turn = three minutes.
Use 1-inch per turn per knot of basic ship speed.
Maximum battleship acceleration = 4 knots per turn if current speed < 1/3 maximum available speed, 2 knots per turn if current speed < 2/3 maximum available speed and 1 knot per turn otherwise. Maximum deceleration is the reverse.
Craft a turn gauge in the shape of a 7in x 7in upside-down "U" with semicircular dome on top, marked off in such a manner that: 4 knots of straight movement must be expended before starting a change of direction, first 45deg change of direction costs 3 knots, 4 knots for the next 45deg and 5 knots for any additional 45deg change of direction.
Place the respective ships on the floor within twenty feet or so of each other, without either player knowing the exact intervening distance.
Start the exercise by writing down movement orders and a range estimate for the first salvo of the turn.
Both ships move simultaneously,then conduct their gunnery in turn.
If the range of the written estimate </= 120 inches, the shooting ship gets a salvo rate of fire of 6 (initial estimate + five corrections); if </= 160 inches, a salvo rate of 5; if </= 200 inches a salvo rate of 4 and a salvo rate of 3 beyond. Deduct 2 salvoes for each 45deg change of direction.
Initial estimated salvo is marked by golf tee at measured distance. Each following salvo is placed 4 inches toward the target until target is crossed or a straddle falls within ½-inch of the target ship. If crossed, reverse direction and halve the interval. If a straddle is achieved, any remaining salvoes in that turn are also considered to have straddled.
Each salvo represents 4 shots from an 8-gun ship. Player may opt to fire by full broadside; if doing so the maximum number of broadsides may not exceed 4 per game turn, regardless of range.

Gotta run. If you're interested, I can post a bit more detail.

B

Even my sub had a hard time changing course by more than 2-3 degrees per second, and I had a huge rudder to work with.

Blutaarski,
I'm on board on everything except the ability to voluntarily fire a variable sized sheaf. Why? If a ship can fire a ladder ranging shot of three rounds from three turrets at different elevations wouldn't it be able to fire a salvo of one round from each of four turrets with each gun at a different elevation?

I understand wanting to fire a grouping as tight as possible but that needs to be balanced with the chance of straddling in undesirable conditions when a range keeping error may be 5% of the range.

If a ships fire control could give corrections as small as 100 yards couldn't it give a different elevation setting to four guns that would land theoretically every 100 yards? So if a typical salvo spread for a WWI 13.5" gun at 12,000 yards was 300 yards shouldn't he be able to lengthen it to 400 or even 500 yards even if the grouping may cluster around the center?

I did some checking in John Brooks book on Naval Gunnery and could not find data to back up my claim and I know you've spent hours going over source documents so I'll defer to you. However, I have seen pictures of gunnery trials of US ships in the 1930's that clearly showed a well spaced salvo at even intervals. Would that be because more modern equipment allowed it?

Thanks,
Wolfhag

Blutarski,
That's an interesting setup. I've done about the the same on a 12-foot floor estimating range and using Seas or War rules. I like the idea of how you walk the salvos over and under.

Personally, I like the idea of simulating building up a ranging plot between opponents because it reflects the differences between the rangefinders of the British and Germans in WWI rather than manual range estimation. Spending more time gives a better initial range estimation but delays opening fire. It's a risk-reward decision for the player.

Seas of War has a good gunnery system but I also liked the system I outlined above.

The player keeps track of each ship's target range keeping error which can be increased by environmental factors, near misses, smoke interference, maneuvering, etc. The rate of fire is determined by range unless going to rapid fire once you straddle but that can only be maintained for a limited number of salvos as each salvo will slightly increase the range keeping error because of poor observation from multiple splashes and target maneuvering.

There is one other way using a plotting board but I have to get that stuff out.

Wolfhag

Hi Wolfie,

"I'm on board on everything except the ability to voluntarily fire a variable sized sheaf. Why? If a ship can fire a ladder ranging shot of three rounds from three turrets at different elevations wouldn't it be able to fire a salvo of one round from each of four turrets with each gun at a different elevation?"

I understand wanting to fire a grouping as tight as possible but that needs to be balanced with the chance of straddling in undesirable conditions when a range keeping error may be 5% of the range.

If a ships fire control could give corrections as small as 100 yards couldn't it give a different elevation setting to four guns that would land theoretically every 100 yards? So if a typical salvo spread for a WWI 13.5" gun at 12,000 yards was 300 yards shouldn't he be able to lengthen it to 400 or even 500 yards even if the grouping may cluster around the center?

I did some checking in John Brooks book on Naval Gunnery and could not find data to back up my claim and I know you've spent hours going over source documents so I'll defer to you. However, I have seen pictures of gunnery trials of US ships in the 1930's that clearly showed a well spaced salvo at even intervals. Would that be because more modern equipment allowed it?"

>>>>> It is my understanding that it was indeed technically possible (for a director-fitted ship) to manipulate the elevations of individual guns to vary their ranges by some pre-determined amount in order to spread the salvo pattern. In fact, based upon BCF experience at Dogger Bank, Beatty wrote to the Admiralty (see "Beatty Papers", volume 1) arguing that an investigation should urgently be undertaken to determine the salvo pattern spread that would provide the optimal chance of scoring a hit when straddling. No evidence exists as to whether or not the Admiralty took up Beatty's proposal. However, it is my strong circumstantial suspicion that the post-Dogger Bank BCF gunnery committee convened by Ernle Chatfield under Beatty's direction (see Chatfield's "Navy and Defence") did develop BCF-specific gunnery rules that incorporated use of artificially spread salvos; I cannot conclusively prove it, as no BCF archival records for the period between Dogger Bank and Jutland have apparently survived.

We DO however know that, a year or so later, the "1916 Spotting Rules" (the first set of gunnery rules promulgated for use on a fleet-wide basis, addressed the practice of artificially spreading salvos as follows:

"The System of Firing and Spotting for Turret Guns.
Individual Shots in Salvos not to be Spread.
9. Spreading individual salvos both for elevation or range should be discontinued. Spreading for range has been tried and failed. The short shots of a salvo spread for range seldom allow any useful spotting information being obtained from it, and, so far as covering a larger area is concerned, it is a case of trying to obtain a chance hit on the enemy instead of taking much the same chance and hoping to hit him heavily.
Spreading individual salvos for direction has been suggested and is condemned, as, should a salvo straddle for elevation, the very shots which may give the most valuable information to the spotter for elevation may be out of line, and an wholly erroneous estimation of the fall of shot may be conveyed."

- – -

Re use of single shots for ranging, it is worth noting that the IGN, who arguably pioneered ranging ladders, employed 4-5 shot salvos and not individual shots. Use of single shots for ranging (at the 7+ mile distances involved) suffered a worse case of the very same problems that discouraged use of salvos less than 3 or 4 shots. The inherent dispersion characteristics of the projectiles at such ranges made it extremely difficult, impossible in the case of a single shot, to accurately judge salvo MPI. Dispersion might cause a single shot, otherwise fired at a perfect elevation/range to achieve a hit, to land 100 yards over or shot. And spotting salvo MPI onto the target was the most important part of good gunnery.

Hope this helps.

B

Blutarski,
Thanks, that's more enlightening. I like the idea of attempting to "walk" salvos back and forth across the target without a lot of die rolling.

I like to start a game out when both opponents come into LOS they start their range plotting. After a certain amount of turns (variable), the player can choose to start his ranging shots by rolling the dice to determine his initial range error in mils. When satisfied they fire their first ranging shots. It's a battle against time to get the first shot off and start straddling. In WWI there is a slight advantage to the German rangefinders. So it's a bit of a battle to find the range an open fire first.

So I can really see the reason for firing the tightest salvos, especially for spotting purposes.

Since you can only be assured of spotting unders, I guess it would incorrect to assume you knew when you straddled unless you observed a hit.

So when you straddle how do you determine a hit? Would you say that if a ship is turning, even at 2 degrees per second, it's guns are offline until the ship straightens out?

Wolfhag

Hi Wolfie,

"Since you can only be assured of spotting unders, I guess it would incorrect to assume you knew when you straddled unless you observed a hit."

>>>>> It was very difficult at battle ranges to observe a hit unless it produced a noticeable secondary effect (fire, explosion) or a visible detonation (non-pen hit); penetrating hits were almost impossible to sense if no secondary effect was visible.

Straddles, on the other hand, could be reliably sensed by counting the number of visible "shorts"; two visible shorts from a four-shot salvo in line with the target functioned as a pretty reliable indicator of a straddle with the target situated in the 50pct zone of the dispersion pattern.

- – –

"So when you straddle how do you determine a hit?"

>>>>> I'm guessing that you're talking about game mechanics here. Going from straddle to hit is a bit complicated; I will send a separate post when I have more time.

- – -

"Would you say that if a ship is turning, even at 2 degrees per second, it's guns are offline until the ship straightens out?"

>>>>> Yes, IMO – on two counts.
1 – Ships of the WW1 era lacked RPC (remote power control) to continuously adjust turret train and elevation as the firing ship was maneuvering.
2 – Fire control computers of the day, generally speaking, were only really capable of solving FC problems when both firing ship and target were moving upon straight courses. There were some "work-arounds" that theoretically accommodated own-ship maneuver while firing, but they were clumsy and not very accurate.

My approach has always been to observe what was done in practice rather than what was theoretically possible. The deciding factor for me was when I compared the Jutland gunnery logs of some of the British BCs to their maneuvering; it became evident (to me at least) that there was a cessation of fire while the ship was actually in process of performing any serious course alteration.

- – -

FWIW, 2 degrees per second is a pretty fast turn rate. We're talking 90deg in 45 seconds. I can provide some turn data later as well.

B

Blutarski,
Here is the latest version of the plotting table I'm using to determine hits from straddles. The salvos are moved towards the target like you describe but when they get within straddling chance I roll dice to determine range and lateral dispersion for the center of the salvo sheaf.

A model/image of the target is placed in the center with the correct angle off. This eliminates much of the hit modifiers that you need to account for in most games.

pix.sfly.com/-Yxq_w

Wolfhag