PongoDeMer
Posts: 81
Joined: 8/22/2008
Status: offline
|
Many years ago, I read about the Battle of Savo Island. I then tried to reproduce it using SSI's Fighting Steel program and immediately ran into problems. The biggest of these was the fact that when the action began, the allied ships were NOT at general quarters/action stations, but were at the second degree of alert. Consequently, there was a significant delay before fire could be returned. An additional complication was the fact that the U.S.N.'s second degree of alert differed from the R.A.N.'S second degree of alert which resulted in an additional delay before fire could be opened. None of this was modelled by SSI and all I could do to simulate it was to calculate the Australian and American delays and to use a stopwatch before allowing the allies to return fire.
Another crucial factor were the levels of experience of the ships involved in the action. SSI and other software manufacturers simply awarded 'green' ratings to all the allied ships and 'veteran' rating to the Japanese, which was a huge oversimplification. This was also (almost) completely wrong. The battle also featured a cloudbank which gave the Japanese an added advantage, plus a number of freakishly improbable occurrences which also had to be taken into consideration.
I recently came across my project and I have pasted it here. It is fully referenced, however Matrix forum threads do not support footnotes. If anyone wants them in Word format, please contact me via personal message.
Pongo de Mer.
The use of computer simulation as an aid to historical research:
The Fighting Steel Program by SSI/NWS
and the Battle of Savo Island.
By Richard Morgan
Copyright A. R. Morgan July 2008
Dedicated to Lieutenant Commander Mackenzie J. Gregory R.A.N. Rtd. and to all World War II Allied service personnel to whom we owe an incalculable debt.
Contents Page
Introduction 4
Converting reality into virtual reality 5
The scenarios 12
Conclusions 22
Bibliography 25
Appendix 1: The Fighting Steel/Fighting Steel Project program 27
Appendix 2: The Scenarios 29
Appendix 3: Maps 42
Introduction
I had never heard of the Battle of Savo Island until I came across the account contained in Dr Robert Ballard’s book The Lost Ships of Guadalcanal. The encounter took place in the small hours of the 9th of August 1942 and is regarded as America’s worst ever blue water naval defeat. Four heavy cruisers, one of them Australian, were lost to the Japanese, and only minor damage inflicted in return. It also emerged that a British Rear-Admiral on attachment to the Royal Australian Navy had been in command of the defeated forces. Ballard, an ex U.S. naval officer himself holds strong views on the subject. He clearly believes that this Admiral was at fault and bore a significant proportion of the responsibility. I was intrigued. What had happened and why? I re-examined his account. I began to look at individual ships. Questions began to emerge. What weapons systems did they carry and how would these vessels have been used - particularly in a night action? I needed more information. I remembered that I possessed a computer simulation called Fighting Steel (FS) which dealt specifically with World War II surface combat. This program also contained hard data concerning such information as speed, armour thickness and weaponry of all the ships involved. I located it and loaded it on to my computer, then upgraded it with the Fighting Steel Project patch v. 10.43 (FS/FSP).
The instruction manual included appendices on tactics, plus the comparative strengths and weaknesses of the United States, Royal (Australian) and Imperial Japanese navies. However still more information was required. A Google search on the internet yielded a number of different accounts of the battle. The best for my purposes were the Navy Department of Naval Intelligence Combat Narrative and the Wikipedia account both of which included plans. These gave a good over-view that was particularly useful in the preliminary stages of my research. Detailed information was found in Samuel Morison’s book The Struggle for Guadalcanal and Bruce Loxton’s work The Shame of Savo.
Converting reality into virtual reality
A brief outline of the battle is required. A Japanese force commanded by Vice Admiral Gunichi Mikawa, consisting of five heavy cruisers, two light cruisers and a destroyer surprised an American and Australian screening force of similar size, guarding transports that were landing a force of U.S. Marines who had been tasked with capturing Guadalcanal from the Japanese. They passed undetected by two American destroyers that were deployed to the seaward side of Savo Island on radar picket duty, plus another damaged destroyer that was in the vicinity; then in a surprise attack destroyed or disabled two separate forces of heavy cruisers piecemeal without serious loss to themselves. Finally they retired before daybreak to avoid the possibility of counter attack from the air. The battle comprised two separate actions: a preliminary contact between the Japanese and the southern American/Australian cruiser force closely followed by a second contact with the northern American cruiser force. Only the Japanese cruisers were engaged. Their lone destroyer was detached as rear protection and took no part in the main action.
I ran Fighting Steel’s Savo Island scenario, only to find that the Allies always reacted much quicker than they had in real life. Quite evidently the ‘caught with pants down’ reality of the actual battle was something that the program had failed to cater for. Even with the allied fatigue levels set to the ‘fatigued’ setting via the scenario editor included with the game, there was no noticeable difference. One felt in this instance that a further setting of ‘on chinstrap’ or ‘extreme fatigue’ was called for. However I had acquired a good feel for what the artificial intelligence (A.I.) could and couldn’t accomplish. I decided that I would have to start again from scratch. I designed a number of my own scenarios taking the limitations revealed into account and will describe them in detail later.
My first task was to assign individual ships with qualitative values as required by the Fighting Steel game engine. All national leanings and sentiment were cast aside. I allowed myself to be influenced by three factors only:
1) Naval training and doctrine of the three nations involved.
2) History and experience of individual ships
3) The performance of the ships in the actual battle.
Japanese naval professionalism and training circa 1942 was first class, with particular emphasis being placed on night fighting. Aoba, Kako, Kinugasa, and Furutaka comprised the Japanese Cruiser Division 6, all of whom had long experience of working with each other as a team. All these, plus Mikawa’s flagship the heavy cruiser Chokai performed well in the actual battle and consequently I awarded their crews ‘veteran’ status and rated them ‘expert’ in night fighting. The light cruisers Tenryu and Yubari had not worked with the other ships before. In the Northern Contact, they formed a second group (followed by Furutaka, which had dropped back after suffering steering damage). As a division, they were noticeably less cohesive in action and consequently I marked these two vessels down, awarding their crews ‘average’ quality and rating them as ‘good’ in night fighting. I gave these values to the destroyer Yunagi who I have included for the purposes of ‘what if’ or hypothetical scenarios. The fatigue level of all Japanese vessels was set to ‘normal’ to reflect that that the crews would have been properly rested and fed before going into action.
The R.A.N. had been involved in wartime operations since September 1939. However, Canberra had had a quiet war, having been employed as an escort for troop convoys to the Middle East, New Guinea and Singapore. While many of her key personnel had remained on board, efficiency had suffered through numerous personnel changes. Following a recent refit, wartime circumstances had prevented the normal working up period that would have followed, thus denying her the opportunity of returning to full efficiency. Therefore despite length of service and the expertise of key personnel, Canberra was not in a properly worked up state and consequently I felt that she did not merit veteran status. Instead I gave her the ‘average’ rating. Australia and Hobart did not take part, but have been included for the purposes of ‘what if’ or hypothetical scenarios. However, they had seen considerable active service and I have therefore awarded them ‘veteran’ status. The R.A.N. had trained at night, but not to the same degree as the Japanese. I therefore awarded all three ships the ‘standard’ night fighting rating.
The Americans were relative newcomers having only been involved in the war since December 1941. At Coral Sea and Midway the U.S.N. had pretty much written the book on carrier warfare, but at that stage its experience of surface action was very limited, and its techniques and doctrines were still largely unchanged from peace-time. A post battle of Savo Island comment by the American Commander Rear-Admiral Kelly Turner is worth quoting in full.
“The (U.S.) Navy was still obsessed with a strong feeling of technical and mental superiority over the enemy. In spite of ample evidence as to enemy capabilities, most of our officers and men despised the enemy and felt themselves sure victors in all encounters under any circumstances. Accentuating this was a mental uncertainty as to the methods and capabilities of fellow members of this heterogeneous force.
The net result of all this was a fatal lethargy of mind which induced a confidence without readiness and a routine acceptance of outworn peace-time standards of conduct. We were not mentally ready for hard battle. I believe that this psychological factor as a cause of defeat was even more important than the element of surprise.”
By far the most seasoned American ship that took part in the action was Chicago. She had been attached to an Anzac squadron (Task Force 44) since its formation in February 1942. Commanded by Captain Howard Bode, she had performed well at the battle of Coral Sea, working alongside the R.A.N. cruisers Australia and Hobart. She was well worked up and had become an established part of the Australian team. Her training with them had even extended to night fighting exercises. However she was short of gunnery practice, not having fired her main armament since October 1941. Her role in the battle of Savo Island has attracted much criticism, but this was due entirely to the decisions taken by her Captain and surely casts no reflection on the experience or ability of her crew. On the basis of her service record prior to the battle, I felt she almost merited veteran status, but I had to mark her down due to the lack of practice of her main gun crews. Consequently I have awarded her ‘average’ crew rating with ‘normal’ night fighting ability.
At that time there seems to have been no standard doctrine for night fighting within the U.S.N and the indications are that lip service had been paid to this particular skill. Chicago, as just mentioned, did have some experience of night fighting exercises. So it transpires did Astoria. Her skipper Captain Greenman had previously identified lack of night battle preparation as a weakness, so:
During the week at Pearl he took the opportunity to conduct firing and battle drills, held almost every night before leaving for Operation Watchtower. The drills (were) conducted in both readiness condition 1 and 2.
Consequently I have also awarded Astoria the ‘normal’ night fighting rating. However I have given all other American ships (Chicago excepted) a night fighting rating of ‘poor’. With regard to rating crew quality, outdated doctrine and complacency coupled with lack of wartime operational experience initially led me to give all the remaining American ships a rating of ‘green’. However, I made two further exceptions based on performance in action. Astoria responded quickly on coming under attack and she fired almost as many main armament rounds as Vincennes and Quincy put together before being silenced. For this I have given her the ‘average’ crew rating. Patterson skippered by Commander Frank Walker gave a performance that completely belied her lack of experience and consequently I have rewarded her with the ‘veteran’ rating.
Radar was a new technology, and at that stage in the war, only the R.N. and the R.C.N. on the Allied side had any significant operational experience with it. In the Pacific theatre, it was regarded as something of a black art. Destroyers Blue and Ralph Talbot had been placed on picket duty because they were the two ships whose radar performance had been deemed the most effective in past exercises and the Japanese still passed un-detected. Inside the sound, clutter caused by proximity to land would have also degraded performance. I have accordingly adjusted the radar quality of all allied ships to poor with the exception of Chicago. Although she was fitted with CXAM air warning radar, it should be noted that this was never designed to detect surface vessels. Also on the night of the battle it had been switched off. Consequently I have set her radar quality to ‘none’.
The Allied vessels had been covering the landings on Guadalcanal prior to the engagement. Most personnel had been a full forty eight hours without sleep. In consequence I have set the crew fatigue of all Allied ships to ‘fatigued’. At the time the engagement commenced, all Allied ships were in the second degree of alert - two excepted - which I will come to later. This turned out to be a key factor in the battle. It is notable that the FS/FSP program has not modelled degrees of readiness, assuming the crews of all ships in any engagement to be at general quarters. Ways had to be found of getting around this very significant omission. It was also found that the second degree of alert differed considerably between the R.A.N. and the U.S.N.
.
“The (Australian) second degree of readiness required all positions to be fully manned and the American condition two did not. In Canberra for example, personnel were at or near their action stations, but only half the guns and gunnery control positions were in an alert state. Personnel in the other positions were relaxed, perhaps sleeping or eating near their action stations. The call to action stations could be obeyed within seconds… Condition two in the American ships was similar… however there was a very significant difference in that those off watch could sleep in their cabins or in their accommodation spaces which could be some distance away from their general quarters stations. The result was a considerable hubbub in a ship which lasted several minutes as men moved to general quarters”
This is reflected in the reaction times of the two nationalities during the battle. The old saying “seconds in combat are like jewels beyond price” was never truer than on this occasion, and like priceless jewels, every one of those seconds has to be accounted for. The Americans, due to their more relaxed second degree alert state took significantly longer to react than the Australians: with the notable exception of Patterson who had spotted the Japanese before they opened up. Her Captain, Commander Frank Walker was already on the bridge. She responded instantly, turning to port immediately after coming under attack. The only completely alert ship on the Allied side, she was effectively in the first degree of alert. Her gun crews must have been at or near their stations, for they opened up with star shell during the turn. Chicago was nominally in a high degree of readiness but her crew was exhausted:
Captain Bode, after consultation with his officers, ordered a modified condition 1 alert to rest his crew after two harrowing days of action. All stations were fully manned, but the length of the watches was reduced to allow his officers and men to get some much needed rest. The entire crew was fully clothed. No lights except standing red lights were allowed, and no bathing was permitted.
Regarding the actions taken by Canberra, there are a number of anomalies in Bruce Loxton’s account that had to be resolved in order to produce the closest possible simulation. In Chapter 17 he states that:
At 01:42 (the Japanese pilot) was ordered to start illuminating. Almost as (this happened) bridge personnel in Canberra heard an explosion off the starboard bow. One of Chokai’s torpedoes had exploded prematurely… The senior officer on watch was Lieutenant Commander Wight. He immediately summoned the Captain to the bridge by voice pipe… Getting had retired at midnight to his sea cabin, a deck below the bridge. He was on the bridge within seconds… (As this was happening) Wight, seeing three ships fine on the starboard bow had alerted the armament… (and) ordered the wheel to be put hard to port as the Captain arrived on the bridge… before the order to turn to port had had any effect, the Captain reversed the wheel to starboard and gave the order for full speed. He had apparently realised at once that a turn to starboard was necessary if Canberra were to remain between the enemy and the transports.
This account would seem to clash with Loxton’s drawings in Chapter 18. These indicate that the alert was given at 01:43; that Wight ordered hard a port one minute later at 01:44 and Captain Getting ordered hard a starboard and full ahead a full minute after that at 01:45. This struck me as unlikely. In order to get to the bottom of this, I visited H.M.S. Belfast, the last surviving WWII big gun cruiser in Europe, to get an indication of how long it would take to get from the Captain’s sea cabin to the bridge. This very conveniently, was also located one deck below the bridge. Including (a generous) 5 seconds added on for waking/getting up, it took me just 18 seconds. The time to reach the bridge from the officer’s bunks two decks below was 30 seconds – which also included 5 seconds waking/getting up time. For the purposes of the simulation, I decided to work from my own time line, which is as follows:
01:42 + 30 secs – Japanese illuminate and torpedo explodes prematurely.
01:42 + 40 secs – Wight orders: “hard a port, alarm starboard, load load load.”
01:42 + 55 secs – Getting reaches bridge, makes snap assessment and orders hard a starboard and full ahead.
01:43 + 00 secs - Lieutenant Commander Mesley arrives on Bridge.
01:43 + 05 secs – Canberra’s main armament loaded and ready to open fire turrets are turning to starboard: target not yet acquired.
01:43 + 25 secs – Canberra’s rudder bites, she starts to come from port to starboard . Turrets are now reversing from starboard to port.
01:46 + 30 secs – Canberra completes emergency turn to starboard.
01:46 + 40 secs – Canberra’s gunners acquire target.
The time taken for Canberra’s – or indeed Chicago’s - gun crews to stand to and load does not appear to have been a factor in the delay in opening fire. The principal cause for the delay – as Loxton states – was emergency manoeuvring, coupled with the increase in speed, which effectively prevented the gunners from acquiring a target. What also seems to have caused a real problem for the whole Allied Southern Force was a heavy cloudbank situated to the south-east of Savo Island. This formed a black backdrop to the Japanese Force which made them very difficult indeed to spot and target. Even worse, the Allies were back-lit by light cast from a burning transport ship. An equivalent situation on land would be an infantry section at the edge of a tree line in shadow, opening fire on an enemy patrol that was in open ground in moonlight. The infantry by the tree line would have every advantage.
Cloudbanks are not modelled in FS/FSP. The best I could do in the case of Canberra was to factor in an added delay in acquiring a target. Bruce Loxton stated of Canberra after she had come to general quarters that:
“The turrets were moving in unison with the director as its crew sought a target… Canberra was ready in all respects to go about the business of engaging the enemy. Only a few seconds more would have been necessary… All that was lacking was an aiming point before opening fire and a little more time because, just as power was lost, the director saw a cruiser on the port beam.”
I have guessed that the cloudbank would have delayed matters for approximately 10 seconds. Canberra will be delayed 25 +10 seconds, if she opens fire while maintaining the same heading. However, if she turns 90 degrees to port or starboard, she cannot open fire until 10 seconds after completing the manoeuvre. These timings are best factored in by using a stopwatch.
It should be noted that Chicago failed to use her main armament at all. The probable causes were lack of illumination due to a batch of defective star shells, the masking effect of the cloudbank plus the lack of practice of her main gun crews. Significantly she also failed to use her searchlight until after the engagement had ended. I have simulated this by applying the following rules.
1) Chicago will be delayed 25 + 10 seconds from coming under attack if she opens fire while maintaining the same heading
2) If she makes a sharp turn to port or starboard, she cannot open fire until 10 seconds after completing the manoeuvre.
3) She is not allowed to use illumination of any kind.
4) She may only engage illuminated targets, including those that have been set on fire.
The U.S.N. second degree of alert has already been described. Loxton states that the process of coming to general quarters could take them up to eight minutes. Up to this point, I had been factoring in delays by the simple expedient of counting the seconds off with a stopwatch. For longer delays I decided to try a different approach. In the FS program, the function of the main armament of a ship is dependent on the supply of electricity. In theory, the scenario editor allows the electrical system of any given vessel to be disabled for any period (measured in minutes) effectively knocking out the main armament for the same period of time. In practice however, this didn’t work as weapons systems that had been ‘disabled’ in this manner before the start of the scenario obstinately continued to function. The owners of the rights to the software stated that although weapons systems were disabled in FS when deprived of electricity, the FSP upgrade modelled auxiliary power to turrets by allowing firing to continue at a reduced rate and with lower accuracy. This meant that further action was required with regard to scenarios played from the Japanese viewpoint; because Allied ships were still returning fire long before they did in reality and inflicting a disproportionate amount of damage as a result. I ended up taking the drastic expedient of pre-destroying via scenario editor the armament and torpedo systems of all Allied ships that had not used them in the actual encounters. Although far from ideal, this at least had the effect of yielding results that were comparable with the actual battle.
These measures were only applied to Allied ships in the southern contact which was fought at much closer range than the northern contact that followed. In any case it would not have been possible to do the same to the American vessels involved in the latter engagement because they all – eventually - returned fire. In the event, no modifications were necessary for the simple reason that the games system models the vastly superior ranges at which the specialist Japanese masters of lookout were able to spot the enemy at night. Consequently in the simulation, the Japanese will still have spotted and opened fire on the northern cruiser group before they themselves have been detected which remains more or less true to what took place in real-life.
With regard to the timings in the northern contact scenario as played from the Allied viewpoint, my start point was 01:47. Vincennes, Quincy and Astoria were all called to general quarters at approximately 01:48. They were illuminated and came under fire at 01:50. Astoria returned fire the quickest at 01:52 , firing two full salvoes before Captain Greenman, appearing on the bridge for the first time, assumed that they were shooting at their own ships and gave the order to cease fire. Firing was resumed at 01:54. Vincennes did not return fire until after changing course and speed. I have taken this time as 01:53 – six minutes from the start of the scenario. Quincy took far longer, with her main battery finally coming into play at 01:58 – eleven minutes from start. With regard to manoeuvring, Captain Riefkohl in Vincennes did not give orders for the division to increase speed or change course until 01:52. This was simulated via scenario editor by disabling the rudders of Vincennes, Quincy and Astoria (manoeuvring damage) for five minutes from start. I also disabled the rudders of destroyers Wilson and Helm but for a longer period, thus ensuring that (as in real life) they played no significant role in the engagement. This works – but not all of the time. An intermittent bug lurks in the program which sometimes results in ships with disabled rudders managing to manoeuvre. When this happens, the correct drill is simply to pull out of the scenario and start again. This will usually result in the relevant vessels behaving as intended and maintaining their bearings. All manoeuvre and opening fire timings for both northern and southern contacts are recorded in the relevant scenario notes. For the simulation played from the Japanese viewpoint to remain accurate, Division 1 should open fire with torpedoes immediately a target presents itself – and all divisions with gunfire at around 01:50. Historically the cloudbank proved much less of a factor here. Tenryu Yubari and Furutaka to the west, might well have derived some protection from it, but the majority of the Japanese force – by then to the east of the Americans, did not. It was not practical to include it in scenarios involving the Northern cruiser group so I left it out completely.
The Scenarios
My original plan was to simulate the southern and northern contacts as closely as possible and follow this with a number of hypothetical or ‘what if’ scenarios. I ran into trouble almost immediately when I tried to re-create the southern contact. The problem was that what actually happened in real-life was completely improbable. Just consider the following: In this initial contact, Chicago was hit twice by Long Lance torpedoes but remained an effective fighting unit. She should have been sunk - or put out of action - at the very least. Loxton presents compelling evidence that Canberra was disabled not by the Japanese, but by a ‘blue on blue’ incident in the form of a torpedo fired from Bagley – Evidently it happened, but the probability of it happening must have been remote. Even more improbably, the Japanese fired no less than nineteen Long Lance torpedoes at Canberra, some at close range after she had been incapacitated and had become a sitting duck – and missed with every one. Finally, the exceptionally alert Patterson was ideally placed at close range to fire a spread of 8 torpedoes at the Japanese line. One or two hits should have been inevitable. However the order was drowned out by gunfire and the attack never took place. The unlikely result of all this was that when the spray had finally settled, the entire southern cruiser force and accompanying destroyer screen were still on the surface – an outcome that I have never been able to reproduce in simulation!
In my initial attempt to simulate the first encounter, I tried to factor as much of this in as possible. However, the unlikelihood of it all had to be faced and I was forced to stop and ask myself what I was trying to accomplish. Catering for elements that the FS/FSP program had missed such as cloudbanks and alert states was a valid exercise but actually trying to re-create all “the slings and arrows of outrageous fortune” was not. In the end, I had to dismiss the historic course and outcome of the southern contact as a series of flukes resulting in a freak result that should never have happened. To try and reproduce this faithfully in simulation would have meant distorting the program parameters to such an extent as to render the whole exercise pointless. Instead, I decided to concentrate on likelihoods and probabilities.
I ran the southern contact (Scenario 1) as Allied, then as the Japanese (Scenario 2) reproducing the manoeuvres used during the actual engagement. On computer, Canberra is always sunk within the first few minutes – either by torpedo strike, or/and the concentrated gunfire of up to five 8” cruisers. However, she is equipped with a bank of four British Mk 1X torpedoes on each beam, an infinitely more effective weapon than the American Mk XV. The ranges are short and consequently if Canberra manages to launch, she will usually achieve at least one hit, either crippling or sinking a Japanese ship in the process. However, it is by no means certain that this will happen, for in simulation, Canberra as lead ship is always subjected to a maelstrom of fire and seldom survives for long. Chicago’s chances are better however she usually gets hit by at least one torpedo. She is often sunk, but her turn to port taking her away from the Japanese column will sometimes result in her surviving – although in a much more damaged state than in reality. If Chicago should follow Canberra around to starboard she is less likely to survive, but as she remains in contact longer she is more likely to inflict damage on the enemy in return. Another piece of “outrageous fortune” that should not have occurred is Commander Frank Walker’s order to fire torpedoes being drowned out. Therefore when running the simulation, Patterson should get to fire a full spread of eight. However, FS/FSP models the diabolical 80% failure rate of the U.S. Mk XV torpedo. Consequently, she will often achieve strikes, but these will not always result in detonation. In the event that an American torpedo both connects and detonates, the result will be a Japanese ship being crippled or sunk.
Scenario 1a was designed to depict Chicago’s part in the engagement commencing from when she suffered a torpedo strike. This proved to be a brief but intense encounter. In simulation it becomes evident that after being torpedoed, her reduced speed would not have enabled her to maintain contact with the Japanese even if she had reversed course and come on to an easterly heading. Her main battery director had been knocked out which would have had a detrimental effect on her gunnery. This was easily simulated (in this scenario only) by pre-destroying Chicago’s fire director 1 via the scenario editor. When the masking effect of the cloudbank, the damage to her bow plus the fact that she had a dud batch of star-shells, are added to the equation, it is difficult to see what else Chicago could have achieved in the action.
In the actual battle, Chicago was credited with damaging Tenryu with her secondary armament. However in simulation this never happens. Tenryu and Yubari breaking to the north-east are always among the first two ships to lose contact and consequently escape unscathed. Kinugasa and Furutaka as the ‘rear markers’ of their groups are in contact the longest and as a result are the most likely to be targeted. Perhaps in real-life Tenryu and Yubari stayed in line longer than claimed and only broke away after the former came under effective fire, closely followed by Furutaka who had latched onto Tenryu’s wake after fixing a steering problem. Possibly the edge of the cloudbank was ragged and uneven, leaving more distant vessels visible while masking those closer. We shall never know for sure. Chicago’s actions immediately after the southern contact have been (and continue to be) the subject of considerable debate, but discussion of them falls outside the remit of this project.
After running the historic manoeuvres, I experimented with some alternatives, in particular the aggressive use of destroyers. Patterson and Bagley have a total of thirty two torpedoes between them. If they expose one beam to the enemy, fire eight torpedoes and then repeat the exercise with the other beam, it will usually (in simulation) give more than one positive result – despite the appalling dud rate of the American Mk XV torpedo.
Figure 1
Figure 1 illustrates such a manoeuvre. The black areas represent torpedo spreads. Patterson and Bagley have presented both beams to the enemy and all their torpedoes have been fired. An additional four have also been launched by Canberra for a total of thirty six torpedoes heading towards the Japanese. In this particular run-through of the simulation, Canberra and Chicago were lost to a combination of Long Lance torpedoes and gunfire. However Aoba and Kako received two torpedo strikes each in return and were disabled as a result. Had this happened in real life, the Japanese would have been two cruisers down and their formation shattered when the (by now) fully alerted northern cruiser group entered the fray. The result would certainly have been victory for the Allies.
In the northern contact (Scenario 3) following the historical manoeuvres results in an outcome that comes very close to the actual result. In real life the three American cruisers were prevented from manoeuvring by the indecisiveness of their commanders, and from immediately returning fire due to delays caused by their gunners assuming general quarters from the American second degree of alert. In simulation this results in an un-nerving period in which the Americans are under fire and taking hits, but are quite unable to do anything about it. The three cruisers, who are caught between two Japanese columns, will eventually succumb to a combination of gunnery and torpedo strikes and in return, will probably inflict only minor damage. Chokai often ends up losing a turret – as happened in real-life - and the odd shell hit is registered on a couple of other Japanese ships – but that is all. The destroyers Wilson and Helm are located well away from the Japanese and if the historical manoeuvres are followed will usually survive the action. In the actual battle, the Japanese commenced by crossing the American T from the rear and then by accident or design, flanked them on either side with the main column to the east and a smaller column to the west. I have included the (by now) disabled Canberra because at 01:47, which represents the start of the scenario, she was still drawing Japanese fire. After the action remarkably, she was still afloat. On computer she always ends up getting sunk by torpedo. In the simulation, the Americans’ real-life turn to port proves to be of little benefit. However, an acutely angled turn to starboard will not only take them away from the western Japanese column, but it should also ensure that any torpedoes that have been launched at them will miss. In addition this will result in all their 8” guns bearing on the eastern Japanese column. It should be stated though that by the time their commanders have finally made up their minds as to what is happening, the three cruisers have usually suffered varying degrees of damage and in consequence are no longer able to function as a cohesive unit. Destroyers Wilson and Helm represent the best hope of the Americans to inflict significant damage, but because they are stationed at the front of the American line and the Japanese are attacking from the rear, it is much harder for them to get into a position where they can take effective action.
From the Japanese viewpoint (Scenarios 2, 4 5a and 5b) the options are much more straightforward. Their doctrine was to launch a barrage of torpedoes at the enemy, and then to open fire with gunnery just before the time of their arrival. This is every bit as effective in simulation as it was in real-life. The torpedoes should be launched in barrages that are so wide and dense that the enemy will be unable to dodge them all. In marked contrast to the ineffective American Mk XV torpedo, the Long Lance is reliable, carries a potent warhead, and the Japanese have them in large numbers.
Fig 2
When running scenarios 1 and 2, without effective Allied destroyer support the Japanese are in a win – win situation. If Canberra and Chicago maintain course, their T stays crossed. The Japanese will have the double advantage of firing first, plus an overwhelming advantage in firepower with thirty four eight inch guns stacked up against ten on the Allied side. The ranges are short and faced with such a volume of shell-fire Canberra’s life expectancy will be brief. Chicago will seldom survive more than a few minutes longer. If they make emergency turns to port or starboard (Fig 2) the prospects are little better as they will take heavy punishment before their guns and torpedo tubes bear and their exposed beans will become vulnerable to torpedo attack. In this particular run-through, Chicago was badly damaged by gunfire, Canberra was torpedoed and sunk.
At Savo Island, Allied un-preparedness gave the Japanese a huge advantage, both in real-life and in the simulation. In the first few minutes of each contact, the Japanese should be inflicting heavy damage on an enemy who will be almost unable to retaliate. Speed is of the essence, and this becomes really apparent when the whole battle is fought (Scenarios 5a and 5b). The Allied southern cruiser force has to be knocked out almost immediately, in order to give enough time to engage the northern cruiser force before the American crews have had the opportunity to finish coming to general quarters. Maximum violence must be generated. Unless the Southern cruiser force is taken out immediately, the Northern force will have enough time to retaliate, and it carries sufficient firepower to inflict serious damage.
The findings of various official inquiries have listed contributory causes for the defeat, and some of these present a number of ‘what if’ or hypothetical scenarios. The failure of communication between Admirals Turner and McCain which resulted in a crucial lack of air reconnaissance has been discussed elsewhere by eminent historians and naval personnel and I have nothing to add. However the High Command decision most relevant to this exercise, was Admiral Fletcher’s premature withdrawal of his carrier force which had been providing air cover for the landings on Guadalcanal. This affects the simulation because it had a knock on effect, leading to the Flag Officer Rear Admiral Crutchley having to attend a conference that resulted in his absence and that of the heavy cruiser Australia from the battle. It should be noted that by this point in his career Crutchley, a holder of both the Victoria Cross and the Distinguished Service Cross, had amassed an outstanding combat record in both world wars and was easily the most battle seasoned naval officer in the immediate location on either side. He was a formidable fighting seaman by any standard and as such would have represented a notable asset had he been present. The possibility that he might have made a difference cannot be discounted. Scenario 6 explores what might have happened had he and his flagship Australia, taken part. In run-throughs, Australia as lead ship is always subjected to the concentrated fire of the Japanese and usually succumbs. However, the eight extra 8” guns plus another four Mk 1X torpedoes can make a big difference. In simulation, without effective Allied destroyer action the Japanese will prevail despite Australia’s presence, with – on average - two out of the three Allied cruisers being sunk and only light damage being inflicted in return. However, if Patterson and Bagley can launch successful torpedo attacks it actually becomes possible to turn the situation around and inflict a defeat on the Japanese.
Admiral Ghormley listed eight reasons for the defeat at Savo, three of them attributable to Rear Admiral Crutchley. This simulation allows me to comment on two of the latter. They are:
1) No special plan was prescribed to cover the possibility of a surface ship night attack (on) 8-9
August
2) The night disposition relied principally on radar to detect the approaching enemy.
With regard to the first point, it should be pointed out that whatever Ghormley’s impression might have been, Crutchley had actually formulated a plan to repel surface attack. However it failed because any chance of a co-ordinated response depended on the radar pickets giving a timely warning of any approaching enemy and alerting his cruiser patrol groups before they entered the sound. This they manifestly failed to do. However, even this particular failure shouldn’t have resulted in complete disaster. In un-planned for situations service personnel will – in extreme cases - fall back on pre-rehearsed drills and procedures that are specifically designed to give the combatants a fighting chance, even if the element of surprise is against them. Significantly, no such drills were evident in this instance. In the southern contact, it was notable that the Allied ships reacted individually and not according to any pre-planned set of manoeuvres. In the northern contact, there was no discernable reaction at all until after the ships had started to take damage, and by then it was too late.
The overall performance of the destroyers was poor. Patterson reacted aggressively, but still failed to fire any of her torpedoes. Bagley turned to port (going full circle in the process) and fired four torpedoes with disastrous effect achieving (if Loxton’s analysis is correct) a direct hit on Canberra. Despite being ordered to attack with torpedoes, Helm failed to do so, and in the confusion neglected to pass on the order to Wilson. In consequence neither of the latter two vessels played any significant part in the action.
Of the American ships under Crutchley’s command, Loxton makes the point that:
Unlike the Royal Navy, they had no tactical doctrine common to all squadrons, fleets or task forces. When an American task force was formed, as on this occasion, the task force commander was expected to write a book instructing commanding officers how to act in a variety of situations.
He also states that:
The capability of the destroyers was still a matter of conjecture. Generally speaking, these had been joining piecemeal since the middle of May and had spent little time with the task force… The opportunity for these ships to carry out gunnery practices had been few, as they had been detached frequently on escort duties and none of them had been able to carry out a torpedo firing for some considerable time.
In other words the destroyers had almost certainly not been versed or practised in any of the tactical responses that would have been required by Crutchley, and to make matters worse there was no common doctrine or standard operational procedure for them to fall back on. Consequently with the elements of surprise and confusion working against them, their captains, would have had no alternative but to act on their own initiative. It was small wonder that their response was so uneven.
It should be noted that using the simulation, I was able to devise a number of very simple contact drills that could have easily have been carried out in real-life if practised beforehand. The key to success was the torpedo, a weapon which in 1940s night time naval warfare ranked with the gun in importance. Even the chronically unreliable American Mk XV could have an effect at close range if enough of them were used, and U.S. destroyers carried sixteen apiece. Delivery of this weapon was entirely dependent on the destroyers as their cruisers were not armed with them at all. Effective destroyer action was the key but in reality it proved non-existent. Admiral Ghormley has stated that no special plan was produced to cover the possibility of surface ship night attack. He is surely missing the point here. The real problem was that no pre-rehearsed drills or procedures existed to deal with surprise surface ship night attack. Also without effective destroyer action the Allies were fighting with one hand tied behind their back. Sixty four torpedoes were carried by their destroyers but only four were used. This must be set against in excess of thirty that were launched by the Japanese. Allied use of torpedoes can make a huge difference in simulation and undoubtedly would have had a similar effect if they had been used en masse in the actual battle.
.
With regard to Admiral Ghormley’s second point concerning over-dependence on radar, I constructed two scenarios to determine when Blue should have detected the Japanese force. (Under the prevailing conditions Ralph Talbot, patrolling the northern entrance to the sound, would almost certainly have been outside effective radar detection range). Scenario 8a was based on the data taken from the Japanese accounts contained on p – 171 of Loxton. 8b was based on the information presented in the 1951 U.S. Naval War College Analysis and illustrated in the map on pp 36 -37 of Morison. No combat occurred in this encounter, and to ensure that this didn’t happen in simulation, all weapon systems were disabled. I have accepted Loxton’s statement that land based clutter would not have been a problem as Blue rarely went within 7 miles of land, and the ranges involved were shorter than that. Consequently in the ‘environment settings’ section of the scenario editor I left the radar conditions on 100%. However, it was a squally night, so to simulate this I turned the rain setting on, which degraded the radar detection range by approximately half. When running the scenarios, even under these less than ideal conditions, Blue will always detect the Japanese at approximately 00:56 at a range of 7500 yards in 8a, and at about 00:58 at 7700 yards in 8b. However, in real life as the Japanese continued to approach, the ranges closed much further. Loxton makes the telling point that:
What should be accepted… is that no one on watch in Blue saw, at a range of about a mile, a column of eight ships, five of which were about 10,000 tons, moving along the line of sight at high speed, and this on a night when Chokai’s lookouts could see a single destroyer, proceeding at 12 knots, at eight miles.
Shortly afterwards the situation was to repeat itself when the whole Japanese column passed Jarvis at an estimated range of less than 1100 yards without any kind of alarm being raised. In each case the situation clearly falls a long way outside the parameters allowed for in the FS/FSP program. Perhaps additional ‘extreme fatigue’ and ‘don’t fire unless fired upon’ settings are required to simulate this kind of encounter successfully. Whether or not over-reliance on SC radar was a contributory factor to the defeat is surely a moot point. The main problem in this case seems to have been a failure of Mk 1 eyeball.
All this begs the question: what would have happened had Blue been successful in warning the main force of the approach of the Japanese? This is explored in Scenarios 7 and 7a. I am assuming that she should have detected Chokai a few minutes after Chokai spotted her – 00:56 for the purposes of this exercise. Using the map on p-173 of Loxton for my base calculations, I calculated the approximate positions of the Allied northern and southern cruiser groups at that time, and also that of the damaged destroyer Jarvis, which was then exiting the sound. I then worked out a point of contact and estimated how long it would have taken each group or individual vessel to reach it. An advantage of the northern cruiser group’s box patrol pattern became apparent. The northern group would have actually been closer to the Japanese than the Southern group if Blue had spotted them at the given time. The linear patrol pattern of the southern group would have placed them a greater distance away to the east. This is why in scenario 7 the Northern group makes contact first. In the event of Japanese ingress through the southern approach into the sound Blue’s warning would have given both patrol groups ample time to come to general quarters. I adjusted the crew fatigue level from ‘fatigued’ to ‘tired’ to allow for the adrenaline rush experienced by men who would by now have been fully alert and in the knowledge that action was imminent. I left Ralph Talbot out of the scenario (still patrolling her sector) but placed Blue near the outer limit of her radar range, tailing the Japanese force. The Japanese ORBAT also includes the destroyer Yunagi. Both scenarios should be played in the ‘divisional’ command mode, with the player commanding one or other of the Allied cruiser groups. Scenario 7 is a wild melee in which either side can emerge victorious. As always, the Long Lance torpedo can prove decisive. However, when running this particular simulation, the Japanese – in marked in contrast to the historical battle – always find themselves in a stiff fight that should result in them taking significant damage, even when they prevail. Transposed to real-life it seems certain that the survival of any crippled Japanese vessels after day-break would have been doubtful, and that their original mission of destroying the transports would have been rendered impossible. Scenario 7a is designed to show what would have happened had the Japanese T been crossed and both Allied cruiser groups had been alerted in time to work in union. It generally results in a resounding defeat for the Japanese.
In its discussion of Admiral Crutchley’s dispositions, the Naval War College Analysis comments on the lack of distance between the radar picket destroyers and the cruiser groups. This was found not to be a factor with the Southern group, who even at the most western extremity of their patrol line would have come no closer to Blue than about 14 miles, which (assuming timely detection of the approaching Japanese force) would have given ample time and space in which to react. However it was found that this was fair comment with regard to the Northern group. The War College Analysis correctly states:
The Ralph Talbot was not sufficiently distant from the Vincennes Group to give adequate warning of enemy approach. With the Ralph Talbot at the easternmost extremity of its patrol line simultaneously with the Vincennes group at the northern point of its square, there was only five miles between them… The Ralph Talbot might possibly have given a warning of about fifteen minutes before an enemy could close to within gun range of the Vincennes Group.
In fact, had Ralph Talbot detected the approaching Japanese whilst the Vincennes group was at the northern point of its square, this would have understated the case considerably. Play-testing under these circumstances revealed that, the warning time was usually considerably less than fifteen minutes. Scenarios 7b and 7ba explore what might have happened had Mikawa taken the northern approach into the sound. In running Scenario 7b, the radar detection range of 7500 yards (taken from Scenario 8a) translates to only a bare five minutes warning before the Vincennes Group became engaged; in other words not even enough time for them to finish coming to general quarters. However in the Author’s opinion most commentators seem to have placed far too much emphasis on the unreliability of S C radar whilst almost completely ignoring the capabilities of the traditional lookout. Let us not forget that conditions to the north were clear enough for the Japanese to spot Ralph Talbot at a range of 16,000 yards. While the Allied lookouts could probably not match the effectiveness of the specialist Japanese masters of lookout it would seem reasonable to assume that they should still have been capable of the timely spotting of an approaching heavy cruiser force in reasonably clear conditions. The War College Analysis would have it that five miles (or 10,000 yards) was “the limit of night visibility in the existent weather conditions” and I have used this as the detection range in Scenario 7ba. When running this, provided the Vincennes Group turns away from the Japanese when Ralph Talbot makes contact, it is still possible for them to avoid being brought to action for 10 minutes. Aggressive use of destroyers Ralph Talbot, Wilson and Helm (carrying 48 torpedoes between them) can buy a few minutes more, and sometimes inflict considerable damage. The Chicago group will usually join the action about 23 minutes after initial contact by Ralph Talbot. Ten to fifteen minutes warning will not give the Vincennes group time to work up to anything like full speed. However the Japanese will still have to pass them in order to get to the transports and with the American gunners fully alerted and closed up they will have a considerable fight on their hands. In simulation The Japanese usually overwhelm the Vincennes group, but only after taking significant damage in return. They are then engaged immediately afterwards by the Chicago Group, approaching from the south, sustaining yet more damage. The net result for the Japanese will usually be ships crippled, and on occasion sunk. If this had happened in real life it would seem most unlikely that Mikawa would have considered pressing home his attack on the transports.
There was one further approach into the sound that had to be guarded. This was the eastern entrance, lying between Guadalcanal and Florida Island and Admiral Crutchley covered this by deploying the light cruisers San Juan and Hobart escorted by destroyers Monssen and Buchanan. This force (The San Juan group) was commanded by Rear Admiral Norman Scott. Although by far the widest of the three entrances it was full of navigational hazards making it “unsuitable for all except light craft, such as destroyers or PT boats, particularly at night.” Had Mikawa pressed home his attack on the anchorages from the west, this group, together with sundry destroyers and destroyer-minesweepers, would have formed the final line of defence. Scenarios 7c and 7d explore what might have happened had such an attack taken place. When designing these scenarios, it became necessary to consider the tactical level communications – or more accurately the lack of them – between the principal elements of Crutchley’s screening force. It should be remembered that this force consisted of elements from both the United States and the Royal Australian Navies, with a British Rear Admiral in overall command. Each navy had its own procedures and methods of communication. The R.A.N. contingent (including the U.S.S. Chicago) used morse code, relying on a combination of the tried and tested H.F radio and signal lamp. The U.S.N. had largely gone over to the much more recently invented T.B.S. system. This was a V.H.F. voice net. Unfortunately due to poor voice procedure, many messages seem to have been lost, probably due to radio operators transmitting simultaneously.
Signals procedure overall seems to have been erratic. On his departure from the screening force in Australia to attend Admiral Turner’s conference Admiral Crutchley sent a message by radio to inform the Commander of DESRON 4 of the situation. This message was transmitted on a frequency monitored by all ships. Both this message and Turner’s original message summoning Crutchley would have been received in Vincennes, yet inexplicably neither seems to have been passed on to Captain Riefkohl. Patterson’s warning that enemy ships were inside Savo Island was also received in Vincennes but again not passed on to the Captain. Captain Bode of Chicago seems to have been extremely reluctant to transmit anything at all, including vital contact and situation reports. Finally, on discovering that elements of his screening force were in action, Crutchley transmitted a message ordering all destroyers to concentrate on Australia. Unfortunately due to either corruption, or human error, the concentration location was not decoded correctly and instead the destroyers proceeded to a previously arranged R.V. northwest of Savo Island.
All this would strongly indicate that had the Japanese pressed home their attack on the transports, communication problems would have seriously compromised Crutchley’s ability to control his forces. It is for this reason that Scenario 7c was designed in the divisional command mode, which denies overall control. On running it, what usually transpires is that the Japanese will encounter the following units in turn: 1) Chicago, 2) Patterson standing by the disabled Canberra, 3) Australia, 4) the San Juan group, and 5) the three destroyer- minesweepers guarding the Guadalcanal transports. None of these individual elements has the power to stop them and they will either be brushed aside or destroyed piecemeal leaving the enemy free to enter the transport area. In contrast Scenario 7d assumes workable communications and has been designed to be played in the standard mode, giving the player overall command. Here, ten destroyers have concentrated on Australia as ordered and are formed in two divisions. Chicago – damaged but still an effective combat unit – has on receipt of orders from Crutchley - also joined Australia and is now in formation with her. All ships are in communication with each other. In consequence the Commander has a good overview of what is happening and can direct his forces accordingly. All this makes a huge difference and when Scenario 7d is run the Allies will usually prevail. While use of the divisional command mode is a simplistic way of modelling communication problems, the Author would argue that the overall effect is not unrealistic. It must be accepted that erratic communications would equate to at the very least a partial loss of control which would severely limit the options open to a commander. The conclusion must be that the Allies possessed the necessary combat assets to stop the Japanese had they pressed home their attack on the transports but that poor communications would have rendered the task an uphill struggle.
Conclusions
Where does the truth lie? In the case of a battle such as this one, that truth can prove a very elusive commodity. Extreme fatigue often engenders a hallucinogenic state of mind. Added to this, adrenaline and tunnel vision can have a dramatic effect on what is or isn’t remembered. Consequently, exhausted men who have been fighting for their lives are not always the most reliable witnesses. In any kind of action involving allies from two different nations, national pride will fog the picture even further. If a blunder occurs that results in a significant number of casualties, the truth will probably be deliberately distorted or even hidden because of the requirements of operational security, the preservation of the morale of both civilian and military personnel and the necessity to maintain cohesion between allies in wartime.
The technique of taking the real-life events and placing them within an artificial framework formed by a computer program has proved an extremely effective way of cutting through the fog and the distortions. This is because in order for the exercise to be valid, the data going in has to be as accurate as possible. This has meant putting the accounts of the battle I have used under a microscope. I have had to analyse and make judgments of my own to resolve clashes and anomalies within and between them and to award the qualitative values (to ships, crews and environmental conditions) that the program requires in order to function. Often this has meant having to read beyond my principal sources. In the process I have found the computer program had not catered for all the relevant factors, which has meant having to bodge a number of extra elements into it in order for an acceptable level of realism to be maintained.
When you possess accounts that have been termed definitive, written by distinguished and high ranking service personnel who were either at the battle or who fought in the campaign, the tendency is to accept everything that they say as gospel. The argument “He was there” can carry an awful lot of weight. This, as I have found, does not apply when you are constructing a computer simulation. Every element that you extract from these accounts has to be reduced to the bare nuts and bolts, and in the process you will be forced to question things that you would have automatically accepted before, and pick up on details that you would have otherwise missed.
Research through reading is usually a passive activity. Here, the research process becomes much more active because you are now being forced to assess, pass judgment and make a decision on every single bit of data that is entered into the program. When the simulation is up and running, you are effectively placed in command and the process becomes entirely active. Your decisions will now result in consequences that you will see unfolding before your eyes. Reading about a decision is one thing; having to take a similar decision yourself entirely another, even when seated in front of a computer monitor.
Once constructed, the main benefit of the computer simulation was having the opportunity of experimenting with tactics and testing hypotheses. The key question in this case was always: Is there anything that the allies could have done that would have produced a different result? In simulation it became apparent that everything hinges on the outcome of the southern contact, and in particular the part played by the two screening destroyers Patterson and Bagley who are ideally placed to fire full spreads of torpedoes at the Japanese line. Patterson should always fire at least eight, but Bagley who is positioned near the head of the enemy line, has ample time to present each beam in turn, and release all sixteen. Add another four from Canberra and this gives you a total of up to twenty eight torpedoes, all fired at close range. Even allowing for an eighty percent failure rate on impact, at least one or two should take affect. In real life, even if no effective hits had been registered, the Japanese reaction on spotting large numbers of incoming torpedoes on the starboard beam would almost certainly have been for individual ship’s captains to make emergency manoeuvres to avoid them. The net result would have been to shatter the Japanese formation, which at the very least would surely have bought sufficient time to allow the Northern patrol group to finish coming to general quarters. However in the actual battle Bagley failed to fire her starboard torpedoes when they bore, and achieved a disastrous result with the four she eventually released from her port tubes. Patterson, should have fired at least eight, but ended up firing none. The lack of practice by the crews on both vessels was obvious. “Train hard, fight easy” was never a truer maxim. One night-time torpedo firing exercise carried out earlier, to bring the torpedo-men of all the destroyers back up to speed, would probably have made all the difference on the night. With hindsight, the fact that Rear-Admiral Crutchley (for whatever reasons) had failed to do this must be considered one of the principal factors behind the defeat.
However inflicting damage on the enemy represents only one part of the equation. The other crucial element is time. It took the Americans in the Northern cruiser group an average of more than six minutes before their gun crews had assumed their action stations and were in a position to return fire. Every extra second that the Southern cruiser group could have bought them would have been vital. In real-life, the Japanese crushed the southern group in minutes and had opened up on the northern group inflicting serious damage before they were able to retaliate. In simulation, if the Southern cruiser group can find a way of staying in the fight for even five minutes more, it will give the Northern group a fighting chance. Co-ordinated destroyer action and effective torpedo use is the best way of achieving this.
Most of the problems encountered by the Northern group stemmed from a combination of indecisiveness by the ships’ captains plus the time taken to come to general quarters due to the American second degree of alert, in which off-duty personnel were not at or near their action stations, but were, in many cases, dispersed to other parts of the ship. It should be stated that at least some of this indecisiveness can be attributed to Admiral Crutchley’s failure to inform Captain Riefkohl of the Southern cruiser group’s plan of operations, despite being asked. It was notable that the initial Northern group assumption was that they were being fired on by on their own ships, a conclusion that they would probably not have reached had Captain Riefkohl been informed of the movements and location of the Southern group! Had this been the case Astoria’s skipper Captain Greenman would almost certainly not have given the order to cease-fire in mid-battle and Astoria might well have registered a number of extra hits as a result, but in the face of overwhelming enemy firepower, this would probably have had little if any effect on the outcome of the battle. In the case of Vincennes and Quincy, it would have made no difference at all due to the time taken by these ships to come to general quarters. Here, the simulation demonstrates the consequences vividly, giving a sense of what it must have been like to have been receiving fire and taking heavy damage, without being able to do anything in return. In this kind of situation seconds are vital, minutes are fatal, and the American response was measured in minutes. While many factors played their part, those vital lost minutes easily comprise the most significant reason for the defeat of the Northern group. After the battle Admiral Mikawa was to write:
“I was greatly impressed, however, by the courageous action of the northern group of U.S. cruisers. They fought heroically despite heavy damage sustained before they were ready for battle. Had they had even a few minutes warning of our approach, the results of the action would have been quite different.”
Hard lessons were learned that night. It is worth noting that they were rapidly put into effect. On the 11th of October 1942, the Americans would encounter the Japanese in another night action, at the battle of Cape Esperance. This time they were to emerge victorious.
Bibliography
Ballard R.D. The Lost Ships of Guadalcanal Warner 1993.
Bates R.W. Naval War College: The Battle of Savo Island, August 9th, 1942. Strategical and Tactical Analysis Part 1. Bureau of Naval Personnel 1950.
Frank R.B. Guadalcanal Penguin Books New York 1990.
H.M.A.S. Canberra
http://en.wikipedia.org/wiki/HMAS_Canberra_%281927%29
H.M.A.S. Australia
http://en.wikipedia.org/wiki/HMAS_Australia_%281927%29
I.J.N.S. Aoba
http://en.wikipedia.org/wiki/Japanese_cruiser_Aoba
I.J.N.S. Chokai
http://en.wikipedia.org/wiki/Japanese_cruiser_Ch%C5%8Dkai
I.J.N.S. Kako
http://en.wikipedia.org/wiki/Japanese_cruiser_Kako
I.J.N.S. Kinugasa
http://en.wikipedia.org/wiki/Japanese_cruiser_Kinugasa
I.J.N.S. Furutaka
http://en.wikipedia.org/wiki/Japanese_cruiser_Furutaka
I.J.N.S. Tenryu
http://www.answers.com/topic/japanese-cruiser-tenry
I.J.N.S. Yubari
http://www.history.navy.mil/photos/sh-fornv/japan/japsh-xz/yubari.htm
I.J.N.S. Yunagi
http://navalhistory.flixco.info/H/79416/8330/a0.htm
Loxton B. with Coulthard-Clark C. The Shame of Savo Allen & Unwin 1994
Morison S.E. The Struggle for Guadalcanal University of Illinois Press 1977
The Navy Department of Naval Intelligence Combat Narrative: The Battle of: Savo Island 9 August 1942
http://www.ibiblio.org/hyperwar/USN/USN-CN-Savo/index.html
Shanks Sandy The Bode Testament Writers Club Press 2001
U.S.S. Astoria
http://en.wikipedia.org/wiki/USS_Astoria_%28CA-34%29
U.S.S. Bagley
http://www.answers.com/topic/uss-bagley-dd-386
U.S.S. Chicago
http://en.wikipedia.org/wiki/USS_Chicago_%28CA-29%29
U.S.S. Helm
http://en.wikipedia.org/wiki/USS_Helm_%28DD-388%29
U.S.S. Patterson
http://en.wikipedia.org/wiki/USS_Patterson_%28DD-392%29
U.S.S. Quincy
http://en.wikipedia.org/wiki/USS_Quincy_%28CA-39%29
U.S.S Vincennes
http://en.wikipedia.org/wiki/USS_Vincennes_%28CA-44%29
U.S.S. Wilson
http://www.destroyerhistory.org/goldplater/danfs408.html
Warner D. and P. Disaster in the Pacific New light on the Battle of Savo Island Naval Institute Press Annapolis, Maryland 1992.
Wikipedia: The Battle of Savo Island http://en.wikipedia.org/wiki/Battle_of_Savo_Island
.
Appendix 1: The Fighting Steel/Fighting Steel Project program
In 1999, the Fighting Steel Program was released by SSI/DBZ, constructed as a tactical level war-game it also shared many of the characteristics of a 3D combat simulation. Originally designed to cover naval surface combat from 1939 - 1942, it was evident that a vast amount of historical research and naval expertise had gone into its creation. Unfortunately, due to financial considerations, it was released to the public in an incomplete state with a large number of bugs still present in the software. The inevitable consequence was that the game crashed on a regular basis rendering it all but unplayable. Not surprisingly, it failed to make much of an impact and apart from a brief appearance as a budget priced re-release it hasn’t been heard of since.
However, it remained the only game of its kind in existence, a fact that didn’t escape the attention of the Naval Warfare Simulation (N.W.S.) design team who applied for and were granted the right to design un-official add-ons, patches and scenarios to the original software. They duly eliminated the bugs and got the program to run. They then extended the period covered to include all of the Second World War, factored in the relevant technological improvements, and added a large number of ships from many different navies. All of this was done without any prospect of financial reward, and the results were released in 2005 as the Fighting Steel Project (FS/FSP) in a free down-loadable patch. Improvements continued to be made, culminating in the final patch, FSP v 10.43, released in 2006. A campaign simulator designed to run with FSP called Thunder at Sea was also produced and this can be purchased directly from N.W.S. FSP is now a closed project.
FS/FSP was designed to simulate a vast number of surface engagements in a period covering the whole of the Second World War, and it performs this task admirably. However, when used to reproduce a unique set of circumstances, such as those encountered at the Battle of Savo Island, it was inevitable that problems would arise. These are listed below, plus the measures taken to overcome them.
1) Alert states
FS/FSP assumes that all ships engaged in a battle are in the first degree of alert. Quite obviously, this was not the case at Savo. In fact, low Allied alert states were found to be a principal cause for their defeat. In terms of the simulation, this translated into factoring in delays in both opening fire and manoeuvring. This was accomplished successfully using a stop-watch.
2) Cloudbanks and Silhouetting
On the night, the Japanese were partially concealed from view by a large inky black cloudbank located south of Savo, which acted as a backdrop. On top of this, the Allied ships were silhouetted by light cast from a burning transport ship. Both of these factors gave the Japanese a significant tactical advantage. Cloudbanks and silhouetting are not modelled at all in FS/FSP and proved impossible to simulate by other means. The only measures I could take – which were far from satisfactory – were to factor in via stopwatch an additional delay to the time taken for allied gunners to open fire, and to decrease visibility levels.
3) The effect of bright lights at night on the human retina
When running the simulation according to the historical manoeuvres, it is notable that Chicago always ends up receiving far more damage than she did in real-life provided that is, she isn’t sunk outright. The probable reason that she got off so lightly in real-life is simply that Canberra was set on fire. The brightness of the flames, apart from drawing even more attention to Canberra, would have burnt into the human retina and rendered the un-illuminated Chicago much more difficult to spot. This does not appear to have been modelled in FS/FSP and it has proved impossible to reproduce by other means.
4) Temporary disablement of ship systems
In theory, the simulation software allows the pre-disablement of ship systems via the scenario editor. For example, disablement of a ship’s electrical system for a set number of minutes should prevent that ship’s main armament from firing for the same amount of time. Rudders can be put out of action in the same manner. This would have enabled the programming in of the exact historical delays suffered by the Allies in opening fire and manoeuvring. In practice, this didn’t happen as weapons that had been “disabled” in this way obstinately continued to function. The only way around this was to pre-destroy via scenario editor, all weapons systems that had not been used in the actual battle. When playing Japanese against the A.I., this at least yielded results that were compatible with what actually happened, however it was a far from ideal solution. With regard to manoeuvring, the approach was more successful. However, an intermittent bug lurks within the program which sometimes results in ships turning before they are meant to. If this happens, the correct drill is to pull out of the scenario and start again. This usually cures the problem.
5) The Artificial Intelligence
The A.I. has limited capabilities that make it impossible to reproduce all the manoeuvres taken by the opposing force. This is what prevented me from simulating the whole engagement from the Allied viewpoint. The best way of getting around this, would be to re-activate ships systems as necessary and then to play the scenarios head to head against a human opponent. In fact I would strongly recommend this approach. Another problem stems from the fact that either the A.I. or the simulation software incorporates a random element that occasionally results in un-realistic situations, such as the exhausted Allied force spotting and illuminating the expert Japanese night fighters first! If this happens, simply pull out of the scenario and start again.
Appendix 2
The Scenarios
This appendix is divided into three sections. The first contains need to know information that relates to all the scenarios. The second consists of tactical briefings for the scenarios plus any additional information that the player needs to know; and nothing else. This is designed to put the player in the position of command. The third contains detailed notes on individual scenarios. Anyone wishing to approach the scenarios “cold” should avoid this section.
Section 1 General information
The Fighting Steel program by SS1/DBZ is no longer on sale in the shops but remains available from a number of sources including EBay. In order to run the scenarios, first install the program and the scenario editor and then upgrade with the FSP v 10.43 patch. This can be downloaded by following the link below.
http://forums.navalwarfare.org/showthread.php?t=6
Finally, copy the numbered Savo scenarios into the ‘Scenarios’ folder located at:
C:\Program Files\SSI\Fighting Steel\Scenarios
With regard to player options, unless instructed otherwise, the scenarios should be played in the ‘Standard’ command option setting with the difficulty level set to ‘Admiral’. The realism settings should be set to ‘Advanced’ but with the ‘unlimited course changes’ option chosen. However players who wish to know what is going on ‘down-range’ should feel free to change the 3D camera setting to ‘un-restricted’. To facilitate the reproduction of historical manoeuvres, players should make frequent use of the ‘pause’ facility whilst giving orders. It is best not to use the divisional gunnery options. Instead all targeting should be done individually via the ‘Single Ship’ panel. The ‘Game Clock’ on the ‘Camera Control Panel’ is a very handy feature for observing any specified manoeuvring and gunnery delays. The 2D replay facility is vital for finding out exactly what happened and should be run as a matter of course at the conclusion of any scenario. For a particularly interesting insight, players should first become thoroughly familiar with the program by practising on night-time scenarios not related to this project. Finally, they should prepare themselves by going forty eight hours without sleep and only then, play scenarios 1, 1a and 3, without using the pause facility. The ‘Divisional’ command option with the cruiser groups selected is recommended in this instance.
Historically the southern contact ended at 01:50, about the same time as the northern one began. In simulation if the Allies can keep the Japanese occupied until 01:55 they are deemed to have won as those extra five minutes would have given two of the three cruisers in the northern group the time to finish coming to general quarters. Either way at 01:55, the player should end southern contact scenarios using the ‘end game’/‘accelerate end’ commands. Northern contact scenarios and those depicting the whole battle should conclude at 02:20.
Section 2 Scenario briefings.
Southern Patrol Group orders Scenario 1
N.B. For Scenario 6, orders are identical except Australia is present and Admiral Crutchley is in command.
General Outline
As Captain Bode of the U.S.S. Chicago, you are the temporary commander of the southern patrol group assigned to protect the transports engaged in landing operations. You are patrolling the southern entrance to the sound. Captain Riefkohl commanding a force of three heavy cruisers, screened by two destroyers, is patrolling the northern entrance. A radar picket consisting of destroyers Blue and Ralph Talbot is deployed to the seaward side of Savo Island. The eastern approach is guarded by Rear Admiral Norman Scott commanding two light cruisers plus destroyer screen.
Area of operation – see Map 1 Appendix 3
Situation Enemy Forces.
Intel. have informed us that a force of two seaplane tenders escorted by three cruisers and three destroyers are approaching, believed to be bound for Rekata Bay, but they are not thought to represent a direct threat. No enemy attack is expected before 11:00L.
Situation Friendly Forces
Your force consists of the heavy cruisers Canberra and Chicago, screened by destroyers Patterson and Bagley.
Attachments and detachments
The Screening Force Commander Rear Admiral Crutchley has been summoned to a conference off the Guadalcanal beach-head and has departed in his flagship Australia. He may or may not be back by dawn.
Mission
1) To guard the southern entrance to sound against surface attack
2) To come to assistance of the northern or eastern cruiser groups as dictated by the tactical situation.
Special Rules (C = contact)
1) Canberra and Chicago may not manoeuvre until C + 55 seconds.
2) Canberra and Chicago may not open fire until C + 35 seconds if they maintain heading.
3) If Canberra and Chicago turn to port or starboard, they may not open fire until 10 seconds after completing the manoeuvre.
4) Chicago and Bagley may not use any illumination.
5) Chicago and Bagley may only fire on illuminated targets. (These may include ships that have been set on fire).
6) At C, Bagley is to commence 360 degree turn to port. She may engage only with port side torpedoes when they bear and with her guns on completing the turn.
No restrictions apply to Patterson at all.
Scenario 1a Savo Island South: Chicago
To be played in divisional command mode as Chicago.
You are Captain Howard Bode commanding U.S.S. Chicago. The southern cruiser force has come under surprise attack and the situation is one of total confusion. You have made an emergency turn to port and are now on a westerly heading. The manoeuvre has not saved you as you have just taken a torpedo strike to the bow. The main battery director has also been knocked out. You are waiting on damage reports however it is certain that at the very least, the wrecked bow will have seriously reduced your speed. Canberra who made an emergency turn to starboard is now abaft of you and on fire. The attacking enemy force, believed to consist of destroyers is heading east at high speed. You are attempting to illuminate, but so far, all star shells fired have been duds.
What do you do now?
Special Rules.
1) Chicago may not use any illumination.
2) Chicago may only fire on illuminated targets. (These may include ships that have been set on fire).
Japanese Orders (Scenarios 2, 4, 5a and 5b).
Quick Battle Orders
Situation Enemy Forces – update.
1) Southern cruiser force consisting of two heavy cruisers plus destroyer screen.
2) A northern group consisting of three enemy cruisers has just been spotted 60 degrees off the port bow at a range of 18,000 yards. Neither group appears to have sighted us so we still retain the element of surprise.
Modified Plan
Cruiser Division 6 (Aoba, Kako Kinugasa and Furutaka) led by Chokai will destroy the southern group by torpedo attack and gunnery and then eliminate the northern group, before proceeding to the anchorage off Lunga Point and destroying the transports. Speed is vital. Any delay in the destruction of the southern group will give the northern group time to react, and they possess considerable firepower. 18th Cruiser Squadron (Tenryu and Yubari) will proceed independently to the Tulugai anchorage as per original plan. Destroyer Yunagi has been detached as rear protection.
Northern Patrol group orders
General Outline
You are the commander of the Northern patrol group assigned to protect the transports engaged in landing operations. You are patrolling the northern entrance to the sound. Rear Admiral Crutchley, the Screening Force Commander is stationed with a force of three heavy cruisers, screened by two destroyers, patrolling the southern entrance. A radar picket consisting of destroyers Blue and Ralph Talbot is deployed to the seaward side of Savo Island. The eastern approach is guarded by Rear Admiral Norman Scott commanding two light cruisers plus destroyer screen.
Area of operation – see Map 1 Appendix 3
Situation Enemy Forces.
Intel. have informed us that a force of two seaplane tenders escorted by three cruisers and three destroyers are approaching, believed to be bound for Faisi, but they are not thought to represent a direct threat. No enemy attack is expected before dawn.
Situation Friendly Forces
Heavy Cruisers Vincennes, Quincy and Astoria, plus destroyer screen consisting of Wilson and Helm.
Attachments and detachments: none.
Mission
1) To guard the northern entrance to the sound against surface attack
2) To come to assistance of the southern or eastern cruiser groups as dictated by the tactical situation.
Special Rules
Scenario start is 01:47. All timings are referred to as ‘from start’
1) No ship may manoeuvre until 01:52 – 5 minutes from start.
2) Astoria and Vincennes cannot open fire until 01:53 – 6 minutes from start.
3) Quincy cannot open fire until 01:58 – 11 minutes from start.
Scenarios 7 and 7a Allies Alerted (Hypothetical).
Play as Allies
General outline
At 00:53 the destroyer Blue deployed to seaward of Savo Island on picket duty, detected a Japanese cruiser force approaching the sound. It is believed that their mission is to attack the transports unloading off Lunga point. Blue has remained in radar contact and is tailing the enemy force.
Area of operations – see Map 1 Appendix 3
Situation Enemy Forces
Five heavy cruisers followed by three destroyers in line-ahead formation with spacing of approximately 1300 yards. Enemy is heading for the southern approach into the sound, between Savo Island and Cape Esperance. Japanese night fighting capability is believed to be poor.
Situation Friendly Forces
Southern patrol group – Heavy Cruisers Chicago (division leader) and Canberra. Destroyer screen Bagley and Patterson
Northern patrol group – Heavy Cruisers Vincennes (division leader) Quincy and Astoria. Destroyer screen Wilson and Helm
Attachments and Detachments
The Flag Officer Rear Admiral Crutchley has been summoned to a conference off the beachhead and has left the southern patrol group with flagship Australia. It is unlikely that they will get back in time to play any part in the engagement. Damaged destroyer Jarvis has been ordered to return to base and has departed, but is believed to be still in the area.
Mission
Destroy the attacking Japanese force.
Execution
Attack will take form of a pincer movement. The northern patrol group will attack from north north-east, the southern patrol group from the south-east. Expected point of contact is the eastern end of the strait between Savo Island and Cape Esperance. Destroyers from the two patrol groups will execute co-ordinated torpedo attacks to the head and flank of the enemy column. After the torpedo attack has been completed, all ships will then engage the enemy with gunfire. The heavy cruisers will take on the enemy cruisers, and the destroyers will concentrate on the three enemy destroyers.
Scenarios 7b and 7ba
Same as Scenario 7a except Ralph Talbot detects enemy and Japanese are taking the northern approach into the sound.
Execution
Ralph Talbot Wilson and Helm to carry out torpedo attack on enemy, supported by fire from Vincennes, Quincy and Astoria. Chicago group to close and attack Japanese from south.
Scenario 7c Transports Attacked
Play as Allied, either Division 1 (Admiral Scott) or Division 2 (Admiral Crutchley) in ‘divisional command’ mode or as Japanese in either mode
Allied Orders
General Outline
We have been hit hard by a Japanese force consisting of cruisers and destroyers. Both the Chicago and Vincennes cruiser groups have been neutralised. Heavy casualties have been sustained and we are expecting a follow-up attack on transport areas X-Ray and Yoke at any time.
Area of operations: Map 1 Appendix 3.
Situation Enemy forces
Gunfire has just been heard in the vicinity of the southern entrance into the sound. Earlier we were attacked by a mixed force consisting of four to five heavy cruisers plus two to three destroyers and they are now returning to attack the anchorages.
Situation Friendly forces
Chicago group
Chicago has taken a torpedo hit to the bow and her top speed has been reduced to 25 knots. She is believed to be pursuing contacts to the west of the sound. Canberra is stopped and on fire with all power lost.
Vincennes group
Vincennes and Quincy have been sunk. Astoria is stopped and heavily ablaze.
DESRON 4 plus screening force destroyers
Heavy gunfire has just been heard from Blue’s location. So far, all attempts to contact her by radio have been unsuccessful. Ralph Talbot sustained heavy damage engaging the retiring Japanese force and is listing 20 degrees. She can only make 5 knots and is heading for Savo to effect repairs. Patterson lost her no 4 turret in the action and is currently standing by Canberra. A message ordering all available destroyers to concentrate has been received from Admiral Crutchley. The message is corrupt and it has proved impossible to decode the location of the concentration. However it has been assumed that Crutchley was referring to the original destroyer RV to the northwest of Savo Island. All destroyers are now either near or at that location.
San Juan Group
The San Juan and Hobart, plus destroyers Monssen and Buchanan are still on station patrolling the eastern approach into the sound.
Destroyer minesweepers
Attachments and Detachments
Hovey has been detached to escort the damaged Jarvis to Efate. Southard is currently engaged in ferrying Major General Vandegrift to Tulagi and is believed to be in the location of the Tulagi landing area (transport area Yoke).
Hopkins, Trever and Zane are still available for orders and patrolling to the immediate north of the Guadalcanal landing area (transport area X-Ray).
Mission
Protect the transports at all costs.
N.B. THE BOMBARDMENT CIRCLE IS TAKEN TO REPRESENT TRANSPORT AREA X-RAY. IF YOU KEEP THE JAPANESE OUT OF THIS YOU WIN. IF THEY GET IN, YOU LOSE.
Scenario 7d The Destroyers Concentrate
Play Allied or Japanese in ‘standard’ mode
As Scenario 7c, except the destroyers have successfully decoded Crutchley’s signal and concentrated on Australia as ordered in two divisions.
Attachments and Detachments
Division 5 consists of Selfridge, Mugford, Hull, Bagley and Helm. Division 6 of Henley, Patterson, Ellet, Wilson and Dewey.
Chicago has also been also been ordered to close on Australia and is now formed up with her in Division 2
Scenarios 7c and 7d Japanese Orders
General Outline
In the recently completed attack, our forces were split up and it did not prove possible to reach the main anchorage. This final attack will accomplish the mission. Time now: 04:30. First Light: 0600.
Area of operations: Map 1 Appendix 3.
Situation Enemy Forces
Situation reports indicate that one enemy light cruiser, eight heavy cruisers and five destroyers have been sunk with an additional five heavy cruisers and four destroyers reported damaged. More enemy warships are expected in the vicinity of the anchorages and it has been reported that aircraft carriers are also within striking range.
Situation Friendly Forces.
Chokai’s No 1 turret has been destroyed and both she and Tenryu have sustained a number of personnel killed and wounded. However, damage caused by enemy return fire has been minimal. All our forces have now regrouped in preparation for the final attack.
Attachments and detachments
Yunagi is too low on fuel to participate in another attack and has been detached to guard the exit route north of Savo.
Mission
1) Eliminate any surviving enemy warships
2) Destroy the transports in the main anchorage.
Execution
1) Ingress into the sound through the channel south of Savo Island destroying any enemy warships encountered.
2) Destroy the transports plus any remaining enemy warships at the Guadalcanal anchorage. Exit through the channel north of Savo Island.
The transports must be destroyed at all costs, even if this means exposing our force to air attack after break of day.
3 Scenario designers’ notes
General points
a) Players should ignore the fact that I have used the ‘Bombardment’ mission setting in many of these scenarios unless specifically mentioned. This was done solely to make the AI manoeuvre the enemy forces in a more appropriate manner.
b) There is a difference of opinion regarding the speed of the Japanese force during the engagements. In his book Guadalcanal , Frank Richard states this to have been 30 knots. Sandy Shanks agrees and has recorded the Japanese speed as such in his account of the action in The Bode Testament . This differs from the speed recorded by Loxton which stands at 26 knots . In email correspondence regarding this issue, Lieutenant Commander Mackenzie Gregory R.A.N. Rtd. (Officer of the Watch on board Canberra during the engagement) comes down on the side of Loxton. I concur and it would therefore seem odd that I have entered the Japanese speed as 30 knots in these scenarios. The reason for this is that FS/FSP does not allow 1300 yard spacing between ships. The nearest setting is 1500 yards. This means that the simulated Japanese column is longer than it should be in real-life and takes correspondingly longer to pass on by. The higher speed of 30 knots goes some way towards compensating for this.
Scenario 1a Chicago
Captain Howard Bode, skipper of U.S.S. Chicago was the only officer who was officially censured for his conduct during and after the battle. He subsequently committed suicide. This scenario attempts to cast some light on his actions. I have already commented on the freakish nature of many of the incidents that occurred and have explained my reasons for not trying to reproduce them in simulation. This scenario is the exception to the rule. The issue here is not what should have happened, but what did, and I have tried to reproduce as closely as possible the unreal situation that Captain Bode found himself in on that night, so that players can explore for themselves the options that would have been open to him.
By the time this scenario commences, all the torpedoes that took effect had already been launched. I have therefore disabled the torpedo tubes on all ships, both allied and Japanese, to ensure that the AI does not stray from the historical course of action by firing yet more. Canberra had been knocked out by this point and was just a burning hulk. To simulate this, all of her systems were disabled or destroyed and she was also set on fire via the scenario editor. The cloudbank was a major factor in this particular engagement. Tenryu, Yubari and Furutaka broke to the north east, and would have passed into it, causing them to lose contact with the remnants of the Allied southern cruiser group. This was simulated by setting the visibility level to 21%.
Scenario 2 Savo Island South Axis
Note Appendix 1 section 4: Temporary disablement of ships’ systems
Scenario 3 Savo Island North Allied
The AI was not capable of reproducing the Japanese manoeuvres taken during the course of this engagement, so a compromise had to be struck. The effect aimed at is to simulate how the American force became caught in the cross-fire between the two Japanese columns. Consequently, this scenario is an approximate depiction of the action only. However a large number of run-throughs have shown that the results mirror reality closely. The American cruiser group is always destroyed, with only light damage being inflicted in return.
Scenario 4 Savo Island North Axis
N.B. The AI always follows my suggested course of action in this instance, which is a sharp turn to starboard, taking the Allies away from the western column, and attempting to cross the T of the eastern force.
Scenario 5 Savo Island Complete
Note Appendix 1 section 4: Temporary disablement of ships’ systems.
If players wish to compare the effects of the Australian and American second degrees of alert, this scenario should be run with all weapons systems re-enabled and without any delays factored in. The results will speak for themselves.
Scenario 5a Savo Island Complete
It was found in play testing that the AI handles destroyers aggressively. In consequence it was decided to see what effect this had on the simulation. All deactivated destroyer weapons systems plus the torpedo banks on Canberra were re-enabled.
Scenario 6 Savo Island s. plus Australia.
This scenario should be played the following ways.
a) Historically, with limited destroyer use.
b) With Bagley and Patterson firing full spreads of torpedoes
In option b this scenario becomes winnable.
Scenarios 7 Savo Allies Alerted, 7a Savo Allies Cross the T, 7b and 7ba Savo Northern Approach
I have assumed in scenarios 7 and 7a that Captain Bode would have been given adequate warning of impending action and in consequence would have performed the ‘ships about’ manoeuvre and taken the lead. I do not believe that he would have had the time to accomplish this in the event of Japanese ingress through the northern approach and for this reason, in scenarios 7b and 7ba, I have left the disposition of the Southern group unaltered.
Scenarios 7c Savo Transports Attacked and 7d Savo Destroyers Concentrate
When attempting to simulate an attack on the anchorages, the limitations of the FS/FSP program were reached. The software could not cope with the numbers of vessels involved and only six divisions per side were permitted. A more focussed approach was required. Historically, the most likely course of action would have been for the Japanese to eliminate or brush aside all opposing forces first, and then to have destroyed the transports in the Guadalcanal anchorage (transport area X-Ray). With this accomplished, heading north and taking out the lightly defended ships at the Florida anchorage (transport area Yoke) would have been a mere formality. For this reason, it was decided to leave transport area Yoke plus related vessels out of the equation completely and concentrate instead on transport area X-Ray. Further streamlining was possible by setting the scenario mission to ‘bombardment’ and substituting the bombardment circle for the transports themselves. Fortuitously the AI was found to handle the attacking Japanese much better this way than in the ‘convoy’ mode in which the scenarios were first attempted.
The FS/FSP program does not include the U.S. class of vessel ‘Destroyer-minesweeper’. After consultation with the owners of the software, these were simulated by using flush decked destroyers with all torpedo tubes disabled. Their historic names were added by changing the short names via notepad and for added clarity adding the letters DMS in brackets.
My default setting for all American vessels were ‘green’ ‘fatigued’ crews with ‘poor’ night training, unless I had historical evidence that proved otherwise. In Scenario 7d I did not possess combat records for all the vessels involved, and on reflection I felt that giving all the unknowns rock bottom ratings would not have presented an accurate picture. Background reading on this and other battles indicated that due to lack of any standard doctrine, the U.S.N. at this stage of the war was somewhat uneven in quality. Much depended on flotilla commanders and of course the ship’s Captain. Some ships performed significantly better than others. For this reason, I selected two destroyers at random (besides Patterson) and assigned them higher values.
Scenario 7c was designed to be played in ‘divisional commander’ mode to simulate the historic problems in communication. Scenario 7d was designed to simulate workable communications and should be played in the ‘standard’ mode.
Scenarios 8 and 8a Radar picket – Jap and U.S. versions.
No briefings were necessary here. These scenarios were created to gauge – very roughly – when Blue should have detected the Japanese column, and at what range. The radar model I used can only be described as very approximate indeed. I tried to create a more accurate representation by seeking informed opinion, but no-one I asked was prepared to commit to a precise answer. The only thing I managed to confirm was that radar is still considered a ‘black art’, even within naval circles.
Possible scenario: single Allied cruiser division. (hypothetical)
When designing these scenarios it struck me that if the two Allied cruiser groups had been formed into a single division, they might have been able to give a much better account of themselves. Rear Admiral Morison concurs. However Commodore Loxton states that Crutchley rejected this option for a number of reasons, not least of which was that a group of more than four cruisers would have been an awkward force to handle at night. This would have been complicated even more by the fact that only three of the ships under his command had worked together at night before, and that two of these were not equipped with the T.B.S. radio system. I came down on the side of Loxton. It may yet be a valid exercise to model the Allied cruiser force as a single division, plus attendant destroyers. However one suspects that the “Sir one of our ships just missed a manoeuvre” message might become tediously repetitive.
< Message edited by PongoDeMer -- 5/31/2020 12:18:33 PM >
|
Printable Version
War on the Sea - Killerfish games - 
New Messages
No New Messages
Hot Topic w/ New Messages
Hot Topic w/o New Messages
Locked w/ New Messages
Locked w/o New Messages
Post New Thread