Analysis of Naval Gunnery Combat

[Tom Hunter]

The Model

One of the points/questions that a number of people have brought up is how much we know about what is going on in the model. The asnwer is some, but maybe not everything we would like to know. Here is a list of what we know.


We know how many shells a ship fired. We have the actaully ammunition loaded on the ship, the game has a number of rounds of ammo on the ship, and by dividing the rounds into the number of actual shells we can calculate the number of shell per round.

You may ask "what if the designer reduced the ammo capacity for play balance." My answer is that he can't. He can put more shells into an ammo factor, but he can never change the fact that HWS Warspite had 800 rounds of big gun ammo on board at full load. I hope this makes sense to everyone. In a historical game, the value of ammo, armor or number of guns is always directly analagous to the historical warship. Even if you decide that in your WitP your going to say "my Warspite has 4" guns for the main battery" you are not changing the historical battery of 15" guns on the real Warspite, instead your are designing a game with the wrong values for 15" guns.

There are 18 ammo rounds on the Warspite for her main guns and 18 goes into 800 44.44 times, or 44 rounds per gun. We know that if Warspite shoots 3 rounds that is 132 shells.

Another thing we know is that when a hit is reported it does damage. If the hit is non-penetrating it does less, little or no damage, and if it does penetrate it does more, but we have never seen a ships sys damage go down as the number of shell hits went up.

We also know that bigger penetrate more and do more sys damage than smaller shells. When I look at a ship that has taken 2 non-penetrating 8" hits the sys damage will have gone up 1-3 % looking at a ship where the shells penetrate it goes up 10s of %.

Finally we know that the combat report does report real damage. Ships with reported penetration of their engine spaces move more slowly in the next movement phase. Ships with reports of AA guns destroyed really do have them missing the next day.

Finally we know that the ships really are firing on the target they list. The combat report does not say Atago firing at Prince of Wales when Atago is actually firing at Repulse.

So the truth is we get quite a lot of information from the combat reports, number of hits, penetration of hits, area penetrated, reports on certain types of damage, ship firing, type of shell fired. If we look at the ammo before and after we can easliy get the number of shells fired as well. By the way I rounded Warspites 133 (3x44.44) down to 132 because BBs with even numbers of guns tend to fire even numbers of shells in a salvo.

From there we can get the % hit rates, find out if the ship fired at all and identify oddities like ships that fire secondaries but not main guns, or ships that fire on one target or many targets (more on that coming) or ships that keep firing and hitting under circumstances that make continued accurate firing very unlikely.

I wanted to put in some defense for the model, since some of the counter arguments seem to come down to "the combat report is all lies." I don't see any evidence that is true, maybe more of it should be lies but that is a different discussion, and right now it reports truth.

[witpqs]

ORIGINAL: Tom Hunter

... 1 divided by 132 = .0075 ...

Tom, I'm behind in reading up on this thread so excuse me if somone has pointed this out already. Many of your percentages are expressed wrong and might be messing up one's train of thought. 1/132 is indeed .0075. But percentage is "(hits/shots) x 100", so the percentage of 1 hit in 132 shots is .75%. Likewise some others (from your original post) 1 in 42 is 2.3% (not .023%), 1 in 21 is 4.7% (not .10% - this one must have been a typo or something), etc.

The reason it matters is that the historical sources you cite are using the standard A/B x 100 formula, so the comparisons wind up being off.

BTW, I do think you are on to something. Problem 1 is the number of times an inferior task force comes back for more. I understand better crews and higher rated leadership will get you more, but that expample seems like too much more.

Problem 2 is the BB main batteries lack of firing. You can be too close for those guns to fire, but that seems to have (again) happened way too much (meaning probably at too great a range).

Problem 3 is the hit percentage. Need to be careful with this one. Due to the other factors (surprise, leadership, crew night experience) and hidden (from us) tactical factors arising from them, the 'correct' number of hits might still be lower than 'historical' because the historical instances were more favorable.

Great ananlysis Tom. You're helping everybody. Thanks. [8D]

PS: Lack of torpedoes I figure can be fixed real easy, so not in the same league as the other issues you've uncovered.

[tsimmonds]

If memory serves in the second battle of Guadalcanal the USS South Dakota was hit 14 times by 8" and 14" shells

IIRC SoDak was hit 39 times by 5", 5.5", 6" and 8" shells; only a single hit was by 14". Many of the shells did not explode (including the 14").

[mdiehl]

Here's an interesting read:

http://www.dcfp.navy.mil/mc/museum/War_Damage/57.pdf

The "crew night experience" is a red herring. It's a rating presumed to matter in real life but that has not been obviously indexed to anything real. Ship crew EXP should be completely eliminated from the game.

[Tom Hunter]

Witpqs,

Sorry for the confusion I have seen this expressed two ways and probably got them muddled. Basically when Warspite shoot 132 shells between 3 and 13 of them should wind up on target depending on a wide variety of factors, but barring unusual circumstances, such as a target that dodges and does not return fire.

You are correct and I will start writing it the right way. [8|]

[Tom Hunter]

irrelevant, do you have the link to the South Dakota's after action report? I have been looking for it but unable to find it. It would be helpful to have along with any other logs or detailed acounts.

mdiehl, I am not certain I agree with you, but my opinion is that the experience system does not do a good job of reflecting training or doctrine. However I really want to keep this discussion on topic, rather than getting into a "this is my opinion about what is wrong with the game." Type thread.

I much prefer a thread that shows where the variance between historical combat and the game combat models.

I am afraid I don't have time to get to the cruisers and Oklahoma today, maybe tomorrow.

[treespider]

ORIGINAL: mdiehl

Here's an interesting read:

http://www.dcfp.navy.mil/mc/museum/War_Damage/57.pdf

The "crew night experience" is a red herring. It's a rating presumed to matter in real life but that has not been obviously indexed to anything real. Ship crew EXP should be completely eliminated from the game.

I haven't read the article yet but how do you propose that a situation like Java Sea be modelled with the command problems inherent in the force and the lack of training together? Would this be handled by the Commanders ratings?

[mdiehl]

Tom there is something like the SoDak's AAR in that link that I provided above. I'd have cut and pasted but that function is blocked in the PDF file they provide.

And I'm not trying to distort this into "what's wrong with the game" but people are trying to claim that your analysis is incorrect on the grounds that crew EXP may be the "overlooked factor" in your analysis. My point is that attributing it to crew EXP does not mean that the model is good. Only that *if* crew EXP is the culprit then the crew EXP assumptions are wrong... which would not surprise me in the slightest since no one has articulated a relationship between anything real and the EXP values assigned to the ships in the OOB.

[treespider]

There are 18 ammo rounds on the Warspite for her main guns and 18 goes into 800 44.44 times, or 44 rounds per gun. We know that if Warspite shoots 3 rounds that is 132 shells.

I would agree with you except for one point. Not knowing what the designer intended.

In some game systems I am familiar with, in any given combat those three rounds may actually represent from 100-150 shells (or some other range) and not 132 every single round.

Over the course of multiple combats you would hope that the three rounds would average out to be 132 shells but in any given combat they may only represent 100 or 150.

If this is the case then the only meaningful way to analyze the BB's would be to look at multiple combats instead of one single combat.

[Erik Rutins]

ORIGINAL: mdiehl
The "crew night experience" is a red herring. It's a rating presumed to matter in real life but that has not been obviously indexed to anything real. Ship crew EXP should be completely eliminated from the game.

I disagree with that. It's a game variable designed to reflect night naval combat performance. Call it "Crew night experience" or "night training" or "night doctrine", either way there's evidence that not all navies were equal in this regard.

Based on your comments (i.e. "presumed to matter in real life"), are you suggesting that crews trained in night combat had no advantage over crews that did not train for night combat at all?

Regards,

- Erik

[mdiehl]

I disagree with that. It's a game variable designed to reflect night naval combat performance. Call it "Crew night experience" or "night training" or "night doctrine", either way there's evidence that not all navies were equal in this regard.

Based on your comments (i.e. "presumed to matter in real life"), are you suggesting that crews trained in night combat had no advantage over crews that did not train for night combat at all?

1. Can you name one Allied ship DD class or larger that was in an engagement in the PTO in 1942 whose crew had no training in night combat?

2. How do you know that the scale of difference between any two ships accurately reflects their relative capabilities at night combat?

3. If you had to throw USS Augusta into the game as an ahistorical "what if," how would you rate her night combat experience in contrast to, say, IJN Mogami, as of 8 December 1941?

4. How would USS Pillsbury fall on that ranking scheme?

[Ron Saueracker]

The problem comes down to experience and training. However, what qualifies as experience or training? Does getting ones ass whooped at Savo have the same impact as achieving a clear victory at say, Cape St. George? Does the poor excuse for training the USN practised prewar equal that the IJN reportedly undertook?

There are so many things which would impact combat ability, from the highest political or military leader to the lowest naval rating. Not modeling ship crew factors was the biggest oversight IMO. Ship crew members were constantly changing, so really, the actual experience level would probably have been more successfully modelled as an overall rating for the entire navy in question, as opposed to individual vessels. By utilizing a "pool" for naval "squads" (term used only to compare it to LCU squads), crew rotation, new construction, hostilities only, loss of crew to combat etc) an overall rating could have been achieved.

[treespider]

There are so many things which would impact combat ability, from the highest political or military leader to the lowest naval rating. Not modeling ship crew factors was the biggest oversight IMO. Ship crew members were constantly changing, so really, the actual experience level would probably have been more successfully modelled as an overall rating for the entire navy in question, as opposed to individual vessels. By utilizing a "pool" for naval "squads" (term used only to compare it to LCU squads), crew rotation, new construction, hostilities only, loss of crew to combat etc) an overall rating could have been achieved.

Isn't this already the case in essence? Everyone starts with a rating some better than others, then each ship earns experience based on their experiences (perhaps I'm wrong on this). The ship experience would be passed on to replacements as they are introduced to the ship (OJT on the job training so to speak) barring a wholesale change of the crew. Just some thoughts from the peanut gallery.

On a side note, Professor Dull was enamoured enough with the night fighting prowess of the Japanese that he references it on three seperate pages of his Battle History of the IJN.

[MkXIV]

The system damage argument needs to be chucked. The game already models sys damage and destroyed mounts. No one here is auguring that destroyed mounts should still fire! Undamaged mounts are considered online and capable of firing (or at least there is NO documentation to state otherwise)

Furthermore yes naval combat is confusing and yes oddball stuff happened but even then ships fired. In the first battle of GC most US ships (and probably IJN ones too) had poor Situational Awareness (most of them I think were trying to figure out exactly what the hell Callahan was doing) but yet they found targets and fired, and hit. Most ships mis ID'ed targets but yet they still fired. and Yes they did check fire when Callahan accidently ordered the entire Task Force to Cease fire, but once the initial confusion settled down the battle continued.

Often surface combat plays the role of South Dakota in Guadalcanal II! Heck even SoDak, which had major issues, did fire her main guns repeatedly and, by the way, did one hell of a job blasting the crap out of her own search planes. Rarely, if ever do I see the model take on the model of Washington. I have had battleship flagged T.F. cross the T and not even fire their main guns.

I can understand the fact that ships sometimes remain unengaged, that I can handle, and I can handle ships checking main gun fire for one reason or another, but the part that bugs me is..... why do secondary guns not behave the same way? Battleships will dish it out with 5in guns, but I haven't had a battle where the 16 inchers were wailing away and the 40 mm's were "off doing something else"

No one is saying all surface battles should be perfect slugfest, I don't want that. However when time after time 2 T.F. of 2-3 BB and 4 CA and 6 DD go at it and someone crosses the T and the range goes from 12,000 Yards to 6 to 2 then back to 8 and the grand total of fire is some long lances and a few CA's trading secondary fire, something stinks

[Ron Saueracker]

ORIGINAL: treespider

There are so many things which would impact combat ability, from the highest political or military leader to the lowest naval rating. Not modeling ship crew factors was the biggest oversight IMO. Ship crew members were constantly changing, so really, the actual experience level would probably have been more successfully modelled as an overall rating for the entire navy in question, as opposed to individual vessels. By utilizing a "pool" for naval "squads" (term used only to compare it to LCU squads), crew rotation, new construction, hostilities only, loss of crew to combat etc) an overall rating could have been achieved.

Isn't this already the case in essence? Everyone starts with a rating some better than others, then each ship earns experience based on their experiences (perhaps I'm wrong on this). The ship experience would be passed on to replacements as they are introduced to the ship (OJT on the job training so to speak) barring a wholesale change of the crew. Just some thoughts from the peanut gallery.

On a side note, Professor Dull was enamoured enough with the night fighting prowess of the Japanese that he references it on three seperate pages of his Battle History of the IJN.

It's not the case unfortunately. Ship crew ratings only go up, despite the real life gutting of experienced crew to man new construction. This was a serious problem for navies expanding like the USN, but we don't see this at all. Most of the new construction start with higher crew ratings than the ships manned by career regular navy types at game start.

[barbarrossa]

ORIGINAL: el cid again

The other thing is you may have the wrong idea about what an ET is (or was)?

I was just pullin' your chain a little cid, and yeah I know what an ET is.[:)]

This was so long ago it was almost before the invention of the alphabet! "Solid state" was oo and ah - and you almost never saw any of it - but if you did it was discrete transistors - no one had invented even the smallest integrated circuit yet.

The Mark 13 radar 9 main battery FC radar) on Wisconsin had only one transistor in it, I believe it dated to the Korean era. In Basic Electricity and Electronics school we troubleshot to the component level after covering all the theory stuff and giving the calculator a work out. That was even before we got close to "A" school.

The AN/SPG-55B mod 9 I worked on dated back to the mid-60's too- big, hot analog cards.

Although I did 2 years in the U.S. Army too as a Patriot FC tech, quite a bit of digital stuff there. The FC computer was only about 8 tiny cards!

What FC system was it that you bailed your diseased Fox Division out of?

Just curious, love to talk shop[:)]

[barbarrossa]

ORIGINAL: mdiehl



The designers assumptions, whatever they are, are incorrect. The results he obtained in that combat are implausible.

I don't know, the IJN was pretty shot up. I think Tom's real beef is that those BB's didn't fire MB in the animation.

That was a 15 ship TF he was using, with 2 other friendly task forces in the same hex. Perhaps there's something in the routine that takes into account the possibility of friendly fire so a "cease fire own ships" might apply? Maybe this is a reflection of the 15 ship TF maximum rule.

I dunno, a TF with one elderly BB escorted solely by a bunch of PT's is a little more unrealistic than the overall results of this combat.

Oh no, I crossed mdiehl..........INCOMING!![:D]

[jsglenn4]

I do find it interesting that a single surface action in a single game has inspired such a lengthy thread about the naval combat model. I doubt anyone will have to think too hard to come up with a historical battle(s) the end result(s) of which would seem "broken" when looked at on paper beforehand. To borrow a cliche from the sporting world, "it's why they bother to play the games." While I am relatively new to the game, I have to say that I've not noticed a pattern of ridiculous end results (let me add that I also don't watch the combat animations). Now, if most everyone else out there seems to be getting wank results time and again, then perhaps there is something that needs looking into. Until then, I am content to accept that this is not a tactical simulation and believe that plausible explanations could be made for the very few clinker results that do come down the pike in my games.

I also must confess I don't really think the IJN force got off all that easy. It looked rather beaten about to me.

[mogami]

Hi, Although they were finished off by air attacks I doubt the IJN ships would have survived long enough to return to port had they not been attacked by the air.

[el cid again]

The designers assumptions, whatever they are, are incorrect. The results he obtained in that combat are implausible.

We have insufficient evidence to make such a conclusion at this point. This analysis is but a single tactical action. It is impossible to draw valid conclusions from less than a statistically significant number of datum points (that is, 30 +). It also it possible that "implausable" is meant in a wholly subjective sense: if you mean "improbable" and are not using the correct term, then you are right; if you really mean "implausable" in the sense of "not believable" that is incorrect: I find them improbable but quite plausable. The difference here is based on a different expectation of the sorts of events possible in a naval battle. At least I have the advantage of having witnessed USN losing a naval battle in which it enjoyed apparently overwhelming technical advantages: this may really help adjust my attitude about the range of the possible. Most modern readers really think the US armed forces are nearly invincible in nearly all situations, whereas the truth is very different: we are regularly surprised tactically, operationally, technically and, in addition, sometimes we have bad luck. While that is not the norm, it is not the 1 in 100 shot or less that posters in internet forums love to think it is.

For an astonishingly ugly story of USN technical problems, look up the US Naval Institute Proceedings article "When the Birds Didn't Fly." It is the story of the "3T" missile program - Terrior, Tartar and Taylos - in its early years. These were SAMs, and for a long time they didn't work at all! By the time of the Viet Nam War, we barely got things working, but the combination of techincal problems and the fact the engagements were not the sort the missiles were designed for (closing targets), our score overall was a whopping 60:1 - 60 missiles fired per kill. In spite of this, we dared to attempt to use SAMs in an anti-missile role, and we had at least three ships which NEVER missed any target with more than 100 shoots - even when firing practice rounds without warheads. One ship never failed to knock the target down even with practice rounds, in spite of the fact hitting a practice target is nominally forbidden and more than nominally impossible to do. While our greatest successes in air defense were electronic warfare based, we did achieve a significant hard kill score on at least one occasion (wether just to demonstrate we can, or because the electronic defense failed, I do not know): USS Sterette engaged and destroyed two MiGs and a Styx in rapid succession in 1972. [See Conways All the Worlds Fighting Ships - originally in the NATO volume - or in the later combined Cold War era volume]. This is an example of the opposite possibility: ships may do a great deal better than the statistical average. In a war in which most ships scored so badly their entire arsenal of SAMs would fail to hit, some ships never missed at all. And at least one ship had a very bad habit of shooting at its friends! I witnessed USS Boston fire a Terrier while it was locked on to USS Waddell - this is testimony since I was on the passive ECM set at the time. Had the missile not jumped the beam, it must have hit. When it did jump the beam the radar REMAINED locked on! Nothing but dangerous incompetence can explain that. The range safety system for Terrier was "turn off the beam." [That way, if it jumps the beam, it self destructs.] In fact the normal "kill the target" signal was "turn off the beam." Leaving the beam on meant it would stay on course until it hit the target - or failed. Much smaller missiles hitting ships have been terrible - see cases of US and RN ships hit by Exocet which didn't detonate. In real world navies, things go very wrong very fast. That they do is entirely plausable.