The USN - optimistic modelling?

[grayst]

Right, so I know this may be controversial and I'm metaphorically buckling on my tin helmet and clearing for action.

I tried a little mini-sim of the USN BBs against an equal number of Koenigs and Kaisers. The US ships won fairly easily.

I tried it without Texas and NY, just the US 12" ships against their German counterparts; they won again.

Is this realistic?

I know that the option for "crappy AP " is being added for US ships in the new update, but I had gathered the impression that when the historic US Batdiv9 joined the GF as 6BS in December 1917, its gunnery standards were found to be be below those of the GF.

That was late 1917; what would they have been like in 1916?

My question is whether there has been a bit of "star-spangled-bannerism" leading to unrealistic optimism about USN gunnery.

Or, of course, whether my one test was just a freak result, or if I've been reading history through union-jack-tinted sources.

[barkhauer]

These results are meaningless without a statisticly significant sample. Say, 100 sims.

[grayst]

Well yes, on the other hand someone like BH can tell us exactly what modelling & percentages have been applied for US ships versus GF and HSF.

[barkhauer]

You're implying nationlism/patriotism affecting the outcomes of a simulation based on 2 tests under different conditions. You have no way of knowing if the US ships are horribly underpowered and you got lucky on the random number generator on the one match you tried.

[saddletank]

That's why I think he said he was interested in hearing about the figures from BH.

I believe the comment about the US battleships poor gunnery was reported by Jellicoe himself. The US ships went on several gunnery training sorties in their early months at Scapa and improved dramatically.

[Invincible]

There are several factors that help the US ships. First they fire all guns in each salvo so they have efffectively double the rate of fire of their German and RN counterparts. Next the spread is greater. Normally a tight spread is associated with better gunnery but I have found the US more likely to get at least one hit on a salvo (but less likley to get multiple hits).

For a similar experience put Westfallen (no central director, all gun salvos) and Posen against similar targets and see who gets more hits over time. I have generally found it to be Westfallen. The ammo does get depleted faster though.

[Bullethead]

My question is whether there has been a bit of "star-spangled-bannerism" leading to unrealistic optimism about USN gunnery.

Nope, it's just what can happen sometimes in any game you play, and at different points in the same battle. Gun accuracy is affected by a huge number of variables that are constantly changing independently of each other. The result is that in any 1 given period of time, individual ships can shoot quite well or quite poorly. The same ship can go through hot and cold streaks in the same battle. Some days, BCF shoots better than I.SG at least for a time, to take an extreme example.

Because of that, it takes lots and lots of test runs, under the same conditions, to determine what the actual hit rate is for any of the several types of fire control systems and methods.. As Barkhauer said, something on the order of 100 runs of the same test over and over.

When we made the original Jutland, I ran such tests for the Germans and British. And I had to do them many times over after each tweak of the fire control systems, until we finally arrived at long-term average hit rates very close to what Campbell says. And let's face it, there's really not a whole lot of other data to work from for 1916. But Campbell is great because that was under combat conditions, not target practice, and the game is about combat conditions.

So then we did the US ships. Given that the USN fought no fleet actions in WW1, there's nothing at all for detailed combat accuracy data for it at that time. We thus had to do a lot of extrapolation and, to be honest, SWAGging. The bottom line is that, over the long haul, USN BB accuracy averages out between RN BBs and BCs. IOW, their shooting is pretty mediocre.

However, as Saddletank said, the peculiarities of USN gunnery methods have a strong effect on any one run of a test shoot. In general, USN ships are slower to find the range and can't keep it as long as RN and KM ships. However, because of their wider spread and more shells per salvo, they can often get hits where the other ships would get near misses or straddles. IOW, USN ships have a wider margin of error when they're close to being on target, but they usually have trouble staying there and once they lose the range, it usually takes them longer to get it back.

As a result, most of the time, when USN ships find the range, they tend to hit with more salvos while they have it, but these relatively short periods are spread out between long intervals of groping for the range. Thus, on average, USN have an overall lower hit rate. However, when a USN ship is on a hot streak, its target is going to suffer many hits in quick succession.

That last point is pretty important. Ships have only so much damage control ability at any given time. Thus, when ships get hit many times in quick succession, damage control is only able to deal with some of the problems at once, not all of them, or perhaps only a little on each of them. And some problems get worse over time if left insufficiently attended to. IOW, a ship that gets hit 10 times over half an hour will usually take less total permanent damage than a ship hit by the same 10 shells in the same places but only over 5 minutes. This factor can make USN ships more dangerous than their rather mediocre overall hit rate would otherwise warrant, especially when they're able to hold the range for longer than usual.

But OTOH, the USN burns through its ammo at twice the rate of the other navies. Thus, cold periods hurt the USN more than other navies because they waste twice as many shells. Then, of course, the higher rate of fire really limits the combat endurance of USN ships, whether they're hitting or not.

The net result is that USN ships are the least predictable of any in the game so far. On average, they're mediocre, but their full broadsides amplify the effects of them going hot and cold. When they're good, they're quite good, but when they're bad they're terrible. But either way, they can't spend all day trading broadsides because they run out of ammo twice as fast as the enemy, regardless of which side they're on.

[martin worsey]

Would it not be fair to suggest that actual data on ship performance under fire is lacking. The suggestion that an in game trial has to be run about one hundred times to establish true values for gunnery would imply that the same should be true for a similar evaluation under combat situations.

Campbell raises this point in the text accompanying the calculations on percentage accuracy and qualifies that the only really meaningful comparison is that the shooting of 1/2BCS left much to be desired when compared to 5BS. Even then, calculation of the figures for Queen Mary, show a very similar level of accuracy to the 5BS whilst she was in action.

Comparisons of the relative effectiveness of the classes of warship will thus by nature be somewhat subjective.

American battleships were generous in displacement when compared to contemporary British and German vessels and the centreline disposition of turrets saved the weight of superfluous turrets and also aided protection in keeping them inboard. Underwater protection was allegedly in advance of contemporary British and German warships but it has been suggested that there were detail deficiencies. Overall armour protection would appear to have been good (although this is not necessarily reflected in the data for the individual ships in game) and this appears consistent with the allocation of displacement to protection. Consequently, it could be expected that these ships would be capable of withstanding a fair amount of punishment.

Staying afloat and continuing to fire will be something of an advantage, even if gunnery is a little below par.

[Bullethead]

Would it not be fair to suggest that actual data on ship performance under fire is lacking. The suggestion that an in game trial has to be run about one hundred times to establish true values for gunnery would imply that the same should be true for a similar evaluation under combat situations.

Yup, and that's been done, too.

Campbell raises this point in the text accompanying the calculations on percentage accuracy and qualifies that the only really meaningful comparison is that the shooting of 1/2BCS left much to be desired when compared to 5BS. Even then, calculation of the figures for Queen Mary, show a very similar level of accuracy to the 5BS whilst she was in action.

In Jutland, QM has the same general accuracy level as RN BBs, which is considerably better than other RN BCs.

Comparisons of the relative effectiveness of the classes of warship will thus by nature be somewhat subjective.

Absolutely. This is true in making any game. While the ideal is 100% historical accuracy, this is never attainable because the data for that simple don't exist or (for games about relatively modern stuff) are still classified. Furthermore, much of the available data always contains at least some anecdotal or subjective component supplied by the person who recorded it, making it extremely difficult to quantify. Thus, game designers inevitably face a large number of points in building game mechanics where they have to make judgment calls based on the data they have and also play balance. Different designers will obviously come to different conclusions, give more or less weight to different pieces of evidence, etc.

Our gunnery system doesn't claim to be 100% historically accurate because that's an impossible dream. It does, however, claim to be a close approximation on 2 different levels. First, under conditions similar to those of the actual battle, it generates generally similar overall results. Second, this is accomplished not by fixing each ship with a constant level of effectiveness, which has been pretty much a fixture of wargames since forever, but by allowing them all to vary over time as they did in real life. This was a difficult achievement of which we're rather proud.

The result, we think, provides a more immersive gaming experience than a more determinative combat system. General expectations of combat results at the big picture level are fairly solid, but the details at lower levels are highly variable. Thus, combat entails a more realistic level of risk than with a more determinative system of game mechanics.

[saddletank]

I wanted to clarify that BH referenced my comments in his first post. I think he meant to reference Invincible's points, not mine. I don't want to be given credit for another forum member's post!