Introduction:

Recent advances in battery capacity, pump technology, and battle style, along with observational experience at MWC events has lead many people to comment that ships are getting harder to sink. Having been in the hobby since only 2008 and noticing these same things, I was curious how much better ship survivability is getting, and if there was a way to measure it. Being a nerd for statistics and data analysis I have pooled some data to try to answer a question that many of us have been suspicious of. Are ships getting harder to sink?

Methods:

I was able to track down scoring data from the past 10 Nats, sort it, and analyze it. For my analysis, I included all class 4-8 ships that sortied in a regular 2 sortie fleet battle conducted at Nats from 2002 through 2011. I included all ships that competed two sorties that didn’t sink as well as ships that sank in the first or second sorties. Ships that were declared sunk or withdrew from combat were excluded. I also excluded night battle because it is only one sortie and battle conditions are atypical. I excluded 1 on 1 battles and Wednesday challenge battles because they do not necessarily exemplify the typical battle style during a general fleet battle. I chose to include only class 4-8 ships for the analysis because smaller cruisers and destroyers in class 3 and lower don’t typically use the same battle tactics and technology that the majority of the ships, the capitol ships do.

Results/Discussion:

Perhaps the hardest part was to decide how to sort the data. I will present several different methods of sorting and interpreting the data.

All Ships.

In this series, I sorted all of the ships by year into two groups; ships that sunk and, ships that did not sink. The average holes and average damage by year is graphed below.

Both of these graphs illustrate a few global points. Though it might seem obvious, both of these graphs illustrate the point that when ships sink, they take on average more holes and more damage than ships that don’t sink. The other qualitative observation is that all of the lines are approximately flat, with just a mild trend upwards, especially if you ignore the 2005 sink data point.

Quantitatively, on average, the amount of total holes and total damage ships take has remained fairly consistent throughout the last 10 years. Overall, number of holes it takes to sink a ship is about 80, with linear regression actually showing a decrease of about 1/2 hole/year. The number of holes ships took that didn’t sink was on average 52, with linear regression showing an increase of just 1/4 hole/year. The amount of damage it takes to sink a ship on average over the last 10 years is 1666, with linear regression showing an increase of only 12 points per year. The amount of damage ships that didn’t sink took was on average 1043, with linear regression showing an increase of 23 points per year.

All Ships
holes	Average	Change/year	r2
Sunk	79.93832	-0.5903	0.012
Not Sunk	52.00857	0.247	0.0106
damage	Average	Change/year	r2
Sunk	1665.846	12.183	0.0116
Not Sunk	1042.51	23.257	0.1405

Now back to the original question, “Are ships getting harder to sink?” These data are a little mixed showing that it is taking around 1/2 fewer hole/year in ships that sink, but actually increasing by 12 points/year. The other trend is that ships that don’t get sunk, ships that are surviving the battle are taking more damage (about 1/4 hole and 23 points more points per year). Another surprising observation is the 2005 sink data. It indicates that of all of the Nats in the past 10 years, ships were sinking with more holes and more damage on average than any other year. Interestingly, the ships that didn’t sink that year didn’t take more damage than in other years. Whether it was due to a high average quality of the ships or captains present, or a high number of ships that had some type of partial mechanical failure right before they got trashed and sunk, or some other factor that I can’t think of, the data indicates that at Nats 2005 in Ionia, MI may have been the hardest year on record to sink ships, and when they did go down it was only after lots of bbs.

So strong is the 2005 data that it may actually be skewing the trend somewhat. If we ignore the 2005 data then the trend line for holes does flip upward and would show that ships that sink in combat are taking on average just 1/4 more holes per year (as opposed to decreasing by 1/2 hole) and increasing by 30 more points per year (instead of 12).

Hard sinks/easy sinks.

A confounding variable that comes to mind when considering how much damage ships are taking is the fact that in any given battle, there are ships that seem to sink without adequate damage. These ships are typically sink because of forgetting to turn on the pump, clogged pump, dumb luck, or a poorly built ship. In the same way, at any given battle there are typically several ships that skirt off the water with hardly any damage, to the effect that some times you wonder if they even fired their guns. In order to try to control for some of the extremes, I decided to sort the data in another way. I somewhat arbitrarily but somewhat guided by experience decided to draw a line at 50 aboves and/or 10 belows. In my approximation, if a ship didn’t take at least 50 aboves or 10 belows, in most circumstances it shouldn’t be considered to have taken enough damage to sink and will for the purposes of this article be considered an easy sink. Conversely, a ship that went down with at least 50 above or 10 belows can at first approximation be considered to have earned that sink, and will be considered a hard sink. Similarily, if a ship didn’t take 50 aboves or 10 belows and didn’t sink, they probably were never really in danger of sinking and shouldn’t have sunk, so I will call these easy floats. Again conversely, ships that took more than 50 aboves or 10 belows and stayed afloat probably were at least pumping some and were probably in danger of sinking if the circumstances were right. I know these are not perfect cutoffs, but they will serve to eliminate some of the low damage outliers.

Table2	Sunk	Not Sunk
>50 above and/or >10 below	hard sink	hard float
<50 above and <10 below	easy sink	easy float

When controlling for ships that did not sustain adequate damage, again the some general trends are noticed. In a logical way, ships that sunk take more damage (measured by both points and by holes) than do ships that don’t sink. And again we notice that ships that sunk in 2005 for some reason sank with much more damage than other years. Here, the trend lines however qualitatively appear to be sloping downward.

Ships with Adequate Damage
holes	Average	Change/year	r2
Sunk	97.2076	-2.0864	0.1667
Not Sunk	73.2152	-1.3727	0.5502
damage	Average	Change/year	r2
Sunk	2034.213	-13.717	0.0215
Not Sunk	1521.741	-1.3114	0.0018

Again we venture back to the original question: “Are ships getting harder to sink?” According to these series in which ships that were easy sinks and ships that were easy floats were excluded so that only ships that “should” have sunk or ships that were in danger of sinking and didn’t (ships with more than 50 aboves and/or more than 10 belows) were included: the answer is - no. In fact it appears that ships are actually getting slightly easier to sink, and that over all ships are taking slightly less damage. Ships that sink with adequate damage are taking on average 2 fewer holes and 13 points less damage per year (average being 97 holes and 2034 points to sink a ship with adequate damage). Ships with adequate damage that don’t sink are taking on average about 1 fewer hole and 1 point less damage per year (average being 73 holes and 1523 points in a ship that didn’t sink with adequate damage).

Percent that Sink

Another interesting trend to look at is of the sorties completed, how many ships didn’t make it off the water.

On average over the past 10 years at nats, 26% in any given fleet battle end up sinking, and the trend is slightly increasing (by only 1%/year). Of those ships that sink, on average 23% sink with damage that could be inadequate to cause a sink (less than 50 aboves and 10 belows). Overall, that’s 1 in 16 ships in any given battle that sink with inadequate damage (pump failure, captain error, poorly built ship). This trend however is decreasing very slightly (about 1%/year). Qualitatively however, it does appear that starting in 2008 through present, the trends may be indicating that more people are sinking at nats and fewer people are sinking without adequate damage.

On the flip side, the percentage of ships that don’t sink that do take enough damage that they could be considered in danger of sinking (more than 50 aboves or 10 belows) is increasing. Again, the trend seems to be fairly clearly strongly ascending when looking at the 2008 through present data.

With these two data sets it appears that the case could be argued that ships are taking more damage.

Discussion/Summary:

Some years are up, some years are down. Overall, it doesn’t appear that ships are getting harder to sink. From that conclusion it is tempting to think that pump and battery technology is not improving, but I don’t think I would go that far.