Death and valor on an American warship doomed by its navy

[DGuller]

I'm not a big believer in ProPublica painting an accurate picture.  I may be judging from a sample of one, but the first time I heard about them, it was when they did a study on how auto insurers racially discriminate.  It didn't matter that some actuaries consulted before publication pointed out how their analysis was littered with rookie mistakes that drove their conclusion, they weren't going to take some expert mansplaining things to them.  After the publication they still dug in their heels and went "of course the industry will not admit that they racially discriminate and will instead attack our conclusions".  It's good to be an investigative reporter, but only if you are passionate about truth rather than criticizing the powers to be.

I'd be interested to know more about the rookie mistakes.

Basically they were trying to do an "all else being equal, do people in minority ZIP codes pay more" kind of analysis, but their "all else being equal" was very selectively chosen so that it was really "some else being equal".  A charitable explanation for their failure to properly do "all else being equal" analysis is that they didn't have the skillset to do it properly.  A less charitable explanation is that they knew what story they were going to publish, and it was just a matter of finding the right apple to compare with just the right orange to get to the conclusion.

[Admiral Yi]

Basically they were trying to do an "all else being equal, do people in minority ZIP codes pay more" kind of analysis, but their "all else being equal" was very selectively chosen so that it was really "some else being equal".  A charitable explanation for their failure to properly do "all else being equal" analysis is that they didn't have the skillset to do it properly.  A less charitable explanation is that they knew what story they were going to publish, and it was just a matter of finding the right apple to compare with just the right orange to get to the conclusion.

So something like they controlled for driving record but not local crime rates?

[grumbler]

A couple things I would love to here grumblers take on:

1. The officer not eating in the wardroom. This is reported as kind of a minor problem, maybe even just an eccentricity. HOwever, as a commander....I think I would generically see this as a really serious issue. My officers don't have to like each other, but they do have to respect each other, and they most certainly MUST have a functioning social (military social but still social) relationship. So much of the non-tangible success factors in the military are based on the relationships working, and one officer not eating in the wardroom regularly seems like a giant fucking canary in the coal mine that *something* is wrong that needs to be adressed. Or am I reading too much into that, in regards how it works in the Navy versus other services (my uninformed gut actually tells me this is MORE important on a warship than other services, not less)?

Yes, this would have been a big deal on any ship I served on.  The XO shuld have been on top of this - if the officer in question had a legitimate complaint of being attacked in the wardroom, the XO needs to put the attackers in their place.  If the complaints weren't deemed legitimate, the officer in question should be directed to dine with the others and given some advice on wardroom etiquette and expectations.  Watchstanding requirements and illness are the only reasons to miss eating with your fellow officers.

2. One thing the article mentioned several times was how many missions the ship was put on, with no real chance for rest and refit. One of those missions mentioned was a "show the force" kind of thing where they went with a CVBG. Now, this strikes me as a mission that might be important overall, but have the Fitzgerald as part of it is almost completely irrelevant. Having one more DDG there or not seems to make no real difference. So given its rather poor readiness, why send it? That hardly seems critical.

I see two possibilities here:

1. Its readiness may look poor, but it is actually pretty typical, and not really as large a problem as the article makes out. And maybe none of the ships there were any better, so if you didn't send ships like the Fitzgerald, you wouldn't send any at all.
2. It isn't so much about the need for the ship to be there, but that nobody wanted to admit clearly that it should not go because of its poor readiness and need for more downtime. So its not that it needs to be there so much, but that someone says "Hey, we need some ships for this, lets send, uhhh who is available....X, Y, and Fitzgerald". And nobody wanted to actually say "Yeah, unless you REALLY need them, that is not such a great idea....".

I suspect that it was a little from column A and a little from column B.  Fitzgerald probably wasn't notably worse off than her sister ships at Yokosuka, and no one wanted to be the guy who said "I know she looks ready, but we have failed to make her as ready as she looks."

Average days at sea for current USN Navy ships are higher now than they were in WW2.

[DGuller]

Basically they were trying to do an "all else being equal, do people in minority ZIP codes pay more" kind of analysis, but their "all else being equal" was very selectively chosen so that it was really "some else being equal".  A charitable explanation for their failure to properly do "all else being equal" analysis is that they didn't have the skillset to do it properly.  A less charitable explanation is that they knew what story they were going to publish, and it was just a matter of finding the right apple to compare with just the right orange to get to the conclusion.

So something like they controlled for driving record but not local crime rates?

It was a lot more basic than that.  What they did was calculate average claim cost for each ZIP code, and compared it against a quoted premium for a single hypothetical driver with certain characteristics in each ZIP code.  These two quantities are not comparable, because among other things the average amount of coverage purchased by drivers differs a lot between ZIP codes.  You either have to compare hypothetical driver claim costs against their quoted premium, or you compare average premium collected in a ZIP code against average claim cost in a ZIP code.

I don't know whether I'm managing to convey how basic of an error that was, but that was a very basic error.  To go forward with the story based on such elementary innumeracy reflects badly on their reputation as journalists.

[grumbler]

Right - but don't three start admirals have a lot of influence on that very problem? I mean, these are the guys calling the shots, right?

No.  The Combatant Commanders (the joint guys, like CINCCENT and CINCPAC) have, via the JCS, control of how and where the forces are used.  The services are just the force providers.  CINCPAC orders five destroyers to the ea of Japan, and Seventh Fleet just sends them the ships. The 3-stars at the Pentagon have no control over optempo.

[mongers]

....
I don't know whether I'm managing to convey how basic of an error that was, but that was a very basic error.  To go forward with the story based on such elementary innumeracy reflects badly on their reputation as journalists.

Did those journalists write the article in the OP that is being discussed by Grumbler et al?

[DGuller]

....
I don't know whether I'm managing to convey how basic of an error that was, but that was a very basic error.  To go forward with the story based on such elementary innumeracy reflects badly on their reputation as journalists.

Did those journalists write the article in the OP that is being discussed by Grumbler et al?

The organization ultimately sets the slider on sensationalism/objectivity.  This article describes the low level of maintenance and the state of disrepair of the radars.  Is that a big thing, or really a non-issue for people in the know, and par for the course on every ship that also doesn't crash into a freighter?  I don't know, but I definitely wouldn't trust ProPublica's assessment.  They have a story to push, and from past experience I know that facts and context wouldn't stand in their way.

[mongers]

....
I don't know whether I'm managing to convey how basic of an error that was, but that was a very basic error.  To go forward with the story based on such elementary innumeracy reflects badly on their reputation as journalists.

Did those journalists write the article in the OP that is being discussed by Grumbler et al?

The organization ultimately sets the slider on sensationalism/objectivity.  This article describes the low level of maintenance and the state of disrepair of the radars.  Is that a big thing, or really a non-issue for people in the know, and par for the course on every ship that also doesn't crash into a freighter?  I don't know, but I definitely wouldn't trust ProPublica's assessment.  They have a story to push, and from past experience I know that facts and context wouldn't stand in their way.

That's not an answer to my question.

[Admiral Yi]

It's a rebuttal of your premise.

[Admiral Yi]

Or maybe not.  Maybe your question was innocent.

[DGuller]

The answer is no.  None of the three authors of the Fitzgerald story were among the four authors of the hit piece I took issue with.

[dps]

Average days at sea for current USN Navy ships are higher now than they were in WW2.

How much, if any, of that is due to nuclear-powered ships having almost unlimited range?

[DGuller]

I just realized that another investigative piece that at the time annoyed me greatly due to wanton disregard for math also came from ProPublica.  I saw an article in reference to the original work, and could spot the obvious math error while reading it, but I didn't realize that the original work came from the same innumerate crew from ProPublica.  It was about recidivism algorithm used by courts for parole decisions being biased against black people.  Their proof that the algorithm was biased was merely a mathematical certainty that resulted from black people having a higher risk score than white people.

[grumbler]

How much, if any, of that is due to nuclear-powered ships having almost unlimited range?

I would say zero.  The only nuclear-powered ships left besides submarines are carriers.  The problem is that CENTCOM needs 2 carrier battle groups in the IO, EUCOM needs one in the Med, and PACCOM needs one and sometimes two in the western Pacific. There are only eleven battle groups, and three or four are in overhaul at any given time.  That means that half of the rest are deployed at any given time, and the other half are working up for deployments (which workups are being cut short when a surge is needed to, say, the South China Sea).  The Navy is trying to support a 600-ship-navy deployment pattern with less than 300 ships.

Of course, this negatively impacts retention, which then negatively impacts readiness, which in turn negatively impacts retention...

That's why the Navy brass in the article were so focused on shipbuilding rather than readiness - because the only way out of the vicious cycle (besides reining in the joint commanders, which won't happen) is to build enough ships to meet the deployment needs.  Sounds like a plan... but then they squandered the construction budget on the DDG-1000 (Zumwalt) class program, which spent $22 billion to get three ships (which cannot deploy and have guns for which no ammo will ever be produced).  That's where the three-star types added to the problem.

[mongers]

I found this a quite interesting listen:

https://play.acast.com/s/warcollege/gettingthenavybacktoreadywillcosttensofbillions

Getting the Navy Back to Ready Will Cost Tens of Billions
13 June 2018 17:48

With just over 270 ships doing the work intended to be done by more than 350, the U.S. Navy is stretched thin enough for the seams to show. In order to keep enough ships at sea, training is getting short shrift, and so is sleep. And that means accidents, some of which have been deadly.

What's to be done about it? Dave Majumdar of National Interest shares a few suggestions, but none of them come cheap.