German tank problem

[poesel]

Raise your hand if you expected a bug report.

Bonus points if you are a beta tester and thought: 'OMG - this will delay 2.01 for another week'!

No, I stumbled over this by accident:

http://en.wikipedia.org/wiki/German_tank_problem

May be that's a well known in English speaking countries but new to me. Quite fascinating how they solved this and what sciences had an influence on the war.

[Seedorf81]

Maybe you can also find the statistical formula for the arrival of the Market Garden module?

[c3k]

I have the answer! Oh, sorry. Guess I'm disqualified. Carry on...

[Duckman]

Conventional intelligence gives an estimate of "pretty soon" based on gut feel and reduced volume of communications that usually precedes a release.

Statistical analysis shows a mean time between modules (MTBM) of approximately one year. This also indicates an imminent release since the CW module was released on March 5th 2012.

[JasonC]

Fun OR.

The bit in the article that struck me the most is that conventional intelligence estimates of German tank production in the early midwar period were 4 to 6 times too high, while the statistical OR method got it right. The conventional estimates undoubtedly looked at allied production rates and industrial capacities, and assumed the German output would be about as high, compared to total industrial capacity, as allied. That is where those figures of 1000 to 1500 tanks a month come from. The Russians were actually producing that many in 1942, and the US likewise once it geared up.

But the Germans were actually producing 250 a month. They don't hit those allied like totals until 1944. They do get there - the capacity calculations weren't far off. But conventional intelligence just could not imagine that the aggressive Germans who started all the fighting and seemed more ready and prepared than anyone, were tooling along a near peacetime production rates, clear into 1941-42, 2-3 years after the war started.

The Russians hit capacity in about 18 months, and the US on a similar time scale. It wasn't crazy that conventional intelligence assumed the same would be true on the other side of the hill. But it wasn't, dramatically. They got it wrong because they couldn't guess that the Germans could be so stupid, as not to fully mobilize the economy for war and focus on armaments output exclusively, until after Stalingrad.

[Duckman]

To be fair to the Germans their industrial base should probably be compared to Britain rather than the US. It's true that there were some glaring efficiencies up until Stalingrad, like still working single shifts in factories, but they were also putting a lot of resources into building plants that did not come online until later in the war.

The bomber offensive should also get credit for shifting a lot of German production capacity towards aircraft. Historians have generally focused on the failures of the bomber war, but that seems to be changing.

As for the Soviets, their totals are impressive but also hide the fact that they were producing a lot of light tanks and obsolete aircraft. That shouldn't detract from the huge success they had in massproducing the T-34 throughout the war, but it does skew comparisons somewhat. Speer's much-touted (not least by himself) production miracle was also partly achieved by keeping older models in production.

[Michael Emrys]

Speer's much-touted (not least by himself) production miracle was also partly achieved by keeping older models in production.

And according to Tooze, much of it, at least the first year or so, was merely programs that had been set in place before he got the job finally coming online.

Michael