3 ms·
Debating whether or not its stupid to do the steel test at -100F misses the point, the performance of the steel is plotable as a curve on a graph, and the cold
by h2odragon 5y ago
Debating whether or not its stupid to do the steel test at -100F misses the point, the performance of the steel is plotable as a curve on a graph, and the cold test is one point of data that goes into the interpolation that makes that curve.
If you're assuming inputs to that equation that dont exist it has implications beyond just "how does that steel perform cold"; it means that you might actually have a quite different curve but won't know it because your data was bad to begin with.
Of course we hackers know how to avoid such obvious and silly mistakes; and make things like Zillow's home buying algorithms work perfectly...