 | | Re: Quality issue - update —Scott Arthur
| | | | (...) If exact brick size variation tends towards(1) a normal distribution there will be equal numbers of oversized and undersized bricks. Measuring the length of a “stack” of bricks will basically cancel out any error. Scott A (1) I say “tends (...) (20 years ago, 5-Jul-04, to lugnet.lego, lugnet.color, lugnet.general, FTX)
| | | | | | | | | | Re: Quality issue - update —Mark Bellis
| | | | | | (...) Have you accounted for common mode error? Many bricks are produced by the same mold. If 12 2x4s are produced per machine cycle, then 1 in 12 will be from the same part of the mold. The randomness of the distribution depends how many machines (...) (20 years ago, 5-Jul-04, to lugnet.lego, lugnet.color, lugnet.general, FTX)
| | | | | | | | | | | | Re: Quality issue - update —Ross Crawford
| | | | | (...) Agreed, with a large enough sample size. I hardly think 10 bricks (or even 40) would be enough to cancel out all such errors. Besides, if my stacks reduced the error, doesn't that mean the older bricks are actually likely to have worse (...) (20 years ago, 6-Jul-04, to lugnet.lego, lugnet.color, lugnet.general, FTX)
| | | | | | | | | | | | Re: Quality issue - update —Scott Arthur
| | | | | (...) You are correct; I should have worded that a little better. ;) Scott A (who thinks this would make a good high school sats project) (...) (20 years ago, 6-Jul-04, to lugnet.lego, lugnet.color, lugnet.general, FTX)
| | | | | | |
