Posted By:
GlennAbsolutely nothing there. That might change if we could look at data across the entire range of grades, but I rather doubt it would. From what we have so far the breakdowns for 150 and 350 series by grade are actually even closer than one would expect if there isn't any difference at all in average condition in the overall populations of 150 and 350 cards.
We needed to get a chi-square value of at least 5.99 to say with any confidence that the hypothesis was supported. Our actual chi-square value is 0.30, which is impressively bad evidence. Even if we're completely wrong, sampling error alone would ordinarily yield a higher chi-square value than that. (It's almost like SGC is specifically trying to make the numbers equal for the two series.)
Breakdown of grades by series (from Ted Z's data):
Grade .............. 5 ............ 5.5 ............. 6
Series 150 ........ 40 ............ 17 ............. 7
Series 350 ........ 114............ 58 ............. 22
This has to be compared to
Expected breakdown (rounded to whole numbers for this chart but not in the actual analysis) if the grades are distributed the same for 150 and 350 series:
Grade ............... 5 ........... 5.5 ............ 6
Series 150 ........ 38 ........... 19 ............ 7
Series 350 ........ 116........... 56 ............ 22
The two breakdowns are really astonishingly close to one another. I doubt someone is actually trying to make them even out (why would they?), but we have much better evidence for that than we do that 150 series cards are in worse condition than 350 series cards. Mind you, they actually should be at least a little bit worse if for no reason other than their slightly greater age, but we haven't a shred of evidence in this thread to suggest that they are.