Great piece. Messi being nearly half the soccer index is like the S&P 500 being half one company — at that point "soccer fell 5%" tells you almost nothing about soccer. And the index history being rewritten retroactively is wild. One question: did you try a volume-weighted average instead of the median for the sales? For the players with 20+ sales it might show whether the gains were driven by the cheap cards or the expensive ones.
Thanks. And good question. Honestly the two probably wouldn't diverge much here: it's one card in one grade, so the spread inside each column is narrow to start with. Four of the eight have 17 to 21 sales going in, which is enough to carry either measure; the rest sit on 3 to 6, where neither holds up. Median just felt like the cleaner fit for this kind of comparison and this card selection. The question underneath yours is the better one though: separating a broad move from one carried by a few trades - that needs several cards per player, not a different statistic on one.
Great piece. Messi being nearly half the soccer index is like the S&P 500 being half one company — at that point "soccer fell 5%" tells you almost nothing about soccer. And the index history being rewritten retroactively is wild. One question: did you try a volume-weighted average instead of the median for the sales? For the players with 20+ sales it might show whether the gains were driven by the cheap cards or the expensive ones.
Thanks. And good question. Honestly the two probably wouldn't diverge much here: it's one card in one grade, so the spread inside each column is narrow to start with. Four of the eight have 17 to 21 sales going in, which is enough to carry either measure; the rest sit on 3 to 6, where neither holds up. Median just felt like the cleaner fit for this kind of comparison and this card selection. The question underneath yours is the better one though: separating a broad move from one carried by a few trades - that needs several cards per player, not a different statistic on one.