I think that you can get your answer from the tooltip on that market cap widget:
> Market capitalization is calculated as the sum of each NFT valued at the greater of its last traded price and the floor price of the collection, respectively.
So their market cap calculation does not attempt to capture a current fair market value for all outstanding CryptoKitties. For a non-fungible asset like this, just using the last sale price like this will always give a lagging indicator. In some sense, you could argue that it lags the fair market value by a possibly infinite amount of time.
Concretely, this method would assume that the Dragon cryptokitty that the article discusses has been worth a constant 600 ETH that hasn't fluctuated by even one iota in over 4 years.
I don't know how much better one could do for a market cap calculation. It certainly wouldn't be practical to individually appraise all 2 million cryptokitties on a regular basis. Perhaps one could look at how prices of more frequently traded cryptokitties have fluctuated over time and use that to generate a scaling factor for the old ones. But even that might have downsides. How do you account for the possibly large percentage of cryptokitties that belong to wallets whose keys have been lost? It doesn't make sense to count those into the market capitalization, because they are no longer part of the market. The method being used at least has the advantage of being clean and objective. It's just not particularly useful, is all.
> Market capitalization is calculated as the sum of each NFT valued at the greater of its last traded price and the floor price of the collection, respectively.
So their market cap calculation does not attempt to capture a current fair market value for all outstanding CryptoKitties. For a non-fungible asset like this, just using the last sale price like this will always give a lagging indicator. In some sense, you could argue that it lags the fair market value by a possibly infinite amount of time.
Concretely, this method would assume that the Dragon cryptokitty that the article discusses has been worth a constant 600 ETH that hasn't fluctuated by even one iota in over 4 years.
I don't know how much better one could do for a market cap calculation. It certainly wouldn't be practical to individually appraise all 2 million cryptokitties on a regular basis. Perhaps one could look at how prices of more frequently traded cryptokitties have fluctuated over time and use that to generate a scaling factor for the old ones. But even that might have downsides. How do you account for the possibly large percentage of cryptokitties that belong to wallets whose keys have been lost? It doesn't make sense to count those into the market capitalization, because they are no longer part of the market. The method being used at least has the advantage of being clean and objective. It's just not particularly useful, is all.