NBA Top Shots…

An NFT collection made proudly happen by Dapper Labs, who used an efficient proof-of-stake blockchain, Flow, to make the project happen. They collaborated with the NBA to capture the highlights, the moments, of each player, each team, each season. The short clip linked to each NFT acts as the digital collectable baseball card of the web3 era.

In this dashboard we are going to examine the Sales Volume of NFT Top Shots overall and specifically the Sales Volume’s relationship with the cards play type attribute.

I’ve collected available data from all marketplaces, taking care to express all sales in USD. In the last 3 months secondary sales volume decreased by ~60% by early July. That is until we consider the volume between July 13 and July 18. 1300 - 5200 USD. 1-5% of the already dwindling volume on July 12. This is abysmally low.

The price of the Flow token itself was stable in this period and even increased by approx. 25% since July 18.Unless this improves soon, holders should consider the figure to the right →

In an effort to try and make sure I’m not putting my feet in my mouth I listed the USD volumes of the inverse of my previous selection - ie. Sales Volume of NFTs with NFT_collection tags not in the Top Shots metadata table. The result confirms that the excluded NFTs belong to different NFT projects. It also shows a similar trend in their volume - miniscule since July 12, non-existent since July 18.

Considering the price action of Flow and the lack of any relevant announcements on FLOW’s twitter in the indicated time-period my prime suspect for the lack of trading data since July 18 is Flipside itself (confirmed on Discord, data stopped capturing due to underlying change in Flow’s representation)… With this out of the way let’s address the connection between the connection between Top Shots Volume and play type attribute on the data we do have.

Most NFTs in the NFT_Sales table are missing from the Topshot Metadata table (NFT_IDs don’t match, even though the sold NFTs are in the same collection) even for transactions in April-May, (ie. this is a separate issue with Flipside’s tables).

I’m restricting further analysis to NFTs in the Topshot Metadata table.

• Tx count distribution is within 1.2% of the NFTs attribute distribution.
• Weekly distribution of transaction counts stays within 15% of the overall attribute distribution.
• largest difference on the week of June 20 with 3 Pointers and Assits being overrepresented by 10-15% and Rims being underrepresent

• Overall Volume distribution is within 6% of the NFT distribution by play type
• Prices dropped by ~80% starting in May, however volume only dropped by 20-50% up to the week of June 13
• Weekly Sales distributions show greater fluctuations than tx counts and can diverge by up to 25% from the NFT distribution
• The average price of most play types is ~5x larger than the median. This is due to a small number of large price transactions that occurred mainly in April

• The average hold time (time between sale transactions) of NBA Top Shots is ~ 20 days for all play types if we include NFTs without any sales in the observed period (assigning the time between their creation and the end of the period to them)
• If we only consider NFTs with at least one observed sale, the average is reduced to ~16 days.
  • The longest hold time is for Steals (18.5 days) a slightly undervalued category.
  • The shortest holds ~14 days are for rarest 3 play types (Handles, Dunks and 2 Pointers); these types were overrepresented in the ‘NFTs without transactions‘ category

Conclusions:

1. Bug 1: Most Flow tables stopped updating
2. Bug 2: The majority of Top Shot NFT Trades registered in the Trades table don’t have an ID in the Metadata table.
3. Tx counts and Volume distributions by play type follow the NFTs distribution within 1.2 and 6 % respectively, but there are large (up to 15 & 25%) deviations from this in the weekly tx tallies.
4. Prices for all categories dropped by ~80% at the beginning of May (for the Mid-range -95%)
5. Traded NFTs were held on to for 14-18 days on average, rare categories were held the least

Overall the Play Type attribute only has a minor effect on the Sales transaction count and Volume of the NBA Top Shots beyond what is explainable by the NFTs category distribution, but we observed significant short lived deviations in price, tx count and volume.