Timing your NFT bids using market emotions



As Weiwu Zhang explains on this article “What I learned by (unsuccessfully) bidding $440,000 USD for the Economist’s NFT”, winning a top rated NTF auction can be extremely hard due to network congestions, delays in the updates of bidding increments, and other issues.

Here at EdgeSeekers we have been developing tools to track the popularity of tokens and collections before they have a price history (more on this in a future post). 
However, the other key components to anticipate demand and possible network issues are actually externalities: the market sentiment about the crypto that is used to issue the tokens, and the emotions expressed by market participants. Here we will demonstrate our approach to timing the market using text analytics and sentiment.

First, we look at the overall network perception: is there a buildup in the sentiment and volume of references (buzz)? Is this coming from mainstream media outlets or just social channels? What are the actual sources (to the YouTube profile and subReddit level)? These indicators inform us about possible upward pressures in general public interest that can in turn create conditions for network congestion.
Then we look at the specific emotions that are being expressed in the interactions of market participants, which are expressed as standardized scores that allow us to compare between different types of sentiment and across different blockchains that are preferred by NFT issuers (Ethereum, Cardano, Solana, etc).

Fear, market risk, long-short forecast…all are important dimensions to consider (specially in the last 5 days close to the auction). We also pay attention to things that break some pattern: for instance here Future vs Past indicates a current backward looking outlook, which may in turn ease some tensions in regards to short-term market expectations:

Here is the full list of emotions tracked:

Metrics definitions:

Created with