I was listening to the Yapline earlier and @R2D2zen was talking about how @wallchain leaderboards are being gamed So i took screenshots of all the leaderboards and ran it through an LLM to find similarities, oddities and see what comes out of it The results will surprise you

The screenshots for the following seven leaderboards were taken today and processed: • wallchain • limitless • idOS • genome • heyelsa • apecoin r.a.i.d. • covalent Focus was limited to looking at Mindshare % across different leaderboards to see if a pattern emerges

Here's what the LLM pulled out: • @GemsScope appeared on 5 leaderboards • @ArtvisionNFT, @MookieNFT, @leaf_swan - in 4 each • @erequendiweb3 - appears 3 times • @zuri_nft - appears 2 times All within a 0.7 - 1.3% mindshare band

Different types of projects have different audiences, yet these same accounts repeatedly hold about 1% mindshare, no matter the leaderboard That is statistically not possible without automation or coordination The mindshare of these accounts also shows concerted moves. When one rises, the rest follow by ±0.1–0.2% showing possible clusters

Accounts like @beijingdou or @GratefulApe_eth show large variances in their mindshare on different projects, jumping between 1% to 10% It's variances like these that are indicative of actual mindshare engagement with a real audience

This analysis is only indicative of a potential problem on Wallchain and it's up to @max_bevza and @kyparus to look into this Obviously, a single-day snapshot of different leaderboards is insufficient to reach a solid conclusion. It's not a top-only issue either You'll probably find more of the same further down the different leaderboards in varying degrees. And this points to an issue with the Wallchain algorithm and penalties for suspicious patterns displayed

Finally... @ripchillpill recommended to compare the Wallchain leaderboards on Wallchain itself and on Xeet Interestingly, only three accounts overlapped. Everyone else was unique to a single platform If this due to a difference in the algorithm or some other factor? I'm not able to make a definitive conclusion for this, but it was an interesting one to look at