I did but looking at it again it isn't that clear: "replicating basic functionality", "some semblance of relevant search results". You're right that emphasizing

I did but looking at it again it isn't that clear: "replicating basic functionality", "some semblance of relevant search results". You're right that emphasizing that more would be clearer. So since the edit horizon for the post has expired I'll add it here. Regarding quality criterion: replicating PageRank is possible ( link graph betweeness centrality ), and AFAIK, it's not this algorithm, although it is useful that yields Google's current edge, it's their proprietary embellishments. What I intended for this question is it is a replication of PageRank algorithm, which is still AFAIK, very useful. We can also assume that we add any low cost improvements to it that help to fight spam or otherwise improve results. Qualitatively, people looking at these search results would say, "well they're not quite as good as Google, but they are good, there's little spam, they're aware of my search history context." So basically, a lightweight version of Google Search.