tannoy blog — the decline of google search

fog

Google search used to be very good at returning results related to the search terms. But this is no longer the case.

I just asked it for “statistics by country sea area controlled”.

The first result returned is the wikipedia page on the West Bank. This page is not a comparison, it is not statistics, it is not about a number of countries, it is not even about one country (the West Bank is part of Palestine, Israel, or even Jordan, depending on your politics). Of course, the West Bank does lie on the Dead Sea, so the hit isn’t 100 percent wrong: Google Search only ignored the primary search terms.

An important point is that my search terms weren’t the best. But, even so, the results returned should be connected to the search terms given. I could then refine my search to get the exact results I wanted. But that doesn’t work when the results are unconnected to the request.

Part of the problem used to be that Google didn’t have serious competition. They had no external pressure to maintain standards. For example, their main competition, Bing, used to return results that had no connection with the search terms.

But that’s changed. When I put the same search terms into Bing, I got directed to NationMaster, a site that specialises in statistics about countries. Admittedly, it didn’t link directly to the results I wanted, something Google might once have done, but, it was a correct result.

Bing also seems to have sorted out its language problem. It used to return results based on where you were searching from, not your preferred language. That’s fixed too. Today, I got results in a language specified by my browser settings.

I wonder if Google is trying to outguess me, to presume it knows what I want better than I do? That would certainly explain why it gets it wrong. Presumptions are a motorway to mistake.

It’s why I often find myself mocking them, by pretending to make silly presumptions and following them through to their insane conclusions. I do it mostly to try and stop myself making the same mistakes. Interestingly, more than a few people don’t suss this. It’s as though they’re comfortable with insane conclusions from silly presumptions, as though they live with them all the time. Perhaps they do.

Anyway, it looks like I now have to take Bing seriously.