Hot damn, I love JavaScript. Have I told you that lately? It’s like this wonderful magic nobody ever told me about. And now that I’m getting a grip on it (sorta-kinda-MAYBE), I can use it to explore all the search problems I’ve been thinking about for the last almost-30 years.

The one I ponder a lot and always circle back to is: how do you ask for what you don’t know?  When your ignorance is just that of a missing element (I know about going to movies, I need to know when a specific movie is playing) search is easy. But when you approach a topic with very little knowledge, your ignorance and knowledge gaps shape your query.

Then Google shapes your result with its algorithm and whatever biases it contains.  So depending on how ignorant your query is and how machine-biased Google is, you can end up with this shiny little result that gives you a shallow answer, confirms whatever unconscious biases YOU might have, and leaves you feeling like you’ve gained expertise or knowledge (and therefore inappropriate confidence.)

It seems to me the best way to counter that is to put enough density in your search that it defies those “shiny-but-shallow” web pages that SEO, especially AI-based SEO, is so great at creating. By adding keywords relevant to your topic, you’re forcing Google to find you more information-rich results. But in order to do that we’d have to know our topic, which takes us back to our original question: how do you ask for what you don’t know?

One possible answer is to rely on someone else’s expertise. And that’s what Clumpy Bounce tries to do: use Wikipedia categories, and pages within those categories, to build successful Google queries.

The premise is simple: the more topic-focused keywords there are on a page, the more likely it is to be a solid page and not a shiny shallow SEO thing. Of course, if you do a search for TOO many topic-focused keywords, then you tip over the other way and all your search results will be regurgitations of Wikipedia data.

Here’s how Clumpy Bounce works, in three steps:

1. Enter a keyword search (anything that might be in Wikipedia) and you’ll get a list of categories associated with that keyword.

2. Pick a category and you’ll get a list of checkboxes showing you the most recently-popular pages in that category.

3. Pick up to three and Clumpy Bounce will “clump” them into a Google search along with some cruft-reducing anti-search and “bounce” you to a new tab of Google search results.

The idea is to direct your search to rich, information-dense results in Google. And in my testing it works pretty well as long as you don’t pick your pages from too general a category. Luckily I’ve got a great example to walk you through: fusion energy!

Using Clumpy Bounce

You might have heard that there’s been a possible breakthrough for fusion energy and now you want to learn more about fusion energy and associated topics. Let’s see how that works with Clumpy Bounce. First enter your keyword, which in this case is fusion energy.

Click on Find Categories For Your Query and you’ll get a dropdown list of categories to which that topic belongs.

In this case there are only three results, but fusion power is one of them. I’m going to choose that option and click on the partially-obscured button underneath, Find Popular Pages In This Category.

Now you’ll get a list of checkboxes that reflect the 15 most recently-popular pages in that category. (Sometimes the list will contain less than 15 items because pages that start with “List of” are filtered out.) Click on up to three of these items. If you don’t recognize anything on the list, try to pick three that you think would go together. In my case I’m going to choose fusion power, Aneutronic fusion, and General Fusion. Then I’m going to click on Build A Google Search For This Topic.

We won’t get anything else on the Clumpy Bounce page this time – instead you’ll be bounced to a new tab with a page of Google search results.

A very manageable 94 results, with some compelling enough that I already peeked at them as you can see. Going through the results I can infer that General Fusion is a company, and I wasn’t really looking to include companies in my initial search. So I swap out General Fusion for Lawson criterion on the Clumpy Bounce checkboxes and click the Google search button again.

While the results are still focused on fusion energy in general, they are very different from the first set, as you can see.

But there are three categories under which fusion energy exists in Wikipedia. A second one is called Sustainable Energy and it generates a very different set of popular pages:

When you’re choosing pages to build into a search query, follow a few strategies:

1. Don’t choose ambiguous terms: “soiling” can be about solar energy or nervous chihuahuas. I’m not using it in my Google search.
2. Sometimes someone or something will be so overwhelmingly popular that it will end up at the top of a popular page list even though they’re only peripherally-related to the topic. Avoid choosing those pages because they’ll hurt your search results.
3. Avoid choosing three items that are next to each other – that can trip a lot of site results which use Wikipedia data.
4. When in doubt, choose a couple of items in the middle.

In this case I’m choosing Geothermal energy, Fusion power, and Solar energy. Again I get a nice rich set of search results.

These Google searches have been good so far because the categories we’re using to find pages are good – they’re not too specific and not too general.

If you choose categories that are too specific you might find that you get too few search results. If you  choose a category that is too general your search results won’t be particularly focused. Let’s look at the last category to which fusion energy belongs, Emerging technologies.

As you can see, that’s a much more topically-ambiguous list. You can choose pages here and while you’ll get fun results, it’s less likely that they’re going to be focused on emerging technologies – it’s too general a topic. Here’s results for 3D Printing, nanonsensor, and CRISPR:

In my testing Clumpy Bounce I found it worked well for all kinds of things, like researching businesses in a particular industry or area:

Or looking up natural science topics:

You can even search people and explore topics that way. This is going to come in handy as my husband is on a huge David Graeber kick.

There is a bug which I have identified but have yet to quash: if a page in the category was created during the timespan being checked for page popularity, the search will fail. I thought I had an easy fix, but the easy fix blew everything up, so I’ll keep pecking at it. In the meantime if you try to get a page list for a category and don’t get any result within a few seconds of choosing the category and clicking the button, try again. It shouldn’t happen too often.

(One of the things about working on your own is you figure it out when you figure it out.)

I think of these different JavaScript skills I get as I build the Gizmos to be like Lego pieces . Being able to make menus based on user choices and API call results is a very useful piece indeed; I might go back and revisit some earlier Gizmos. Maybe Gender Scanner should have a keyword-based category search?