Limits in your searching can be so annoying...
 |
| Scales to get relative and absolute weights. P/C Jean Poussin, from Wikipedia. (CC BY-SA 3.0) |
... and there are times when you'd like to look at the data, but the tools just don't quite let you do the analysis you'd like to do.
This week, I happened to be looking at which of the most common holidays in the US created the most search traffic. Why? As you know, many people look to Google search traffic as an approximation for the level of interest in a topic, and I wanted to know which of the holidays occupies more mental space in the brains of US citizens? Christmas? Thanksgiving? The Super Bowl? Or.. what?
To figure this out, I really needed to compare multiple search queries at the same time. Sounds great.
But there's a problem: How can I create a chart like this, with more than 5 variables? (Google Trends only allows you to compare 5 different queries!)
As you can see, here I'm showing 6 different holidays on the same chart. This is a "Relative Search Volume" chart, and not the absolute numbers. When Google computes such charts, the measure with the largest volume will be normalized to 100%. All of the other data is based off of that.
So... the obvious Challenge is this:
1. How can you get more than 5 search volumes compared side-by-side?
And a corresponding question:
2. Since these are all relative values, are these BIG numbers, or relatively small ones? How would you know?
Tell us what you think the answers to the Challenge are in the comments. What clever method did you use to get to 6, 7, or more variables on the same Google Trends chart?
Keep Searching!
I searched [google trends more than 5 words] a Keyword is the key. Different articles and how to use them.
ReplyDeleteYou make the comparison and then you use sheets to make those 2 or more 5 items comparisons in just one.
Maybe there's a better new solution. I'll keep Searching!
Also thinking. In Dr. Russell's list: The Super Bowl is a different comparison. That is Christmas and Thanksgiving are Holidays. Super Bowl is not so maybe comparison is searching using only those three a good way to find out answer? I'll try that too
¯\_(ツ)_/¯ I have NO frame of reference for this - export to spreadsheets?
ReplyDeletehttps://youtu.be/Kkd2PkbCdSc?si=qTTF2OpOqO6exxH-
see comments too -
https://digitaljobstobedone.com/2017/07/10/how-do-you-compare-large-numbers-of-items-in-google-trends/
#2 - LLN -
https://www.investopedia.com/terms/l/lawoflargenumbers.asp
https://en.wikipedia.org/wiki/Law_of_large_numbers
CLT -
https://www.investopedia.com/terms/c/central_limit_theorem.asp
https://www.embibe.com/exams/comparing-very-large-and-very-small-numbers/
trend tutorials -
Deletehttps://www.youtube.com/@GoogleSearchCentral/search?query=trends
cockle shells & corals & light management -
ReplyDeletehttps://www.npr.org/2024/11/23/nx-s1-5199634/heart-shaped-mollusc-shell-resembles-fiber-optic-cables
oh, & on the heels of judicial rumblings - Open AI & possible search disruption
https://wccftech.com/after-shaking-up-search-engine-market-with-searchgpt-openai-is-now-gearing-up-to-challenge-google-with-its-own-web-browser/
Ryugu & Bennu - corruption limits
https://gizmodo.com/rare-asteroid-sample-contaminated-by-microorganisms-despite-scientists-best-efforts-2000528691
from your scale image… see description –
ReplyDeletehttps://en.wikipedia.org/wiki/Weighing_scale#/media/File:Balance_%C3%A0_tabac_1850.JPG
"The Ancient Egyptian Book of the Dead depicts a scene in which a scribe's heart is weighed against the feather of truth."
https://en.wikipedia.org/wiki/File:El_pesado_del_coraz%C3%B3n_en_el_Papiro_de_Hunefer.jpg
https://en.wikipedia.org/wiki/Maat
https://en.wikipedia.org/wiki/Hunefer
https://timetrips.co.uk/papyrus_of_hunefer.htm
https://timetrips.co.uk/book%20of%20the%20dead.htm
https://en.wikipedia.org/wiki/Ammit
many forms,all badass…
https://en.wikipedia.org/wiki/File:Ammit_BD.jpg
Wonderful images from the Egyptian pantheon.
Deleteimagery will become everything, only the eyes lie
ReplyDeleteEugene Goostman - " a 13-year-old Ukrainian boy"
well, at least he isn't Ammit… yet
https://scottaaronson.blog/?p=1858
https://en.wikipedia.org/wiki/Eugene_Goostman
https://en.wikipedia.org/wiki/Computing_Machinery_and_Intelligence
https://en.wikipedia.org/wiki/Philosophy_of_artificial_intelligence
was Eugene in Duat? " star-in-circle: 𓇽." is Turing Thoth? AI Osiris?
https://en.wikipedia.org/wiki/Ancient_Egyptian_afterlife_beliefs
Ammit reappears -
https://en.wikipedia.org/wiki/Duat
https://en.wikipedia.org/wiki/Osiris_myth
"Eugene is a common male given name that comes from the Greek εὐγενής (eugenēs), "noble", literally "well-born", from εὖ (eu), "well" and γένος (genos), "race, stock, kin". Gene is a common shortened form."
https://www.google.com/search?q=source+of+the+name+Eugene&rlz=1CAACAC_enUS1032&oq=source+of+the+name+Eugene&aqs=chrome..69i57j0l5.14818j0j7&sourceid=chrome&ie=UTF-8
https://en.wikipedia.org/wiki/Io_(moon) (also a calico)
https://www.britishmuseum.org/collection/object/Y_EA6752
https://www.nytimes.com/2015/10/02/arts/design/animal-mummies-unwrapped.html
fancy feast tins? yet to be found by Zahi Hawass
https://www.npr.org/2018/11/11/666704847/archaeologists-discover-dozens-of-cat-mummies-100-cat-statues-in-ancient-tomb
no wonder the fascination -
https://en.wikipedia.org/wiki/Quay_with_Sphinxes
https://en.wikipedia.org/wiki/Bonaparte_Before_the_Sphinx
https://www.napoleon-series.org/faq/c_sphinx.html
https://www.thebritishacademy.ac.uk/blog/what-is-egyptology/
https://www.google.com/search?q=british+egyptologist&rlz=1CAACAC_enUS1032&oq=british+egyptologist&aqs=chrome..69i57j0l5.18530j0j4&sourceid=chrome&ie=UTF-8
Happy T day to all sRs readers & Dan (not "T" testosterone)
another scale -
https://www.gocomics.com/pearlsbeforeswine/2024/11/25
https://www.gocomics.com/getfuzzy/2024/11/25
I tried asking Gemini how to get around the comparison limit on Google Trends. It can back with downloading the data from different trend searches and learn how to write a python script. It also suggested a Chrome extension called Glimpse Glimpse Everything I read sounded like it would work. In the short time I tried to make it work, I couldn't figure out how to get more than five comparisons.
ReplyDelete¯\_(ツ)_/¯
I had a rethink about this and went back to the A.I.'s and asked them to create a comparison chart of the major U.S. holidays.
DeleteGemini created a chart in python, but used hypothetical data and told me it couldn't access the actual data in Google Trends. It also included the code so a user could plug in the real data.
Perplexity gave me a table and very fuzzy comparisons and focused it on marketing possibilities around different holidays. Perplexity appears to be getting more into sales and shopping than it was initially.
Sobek & Taweret (don't play) - life and death on the Nile…
ReplyDeleteas a team:
"Ammit was the Egyptian idea of the punishment of the soul. The name means "devourer" or "soul-eater." Ammit was usually known as 'The Devourer of the Dead' or the 'Eater of Hearts'. Ammit was believed to eat any souls found to have sinned. They would then be digested for eternity in acid."
https://marvelcinematicuniverse.fandom.com/wiki/Ammit
https://anthropologyreview.org/history/ancient-egypt/egyptian-god-ammit/
http://thenilerivertas.weebly.com/the-dangers-of-the-nile.html
https://www.youtube.com/watch?v=zBnchVmArKE
https://www.youtube.com/watch?v=OOCb1tmx6Mc
https://www.google.com/search?q=crocodile+hippo+battle&client=firefox-b-ab&sca_esv=9784fd51e0900df8&tbm=vid&sxsrf=ADLYWILySbyXDTnlK9Tga6LipVxpIWjWsQ:1732560482912&source=lnms&sa=X&ved=0ahUKEwjh14rhkviJAxWSHjQIHfsWC7AQ_AUIBygC
https://www.readingmuseum.org.uk/blog/sacred-animals-ancient-egypt
"The hippo is also often found within grave goods, in the form of a blue faience model decorated with examples of river plants such as the lotus flower. This promotes a connection with growth, new life and cosmogony; rather than an image of chaos and ferocity. Hippos have the ability to continuously submerge themselves underwater for several minutes before resurfacing – a wonderful metaphor for rebirth and regeneration. When underwater, sometimes only their back is visible. This was reminiscent of the Egyptian creation myth where the first primeval mound rises up from the chaotic waters of Nun. The ancient Egyptians believed that placing hippopotamus models in their tombs would provide them with this renewing power and would guarantee their rebirth, by magically passing over these qualities. Interestingly, many hippo statues have been discovered with broken legs – it is possible that this was a deliberate attempt by the ancient people to avoid any unfortunate incidents with the animal after death, as it was believed that depictions in the tomb could magically come to life. For them, it was certainly better to be safe than sorry."
'Nile' not 'Nihil'… ;^P
Nihilism is the belief that all values are baseless and that nothing can be known or communicated. It is often associated with extreme pessimism and a radical skepticism that condemns existence. A true nihilist would believe in nothing, have no loyalties, and no purpose other than, perhaps, an impulse to destroy.
Field of Reeds didn't Kevin Costner/W. P. Kinsella make that movie?
ReplyDeleteAaru — Twenty-One Secret Portals
https://en.wikipedia.org/wiki/Aaru
Neorxnawang:
https://en.wikipedia.org/wiki/Neorxnawang
Thoth:
https://www.huntmuseum.com/stories/objects-in-focus/thoth-baboon-and-ibis-god-of-the-moon/
related - Cahokia - notions still in flux
ReplyDeletehttps://youtu.be/ruWuAas8T7Y?si=z7gRMnHkdCWP3mFQ
in the comments -
"In 1904, 16 mounds were destroyed to build out St Louis Forest Park for the World’s Fair. St Louis used to have a nick name of mound city because of all the mounds which is hard to believe given that there are is just one mound remaining on the StL side of the Mississippi"
https://www.youtube.com/@pbsterra/videos
https://en.wikipedia.org/wiki/Portal:Ancient_Egypt
DeleteFirst, I tried adding to the value of the q parameter in the URL: https://trends.google.com/trends/explore?geo=US&q=Thanksgiving,World%20Series,Easter,Christmas,Super%20Bowl,New%20Year%27s&hl=en-US
ReplyDeleteNo luck: Oops! There was a problem displaying this page.
I tried the same technique again, but using the non-”search term” entities like "holiday" or "festival" that look like this in the URL: %2Fg%2F11fldyv50_
That didn't work either.
I figured the answer would be to build a chart by combining the results of multiple queries, and I see Ramon Gonzalez figured that out too.
For the normalization to make sense, I knew I'd need to keep some parts of the query consistent across multiple searches, so I could fit the pieces together on the same scale. I wasn't sure how to go about this until I noticed what looked like a hint in your post: "the measure with the largest volume will be normalized to 100%. All of the other data is based off of that.”
That told me I only need to have the largest term (of the whole set) in each comparison.
We already know that Christmas is the biggest from your set. But how do we figure that out? Let’s try a different set:
decorations, destinations, flights, sales, gifts, recipes
Starting with the first five, [flights] is normalized to 100.
Now let’s swap out one of the others for recipes:
Flights is still at 100, so we know that’s the biggest.
Before going to the CSV export, I tried one more hack since it was right next to the CSV download button. Similar to the URL-based attempt, I tried adding additional terms the the embed code. That didn't work either.
So, I downloaded CSV for the first 5 terms, then did a second query with only [flights] (the largest) and the extra term.
I created a new sheet in Google Sheets and followed these steps:
1. Import the first sheet as a new spreadsheet.
2. Select the top row of the first empty column and import the second sheet, selecting "replace data at selected cell."
3. Delete duplicate columns.
4. Select the data.
5. Insert -> Chart.
That was it!
One issue I noticed before deleting the duplicate [flights] columns was that they were very close, but not exactly the same. Using the MINUS function I could quickly see that the biggest difference was 9 points, but most differences were much smaller with an average of 1.8. I don't know why that happened. Maybe I should try different combinations of queries to find a pattern.
Nice! (And great to see you on the blog!)
Delete