2930: Google Solar Cycle

Explain xkcd: It's 'cause you're dumb.
Jump to: navigation, search
Google Solar Cycle
From Google Trends, it looks like the lag between people Googling cocktail recipes and 'hangover cure' is 14 hours.
Title text: From Google Trends, it looks like the lag between people Googling cocktail recipes and 'hangover cure' is 14 hours.


Solar flares appear on the Sun's surface in different numbers at different times. A solar cycle is the amount of time that elapses between times of maximum (or minimum) solar flare activity. This period is approximately 11 years.

Solar flares can generate interest, and 'news' stories at times of high activity, because they can sometimes cause aurorae, as well as power outages and other similar issues. Hence people will be more aware of them, and search for "solar flares" on Google to learn more about them. This trend somewhat matches the solar cycle since people will be more interested/concerned about solar flares during the times they are abundant and search for them more often. Randall notes that Google has existed for long enough to see the trend in searches for "solar flare" over a full solar cycle.

The title text comments that people usually Google "hangover cure" 14 hours after they search for cocktail recipes. This suggests that people decide they want cocktails, look for ways to make cocktails, make (and presumably drink) the cocktails, wake up with a hangover and look for ways to get rid of the hangover. Google trends does indeed suggest that there is roughly a 14-hour difference in peaks between these searches. However this does not mean (as the title text implies) that the people searching for cocktail recipes are the same people that are searching for hangover cures later. It may represent the (not unreasonable) assumption that people who are exploring the idea of cocktails (for themselves or others) generally start to do so from the early evening onward, whilst those who find themselves freshly under the weather (not just from cocktail consumption) are likely to be finally provoked to look up a solution from around mid-morning.

One can illustrate the pitfalls of assuming such causation by substituting "chicken nuggets" for "cocktail recipe" which shows a very similar relationship. Chicken nuggets are not known to cause hangovers.[citation needed]

Looked at another way, there is a 10-hour 'lag' from searching for "hangover cure" to searching for "cocktail recipe"; this does not imply that having a hangover is causing people to be interested in drinking cocktails! (Though they may lead to chicken nuggets.)

Another very important problem with looking at scales of less than a day is that the Earth has 38 time zones, and people drink cocktails and wake up all the time around the world. Since everyone is searching the same Google, it is impossible to deduce anything on a daily basis by looking at global data. It is possible to see only searches from a particular country, but even then, in the case of the US there are several hours difference from east to west that may smear out any such direct observations. This also applies to other countries/regions, being potentially at its most extreme in Russia. This might be moot if the data is preprocessed to 'local' time, although the opposite issue might arise in a case such as Chinese data (though any examination of Google Trends for China may not be fruitful).

The sunspot cycle was 'explained' in 2725: Sunspot Cycle.


Ambox notice.png This transcript is incomplete. Please help editing it! Thanks.
[The label of the graph in the top left corner of the image:]
Google Trends search traffic for "Solar Flare"
[A graph starting in 2005, peaking in 2013. It rises until 2024 when the graph cuts off. There is an arrow in between the two peaks labeled "11 years".]
[Caption below the panel:]
I like that Google has existed almost long enough for us to observe the solar cycle using Google Trends.

comment.png add a comment! ⋅ comment.png add a topic (use sparingly)! ⋅ Icons-mini-action refresh blue.gif refresh comments!


When I opened the page I found a bunch of vague misspelled content (maybe spammers or bots? I don't know) so I updated the explanation. Not very knowledgeable about most things, though, so it probably still is not good enough. DNA diva (talk) 01:10, 9 May 2024 (UTC)

Looking at the first couple of user-edits ("first" Explanation and subsequent Transcript edit), I think they were 'rushed' in order to establish the whole "first"ness and/or escape possible Edit Conflicts (probably needlessly, looking at timings). Not precise enough to be 'AI bot' text (accuracy being different, of course), and no obvious spam-intent.
Could have done with more care to be taken, so as not needing to be entirely revamped by the likes of yourself, somewhat diluting the 'kudos' and making the little effort taken almost entirely pointless. But I like eagerness. Next time, perhaps, a bit of care will go into the mix and we'll be on the way to another useful contributor. 04:58, 9 May 2024 (UTC)

This comic got me to search about Danger Kitty (talk) solar flares too.

Contrary to the current explanation, it appears you can search for cocktail recipe searches vs hangover cure searches at a sub-daily resolution. 07:51, 9 May 2024 (UTC)

As of my check, just now, I find it very funny that the (current) state-map attached to that result seems to indicate that every state trend predominantly is for searching for cocktails, between the two options. Except for Wyoming, which is solidly the one and only "hangover" state. 10:35, 9 May 2024 (UTC)~
It's likely because there's only like 7 people living there. Looking worldwide, Nigeria is the only "hangover country". Tcf (talk) 13:51, 9 May 2024 (UTC)
And Russians drink but don't have hangovers. 16:04, 9 May 2024 (UTC)
They do. But they probably don't need to search for cure as they already know. -- Hkmaly (talk) 19:30, 9 May 2024 (UTC)

Google solar cycle? Ok. 13:46, 9 May 2024 (UTC)

If cocktail and hangover cure searches are both purely diurnal, it would be difficult to tell which comes first. But there's actually a larger cycle of the daily peak volumes. When you compare these two graphs, you can see that the hangover line follows the cocktail line. This strongly suggests a cause and effect. Barmar (talk) 14:13, 9 May 2024 (UTC)

Is it a coincidence that a major solar flare occurred on the date of this comic, for which NOAA issued a geomagnetic storm warning for Friday through Sunday? Ianrbibtitlht (talk) 19:57, 10 May 2024 (UTC)

Spike of March 8, 2012[edit]

As far as the spike about 12 years ago, there was this. 00:54, 9 May 2024 (UTC)

Speaking of spikes, what if the internet and Google Trends had existed in 1859? 00:59, 9 May 2024 (UTC)
...we'd still be repairing the infrastructure damage in 1860? ;) 04:58, 9 May 2024 (UTC)
...our entire society as we know it would've collapsed long ago. Psychoticpotato (talk) 12:36, 9 May 2024 (UTC)
Maybe it did and we actually had the internet in 1859 but the solar storm was so bad that it wiped out any traces of it as well as people's memories of it. 19:21, 9 May 2024 (UTC)
Dementia Storm? Woah... Psychoticpotato (talk) 20:14, 9 May 2024 (UTC)

Correct URL?[edit]

Is this the correct URL for the query? It's a lot more jagged than I was expecting. https://trends.google.com/trends/explore?date=all&q=%2Fm%2F0f81b&hl=en Thanks~

I second this comment, and wish to ask what exactly Randall's settings were on, because "worldwide" and "2004-present" don't match the comic. 13:45, 9 May 2024 (UTC)
Neither do "United States" and "2004-present". For future record, the worldwide graph has a peak of 100 in 2011 and other peaks in the low dozens in 2014, 2017, 2021, and a mere 25 in 2024. Tcf (talk) 13:51, 9 May 2024 (UTC)
The actual peak is in March 2012, not 2011. If Randall's "y axis" is set to 2005, then the peak is correctly shown, and the "11 year gap" appears to be 2013-2024. But the Y-axis doesn't appear to be to scale, at least not a linear one. But hey, it's a freehand-drawn comic, so I'm okay with a little imprecision. 16:04, 9 May 2024 (UTC)