A/B testing is a form of controlled experiment in which test subjects are randomly split into two groups, A and B, and each group is shown a slightly different version of the same thing. This is most often used for market research, as it allows researchers to discover which of two options are received more favorably by consumers. For example, a website might employ A/B testing by randomly showing 50% of visitors a version with a different font. By checking their site traffic analytics afterward, the site operators can see which version of the site received the most user engagement, which might tell them that the alternate font is a better choice.
Linear A is an as-of-yet undeciphered writing system of the ancient Minoan civilization (a civilization based on the island of Crete). It appears similar to the deciphered Linear B writing system, but if the pronunciation rules of Linear B are applied to Linear A, it produces a language unrelated to any known language.
Linear B, on the other hand, has been deciphered. It is a syllabic script that was used for writing Mycenaean Greek, the earliest form of Greek for which we have evidence. It predates the Greek alphabet by several centuries and likely evolved out of the earlier Linear A writing system.
While not completely consistent with the definition of A/B testing presented above, the comic jokingly suggests that the choice of writing system could be decided through A/B testing, with the "A" and "B" literally being Linear A and Linear B. The test subject apparently can read Linear B (which encodes Mycenaean Greek), but not Linear A (which produces what's seemingly gibberish when read through the rules of Linear B). It is also a pun on the common phrase "[it's] Greek to me", which people use to refer to something as gibberish, but here, it is the Greek text which is comprehensible to Cueball, while instead the other one isn't.
|| This transcript is incomplete. Please help editing it! Thanks.
- [Cueball is sitting behind a computer desk, facing to the right, gesturing at the screen. Ponytail stands behind him and Hairy stands in front of him, both taking notes in a pad.]
- Cueball: I like this one more because it encodes Mycenaean Greek. The other one just looks like gibberish.
- [Caption below the panel:]
- Linear A/B testing
Versions of the script engine used in early versions of the Opera web browser were named after ancient writing scripts: Linear A, Linear B, Futhark (the oldest form of the runic alphabets used by Germanic tribes), and Carakan (Javanese script known as Aksara Jawa, a modern variant of Aksara Kawi).
add a comment! ⋅ add a topic (use sparingly)! ⋅ refresh comments!
Came straight to this site after trying to read today's comic18.104.22.168 16:33, 17 May 2019 (UTC)
- Yeah, no kidding! Usually I get the majority of the comic and I come here for the finer points and title text (I browse these sites on my iPad, no mouse with which to see the "MouseOver text"). But this one, *whiff!*, right over my head! LOL! NiceGuy1 (talk) 04:09, 18 May 2019 (UTC)
- @NiceGuy: You can touch and hold on the image to see the mouseOver text. --22.214.171.124 06:47, 18 May 2019 (UTC)
- Years now of reading XKCD (including a couple of years before I joined this site), over two different iPads, and I never knew that! Thanks! I only know holding pictures for if I want to save them, and if I want to do that (which I often do to share on Facebook, if I share the link Facebook insists on picking the wrong picture for the preview), I open the Full Size link listed on the comic first (which obviously doesn't have the text). NiceGuy1 (talk) 04:21, 24 May 2019 (UTC)
Barmar (talk) 16:51, 17 May 2019 (UTC)
I'm not seeing any relevance to "It's Greek to me." I think that's just a coincidence HisHighestMinion (talk) 17:58, 18 May 2019 (UTC)
Is this actually an example of A/B testing? It seems like Cueball is in both the A and B groups (ie. he got to see both Linear A and Linear B, and compared the two). That doesn't seem to fit the definition of A/B testing which is about showing each of two random groups a different thing. Hawthorn (talk) 19:55, 18 May 2019 (UTC)
"Linear A, on the other hand, . . . ." doesn't make much sense as the first item of a list. I would expect "Linear A, blah, blah blah . . . . Linear B, on the other hand, blah, blah, blah . . . ." 126.96.36.199 23:34, 18 May 2019 (UTC)
- I edited the explanation to try to address these last two comments from other readers. I also reworded/reordered the title text explanation to highlight the "Java script" pun a bit, as well as explain how performance testing determined that Linear A script rendered faster than Aksara Kawi script. Ianrbibtitlht (talk) 00:11, 19 May 2019 (UTC)
I wonder if it's worth mentioning that per the title text the site is written in Linear A despite the fact that Cueball described it as gibberish in the main comic panel. This may be a reference to how programming languages usually appear as gibberish to normal users (i.e. non-programmers). Ianrbibtitlht (talk) 00:15, 19 May 2019 (UTC)
https://web.archive.org/web/20190516074849/https:/phys.org/news/2019-05-bristol-academic-voynich-code-century-old.html [Comet] 20:15, 23 May 2019 (UTC)
I don’t see any meaningful connection to the Voynich manuscript. It is not true that they used “a stochastic methodology,” they [] to have realized it was abbreviations. Furthermore the claim that they decided it at all is widely dismissed as untrue. Finally, I don’t see any obvious reason to believe that these Linear A / Linear B testers are using a stochastic technique. I would recommend removing the tag for further elaboration on this connection unless the claim of similarity can be substantiated. Stellaathena (talk) 13:51, 15 June 2019 (UTC)
Can someone please make a re-mix of xkcd #435 with this #2151 embedded inside it, two spaces over on the left (like math is one space over on the right of the scale) with the caption "Linguistics" under it? 188.8.131.52 18:48, 19 May 2019 (UTC)