I propose deleting the "Timeline of the best-selling albums" section. The unfortunate reality is that we do not know what were the all-time best-selling albums in the pre-Thriller era. Record sales tracking for much of the 20th century was very crude at best. Determining how much an album sold was difficult enough in the US, let alone globally. The sources in the section are not based on certifications and contain very dubious or contradictory assertions. The Sheila Weller article cited for Carole King's Tapestry says that by 1976, "global sales of her eight solo albums totaled 20 million", even though as of 2024, her certified sales are less than 19 million, despite some very recent sales audits in the US, King's largest market by far. Many of the other sources are also media articles (or even less reputable sources like Barnes & Noble), which, as we know, rely on the numbers provided by the artist's representatives or record company, which are often inflated for promotional purposes. For instance, the claim that The Sound of Music soundtrack sold 15 million copies is based on an Allmusic article which says clearly that the record company (RCA) was their source.

Relying primarily on non-certified numbers opens a Pandora's box. Many of you are probably familiar with the claims that Iron Butterfly's 1968 album In-A-Gadda-Da-Vida sold 8 million copies in its first year and 30 million copies overall. While these claims are not supported by certification data, they have been repeated by journalists and made their way in various articles and books, so why not then include the album in this article, both in the main section and in the timeline section?

I'm a music stats geek as much as the next person, but I think sometimes it's better to just admit that we don't don't have good data and therefore cannot make certain statistical rankings. --NicolasJz (talk) 17:06, 1 April 2024 (UTC)Reply "Certification data" is a minefield. As noted above, albums can be OVER-certified. But, many albums are under-certified. Many record labels simply could not care less about paying an organisation to "verify" their sales, then give them Gold, Platinum, or whatever records. Secondly, many records WERE certified years or decades ago, but no attempt was made, or will ever be made, to update those certifications. And, of enormous importance, many record labels deliberately under-certify releases. Because if they properly certified them, they would have to pay more money to the musicians, producers etc. That extra money goes into the accounts of the executives. Of course, there is also the more innocent fact that music sales were recorded on countless thousands pieces, slips etc. of paper, a very large percentage of which simply no longer exist. Keeping rooms full of sales receipts from the 60s was not something that any record company had any intention of doing. In short, "Certification Data" is fun, and you can make tables and lists of "Album with highest Certification", "Artist/Band with highest combined Certifications" etc. But it is laughable to use those Certifications as definitive information. Especially as entities like RIAA, SNEP, and Music Canada don't even keep complete records of their own Certifications that they have verified and awarded. — Preceding unsigned comment added by 197.83.246.171 (talk) 07:51, 4 April 2024 (UTC)Reply

I certainly don't disagree that certification data are very imperfect and I am not advocating for replacing the current list with a list based on what album had the highest certifications at any point in time. But relying on media articles that mostly just parrot whatever record companies and publicists claim is even more problematic. At least certifications can assist in corroborating some of these claims. When it comes to pre-Thriller "best-selling album of all time" claims, we just don't have enough reliable information to back up these claims. I wish there were a good alternative to certifications, but media articles are not it. As I said previously, multiple articles have relayed the dubious claims about Iron Butterfly's In-A-Gadda-Da-Vida, so one could claim that this album should be included on this page as well. NicolasJz (talk) 13:38, 4 April 2024 (UTC)Reply Even Thriller etc. are problematic. Go to the RIAA website, search for Michael Jackson Thriller under Gold&Platinum. Notice the way the Certifications rise, how many million at a time, how long between re-certifications etc. Anyone can tell how unlikely and incredible(as in not credible) that is. Other noteworthy unlikely recertifications include, but are not limited to, Led Zeppelin IV, and Their Greatest Hits by The Eagles. Look at the timeliness between release, and how far along each album's timeline each Big Certification occurs. It becomes obvious that there's "inflated certifications". So, should we just accept these enormous numbers just because the RIAA says so? How is that more "reliable" than something in a respected publication? Isn't it Wikipedia's policy to exclude sales claims that are based only on media articles and not also on certifications? The second sentence of the article, for instance, says albums must have at least 10M certified units. I'm just curious about what the policy is exactly. When I look at various Wikipedia lists of best-selling albums and artists, certifications are the basis of the lists. The main section of this article is based on certifications and specifically excludes albums whose claimed sales are completely at odds with their certification numbers, such as Iron Butterfly's In-A-Gadda-Da-Vida (30M claimed), the Who's Tommy (20M claimed), Blondie's Parallel Lines (20M claimed), Barbra Streisand's A Star Is Born (15M claimed) and Frampton Comes Alive! (16-20M claimed). I am trying to understand why this one subsection uses a different methodology, the result being that In-A-Gadda-Da-Vida (with 8M claimed sales by 1969) should be included in here but is excluded from the main section (despite 30M claimed sales in total). Likewise, according to media articles, Frampton Comes Alive! was the album that overtook Tapestry as the best-selling of all time, not Saturday Night Fever (https://www.upi.com/Archives/2001/01/04/Peter-Frampton-still-very-much-alive/8977978584400/). If album sales claims do not need to be backed up by certification numbers, shouldn't we include on the list all of these albums I mentioned? Shouldn't we get rid of the "at least 10 million certified units" rule at the beginning of the article and change the list accordingly?

I don't want to make a big deal out of this because obviously it's not the most important thing on Wikipedia. For my own knowledge and for future editing, I would just like to understand what is the favoured methodology to use and to see it applied consistently. Best --NicolasJz (talk) 19:18, 4 April 2024 (UTC)Reply We have information from the CBS Manufacturing plant responsible for shipping all copies of Thriller, they stated within the first 2 years 40 million copies of Thriller were shipped, at the end of the 1980s we have a physical plaque showing that it's sales stood 47-48 million and by 2001 a total of 68 million units were shipped worldwide at the time of the Plant's closure when the Sony Music bought them out.Historic album was 'Thriller' at Ga. plant.

Jackson himself stated it sold 60 million by 2003. This would mean that it shipped over 70 million at the time, but the album had only sold 60 million, since then it's sold at least 15 million more in the US. The RIAA counted shipped units to retail for certifications, not copies sold specifically and there was a period from 1985-1991 where Thriller wasn't certified at all, which would be significantly higher if it was Audited like the Eagles had done with their album. So considering this logic, The 100 million figure for Thriller is how much was shipped to retailers worldwide, the actual number of sales would be between 75-80+ million.

The claimed album sales for Thriller and other albums like Dark Side of the Moon is no less inflated than the record sales of the Beatles Never17 (talk) 23:00, 21 April 2024 (UTC)Reply

Unfortunately there isn't a perfect method. It is true that people inflate numbers to make albums appear far more successful than they really are. But it is also true that many albums are enormously under-certified. Many big-selling pre-1976 albums were only ever certified "Gold", and were never re-certified when multi-Platinum Certification was introduced. Does this mean that none of them reached US sales of 1 million units? Then, as stated, there is the distinct possibility that some albums have been over-certified. The Eagles themselves have stated that the RIAA Certification for "Their Greatest Hits" is too high. If Wikipedia wants a "highest certified albums of all time" article, good. But calling it "List of Best-Selling Albums" is inaccurate and misleading. "Most Certifications" and "Biggest Selling" are NOT the same thing. — Preceding unsigned comment added by 197.83.246.171 (talk) 07:53, 5 April 2024 (UTC)Reply

Just go to the RIAA's website, click on "Gold & Platinum". then enter "Thriller" into the search box. Here's what we get..

• Thriller was released on December 30, 1982.

• It was first certified Gold and Platinum on January 31, 1983. (Which actually violated the RIAA's own Certifcation Laws of the time. As no record could apply for certification until at least 60 days after date of release. So already we have a MAJOR red flag. One month in.)

• Multi-Platinum certification was introduced in 1984. And on October 30, 1984, Thriller is certified 20xPlatinum. Not 18 or 19 or 21 or 22. Exactly 20 million. Isn't that remarkable? They should have waited a day and certified it on Halloween. But that may have been too obvious.

• Now here's one of the Big Clues. Having been certified 20-times Platinum in October 1984, Thriller finally gets certified 21xPlatinum on..May 24 1990. So it sold over 20 million copies in less than 2 years, but then needed over five-and-a-half years to sell the 21st million? And, we're just getting started..

• It's 22xPlatinum on August 25 1993. So, that must mean that it sold about twice as fast in a few years of the 90s as it had for most of the 80s.

• 24xPlatinum on September 29, 1994. The gaps are getting smaller now. Which anyone can tell you is the exact opposite of what really happens.

• 25xPlatinum on July 7, 1997. 7/7/97.

• 26xPlatinum on October 30 2000. So, the FIRST multi-Platinum was on October 30. And here in 2000, it's getting certified again on...October 30.

• 27xPlatinum on April 25 2005.

• 28xPlatinum on March 6 2009. 03/06/09. By the way, surely anyone who wanted Thriller would have bought it by now. or, just listened to it free online.

• 29xPlatinum on August 21 2009. Yup, 5 months. You will say that Michael Jackson died in June 2009. But anyone who was a Michael Jackson fan already owned Thriller. Were they so upset by his death that they raced out and bought another copy of an album they already owned?

• 30xPlatinum on December 16 2015. A big gap among the quick-fire certifications. Has streaming taken over?

• 31xPlatinum on February 1 2016 AND...

• 32xPlatinum on February 2 2016. So, having taken over 6 years to be certified ONE more million, it now does TWO million in under 7 weeks? I think not.

• 33xPlatinum on February 16 2017.

• 34xPlatinum on August 20, 21.

And, as of April 22 2024, no further updated certifications. Maybe they're waiting for October 30 again. It's obvious that this is not real. It's promotion. It's hype. It doesn't represent actual albums sales. It represents the Cult of Michael Jackson, not how many albums were manufactured or sold. And, that opens the door to "certifications" in a much broader sense. — Preceding unsigned comment added by 197.87.135.139 (talk) 08:16, 22 April 2024 (UTC)Reply Wrong Establishment Dates edit