Save Page:

The World's Music Charts

Songs	Song Years	Song Artists	Song Titles
Albums	Album Years	Album Artists

These lists bring together 229,151 individual chart entries to list 109,685 of the most popular songs and albums of all time. It combines 83 song and album charts from Australia, Austria, Belgium, Canada, Eire, France, Germany, Holland, Italy, Japan, New Zealand, Norway, Poland, South Africa, Sweden, Switzerland, the UK and the US as well as some global and internet based charts to give a list of the songs and albums that have topped the charts around the world.

It also provides an answer to the question of who were the world's greatest song and album chart acts of all time.

What was the greatest chart song of all time?

I have a personal opinion on that question, as everyone does. The data in these tables contributes to that debate but cannot decide the matter. The positions in any combined chart reflect both the input data and the scoring method. As is explained below a reasonable level of effort has been taken to ensure the scoring algorithm generates a "median" result. However the results cannot help but reflect the input data.

The chart above shows the number of entries in the 57 source song charts. The preponderance of entries after 1960 means that any acts active after that time would be over represented in any combined results. Any song that had its first chart entry after 1990 will not have had enough time to accumulate a sufficient score to top the charts.

As the profile above shows the number of entries has a slight drop-off at the year 2000. The results, that is the year ranking, are therefore also not reliable after that date (but since there are lots of other sites with charts covering that period I don't care).

Who was the greatest chart act of all time?

Unlike the question about chart songs this question can be answered for songs and albums. By comparing acts within the same year the effect of the differing number of entries can be cancelled out. The answer for songs is, marginally, Elvis Presley (and Bing Crosby comes 10th), for albums the answer is clearly The Beatles. The overall greatest chart act of all times is The Beatles.

How the site is generated

Each of the 83 source charts is held separately, the charts must always have entries for "artist" and "title" usually have entries for "position" and "date" and may also contain extra information such as "duration", "written by", "web page" or "film".

The 229,151 entries in these charts are consolidated, to provide a complete set of attributes for each of the 109,685 items (78,711 songs and 30,974 albums). The most difficult aspect of this task is matching names, they are often misspelt in the source charts, punctuation is usually inconsistent and the list of "featured" artists is always in a different order. Programming a system to recognise that "Uncle Albert" by "Paul McCartney" and "Admiral Halsey" by "Wings" are actually the same song is not trivial.

In general the approach that has been taken is to consolidate entries if they appear similar, having too many false connections is usually better than missing out on them. For this reason there are quite a few places where strict accuracy has been sacrificed to bring things together, for example all of Prince's songs are listed under the name Prince.

The next step is to generate consolidated a score from the chart entry information. There are a variety of possible ways this could have been done, in this case it was decided that the simplest approach was to generate scores from each entry and sum them. The individual entry scores clearly should depend on the position within the source chart, a number one song getting more "points" than the second placed one ad so on.

One reasonable way to generate a score is to use a power law. The score is set to XXX+YYY^position and then each chart is weighed by a factor that takes into account which chart was the source. The weighting can emphasise charts that are solidly based on sales, attempt to match music revenue in each country or follow any number of other reasonable strategies. Using a set of apparently reasonable variations of the weighting, the XXX and YYY values generates the following top 10 lists.

As the wide range of different results shows the parameter values have a big influence on the resulting chart. Using this approach suitable parameters have to be picked to generates a "medium" chart.

Some reviewers of music charts have claimed that Zipf's distribution is a better fit to music charts than a power law, that is the second placed song has half the sales, the third has a third etc. This suggests a different, simpler scoring algorithm, if each entry's score is 1+1/position then each song gets credit for having an entry and the number one song gets most points. The simplest chart weighting is to give equal values to every chart. When these simple parameters are used the result is towards the middle of the range demonstrated above. This is the scoring algorithm that has been used here.

Working out which year to assign an entry to is also surprisingly hard. The year of each song is deduced directly from the chart entries, rather than relying on some kind of unreliable external source. The year is extracted from the date in all the song's chart entries and the song's year is set to the median of these values. This usually generates a reasonable estimate of the year.

Once the individual song scores have been calculated they are processed to generate the various web pages and the links between them.

Frequent Questions

There is a mistake in one of the entries

The complete set of charts contains 200,000 entries. There are bound to be places where names have been misspelt, artists have been wrongly combined or entries missed out. If you think you have seen such a case contact us.

You should add the chart at http://***.com/

There are many charts out there that could potentially be added. In order for me to invest the time adding a chart it must meet the following criteria:

• The chart must start before about 1985, there is enough coverage already for the years 1985-2005 from charts that begin earlier.
• The chart should either come from a region not already covered, or provide more detail than the existing coverage for a country.
• Any chart that requires payment to access will be ignored. Not are there no funds available (since we make no profit) but also as the vendor of the chart clearly sees value in the data as presented and the exchange of value would indicate a contract we would have no desire to infringe their rights. Under these circumstances the partial summary in these lists should probably be viewed as 'fair use' but we don't want to be put into a position where we can be forced to prove it.

If you think you know a site that has such a chart contact us.

I have a chart for XXX, can you incorporate it?

Yes. Send it to us in CSV format with details of where it came from and we can add it. The best charts are those that either cover periods before 1960, or places we have no entries for, charts that don't appear to be objective (e.g. any chart with "Green Day" in more than half the positions 1-20) will be ignored.

The consolidation process requires that each entry has columns for "peak position", "artist" and "title". However we have tools for converting charts that don't. For example a weekly chart of positions needs to be summarised so that each song has a single entry.

Two songs seem to have the same position

When a single is listed as a double 'A' side then, for the purposes of this analysis, the two songs are both independently treated as achieving that position. In this case two different songs can both have entries, for example, stating that they are "Germany 6 of 1978".

The same song has different names

Typically the 83 source charts don't ever use exactly the same name for a song title, album name or artist name. The consolidation process attempts to bring these entries together, but it is hard to find all the changes needed, as a result some titles that to a person are clearly the same (like "Dream Number Nine" and "Dream No. 9") will be listed as distinct songs. Normally this is a mistake and should be fixed in a future version (if you see an example tell us about it).

The artist name is wrong

In order to bring together as many entries as possible the names of the artists have been made more consistent. Clearly "Elvis Costello", "Elvis Costello & the Attractions" and "The Attractions" are three different combinations of musicians, however calling them all "Elvis Costello" makes it easier to see connections. Two song entries wrongly listed with the same artist could be explained by this merging (or it could be a mistake).

If two entries are listed as having different artists which actually should match then tell us about it and we will fix the mistake.

Why don't you have any of the VHS top 100 Charts?

There are a number of TV shows that count down the 100 greatest songs about... dentists, or whatever. In the UK these are shown on Channel 4 while in the US I understand they are on VHS1. After the programs are shown the charts are sometimes available on the internet. I have not included any of these lists, despite the fact that they claim to solicit votes from "the public". The reason is that most of these actually programs post some list of, say 130 candidates (that they have footage for) and the voting only affects the order. In addition, judging by the results, the participants are hardly an unbiased sample.

Where can I get the mp3 files?

This site just lists the charts, it contains no music data at all. If these lists inspire you to expand your collection you can visit iTunes, Amazon or any of the other web sites that will sell you music.

What's with the version numbers?

The sources for the charts are of such varied levels of quality that there are bound to be mistakes in the way the data is assembled. There is a continual effort to fix the most obvious problems in the data, and to add in additional charts as they become available. The version number tracks these changes, so if an issue is identified it can be tied to the version of the data that it was fixed in.

Where can I vote for my favourite?

You can't. There are lots of sites that encourage participation, these seem to suffer from record promoters "carpet bombing" sites with votes to distort results. This site only includes complete charts and provides no way to modify them.

On this site no one gets a vote, not even the administrator.

Surely XXX is too high, can't you fix that?

These lists are produced by a completely automated process that takes the input and generates the web site, once the scoring approach is defined the result cannot be "adjusted".

Just to repeat: On this site no one gets a vote, not even the administrator.

Can I spider the site?

The information here can be used in your own project. If you do that you must prominently note where the original data came from and provide an active link to the site (at the http://tsort.info/ address). If you want to use this data without meeting these requirement you must pay the licence fee (contact us for details). The real reason for placing this restriction is that the data is continually being fixed, new charts are being added and so on.

At the moment you may scrape the site, provided your spidering does not impose too heavy a burden on the server (and remember the bandwidth is quite limited). If the extremely limited resources start being stretched then this position may change.

However you would probably find it easier and more useful to grab the experimental CSV File. If making this file available imposes too heavy a burden on the server it will be removed.

Why do songs move in the charts?

As has already been mentioned, there is a continual effort to fix places where the same song or artist is listed under multiple versions. For example one user recently pointed out that "Riders in the Sky" and "Ghost Riders in the Sky" were actually the same song (for which thanks). The source charts were edited to ensure that the same name was used everywhere, as a result the song went from being at numbers 6 & 7 of 1949 to being the number 1 song of 1949.

You may also notice some artists come and go for the same reason.

Why is the year for XXX listed as 19XX?

The year allocated to a song comes from the source charts. If a chart covers a particular year (for example the CashBox end of year chart) then that obviously provides the year. For weekly charts the normal convention is to list the month during which the song entered the chart, so a song that was in a chart for 16 weeks starting "Dec 1949" will be allocated to 1949 despite the fact that most of its run was during 1950.

The years are extracted from all the charts that the song is in, the median (i.e. middle) year is then assigned to the song. This normally gives a reasonable result but can be fooled by strange cases (like "Stand By Me").

As with the final position there is no room for adjustment. If I don't agree with the assigned year I just have to live with it.

© 2007, Steve Hawtin. Generated 10 May 2008 06:25 GMT. This data may be freely copied provided that the source is acknowledged, a link to the tsort.info site is prominently incorporated and the version number is attached (this data is version 1.3.0069).