Search

Feedback

Back of the envelope percentages? (RE: FAQ Q: Are your statistics reliable?)

Charles Mote Jr

Member
Posts: 17

Charles Mote Jr @ 2020-10-09 15:40:49 UTC

With regard to:


"Q: Are your statistics reliable?


A: No. As said before, given the enormous work ahead of us, our site is far from comprehensive and doesn't claim to be. Also, the entries are heavily biased by the preferences of the editors and our visitors. Conclusion: Never ever claim Summertime] is the most covered song with more than 750 covers according to SecondHandSongs."


Has anyone at SHS ever done rough numbers to see what percentage of "modern" "hits" (20th century popular music) are possibly represented so far (clearly not all originals that have charted / sold have covers)?


Since no one else is even trying to keep track so thoroughly, it's fair to say, for what's known, SHS certainly is reliable. Complete? Obviously, no. But, the depth of care for citation and "shepardizing" is beyond anything any other repository of knowledge is even attempting, let alone achieving.


Just 2 cents from the peanut gallery.

VirileVagabond

Certified Contributor II
Posts: 4885

VirileVagabond @ 2020-10-10 02:39:39 UTC

Sounds like a future editor to me! Smile


I've always heard that "Yesterday" was the most covered song, perhaps from the 20th Century. (I actually won a trivia contest with that answer.) Personally, I doubt no one can really say definitively as there are so many variables. The biases and preferences of the archivists (as mentioned), loss of historical records is another, and there's the definition of what counts as a cover. SHS only treats some performances as a separate cover and disqualifies others. Other sites may define differently.


I'd say that SHS statistics are evidence but not proof.


Bastien I'd say the statement below could use some cleaning up:

No. As said before, given the enormous work ahead of us, our site is far from comprehensive and doesn't claim to be. Also, the entries are heavily biased by the preferences of the editors and our visitors. Conclusion: Never ever claim Summertime] is the most covered song with more than 750 covers according to SecondHandSongs.


a) I'd strike "As said before, given the enormous work ahead of us". This part assumes that a visitor read the whole page from top to bottom when many will just read the portions of interest.

b) "Summertime]" has an orphaned close [bracket] that needs deletion.

c) Visitor preferences have no effect on the most material statistics. I'd assume only "most viewed", etc. I'd either change "visitor" to "contributor" or change to "preferences of the editors, contributors and our visitors".

Charles Mote Jr

Member
Posts: 17

Charles Mote Jr @ 2020-10-10 16:56:35 UTC

Thanks!


Not having access to the back end data, can't even begin to devise work-flow to try to calculate. I don't think I've missed that there isn't any chart info here. Certainly could run "title by year" and see the total. Billboard / other "charting" services give a sense of how many "hits" total per year. I'd think even a "guestimate" would allow increasing SHS' claim to authenticity (" Covers of 50% of all modern, charted, music", etc.)


Second Hand Songs has (in less than 48 hours) already been a tremendous help in my own search and collection, certainly contributing to improvement in my quality of life.


The amount of care and dedication here certainly qualifies SHS as a quality archive. I don't know why it took so long for me to find, but gladly finally did.


Thanks again!

sebcat

Managing Editor
Posts: 5480

sebcat @ 2020-10-10 17:28:23 UTC

Second Hand Songs has (in less than 48 hours) already been a tremendous help in my own search and collection, certainly contributing to improvement in my quality of life.

The amount of care and dedication here certainly qualifies SHS as a quality archive. I don't know why it took so long for me to find, but gladly finally did.

Great to hear. Hope you stick around some too.

Bastien

Manager
Posts: 29087

Bastien @ 2020-10-20 08:53:57 UTC

Has anyone at SHS ever done rough numbers to see what percentage of "modern" "hits" (20th century popular music) are possibly represented so far (clearly not all originals that have charted / sold have covers)?

No, we've never done that, but I see the relevance. Any methodology in mind Charles Mote Jr ?

Since no one else is even trying to keep track so thoroughly, it's fair to say, for what's known, SHS certainly is reliable. Complete? Obviously, no. But, the depth of care for citation and "shepardizing" is beyond anything any other repository of knowledge is even attempting, let alone achieving.

Thanks Smile

Bastien

Manager
Posts: 29087

Bastien @ 2020-10-20 09:03:02 UTC

I'd say the statement below could use some cleaning up

Updated to:

A: They are evidence but not proof. Despite over 900.000 cover songs indexed, our site is far from comprehensive and doesn't claim to be. Also, the entries are biased by the preferences of our editors and contributing visitors.

Conclusion: With almost 3.000 cover versions listed, Stille Nacht! Heilige Nacht! , is not 100% certain the most covered song, but it likely is.

Bastien

Manager
Posts: 29087

Bastien @ 2020-10-20 09:09:37 UTC

I don't think I've missed that there isn't any chart info here. Certainly could run "title by year" and see the total. Billboard / other "charting" services give a sense of how many "hits" total per year. I'd think even a "guestimate" would allow increasing SHS' claim to authenticity (" Covers of 50% of all modern, charted, music", etc.)

No, there are no chart data. We'd have to cross-reference with https://www.billboard.com/ for examples. It's feasible, they have entries per song, e.g.:

https://www.billboard.com/music/barenaked-ladies/chart-history/HSI/song/56960

We also have a ticket in the pipeline to include chart placings for performances: https://secondhandsongs.com/trac/ticket/2973

Second Hand Songs has (in less than 48 hours) already been a tremendous help in my own search and collection, certainly contributing to improvement in my quality of life.

The amount of care and dedication here certainly qualifies SHS as a quality archive. I don't know why it took so long for me to find, but gladly finally did.

Thank you for the wonderful words Charles Mote Jr ! Do you mind if I add this our Testimonials? Smile

sebcat

Managing Editor
Posts: 5480

sebcat @ 2020-10-20 10:59:13 UTC

No, there are no chart data. We'd have to cross-reference with https://www.billboard.com/ for examples. It's feasible, they have entries per song, e.g.:

https://www.billboard.com/music/barenaked-ladies/chart-history/HSI/song/56960

We also have a ticket in the pipeline to include chart placings for performances: https://secondhandsongs.com/trac/ticket/2973

We do have a hit song tag that we could use for some stats in the meantime, e.g. songs that have had the most hits as covers, artists who have had the most hit covers.

VirileVagabond

Certified Contributor II
Posts: 4885

VirileVagabond @ 2020-10-20 14:56:26 UTC

We do have a hit song tag that we could use for some stats in the meantime, e.g. songs that have had the most hits as covers, artists who have had the most hit covers.


I doubt we have objective constraints on the "hit song" tag to make any resulting stats meaningful or reliable, and that's assuming we have enough performances actually and appropriately tagged as such.

sebcat

Managing Editor
Posts: 5480

sebcat @ 2020-10-20 19:29:53 UTC

I doubt we have objective constraints on the "hit song" tag to make any resulting stats meaningful or reliable, and that's assuming we have enough performances actually and appropriately tagged as such.

A Top 20 hit. Only 1767 performances tagged so far. But the more we use it the more relevant it will become. And the chart feature will mean this eventually gets automated.

VirileVagabond

Certified Contributor II
Posts: 4885

VirileVagabond @ 2020-10-20 20:00:16 UTC

A Top 20 hit.


...in any chart worldwide. The relevancy of the chart in question is up to the editor's discretion.


That's not much of a constraint. A top 20 performance on a major US chart deserves an objective tag; however, I don't recall ever seeing an (e.g.) Euro-wide chart, only country specific. Does a performance that went top 20 only in Finland, New Zealand or Japan objectively deserve a tag? Would those tags be of interest to an average user sort for hits?

sebcat

Managing Editor
Posts: 5480

sebcat @ 2020-10-20 22:25:43 UTC

That's not much of a constraint. A top 20 performance on a major US chart deserves an objective tag; however, I don't recall ever seeing an (e.g.) Euro-wide chart, only country specific. Does a performance that went top 20 only in Finland, New Zealand or Japan objectively deserve a tag? Would those tags be of interest to an average user sort for hits?

VirileVagabond if you disagree with the way that a tag is defined, please start a new thread. Personally I’m okay with the current definition - it’s simple and clear. I normally add which countries the song charted and the chart placing in the comments in any case. You are welcome to propose an alternative definition for the tag of course and see if others agree with your suggestion.

VirileVagabond

Certified Contributor II
Posts: 4885

VirileVagabond @ 2020-10-20 23:09:15 UTC

That's not much of a constraint. A top 20 performance on a major US chart deserves an objective tag; however, I don't recall ever seeing an (e.g.) Euro-wide chart, only country specific. Does a performance that went top 20 only in Finland, New Zealand or Japan objectively deserve a tag? Would those tags be of interest to an average user sort for hits?

VirileVagabond if you disagree with the way that a tag is defined, please start a new thread. Personally I’m okay with the current definition - it’s simple and clear. I normally add which countries the song charted and the chart placing in the comments in any case. You are welcome to propose an alternative definition for the tag of course and see if others agree with your suggestion.


sebcat I have repeatedly stated in dedicated and related threads that tags need to be few and concise. Since I was unsuccessful in those efforts, when the risks/faults of how tags have evolved come to fruition (as predicted) I simply point this out. As it will take repeated failures for anything to be done to remedy and any dedicated thread will be ignored, my documenting my position here should suffice for now.

Charles Mote Jr

Member
Posts: 17

Charles Mote Jr @ 2020-10-21 01:20:57 UTC

Don't mind one bit. Please feel free to quote publicly as useful!



I don't think I've missed that there isn't any chart info here. Certainly could run "title by year" and see the total. Billboard / other "charting" services give a sense of how many "hits" total per year. I'd think even a "guestimate" would allow increasing SHS' claim to authenticity (" Covers of 50% of all modern, charted, music", etc.)

No, there are no chart data. We'd have to cross-reference with https://www.billboard.com/ for examples. It's feasible, they have entries per song, e.g.:

https://www.billboard.com/music/barenaked-ladies/chart-history/HSI/song/56960

We also have a ticket in the pipeline to include chart placings for performances: https://secondhandsongs.com/trac/ticket/2973

Second Hand Songs has (in less than 48 hours) already been a tremendous help in my own search and collection, certainly contributing to improvement in my quality of life.

The amount of care and dedication here certainly qualifies SHS as a quality archive. I don't know why it took so long for me to find, but gladly finally did.

Thank you for the wonderful words Charles Mote Jr ! Do you mind if I add this our Testimonials? Smile

Charles Mote Jr

Member
Posts: 17

Charles Mote Jr @ 2020-10-21 01:33:42 UTC

I unfortunately don't have any methodology in mid.


However, Virile Vagabound, in response to my post


https://secondhandsongs.com/topic/76207?offset=0#m203490


directed me to


https://secondhandsongs.com/statistics/stats_most_covered_work_per_original_year


I know it's not official, but, I think there's a reasonable argument that if a work is covered extensively, it would fit at least the loose definition of "popular" (and the more recent the release and higher number of covers might correlate to officially "charting".)


While I haven't been able to find where it might be available publicly the data for "most covered work per original year" can't be that hard to get at.


That's essentially what I was thinking about for a rough, rough statement about how much of "popular" music is covered by SHS. A quick check of Top 100 charted by year against most covered work per original year might bear out the general idea that being highly covered does correlate with charting (if, in fact, it does, I obviously don't actually know.)


Before SHS, like most, my go-to on info on a particular song was Wikipedia and SHS' attention to detail and primary sources easily beats what is normally found there. Positing a back-of-envelope calculation was just me spit-balling a way for SHS to make a legitimate claim about the depth of the site and make it a bit more obvious on the landing page.


Has anyone at SHS ever done rough numbers to see what percentage of "modern" "hits" (20th century popular music) are possibly represented so far (clearly not all originals that have charted / sold have covers)?

No, we've never done that, but I see the relevance. Any methodology in mind Charles Mote Jr ?

Since no one else is even trying to keep track so thoroughly, it's fair to say, for what's known, SHS certainly is reliable. Complete? Obviously, no. But, the depth of care for citation and "shepardizing" is beyond anything any other repository of knowledge is even attempting, let alone achieving.

Thanks Smile