PEAK COMPARISON OF ALL TOP 10 RECORDS ACROSS THE 5 MAJOR WEEKLY RECORD CHARTS
MAR 1960 TO FEB 1969
I apologize in advance, this could be information overload, too many numbers for the average Joe, yada yada. And I admit it’s wacko to a degree, but I needed to do this just to see the lay of the land once and for all, kinda, sorta. So here goes…
I’ve done some more statistical analysis on the 60s UK charts, a step beyond my similar #1’s comparison in posts elsewhere, this time for records that made it into the Top 10. The idea came to me one day after observing the online Tiscali spreadsheet of each record’s peaks across various UK (and 1 US) charts, and I thought some comparison stats could be cranked out fairly quickly in a spreadsheet. I was interested in seeing how closely, or far away, the 5 major 60s charts tracked each other in terms of individual record peaks. Were they all relatively close plus or minus, or did some track the averages consistently better or worse than others? I decided to give it a go, but it turned out to be more work than I had bargained for. Nonetheless, the results were most interesting.
I took the Tiscali spreadsheet ‘by record title,’ sorted it by record date, sectioned it off from the start of the Record Retailer chart March 1960 until the start of the BMRB/’official’ chart Feb 1969. Then also broke that up into 3 separate time periods when there were 5, 4 and 3 charts, then calculated the average chart peak for every record that made the Top 10 on any of the 5 major charts NME, Record Mirror, Melody Maker, Disc, and Record Retailer. Then I calculated the difference between each record’s average chart peak and its individual chart peaks.
I had to do some manual data checking, as Tiscali did not have a pure Record Retailer column, but instead a Guinness column with NME, RR, and BMRB data; and a Record Mirror column with RM and RR data. So I checked for specific pre-BMRB Record Retailer chart peaks during the critical chart change out periods for these charts, and also when the RM and Disc charts ceased. Likewise, I screened for NME and MM record peaks at the beginning and end of the entire period. Thus if a record did not peak on a given chart due to timing or chart startups/cutoffs, then that chart was not analyzed for the record peak comparison. Also, EPs and B-sides that only appeared on a minority # of charts (mostly NME) were not considered. A few times, different sides of the same record peaked on different charts, and in those cases I made a manual calculation of the best peak of either side to count as one entry.
Once the data was complete, I also double checked the records with the greatest peak differences, there were about 20 worst outliers that deserved extra scrutiny. I should point out that I did not double check every Top 10 record peak position against every chart to see if Tiscali was 100% correct. Hopefully the bugs are few (I did find a few), and spread around equally, so as not to affect the overall results. Perhaps one day I’ll get around to analyzing positions 11 thru 20. And the 50s period.
OK. So I calculated comparison results for 4 time periods:
--the start of Record Retailer till the end of Record Mirror, Mar 1960 to Mar 1962, 5 charts NME, RM, MM, Disc, RR
--the end of Record Mirror till the end of Disc, Mar 1962 to Aug 1967, 4 charts NME, MM, Disc, RR
--the end of Disc till the start of BMRB/’official’, Aug 1967 to Feb 1969, 3 charts NME, MM, RR
--the entire period Mar 1960 to Feb 1969, all 5 charts, stats compiled and compared together across all 3 sub-time periods, which is meaningful for NME, MM, and RR, not so much so for RM and Disc
CRUNCHED NUMBERS FOR TOP 10 RECORDS, ENTIRE PERIOD MAR 1960 TO FEB 1969
There were 969 records that peaked in the Top 10 on these 5 charts during this entire period. There were 556 occurrences when a record peak on 1 chart was a standalone farthest outlier (worst distance) from that record’s avg peak across all charts. Here’s the breakdown, the fewer the better:
Chart (# years) . . . . . . . . . . . . . . . . . NME(9) . . . RM(2) . . . MM(9) . . . Disc(7) . . . RR(9)
total # outliers. . . . . . . . . . . . . . . . . . . 131 . . . . . . 18 . . . . . . 106 . . . . . .89 . . . . . . 212
# outliers lower peak than avg. . . . . . 87 . . . . . . . 9 . . . . . . . 49 . . . . . . 38 . . . . . . . 88
# outliers higher peak than avg . . . . . 44 . . . . . . . 9 . . . . . . . 57 . . . . . . 51 . . . . . . 124
avg pos dist, peak to avg pk . . . . . . . 1.63 . . . . . 2.36 . . . . . 1.56 . . . . 1.80 . . . . . 1.76
Likewise, there were 231 occurrences when a record peak on 1 chart was a standalone closest inlier (best distance) to that record’s avg peak across all the charts. Here’s that breakdown, the more the better:
Chart (# years) . . . . . . . . . . . . . . . . . NME(9) . . . RM(2) . . . MM(9) . . . Disc(7) . . . RR(9)
total # inliers . . . . . . . . . . . . . . . . . . . . 57 . . . . . . . 14 . . . . . . 68 . . . . . . . 54 . . . . . . 38
avg pos dist, peak to avg pk. . . . . . . 0.44 . . . . . 0.39 . . . . . 0.33 . . . . . 0.30 . . . . .0.34
So those stats are for the worst and best cases. Here are some stats for all 969 records across this entire time period:
Chart (# years) . . . . . . . . . . . . . . . . . . . NME(9) . . . RM(2) . . . MM(9) . . . Disc(7) . . . RR(9)
# recs, peak lower than avg pk. . . . . . . 399 . . . . . . 87 . . . . . . 339 . . . . . .283 . . . . . 295
# recs, peak higher than avg pk . . . . . . 299 . . . . . . 77 . . . . . . 353 . . . . . .297 . . . . . 410
# recs, at avg peak. . . . . . . . . . . . . . . . . . 271 . . . . . . 45 . . . . . .277 . . . . . .230 . . . . . 264
total # of records . . . . . . . . . . . . . . . . . . 969 . . . . . .209 . . . . . 969 . . . . . .810 . . . . . 969
avg pos dist, peak to avg pk . . . . . . . . . 0.66 . . . . . 0.82 . . . . . 0.63 . . . . . 0.65 . . . . .0.78
Breaking down the chart position distance ‘peak to avg peak’ into ranges, you get this:
Chart (# years) . . . . . . . . . . . . . . . . . NME(9) . . . RM(2) . . . MM(9) . . . Disc(7) . . . RR(9)
# recs w/pos dist 0 to <1. . . . . . . . . . 712 . . . . . 140 . . . . . 728 . . . . . . 606 . . . . .663
1 to <2 . . . . . . . . . . . . . . . . . . . . . . . . . 177 . . . . . . 43 . . . . . .169 . . . . . .140 . . . . .200
2 to <3. . . . . . . . . . . . . . . . . . . . . . . . . . 54 . . . . . . .20 . . . . . . 52 . . . . . . . 43 . . . . . .66
3 to <4. . . . . . . . . . . . . . . . . . . . . . . . . . 17 . . . . . . . 4 . . . . . . .15 . . . . . . . 12 . . . . . .29
4 to <5 . . . . . . . . . . . . . . . . . . . . . . . . . . 7 . . . . . . . .0 . . . . . . . 1 . . . . . . . . 7 . . . . . . . 6
5 to <6. . . . . . . . . . . . . . . . . . . . . . . . . . .2 . . . . . . . .2 . . . . . . . 3 . . . . . . . . 1 . . . . . . . 1
6 to <7. . . . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . .0 . . . . . . . 1 . . . . . . . . 1 . . . . . . . 2
7 to <8. . . . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . .0 . . . . . . . 0 . . . . . . . . 0 . . . . . . . 1
8 to <15. . . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . .0 . . . . . . . 0 . . . . . . . . 0 . . . . . . . 0
15 to <16. . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . .0 . . . . . . . 0 . . . . . . . . 0 . . . . . . . 1
total # of records . . . . . . . . . . . . . . . . 969 . . . . . .209 . . . . . 969 . . . . . 810 . . . . . 969
THE 9 RECORDS WITH THE GREATEST CHART POSITION DISTANCE, PEAK TO AVERAGE PEAK, + STATS
Here are the 9 records with the greatest distance, peak to average peak, between 5 to 15 chart positions. I’ll throw in the stats of these 9 as an example to show how the stats for all 969 records (above) were calculated.
--------------------------------------------------
Billy Fury - Colette, 1960 Mar 12
peaks . . . . . . NME 19 . . . RM 18 . . . MM 15 . . . Disc 18 . . . RR 9 . . . avg peak = 15.8
dist from avg . . 3.2 . . . . . . 2.2 . . . . . . . 0.8 . . . . . . 2.2 . . . . . 6.8
---------------------------------------------------
Russ Conway - Lucky Five, 1960 May 28
peaks . . . . . . NME 13 . . . RM 9 . . . . MM 20 . . . Disc 18 . . . RR 14 . . . avg peak = 14.8
dist from avg . . 1.8 . . . . . . 5.8 . . . . . . . 5.2 . . . . . . 3.2 . . . . . . 0.8
--------------------------------------------------
Charlie Drake - Mr. Custer, 1960 Nov 5
peaks . . . . . . NME 10 . . . RM 14 . . . MM -NC- . . . Disc 19 . . . RR 12 . . . avg peak = 15.2
dist from avg . . 5.2 . . . . . . 1.2 . . . . . . . . 5.8 . . . . . . . 3.8 . . . . . .3.2 . . . . (est MM 21)
--------------------------------------------------
Karl Denver - Marcheta, 1961 July 1
peaks . . . . . . NME 15 . . . RM 16 . . . MM 12 . . . Disc 15 . . . RR 8 . . . avg peak = 13.2
dist from avg . . 1.8 . . . . . . 2.8 . . . . . . . 1.2 . . . . . . 1.8 . . . . . 5.2
--------------------------------------------------
Bobby Darin - You Must Have Been A Beautiful Baby, 1961 Oct 14
peaks . . . . . . NME 10 . . . RM 18 . . . MM 13 . . . Disc 14 . . . RR 10 . . . avg peak = 13.0
dist from avg . . 3.0 . . . . . . 5.0 . . . . . . . 0.0 . . . . . . 1.0 . . . . . 3.0
--------------------------------------------------
Helen Shapiro - Let’s Talk About Love, 1962 May 12
peaks . . . . . . NME 16 . . . MM 18 . . . Disc 10 . . . RR 23 . . . avg peak = 16.75
dist from avg . . 0.75 . . . . . 1.25 . . . . . . 6.75 . . . . 6.25
--------------------------------------------------
Yardbirds - Evil Hearted You / Still I’m Sad, 1965 Oct 16
peaks . . . . . . NME 10/9 . . . MM 2/2 . . . Disc 2/NC . . . RR 3/3 . . . avg peak = 4.0
dist from avg . . . 5.0 . . . . . . . . 2.0 . . . . . . . . 2.0 . . . . . . . 1.0
--------------------------------------------------
Young Idea - With A Little Help From My Friends, 1967 July 22
peaks . . . . . . NME 29 . . . MM -NC- . . . Disc 30 . . . RR 10 . . . avg peak = 25.0
dist from avg . . 4.0 . . . . . . . . 6.0 . . . . . . . 5.0 . . . . . 15.0 . . . . (est MM 31)
--------------------------------------------------
Simon & Garfunkel - Mrs. Robinson EP, 1969 Feb 8
peaks . . . . . . . NME 19 . . . . MM 22 . . . . RR 9 . . . . avg peak = 16.67
dist from avg. . . 2.33 . . . . . . . 5.33 . . . . . 7.67
--------------------------------------------------
Some average stats for just these 9 records, as an example for how the stats for all 969 records were calculated:
avg pos dist, peak to avg pk . . NME 3.009 . . . RM 3.4 . . . . MM 3.065 . . . Disc 3.219 . . . RR 5.435
# of outliers . . . . . . . . . . . . . . . . . . . 1 . . . . . . . . . . 2 . . . . . . . . . . . 1 . . . . . . . . . . 1 . . . . . . . . . . . 4
# of inliers. . . . . . . . . . . . . . . . . . . . . 3 . . . . . . . . . . 1 . . . . . . . . . . . 3 . . . . . . . . . . 0 . . . . . . . . . . . 2
avg outlier dist pk to avg pk . . . . . 5.0 . . . . . . . . 5.4 . . . . . . . . . . 5.8 . . . . . . . . 6.75 . . . . . . . 8.667
avg inlier dist pk to avg pk . . . . . 2.361 . . . . . . . 1.2 . . . . . . . . . 0.667 . . . . . . . NA . . . . . . . . . 0.9
RESULTS SUMMARY
For the full 9 year period Mar 1960 to Feb 1969, of NME vs. MM vs. RR, I’m seeing the following. RM and Disc, at 2 and 7 years, will occasionally show something interesting, but more is revealed with them when breaking down the 9 year period into 3 sub-periods. I’ve done that, and will show those results in separate postings. So looking at the results of the big 3 NME, MM, RR:
--RR has the worst/most outliers at 212, or 22% of the 969 records, with the highest ‘peak to avg peak’ distance of 1.76 chart positions
--MM has the best/fewest outliers at 106, or 11% of the 969 records, with the lowest ‘peak to avg peak’ distance of 1.56 chart positions
--RR has the worst/fewest inliers at 38 (= 4%), avg peak distance 0.34 chart positions
--MM has the best/most inliers at 68 (= 7%), avg peak distance of 0.33 chart positions
--RR has the worst/highest avg peak distance across all 969 records at 0.78 chart position
--MM has the best/lowest avg peak distance across all 969 records at 0.63 chart position
--RR has the worst/fewest # of records at 663 in the 0 to <1 dist range ‘peak to avg pk’
--MM has the best/most # of records at 728 in the 0 to <1 dist range ‘peak to avg pk’
--RR has the worst/most # of records at 295 in the 1 to <4 dist ranges
--MM has the best/fewest # of records at 236 in the 1 to <4 dist ranges
--RR has the worst/most # of records at 11 in the 4 to 15 dist ranges
--MM has the best/fewest # of records at 5 in the 4 to 15 dist ranges
--NME takes the middle ground between MM and RR in most all stats
--RM and Disc are not too shabby either, if their stats are pro-rated to 9 years, they come off very well, most times looking better than RR
--I’ll throw out that the NME peaks are probably skewed/hurt a little bit by the EPs and separate B-sides they charted. Not only could the separate B-sides have prevented a higher position for the A-sides had they been combined, but the B-sides and EPs could have also blocked records below them from peaking higher. So adjusting for these situations, NME should come out looking slightly better than it does. But I don’t have time to monkey around with all that right now, ha.
CONCLUSION (man, this feels like a term paper…)
So knowing that MM sampled 3 to 3.5 times more record shops than RR during the 60s, and considering all of the above stats, one can only conclude that MM is the best chart for this time period, and RR is the worst. NME being the middle ground. I don’t see any other way around it. I didn’t know what I was going to find at the beginning of this task, but there it is in crunched numbers.
As it has been said before, all 5 charts generally track together fairly well. The records towards the top of 1 chart are usually at the top of all charts, just in a slightly different order. The big hits are the big hits. The problem comes into play when someone in an article, TV show, or contest, starts claiming how many #1 records such-and-such an artist achieved, and they go quoting from the least accurate chart, that was not ‘official’ during the time period in question.
So which chart SHOULD be designated and used as the ‘best’ for the 60s when there was no ‘official’ chart? The most accurate chart (MM) that has 50 weekly chart positions for 4.6 years, 30 positions for 2.3 years, and 20 positions for 2.1 years? Or the chart (RR) with 50 weekly chart positions over that entire time period, but is not the 1st or even the 2nd most accurate chart over that time period, or any piece of that time period?
How about a compromise solution, such as use MM when it has 50 weekly chart positions, but when it has less than 50 then use MM for the higher chart positions, and fill in the lower chart positions with NME and RR records? That would be the best of both worlds: better accuracy, and the most positions.
It could be that there is a better ‘NME to MM’ swap over date than Mar 1960, which can be statistically determined. There are actually about 15 better chart alternatives for the 60s than RR that I’ve thought up, some are rather interesting. I’ll post them for you soon (he threatens).
So that’s it for this go ‘round. Like I said, I’ve got stat data for the 3 separate time periods of the 60s when there were 5, 4, and 3 charts, and I’ll get that to you someday soon.
I sure could use a Zoom ice lolly right about now, a tube of Smarties, some McVitie’s dark choc digestives, a Corgi car…
Any thoughts, comments?
END END END END END END END END END END END END END END END END END END END END END
MAR 1960 TO FEB 1969
I apologize in advance, this could be information overload, too many numbers for the average Joe, yada yada. And I admit it’s wacko to a degree, but I needed to do this just to see the lay of the land once and for all, kinda, sorta. So here goes…
I’ve done some more statistical analysis on the 60s UK charts, a step beyond my similar #1’s comparison in posts elsewhere, this time for records that made it into the Top 10. The idea came to me one day after observing the online Tiscali spreadsheet of each record’s peaks across various UK (and 1 US) charts, and I thought some comparison stats could be cranked out fairly quickly in a spreadsheet. I was interested in seeing how closely, or far away, the 5 major 60s charts tracked each other in terms of individual record peaks. Were they all relatively close plus or minus, or did some track the averages consistently better or worse than others? I decided to give it a go, but it turned out to be more work than I had bargained for. Nonetheless, the results were most interesting.
I took the Tiscali spreadsheet ‘by record title,’ sorted it by record date, sectioned it off from the start of the Record Retailer chart March 1960 until the start of the BMRB/’official’ chart Feb 1969. Then also broke that up into 3 separate time periods when there were 5, 4 and 3 charts, then calculated the average chart peak for every record that made the Top 10 on any of the 5 major charts NME, Record Mirror, Melody Maker, Disc, and Record Retailer. Then I calculated the difference between each record’s average chart peak and its individual chart peaks.
I had to do some manual data checking, as Tiscali did not have a pure Record Retailer column, but instead a Guinness column with NME, RR, and BMRB data; and a Record Mirror column with RM and RR data. So I checked for specific pre-BMRB Record Retailer chart peaks during the critical chart change out periods for these charts, and also when the RM and Disc charts ceased. Likewise, I screened for NME and MM record peaks at the beginning and end of the entire period. Thus if a record did not peak on a given chart due to timing or chart startups/cutoffs, then that chart was not analyzed for the record peak comparison. Also, EPs and B-sides that only appeared on a minority # of charts (mostly NME) were not considered. A few times, different sides of the same record peaked on different charts, and in those cases I made a manual calculation of the best peak of either side to count as one entry.
Once the data was complete, I also double checked the records with the greatest peak differences, there were about 20 worst outliers that deserved extra scrutiny. I should point out that I did not double check every Top 10 record peak position against every chart to see if Tiscali was 100% correct. Hopefully the bugs are few (I did find a few), and spread around equally, so as not to affect the overall results. Perhaps one day I’ll get around to analyzing positions 11 thru 20. And the 50s period.
OK. So I calculated comparison results for 4 time periods:
--the start of Record Retailer till the end of Record Mirror, Mar 1960 to Mar 1962, 5 charts NME, RM, MM, Disc, RR
--the end of Record Mirror till the end of Disc, Mar 1962 to Aug 1967, 4 charts NME, MM, Disc, RR
--the end of Disc till the start of BMRB/’official’, Aug 1967 to Feb 1969, 3 charts NME, MM, RR
--the entire period Mar 1960 to Feb 1969, all 5 charts, stats compiled and compared together across all 3 sub-time periods, which is meaningful for NME, MM, and RR, not so much so for RM and Disc
CRUNCHED NUMBERS FOR TOP 10 RECORDS, ENTIRE PERIOD MAR 1960 TO FEB 1969
There were 969 records that peaked in the Top 10 on these 5 charts during this entire period. There were 556 occurrences when a record peak on 1 chart was a standalone farthest outlier (worst distance) from that record’s avg peak across all charts. Here’s the breakdown, the fewer the better:
Chart (# years) . . . . . . . . . . . . . . . . . NME(9) . . . RM(2) . . . MM(9) . . . Disc(7) . . . RR(9)
total # outliers. . . . . . . . . . . . . . . . . . . 131 . . . . . . 18 . . . . . . 106 . . . . . .89 . . . . . . 212
# outliers lower peak than avg. . . . . . 87 . . . . . . . 9 . . . . . . . 49 . . . . . . 38 . . . . . . . 88
# outliers higher peak than avg . . . . . 44 . . . . . . . 9 . . . . . . . 57 . . . . . . 51 . . . . . . 124
avg pos dist, peak to avg pk . . . . . . . 1.63 . . . . . 2.36 . . . . . 1.56 . . . . 1.80 . . . . . 1.76
Likewise, there were 231 occurrences when a record peak on 1 chart was a standalone closest inlier (best distance) to that record’s avg peak across all the charts. Here’s that breakdown, the more the better:
Chart (# years) . . . . . . . . . . . . . . . . . NME(9) . . . RM(2) . . . MM(9) . . . Disc(7) . . . RR(9)
total # inliers . . . . . . . . . . . . . . . . . . . . 57 . . . . . . . 14 . . . . . . 68 . . . . . . . 54 . . . . . . 38
avg pos dist, peak to avg pk. . . . . . . 0.44 . . . . . 0.39 . . . . . 0.33 . . . . . 0.30 . . . . .0.34
So those stats are for the worst and best cases. Here are some stats for all 969 records across this entire time period:
Chart (# years) . . . . . . . . . . . . . . . . . . . NME(9) . . . RM(2) . . . MM(9) . . . Disc(7) . . . RR(9)
# recs, peak lower than avg pk. . . . . . . 399 . . . . . . 87 . . . . . . 339 . . . . . .283 . . . . . 295
# recs, peak higher than avg pk . . . . . . 299 . . . . . . 77 . . . . . . 353 . . . . . .297 . . . . . 410
# recs, at avg peak. . . . . . . . . . . . . . . . . . 271 . . . . . . 45 . . . . . .277 . . . . . .230 . . . . . 264
total # of records . . . . . . . . . . . . . . . . . . 969 . . . . . .209 . . . . . 969 . . . . . .810 . . . . . 969
avg pos dist, peak to avg pk . . . . . . . . . 0.66 . . . . . 0.82 . . . . . 0.63 . . . . . 0.65 . . . . .0.78
Breaking down the chart position distance ‘peak to avg peak’ into ranges, you get this:
Chart (# years) . . . . . . . . . . . . . . . . . NME(9) . . . RM(2) . . . MM(9) . . . Disc(7) . . . RR(9)
# recs w/pos dist 0 to <1. . . . . . . . . . 712 . . . . . 140 . . . . . 728 . . . . . . 606 . . . . .663
1 to <2 . . . . . . . . . . . . . . . . . . . . . . . . . 177 . . . . . . 43 . . . . . .169 . . . . . .140 . . . . .200
2 to <3. . . . . . . . . . . . . . . . . . . . . . . . . . 54 . . . . . . .20 . . . . . . 52 . . . . . . . 43 . . . . . .66
3 to <4. . . . . . . . . . . . . . . . . . . . . . . . . . 17 . . . . . . . 4 . . . . . . .15 . . . . . . . 12 . . . . . .29
4 to <5 . . . . . . . . . . . . . . . . . . . . . . . . . . 7 . . . . . . . .0 . . . . . . . 1 . . . . . . . . 7 . . . . . . . 6
5 to <6. . . . . . . . . . . . . . . . . . . . . . . . . . .2 . . . . . . . .2 . . . . . . . 3 . . . . . . . . 1 . . . . . . . 1
6 to <7. . . . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . .0 . . . . . . . 1 . . . . . . . . 1 . . . . . . . 2
7 to <8. . . . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . .0 . . . . . . . 0 . . . . . . . . 0 . . . . . . . 1
8 to <15. . . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . .0 . . . . . . . 0 . . . . . . . . 0 . . . . . . . 0
15 to <16. . . . . . . . . . . . . . . . . . . . . . . . .0 . . . . . . . .0 . . . . . . . 0 . . . . . . . . 0 . . . . . . . 1
total # of records . . . . . . . . . . . . . . . . 969 . . . . . .209 . . . . . 969 . . . . . 810 . . . . . 969
THE 9 RECORDS WITH THE GREATEST CHART POSITION DISTANCE, PEAK TO AVERAGE PEAK, + STATS
Here are the 9 records with the greatest distance, peak to average peak, between 5 to 15 chart positions. I’ll throw in the stats of these 9 as an example to show how the stats for all 969 records (above) were calculated.
--------------------------------------------------
Billy Fury - Colette, 1960 Mar 12
peaks . . . . . . NME 19 . . . RM 18 . . . MM 15 . . . Disc 18 . . . RR 9 . . . avg peak = 15.8
dist from avg . . 3.2 . . . . . . 2.2 . . . . . . . 0.8 . . . . . . 2.2 . . . . . 6.8
---------------------------------------------------
Russ Conway - Lucky Five, 1960 May 28
peaks . . . . . . NME 13 . . . RM 9 . . . . MM 20 . . . Disc 18 . . . RR 14 . . . avg peak = 14.8
dist from avg . . 1.8 . . . . . . 5.8 . . . . . . . 5.2 . . . . . . 3.2 . . . . . . 0.8
--------------------------------------------------
Charlie Drake - Mr. Custer, 1960 Nov 5
peaks . . . . . . NME 10 . . . RM 14 . . . MM -NC- . . . Disc 19 . . . RR 12 . . . avg peak = 15.2
dist from avg . . 5.2 . . . . . . 1.2 . . . . . . . . 5.8 . . . . . . . 3.8 . . . . . .3.2 . . . . (est MM 21)
--------------------------------------------------
Karl Denver - Marcheta, 1961 July 1
peaks . . . . . . NME 15 . . . RM 16 . . . MM 12 . . . Disc 15 . . . RR 8 . . . avg peak = 13.2
dist from avg . . 1.8 . . . . . . 2.8 . . . . . . . 1.2 . . . . . . 1.8 . . . . . 5.2
--------------------------------------------------
Bobby Darin - You Must Have Been A Beautiful Baby, 1961 Oct 14
peaks . . . . . . NME 10 . . . RM 18 . . . MM 13 . . . Disc 14 . . . RR 10 . . . avg peak = 13.0
dist from avg . . 3.0 . . . . . . 5.0 . . . . . . . 0.0 . . . . . . 1.0 . . . . . 3.0
--------------------------------------------------
Helen Shapiro - Let’s Talk About Love, 1962 May 12
peaks . . . . . . NME 16 . . . MM 18 . . . Disc 10 . . . RR 23 . . . avg peak = 16.75
dist from avg . . 0.75 . . . . . 1.25 . . . . . . 6.75 . . . . 6.25
--------------------------------------------------
Yardbirds - Evil Hearted You / Still I’m Sad, 1965 Oct 16
peaks . . . . . . NME 10/9 . . . MM 2/2 . . . Disc 2/NC . . . RR 3/3 . . . avg peak = 4.0
dist from avg . . . 5.0 . . . . . . . . 2.0 . . . . . . . . 2.0 . . . . . . . 1.0
--------------------------------------------------
Young Idea - With A Little Help From My Friends, 1967 July 22
peaks . . . . . . NME 29 . . . MM -NC- . . . Disc 30 . . . RR 10 . . . avg peak = 25.0
dist from avg . . 4.0 . . . . . . . . 6.0 . . . . . . . 5.0 . . . . . 15.0 . . . . (est MM 31)
--------------------------------------------------
Simon & Garfunkel - Mrs. Robinson EP, 1969 Feb 8
peaks . . . . . . . NME 19 . . . . MM 22 . . . . RR 9 . . . . avg peak = 16.67
dist from avg. . . 2.33 . . . . . . . 5.33 . . . . . 7.67
--------------------------------------------------
Some average stats for just these 9 records, as an example for how the stats for all 969 records were calculated:
avg pos dist, peak to avg pk . . NME 3.009 . . . RM 3.4 . . . . MM 3.065 . . . Disc 3.219 . . . RR 5.435
# of outliers . . . . . . . . . . . . . . . . . . . 1 . . . . . . . . . . 2 . . . . . . . . . . . 1 . . . . . . . . . . 1 . . . . . . . . . . . 4
# of inliers. . . . . . . . . . . . . . . . . . . . . 3 . . . . . . . . . . 1 . . . . . . . . . . . 3 . . . . . . . . . . 0 . . . . . . . . . . . 2
avg outlier dist pk to avg pk . . . . . 5.0 . . . . . . . . 5.4 . . . . . . . . . . 5.8 . . . . . . . . 6.75 . . . . . . . 8.667
avg inlier dist pk to avg pk . . . . . 2.361 . . . . . . . 1.2 . . . . . . . . . 0.667 . . . . . . . NA . . . . . . . . . 0.9
RESULTS SUMMARY
For the full 9 year period Mar 1960 to Feb 1969, of NME vs. MM vs. RR, I’m seeing the following. RM and Disc, at 2 and 7 years, will occasionally show something interesting, but more is revealed with them when breaking down the 9 year period into 3 sub-periods. I’ve done that, and will show those results in separate postings. So looking at the results of the big 3 NME, MM, RR:
--RR has the worst/most outliers at 212, or 22% of the 969 records, with the highest ‘peak to avg peak’ distance of 1.76 chart positions
--MM has the best/fewest outliers at 106, or 11% of the 969 records, with the lowest ‘peak to avg peak’ distance of 1.56 chart positions
--RR has the worst/fewest inliers at 38 (= 4%), avg peak distance 0.34 chart positions
--MM has the best/most inliers at 68 (= 7%), avg peak distance of 0.33 chart positions
--RR has the worst/highest avg peak distance across all 969 records at 0.78 chart position
--MM has the best/lowest avg peak distance across all 969 records at 0.63 chart position
--RR has the worst/fewest # of records at 663 in the 0 to <1 dist range ‘peak to avg pk’
--MM has the best/most # of records at 728 in the 0 to <1 dist range ‘peak to avg pk’
--RR has the worst/most # of records at 295 in the 1 to <4 dist ranges
--MM has the best/fewest # of records at 236 in the 1 to <4 dist ranges
--RR has the worst/most # of records at 11 in the 4 to 15 dist ranges
--MM has the best/fewest # of records at 5 in the 4 to 15 dist ranges
--NME takes the middle ground between MM and RR in most all stats
--RM and Disc are not too shabby either, if their stats are pro-rated to 9 years, they come off very well, most times looking better than RR
--I’ll throw out that the NME peaks are probably skewed/hurt a little bit by the EPs and separate B-sides they charted. Not only could the separate B-sides have prevented a higher position for the A-sides had they been combined, but the B-sides and EPs could have also blocked records below them from peaking higher. So adjusting for these situations, NME should come out looking slightly better than it does. But I don’t have time to monkey around with all that right now, ha.
CONCLUSION (man, this feels like a term paper…)
So knowing that MM sampled 3 to 3.5 times more record shops than RR during the 60s, and considering all of the above stats, one can only conclude that MM is the best chart for this time period, and RR is the worst. NME being the middle ground. I don’t see any other way around it. I didn’t know what I was going to find at the beginning of this task, but there it is in crunched numbers.
As it has been said before, all 5 charts generally track together fairly well. The records towards the top of 1 chart are usually at the top of all charts, just in a slightly different order. The big hits are the big hits. The problem comes into play when someone in an article, TV show, or contest, starts claiming how many #1 records such-and-such an artist achieved, and they go quoting from the least accurate chart, that was not ‘official’ during the time period in question.
So which chart SHOULD be designated and used as the ‘best’ for the 60s when there was no ‘official’ chart? The most accurate chart (MM) that has 50 weekly chart positions for 4.6 years, 30 positions for 2.3 years, and 20 positions for 2.1 years? Or the chart (RR) with 50 weekly chart positions over that entire time period, but is not the 1st or even the 2nd most accurate chart over that time period, or any piece of that time period?
How about a compromise solution, such as use MM when it has 50 weekly chart positions, but when it has less than 50 then use MM for the higher chart positions, and fill in the lower chart positions with NME and RR records? That would be the best of both worlds: better accuracy, and the most positions.
It could be that there is a better ‘NME to MM’ swap over date than Mar 1960, which can be statistically determined. There are actually about 15 better chart alternatives for the 60s than RR that I’ve thought up, some are rather interesting. I’ll post them for you soon (he threatens).
So that’s it for this go ‘round. Like I said, I’ve got stat data for the 3 separate time periods of the 60s when there were 5, 4, and 3 charts, and I’ll get that to you someday soon.
I sure could use a Zoom ice lolly right about now, a tube of Smarties, some McVitie’s dark choc digestives, a Corgi car…

Any thoughts, comments?
END END END END END END END END END END END END END END END END END END END END END
Comment