Author |
Topic Search Topic Options
|
Gerinski
Prog Reviewer
Joined: February 10 2010
Location: Barcelona Spain
Status: Offline
Points: 5093
|
Posted: December 31 2014 at 03:21 |
Kazza3 wrote:
Nice stuff Gerinski! I'd be interested to see a version of the albums over 4.00 chart, where rather than total number of albums above, it was the percentage of albums from that year that were above, if you've got access to the data to do that without manually counting them all!
Also, the subgenre graph might be easier to read if it was displayed as a percentage area chart, like this (just a random example):
|
Yes I have the data and of course I do not have to count anything manually! I copied the whole contents of the Top 100 page of each year as text in excel and then it was just a matter of getting it ordered properly into a table form.
Will check that. BTW I tried to make the subgenres evolution graph in this style and it gets too messy, the jumps from one year to another are too abrupt. Showing it as % bars instead of areas is a bit more clear (see post below). Will see which graph style fits best for this one.
|
|
Kati
Forum Senior Member
Joined: September 10 2010
Location: Earth
Status: Offline
Points: 6253
|
Posted: December 31 2014 at 03:23 |
that graph is test dummy crash, downhill scary Awww group hug
|
|
Gerinski
Prog Reviewer
Joined: February 10 2010
Location: Barcelona Spain
Status: Offline
Points: 5093
|
Posted: December 31 2014 at 03:26 |
Kati wrote:
|
Considering I am a nincompoop, these statistics need more details for me the x y axis are not very clear to me, however I bet it will make perfect sense to others big hug to you, Gerinsky [/QUOTE]
No problem Kati the horizontal axis are the Top 100 albums of the year, from # 1 at the left up to # 100 at the right, and vertical is their rating. Since I used the 'normal' rating and not the QWR, it is possible that an album with, say, a rating of 3.90 is higher in the position rank than an album with rating 4.00, that's why you see a "saw" with ups and downs and not a simple slope line getting lower and lower as you move to lower positions in the ranking. I hope it's a bit more clear now?
|
|
Gerinski
Prog Reviewer
Joined: February 10 2010
Location: Barcelona Spain
Status: Offline
Points: 5093
|
Posted: December 31 2014 at 03:31 |
Hahahaha, don't worry it's not a profits graph We look at the albums ratings from the top # 1 album (at the left) to the last in the top 100 (at the right) so logically the trend goes downward
|
|
Kati
Forum Senior Member
Joined: September 10 2010
Location: Earth
Status: Offline
Points: 6253
|
Posted: December 31 2014 at 03:31 |
Gerinski wrote:
Kati wrote:
|
Considering I am a nincompoop, these statistics need more details for me the x y axis are not very clear to me, however I bet it will make perfect sense to others big hug to you, Gerinsky |
No problem Kati the horizontal axis are the Top 100 albums of the year, from # 1 at the left up to # 100 at the right, and vertical is their rating. Since I used the 'normal' rating and not the QWR, it is possible that an album with, say, a rating of 3.90 is higher in the position rank than an album with rating 4.00, that's why you see a "saw" with ups and downs and not a simple slope line getting lower and lower as you move to lower positions in the ranking. I hope it's a bit more clear now? [/QUOTE]
Gerinsky, you are so awesome really! Yep, got it now, thus average rating is 4 or between that and 4.10 high scores indeedy another big hug to you
|
|
Barbu
Forum Senior Member
Joined: October 09 2005
Location: infinity
Status: Offline
Points: 30845
|
Posted: December 31 2014 at 03:39 |
Gerinski wrote:
I hope it's a bit more clear now? |
|
|
Gerinski
Prog Reviewer
Joined: February 10 2010
Location: Barcelona Spain
Status: Offline
Points: 5093
|
Posted: December 31 2014 at 03:41 |
I remade the graph for the evolution by sub-genres in % bars form, maybe it's a bit clearer this way? (don't get scared Kati ). The sub-genres are arranged in the order they appear along history, so newer sub-genres get added at the top of pre-existing sub-genres. So for example the first at the bottom of the pile was JR-F, then came RIO-Avant, Psychedelic etc. RPI appeared in 1971, Neo in 1982, Tech-Extreme in 1985 etc.
A curiosity, RPI completely disappears in 1980 and stays absent until its comeback in 1988, after which it has remained present every year.
Sure you will need to show this at full size in order to see something (unless you have a huge monitor)
Edited by Gerinski - December 31 2014 at 03:42
|
|
Barbu
Forum Senior Member
Joined: October 09 2005
Location: infinity
Status: Offline
Points: 30845
|
Posted: December 31 2014 at 03:43 |
|
|
Gerinski
Prog Reviewer
Joined: February 10 2010
Location: Barcelona Spain
Status: Offline
Points: 5093
|
Posted: December 31 2014 at 03:44 |
wait until I make the one suggested by Kazza
Edited by Gerinski - December 31 2014 at 03:45
|
|
Kati
Forum Senior Member
Joined: September 10 2010
Location: Earth
Status: Offline
Points: 6253
|
Posted: December 31 2014 at 04:58 |
hahahaha I love Grumpy cat he's so obnoxious funny! hahaha xxx
|
|
tboyd1802
Forum Senior Member
Joined: June 01 2012
Location: USA
Status: Offline
Points: 898
|
Posted: December 31 2014 at 10:22 |
Gerinski wrote:
tboyd1802 wrote:
Ok, by this analysis how does 2013 compare to 2012? In a previous post, you proposed that 2014 metrics may be down because of the shorter time window for which 2014 reviews could have been posted. 2013 is clearly anomalous to both years prior and after 2013. Comparing 2012 to 2013 would be one way to test your hypothesis.
|
Here are the ratings for the Top 100 in 2012, 2013 and 2014 (when trying to add more years the graph gets too messy). We see that the anomaly of 2013 is that a lot of the Top 100 albums stay over the 4.00 mark (red line). Even the albums around the 70th rank are still above 4.00. For 2014 after the 20th rank practically all albums are already below 4.00. 2012 was a bit in between, with albums starting to drop below the 4.00 mark around rank 55th.
In the small box I mention the average number of ratings per album, where we see that indeed for 2014 albums the number of ratings is still quite low compared to previous years.
There are too many unknown factors, for example I guess that reviewers activity has a significant impact since their reviews have a high weight. In years were reviewers are very active reviewing the new albums it is sound to imagine that ratings get higher, while if, just for example, in 2014 they were relatively inactive reviewing new albums, the ratings will be mostly by normal members and the total ratings will be lower.
|
Thanks for running the 2012-2013 comparison. Based on these data, 2013 to 2014 did mark quite a change.
Agreed, the number of ratings is quite low for 2014 compared to other years. Based on my anecdotal observations of reviewer behavior, however, I would expect that as more reviews come in that the 2014 ratings would actually go down - not up. In watching rankings for individual offerings evolve over time it is pretty rare to see these ratings go up over time (yah, yah in the short term they go up - but, I am speaking longer term), but quite common for them to fall over time. I'm thinking specifically of the overall ratings for say the Steven Wilson offering (the Raven), or as we are now seeing for the most recent IQ offering. It seems as though the true believers offer their opinions first and that these get tempered latter with a more diverse set of reviews.
Another great data mining activity there - if we only had access to the data !-) Do we know what data are actually archived at the server end?
|
|
Gerinski
Prog Reviewer
Joined: February 10 2010
Location: Barcelona Spain
Status: Offline
Points: 5093
|
Posted: December 31 2014 at 11:09 |
I believe this is more or less the one suggested by Kazaa, showing in which range of rating values the Top albums are in each year. So just as example, of the Top 100 albums of 1979, 8% of them have a rating lower than 3.4, 17% between 3.4 and 3.6, 21% between 3.6 and 3.8, 19% between 3.8 and 4.00 etc.
We see clearly that the presence of albums with ratings lower than 3.6 was significant during the 1980's and 1990's, but has completely disappeared in recent years. As from 2006 we do not have any album with rating lower than 3.61 in the Top 100.
Only 5 years have albums rated over 4.6 = 1972, 1973, 1975, 1997 (Art Zoyd's Haxan) and 2010 (Vezhlivyj Otkaz's Geese and Swans, this one perhaps a bit weird since it has only 1 review (5 stars by a collab) and only 23 ratings in total).
|
|
Gerinski
Prog Reviewer
Joined: February 10 2010
Location: Barcelona Spain
Status: Offline
Points: 5093
|
Posted: December 31 2014 at 11:15 |
OK I don't have to cook for this evening so a last one before closing the year .
This is not a graph but just a table showing the sub-genre of the Top 25 albums for each year, with each sub-genre colour coded. To keep it in a reasonable size I have split it in two images, one from 1965 to 1990 and another from 1991 to 2014.
(again you will have to "show image in new tab" to see it big enough)
I wish you all a nice New Year's eve and a healthy and happy 2015!
Edited by Gerinski - January 01 2015 at 03:14
|
|
infocat
Forum Senior Member
VIP Member
Joined: June 10 2011
Location: Colorado, USA
Status: Offline
Points: 4671
|
Posted: January 01 2015 at 01:00 |
|
-- Frank Swarbrick Belief is not Truth.
|
|
Gerinski
Prog Reviewer
Joined: February 10 2010
Location: Barcelona Spain
Status: Offline
Points: 5093
|
Posted: January 01 2015 at 08:10 |
Gerinski wrote:
|
A couple of thoughts about this one. First is how strong was the RPI released in 1972 and 1973. One might tend to think that it was the 'genuine' British Symphonic which ruled on those years, but it is clear that PA fans love more their Italian counterparts: in both years there are 7 RPI albums in the Top 25 for only 4 of Symphonic from other countries.
The second thought is the apparent decline in recent years of Prog-Metal, Tech-Extreme, Experimental Post-Metal and Post-Rock/Math. These subgenres occupy a significant share of the Top 25 of the 1990's and 2000's but their presence in the last 5 years is quite diminished.
|
|
terramystic
Forum Senior Member
Joined: February 02 2005
Status: Offline
Points: 776
|
Posted: January 01 2015 at 17:33 |
Nice work G! Thanks!
As for subgenres, I don't see any big musical difference between todays symph, RPI and neo.
|
|
floflo79
Forum Senior Member
Joined: March 21 2013
Location: France
Status: Offline
Points: 245
|
Posted: May 07 2015 at 07:43 |
Sorry, but on the sub-genres percent in the top of 1970, the total makes 104% and not 100%...
|
|
|
Donate monthly and keep PA fast-loading and ad-free forever.