Actionable Rating System
I structured this rating system around what I practically DO with a record... do I put in my all-time chart? that's a 9. do I enjoy once with no follow-up? that's a 5. Etc as below.
10: A 10/10 record has left a profound, deep, lasting impression on me personally. These records represent the entire work of an artist who work has effected my expectations of what music is or can be. As of today, I have experienced all of my 10/10 artists live, and this experience is included in the overall impression. Because of this cumulative effect, it would not be possible for me to rate a record as 10 after a first (or even second) listen. Only after I've had a chance to dig into the artists catalog and experience what they have to offer, and listen to the record multiple times, only then it it eligible for this mar. 10 is a very rare designation. Today I have eleven such records.
9: A 9/10 is an all-time personal favorite. These are records I put on my big chart. These are records that stand out for me across all time. While I don't keep an exact count, this is approximately top-100 tier. In a typical year I might find one or two such records among all new releases.
I typically only show half points above 9, but in my spreadsheet I use tenths to keep them in order.
8: An 8/10 is a personal favorite. It is not to the level of my big chart, but it might be close (that's where the tenths come in.) Not only do I consider an 8 to be a great record, but it also resonates with me personally, as "my style", or a sound that I strongly relate to personally. Records that I consider great, but that do not resonate with me personally, will fall somewhere in the 7/10 tier. Thus, everything in the 8, 9, and 10 tiers are special to me personally, and subject to all my indiosyncratic preferences. I will typically find five to ten "8s" in a year's worth of new releases, so this is a best-of-the-year tier.
7: The 7/10 tier can have three meanings. For an old release, it is a record I acknowledge is very good, but that doesn't resonate with me personally. A lot of old classics fall somewhere in this tier.
For an old release that I'm hearing for the first time, 7.x is a record that does strongly resonate with me, and that I intend to listen to again. There is a good chance such a record will enter the 8 tier, if and when I actually go back to it. It is very rare that I would score a record 8/10 on a first listen, because, by definition, an 8 involves multiple listens. Hence, 7.x means a record I really liked, plan to come back to, and may one day promote to an 8. The tenths represent how strongly compelled I am to move it to 8.
For a new release, it is best-of-the month tier. 7/10 is a tipping point, where a new record automatically lands on my top picks of the month list, and serves as a reminder to myself that I must listen to it again in the coming weeks.
6: A tipping point for a record I consider worth a second listen. I'm an explorer by nature, always listening to new music, whether a new release, or new to me. A 6 is something that stands out as not only good, but good enough to listen to at least once more. Because I am constantly checking out new records, this mark is not so easy to hit.
For new releases, 6/10 is best-of-the-week tier, I typically find two or three of these in a week's worth of new releases. 6 does not automatically mean I will listen to it again soon, but just that I deem it worth another listen some day. Tenths of a point are very important between 6 and 7. The more tenths a 6 gets, the more likely it is to get a repeat listen the month it is released. In a given month, I typically don't have time to play all the 6's again. So I'll be sure to hit all the 7's and then catch the higher rated 6's. A high 6 is very likely to get a second listen, and best-of-the-month typically starts around 6.5, depending on how many high 6's I found that month.
5: 5/10 tier is something I liked enough to listen to it all the way through without getting frustrated or bored with it, but that I would not go out of my way to listen to again. This describes a lot of the records I listen to, and 5 contains more records than any other single tier. (the majority of records I hear fall in the 5 to 6 range.)
My 5 is best understood in the context of an explorer. I am constantly checking out new records. In a given week, I typically listen to more records that are new to me than not. This is just the way I enjoy my music. I find it more interesting to hear something new than to listen to the same old favorites over and over. With the advent of modern streaming services, this has become easy to do. Obviously, not all of the new records I hear will be great, and I don't expect that. I enjoy them, not because they are top tier, but because they are new. As an explorer, 5/10 means I enjoyed the record one time through, in the way I enjoy new music. So 5/10 is "good". But, since I spend so much time on new music, I can't possibly replay all my 5's (I find it difficult to work my way back around to all my 6's). In a slow month, I might have time to replay a few 5's, if so, the tenths will matter. I'll go back to anything score 5.9 first, then 5.8, and so on. Anything over 5.5 had some pretty good stuff that I enjoyed while it lasted.
4: 4/10 tier are records that are a chore to listen to all the way through, or otherwise boring. I might not call these "bad" records, but I find parts frustrating for hard to stick with. Many times, if I realize a record is tending toward 4, I stop listening after a few tracks. So there are many 4's released every week, but I typically only listen to a few, maybe one or two, maybe not even that. Typically, if I listen to a 4 all the way through, it is because it was recommended, or very popular, and I want to be able to take an opinion. (I only rate records I've listened to all the way through at least once.)
4.9 would mean I mostly enjoyed it for one play, except for a few parts that were frustrating. 4.1 would mean it had a few redeeming parts, but was mostly a chore, and so on in between.
I use the 3 and 4 tiers for records I find very boring, even if they may have some technical merit.
3: Here is where things get bad. 3/10 is something I find not just hard to listen to, but without any redeeming value, or just plain boring. But it is not especially bad. Nothing about it stands out from other bad records I've heard. Listen once and forget it.
2: It takes a special kind of bad to score below 3. To be a 2, a record has to do something special to piss me off. Quoth Butthead: "Shut up fartknocker"
1: Amazingly bad. So bad, it is almost brilliant in a way. These are worthy of worst-record ever list. Something about them is so terribly bad, that I could tolerate to listen to it again just to marvel on how very much something can suck. So bad I can laugh out loud as I listen. Quoth Butthead: "this sucks more than anything that has ever sucked before." Very few records earn this distinction (1/1000 approximately).
Alternate Statistical Rating System
A mathematical approach to rating records...
4 standard deviations will approximately include the variation of a normal (Gaussian) distribution for a population of 10,000. My last.fm library currently contains about 10,000 records, so I will use this number for a baseline population (whether this number is 5, 10, or 15 thousand is not important to the point I am trying to make, I'm talking order of magnitude)
assuming +/- 4 standard deviations and 10 tiers, each tier should occupy 8/10 (4/5) of a standard deviation. This works out to approximately the following distribution
10: 0.1%
9: 1.0%
8: 10% (technically more like 8%, but I'm using round numbers)
7: 25% (image shows 20%, it is one revision old)
6: 50+%
5: 50-%
4: -25%
3: -8%
2: -1.0%
1: -0.1%
for my library size, this means 10 records in the 10/10 tier, distributed as follows:
10: 10 records
9: 90 records (100 - top10)
8: 900 records (1000 - top100)
7: 1500
6: 2500
5: 2500
4: 1500
3: 900
2: 90
1: 10
to assess a record, I think of it relative to my all-time favorite chart. I ask, would I put this record on my top 10 chart? If not, it is a 9 at best.
would I put it on my top 100 chart? If yes, it is a 9. if maybe, 8.5. If no, 8 at best.
8/10 would be my top 1000 chart (not that I have the spare time to compile such a thing)
7/10 is entry level in the "Good-to-Great tier" - top quarter.
6/10 is a good record that is worth a second play. Better than 50% of records I've heard
5/10 is not bad. there is usually something redeeming in any record, and I can find some good in a 5. But there is not a lot worth a second play. 50% of records I've heard are better.
4/10 is entry level for the "Bad-to-Suck tier". Might still have some interesting songs/sounds, but those are rare.
3/10 is just foregtabby bad. I've heard hundreds of bad records that were just bad, and I don't even remember hearing them. Nothing stands out as memorably bad
2/10 is bad, with enough going on to be annoying. Quoth Butthead: "Shut up fartknocker"
1/10 is so bad, so bad, it is almost brilliant in a way. These are worthy of worst-record ever list. Something about them is so terribly bad, that I could tolerate to listen to it again just to marvel on how very much something can suck. So bad I can laugh out loud as I listen. Quoth Butthead: "this sucks more than anything that has ever sucked before." Very few records earn this distinction (1/1000 approximately).
full image:
end of post
