Scores updated to reflect the surge in positive reviews for Black Sabbath / Ozzy Osbourne albums following the loss of Ozzy. RIP

Wow, a lot of work here. Thx for the list. Looking better then the Top 200 here.

The 2026 update is taking longer than I anticipated... turns out a lot of scores change within a year! Hahaha. Hopefully I can update the 2026 version before 2027...

The 2026 update is here after a lot of number crunching!

Of course most albums moved around a little bit. See the comments next to albums for albums that moved significantly from the 2025 ranking.

>>> New Additions:
Vektor - Terminal Redux (released in 2016)
Insomnium - Winter’s Gate (released in 2016)
Moonsorrow - Jumalten Aika (released in 2016)
Trees of Eternity - Hour of the Nightingale (released in 2016)
Ozzy Osbourne - Blizzard Of Ozz (rating improved)

>>> Honorable Mentions - Albums that Barely Missed the Cut:
Sentenced - The Cold White Light
Dying Fetus - Destroy The Opposition
Saxon - Strong Arm Of The Law
Riverside - Second Life Syndrome
Riverside - Out Of Myself
Saturnus - Paradise Belongs To You
Haggard - Eppur Si Muove
Therion - Sirius B
Rhapsody - Symphony Of Enchanted Lands
Sanctuary - Into The Mirror Black

Good idea for a list.
Maybe you could explicitly add to the restrictions the genres you were including (or those you were excluding), for transparency. Wasnt directly obvious for me that you are limiting the list to metal (though it does say so in the list title).
And I was wondering how exactly you calculated the average score over three different scales (a 10 point scale ranging from 1 to 10, a 10 point scale from 1 to 5, and a 5 point scale from 1 to 5, right?).

Sometimes it is controversial if a band or album is considered "metal". Are there bands or albums that you consider "metal" that are not on this list? It is possible that their overall score was calculated but just was not high enough. The top 400 cutoff was not arbitrary - it turned out that only albums with an overall score of at least 8/10 were eligible to be included in the top 400.

Yes, I calculated average score over three different scales the exact way you described. The only time some other alteration was required was for an album with less than 10 reviews/scores on metal-archives. In that case the Bayesian average was used.

For example, suppose an album had 8 scores on metal-archives and an average score on metal-archives of 95%. I used 10 scores and 75% as baselines. So I calculated ( (8*95) + (10*75) ) / 18 and rounded the resulting percent to the nearest whole number. Then converted the percentage to the score out of 10. For the numbers I gave above the Bayesian average is 83.888... so that would round up to 84%, so 8.4/10.

Bayesian averages had to be used for around 50 albums or so. It is not a perfect system, but obviously if an album has an average score of 100% on metal-archives but this average is only out of 2 scores then counting the metal-archive score as 10/10 is not reliable.

Thank for elaborating on your calculation of a Bayesian average in case of low sample size on MA. I think it is a reasonable approach given the context, and its limitations are acceptable for the overall aim of a relative ranking.
With my question I was rather addressing the averaging of the three different scales and integrating them into one overall score, which determines the final ranking, if I understand correctly. Have you applied any kind of weighting algorithm, addressing potenial differences between the three scales and their empirical distribution in the actual data? For example, is 80% on MA comparable to 8/10 on MS, or if we cannot assume it is, how could we address the difference in a weighting scheme?

Written by Redel on 01.03.2026 at 19:03

Have you applied any kind of weighting algorithm, addressing potenial differences between the three scales and their empirical distribution in the actual data? For example, is 80% on MA comparable to 8/10 on MS, or if we cannot assume it is, how could we address the difference in a weighting scheme?

No, I did not apply any weighting algorithm. I would not know where to begin. 9/10 on MS is more rare than 9/10 on MA, and 9/10 on RYM is extremely rare. This makes sense considering the overall respective popularity of these different websites and the typical musical preferences of the users that frequent them.

Written by gckator on 01.03.2026 at 22:56


No, I did not apply any weighting algorithm.

And I guess I wouldnt see any problem at all with this approach if it werent for:
"If a band was not listed on metal-archives then their albums' overall mean scores were calculated WITHOUT considering metal-archives."
If I understand correctly, albums not listed on MA got their overall score calculated as an arithmetic mean over three average ratings, while it is four ratings for all other albums. This is potentially unfair if average ratings on MA are significantly different from the other three sites, right?

Written by Redel on 03.03.2026 at 21:09

Written by gckator on 01.03.2026 at 22:56


No, I did not apply any weighting algorithm.

And I guess I wouldnt see any problem at all with this approach if it werent for:
"If a band was not listed on metal-archives then their albums' overall mean scores were calculated WITHOUT considering metal-archives."
If I understand correctly, albums not listed on MA got their overall score calculated as an arithmetic mean over three average ratings, while it is four ratings for all other albums. This is potentially unfair if average ratings on MA are significantly different from the other three sites, right?

Potentially, sure. Some albums not listed on MA might have benefited from having MA scores to pull from. Some albums that ARE listed on MA would benefit from NOT having MA scores affecting their overall mean (e.g. "Reign in Blood" by Slayer).

If you have a suggestion for dealing with the albums not listed on MA then I am all ears.

Written by gckator on 04.03.2026 at 01:30


If you have a suggestion for dealing with the albums not listed on MA then I am all ears.

Well, I'd say this potential distortion is kind of inherent to the approach, and it is maybe not too much of an issue quantitatively. But we dont know for sure. Thus, I'd suggest to put a note at all albums that didnt get an MA rating calculated into the score, for transparency. If you have the time you could go one step further and even put the average ratings of the individual sites next to each album, or at least the final integrated score, but since you seem to be updating the list frequently that might be too much work to do, and the list is also fine without the exact numbers.

In the comments I put "not listed on metal-archives" next to those albums (Can you see these? Hopefully there is not a glitch preventing them from being visible.)

I might include the "overall mean" scores next to albums. I only plan on updating this list once a year.

Written by gckator on 06.03.2026 at 06:42

In the comments I put "not listed on metal-archives" next to those albums (Can you see these? Hopefully there is not a glitch preventing them from being visible.)

Yes, I can see them. That's good for transparency, I'd say.

Update: Exact OMs ("Overall Means") are listed for the very top albums while approximate OMs (rounded to the nearest hundredth) are listed for all other albums.

Very cool project and one that requires/required a lot of work, much respect. I would like to point out that albums by the band The Ocean can be found in the Metal Archives under the name The Ocean Collective, they are on like the 7th position in the search because of this name changing business.

Written by Jaro on 20.04.2026 at 05:56

Very cool project and one that requires/required a lot of work, much respect. I would like to point out that albums by the band The Ocean can be found in the Metal Archives under the name The Ocean Collective, they are on like the 7th position in the search because of this name changing business.

Ah, good to know! Thank you for the info. That might shift the ranking of the album "Pelagial" by a little bit.

Quick Update: It was pointed out that the "The Ocean" actually is on metal-archives. So the Overall Means for the albums "Precambrian" and "Pelagial" were recalculated. This resulted in both albums being higher on the list.