AutoRateAccurate - Release 1.61 (10/11/2006) [MM2]

[Big_Berny]

The developers fixed the incorrect ratings on the ipod when using AutoRateAccurate: "2691 Fixed iPod synchronization: when some scripts are used, ratings appear off by 1 star"

Thanks!

Big_Berny

[TJOHO]

Hi,
I run MM off an external HDD, switching between my office and home computer. Everything used to work fine for those two computers, however:
Last night, I used in on a third computer, which also worked fine. But now something is messed up with your script, regardless of what computer I use.

Whenever a new track starts, an error message pops up.
My OS is Korean XP, so I don't know exactly what the error message says, but the English part is:

Error 13 - Microsoft VBScript (something in Korean) [string: "X"]
File: "Y", Line: 50, Column: 5.

where X is the time/date when I closed MM on the computer I used last night and Y is the complete path of the script.

If I OK that error message, a new one pops up saying "Error happened during script execution" and repeating the part in Korean and the [string: "X"].

The script does not rate any songs right now, although it is enabled (edit: sorry, it does rate. It doesn't rate as you play, but if you hit the Rate Now button, it rates them. I don't know if the skip detection works or if it just goes by play history).

Any ideas?

[Big_Berny]

Oh sorry, I din't saw your post. I'll check it as soon as possible.

Big_Berny

[Big_Berny]

Hi TJOHO,
very strange what happens...

Could you send me a mail, with your mediamonkey.ini (which is stored in the same directoy like the database; normally My Music\MediaMonkey)?

Oh, and could you check if ti works when you disalbe SkipDetection?

Thanks!
Big_Berny

[TJOHO]

Hi Big_Berny,

I've PM'ed you the ini file and checked: yes it does work when I disable SkipDetection.

[MarS]

I've tried both AutoRating scripts, and my preference is for this one, but I feel that the algorithm is too agressive, to the point of being "unstable", i.e. it takes a while for it to converge to a meaningful value, which may never happen on a very large collection. Why not start with a "neutral" point for all "not seen" files, and then add/subtract according to history of use, if I could suggest?

Consider this; in a large collection (mine's hiting 5K albums), most songs are never played, but not because they're bad, but because it's not "their time", so they don't get a chance to get rated, but when they do get played, they receive high rating, and then they don't get played again for a very long time, and their rating plumets, and stays there, but that rating does not reflect their "value". I'd say, that if a song that's been in a collection for a very long time gets played occasionally, that that would be suggective of it's rather high "value", but the script assigns it a low one.

Another example; I get a new album, add it the the library, and then, because it is a new album, I give it a couple of listens. It not necessarily a good album, but I listen to it to find out if anything in there "rings a bell". The script will assign those songs a high rating, and that rating is quite persistant because those were consecutive listens in a short time span. It maybe a mediocre album, and I may never listen to it again (yet, often for completness, I will not delete it from disk), but in the mean time, it will show as a high rated album (and this "missleads" scripts as ScrobblerDJ). Were the script to start from a value like 2, or 2.5 and add a half a star for every couple of listens, a rating for an album in this example would be more realistic accross it's lifetime in the collection, i.e. go up reflecting a temporary increase in interest, and slowly decay as time progresses.

Just some thoughts.

Thank you for your work so far, and for sharing it.

[TJOHO]

I have been using this script for a short while and like it, but have been waiting patiently for it to distribute ratings evenly along the percentages set in the options panel.

As of now, all the music it has rated is from three stars and up.

I figured it just takes a while as I listen to more and more in my library (my library is only 3 weeks old, I'm a MM newbie), but if MarS is right in these ideas, ratings may never accurately reflect a song's value in my collection.

His examples seem valid, so I'd be interested to hear what other users and Big_Berny himself think.

[JuBe]

Sup guys, I'm not quite clear on what SkipCalculation does. Could someone please explain?

[Big_Berny]

Hi MarS,
I think what you suggest is exactly what the other Auto Rating scirpt of DiddeLeeDoo does. It uses a standard rating and increases it when you listen to a song.

I don't like this method because it's no the same thing when you listen to a song two times in two years or two times during the last day. Bye the way: My script doesn't rates unlistend songs bad, it just doesn't rate them.

But I know what you mean and the ratings could be rated much better when you had a table which shows exactly when you listened to a song. (There is but not it doesn't count the plays on your portable, so it's uselessfor me).

It's always difficult if not impossible to predict a rating of a song just with these simple numbers the scripts can use. For my my algorithm works quite good but I already said somewhere that for others other algorithms wor better maybe. Because of that you also can tune the formula. Ath the moment I use this one:

Code: Select all

Points = Int(100000 * (Song.PlayCounter-2*Skip)^3 / (DateDiff("d", Song.DateAdded, Now) + 1) / 1.01^(DateDiff("d", Song.LastPlayed, Now)+1))

Unfortunately I don't have time today because I have to organize a little party for tonight. But tomorrow I'll post what the formula exatly means and how you can tune it for your needs.

Regards
Big_Berny

PS: Sorry for don't answering quicklier but I didn't see your post! Please write me a PM when I don't react on your questions.

[junglemonkey]

that is just what diddeleedoo's autorate script does.
by history count of the number of times played total not the weekly count that can be seen in the list view.

but i just can not get my big berny's script ? just can't see something in it?

but diddeleedoo's i got nailed down good.
which just confirms big berny's point of other scripts for other folks.

IMHO i think diddeleedoo's is better. but to each his own.

roving cowboy / keith hall.

[Spazz]

One teensy request. Could you add an option to auto rate all unplayed and unskipped songs at a rating of your choice? I ask this because I use the Radiofreemonkey and it boosts new songs above 1.5 stars to play more often. However the autorating seems to like putting unrated/unplayed songs to .5 stars. I think if you could set it to whatever you want a base rating to be would get your songs rated more efficiently. Oh, and that script doesn't like unrated songs either :/

[Big_Berny]

@MarS,
The formula I posted means that it calculates the Points (value) of a song like this: PlayCounter*PlayCounter*PlayedPerDay (with some small modifications). Because of this new songs which you listened once will have high ratings for the first days. Although it will quickly decrease if you don't listen to it again.

If for you just the playcounter is important you could simplay use:

Code: Select all

Points = Int(Song.PlayCounter)

of something like: (which decreases the weight of PlayedPerDay)

Code: Select all

Points = Int(100 * (Song.PlayCounter-2*Skip)^5 / (DateDiff("d", Song.DateAdded, Now) + 1) / 1.01^(DateDiff("d", Song.LastPlayed, Now)+1))

If you need some help to modify the script let me know! (Hint: You just have to change the formula in the AutoRateAcc.vbs and Auto/AutoRateAccStarter.vbs)

@spazz:
Thank for that request. I'll add it when I have time, shouldn't be too difficult. Unfortunately I've much to do for the university now but in about 6 weeks I have some free time.

Regards
Big_Berny

[Spazz]

That's alright. I'm in college too so I know how it goes.

[TJOHO]

Thanks for the tips on how to modify the ratings (feature request: putting these modification options in the option page, maybe as sliders)!
I'll definitely put less weight on PlayedPerDay.

I'm wondering, though, if it is possible to make the base rating lower? I just listened to a song for the first time that was added to the library two months ago.
It is autorated at four, while something like two seems more suitable.

[Big_Berny]

Hi TJOHO

Thanks for the tips on how to modify the ratings (feature request: putting these modification options in the option page, maybe as sliders)!

I already wanted to to implement a optionfield where you can modify the formula more easily. Unfortunately it's not to easy to keep it simple.
The idea with the slider sounds really intersting to me. This way you only can change the weight of PlayedPerDay and not the whole formula but it's much simpler. I'll have a look when I have time again!

I'm wondering, though, if it is possible to make the base rating lower? I just listened to a song for the first time that was added to the library two months ago.
It is autorated at four, while something like two seems more suitable.

Well the script doesn't has a "base rating" or something like that at the moment. It uses the same formula (condition) for all songs. So my advice to you: Decrease the weight of PlayedPerDay.

This formula should be better for you:

Code: Select all

Points = Int(100 * (Song.PlayCounter-2*Skip)^5 / (DateDiff("d", Song.DateAdded, Now) + 1) / 1.01^(DateDiff("d", Song.LastPlayed, Now)+1))

By the way: When I started this script I asked you (MediaMonkey-forum-visitors) to help me tweaking the formula. Unfortunately nobody did.
But as I said one of the favourite parts of my script is the AutoCalibration which allows you to change the formula very fast.

Big_Berny