JJ2 Level Trivia
 

I've written a simple script to get these stats out of a level:

FileName, date, LevelName, Tileset, Music, Next Level, Layer 4 Width, Layer 4 Height

Now that I have this information for my 1.23 level cache it's time to play a game! Guess the overall statistics on the following:

1. What are the 5 most common tilesets?
2. What are the 5 most common music files?
3. What year has the most number of levels?
4. What month has the most number of levels?
5. What's the most common level size (x,y)?
6. What's the most common level name?
7. What's the smallest level area?
8. What's the largest level area?

Note that I'm getting statistics from my 1.23 jj2 level cache. Here's an old version: http://www.jazz2online.com/jcf/showthread.php?t=17060

Results will be posted in a day or two so make your predictions soon! And if anyone wants the code for getting raw levels data let me know and I can give it to you.

3. 2004
4. December
6. Battle Game

1.
Castle
Diamondus
Tubelectric
Medivo
Carrtous
2. I haven't heard a song so many times to consider it common, except for the Fooruman theme
3. hm. 1999
4. December
5. 256x64

Not really sure on 6-8

1) Castle, Top secret 3, Diamondus, Carrotus, Cool Day
2) Haunted.it, Tubelectric, Credits, (-), Medivo
3) 2001 (didn't you lose your jj2 cache at about this spot in time?)
4) August
5) 256x64
6) Battle game
7) ugh... 12 x 12?
8) 1024 x 1024

Carrotus1, Castle1, Cheq, Diamondus1, Mez01
Castle, Dang, Fastrack, Medivo, Order
2001
April
256x64
Battle Game
3x3
700x700

is "" counted as a music file?

Could you place the script in this thread? :)

More interesting than the stats above however, I downloaded all of the files that were ever uploaded to J2C, UJ, or J2O. I then removed all duplicates (same MD5 hash).

The result is 9257 j2l files.

Fun stats:

[IMG]http://www.jazz2online.com/junk/bobby/levelstats/yearmonth.PNG[/IMG]
[IMG]http://www.jazz2online.com/junk/bobby/levelstats/tileset.PNG[/IMG]
[IMG]http://www.jazz2online.com/junk/bobby/levelstats/music.PNG[/IMG]
[IMG]http://www.jazz2online.com/junk/bobby/levelstats/levelname.PNG[/IMG]
[IMG]http://www.jazz2online.com/junk/bobby/levelstats/nextlevel.PNG[/IMG]

And if you are curious about other stats just ask and I can probably figure it out.

An Excel file with all of the data.

Code I used to get this data.
Notes: The code I used to get the data is very rough, I wrote it quickly, may have bugs and is very ugly. You will need to change variables in the main method to point to the directories that you want to search.
To run:
javac *.java
java JJ2LevelStats 1> results.csv 2> errors.txt

And if you have any questions about the code/need more info on how to run it I'd be happy to answer them, I doubt anyone other than Stijn will care though. :)

wow, nice! i'm surprised to see tube is #1, i hate that tileset 8(
and lol @ CC ragon's levelnames :P

still lolling: at the German levelnames in Bob's list .DDD

That's most interesting.

Campaigns1.j2t... 102? That's insane.

2009?

You have to take the dates on the levels with a grain of salt because if people set the computer clock to the wrong date then that's when the level was made. I can assure you that no level was really made in 2009 yet. :)

And I agree, Violet. For a tileset I've hardly heard of I was shocked it see it so high.

Whahah I made the smallest level =D

In 3/2003?

A few observations (and questions):

1. Where's Top Secret 3 in the list of tileset frequency?
2. Looking at the trend of level production, one can note a major boom of levels made in 2002, a year after Jazz 2 Online first opened.
3. While we are heading in a general downward slump of level production, 2007 was definitely more busy than some other years. Everybody stop nay-saying and start making more levels.
4. If possible, could you compare the frequency of tileset use against the size of the tileset? I wonder if tileset flexibility and eyecandy (and thus a larger tileset) is correlated to use.

Interesting.

#14.
Glancing... a lot of frequency of course comes from the tileset being official, Mez01 and top3 being the only ones to really make an impact on the higher levels, as well as having a straight-on perspective. Diamondus is the most 3D of any of the frequently-used tilesets, and then Jungle, but for the most part they're very blockish.

What is Campaigns1.j2t, exactly?

This is very cool...I did something like this back in 2005, but without scripts and gave up 70% of the way through. I got fairly similar results.

Bobby, what was your method of downloading all those levels?

I mass-downloaded all of the levels that were linked on J2C. For the UJ and J2O levels I used the upload directories on J2O. It would be a little harder for you to get all of the J2O levels like I did because you don't have FTP access.

all I wonder is how big is your Jazz2 directory now

I think I saw something about "some valuable gigas" on the "Did You Know".

I wonder if this is anyhow 'classified information'
 

Bear in mind that a part of that 1.43GB is pictures, voice clips, all the HTML/CSS/PHP/etc, and screenshots as of recently. If someone could check the size all .zip files from the /downloads/ dir it would return a more accurate result.

How old is this unknown post of unaccessible forum?

March 10th, 2007, 03:42 AM.

I downloaded about 1600MB of zip files for the download pages. But keep in mind that there will be some duplicates/older versions and perhaps some deleted files.

I'd estimate about 1420 MB or something of that sort.

By the way, note that Bobby's stats are case sensitive, so "Tube1.j2t" is regarded as another tileset than "tube1.j2t". Disregarding the case, stats would be like this:

Most used tilesets

1. tube.j2t - 311
2. mez01.j2t - 306
3. carrot1.j2t - 301
4. diam1.j2t - 254
5. castle1.j2t - 240

Most used music

1. diamond - 339
2. tubelec - 300
3. castle - 295
4. carrotus - 278
5. medivo2 - 250

Level name

1. untitled - 113
2. the end - 35
3. battle game - 29
4. |h|o|l|i|d||a|y| h|a|r|e| |98 - 26
5. battle level - 22

If you look at the stats you'll see that it should be identical (or very very close to that). I *think* that Excel does string comparison as case-insensitive by default.

Well, the numbers are quite different at least...

I think you're looking at the numbers from my JJ2 cache, not the ones from overall J2O that I posted in images. (But let's get back to the real topic.)