Lies, Damned Lies, and Statistics

When I first started Doom Underground, I knew that since I was keeping the information very organised and doing things like generating indices automatically, one really cool thing I could do was generate some statistics on the levels reviewed. At the time I had other priorities, like actually getting the site launched, so the idea got shelved... until now. So here it is, automatic statistics on the WADs reviewed on DU.

Before anyone thinks about drawing any conclusions from this data about Doom WADs and editing in general, I should point out that:

• With only around 250 WADs catalogued here, this isn't a large enough sample to draw any strong conclusions about the wider body of Doom WADs.
• This is absolutely not a random sample - it's based on stuff I've reviewed, which is heavily skewed towards Boom levels, levels from authors I know, and classic levels. So there's no way it is random enough to be considered representative of Doom WADs in general.

Having said all that, I consider the results interesting, in particular the comparison of Doom editors and tools further down.

Wads by Year

The year in which the WADs were written. The heavy bias towards '98 reflects the year when I was most active at reviewing, and the year in which most Boom WADs were made.

Year	Wads	= Doom, = Doom 2 = Ultimate Doom
1993	0
1994	47
1995	35
1996	35
1997	24
1998	78
1999	65
2000	9
2001	8
2002	6
2003	4
2004	4

Tools

I've religiously recorded the tools used to make every WAD that has been reviewed, or at least as well as I can using the information provided in the text files. I've listed the editors first, then the other tools. I've only listed the most popular tools over the archive I have, and since I review a lot of old levels those tend to be the classics like DEU and BSP.

Tool	Wads	= Doom, = Doom 2 = Ultimate Doom
DCK	60
DeeP	37
DETH	43
DEU	69
DoomCAD	21
DoomEd	11
EdMap	11
WadAuthor	23
WadEd	31
Doom Builder	3
BSP	86
DEUTex	8
NWT	34
RMB	45
WARM	46
WinTex	76
ZenNode	14

I should also point out some limitations of the data itself:

• The counts are of WADs, not levels, so it's biased against episodes and megawads. Unavoidable, as the text file rarely gives a detailed breakdown of which tool was used where.
• Nor do they give an indication of how much a particular tool was used - a megawad might have only used an editor for one level, and another editor for the other 31, but we count them equal. Not enough information to do otherwise.
• This all depends on the author accurately reporting what tools they used. Some authors and projects, who shall remain nameles, consider themselves above the idgames WAD authoring template and don't provide that information. Bastards.