Recently I seem to have come across many ideas related to farting or guns - although rarely both in the same idea (fortunately). I had formed the impression that the HB crowd were either suffering from a collective bowel disorder syndrome, or were all 14 year old boys. I decided to do some research and discovered to my surprise that this is not in fact the case. I became interested and decided to test a number of propositions. I did so by counting the number of search returns for key words associated with my proposition. Here are my results (stop complaining - the invention is at the end!).

Proposition 1: Half Bakers are obsessed with violence.
FOR: hate (1102); murder (164); kill (868); torture (92); smash (121); hurt (484); cruel (194).
Total (3025)

AGAINST: love (2575); peace (309); calm (150); happy (2051); sweet (533); nice (3510); smile (239),
Total (9367)
Proposition 1 defeated by approx 3:1

Proposition 2: Half Bakers are obsessed with bodily evacuation functions – even more than sex!
FOR: turd (50); shit (496); arse (201); crap (698); toilet (550); piss (291); vomit (79).
Total (2365)
AGAINST: sex (558); penis (135); vagina (63); orgasm (28); coupling (42); arousal (9); foreplay (12).
Total (847) Proposition 2 confirmed by approx 3:1

Proposition 3: Half Bakers care more for guns than for drinking
FOR: gun (680); weapon (302); rifle (89); shoot (589); bullet (156); target (549); armed (215).
Total (2880)
AGAINST: beer (754); alcohol (341); wine (266); drink (912); drunk (398); booze (88); bar (935).
Total (3694) Proposition 3 – too close to call!

Finally I notice a few unusual comparisons which (I think) are quite revealing about the half bakers:

Custard (450) vs Marmite (41)
Human (1544) vs Alien (202)
Electronic (642) vs Genetic (250)
Car (2792) vs Spaceship (41)

Now before [galukalok] gets too excited – I know he already published an idea for graphing the changing quality of HB ideas over time (see link). And [UnaBubba] – hold your horses! I am aware that you published the “Useless HB Statistics Crawler” which was about gathering statistics about HB content (see link) but this is different.

Firstly I propose a psychological profiling technique which can be applied to any website which is primarily composed of user-submitted content (such as the HB). It works on the principle that the things one chooses to discuss and the vocabulary one employs can be more revealing than the meaning of what one actually says (“Its not the reporting – its the agenda which matters”). The goal of the technique is to establish a profile of the average contributor to the site.

The technique involves the preparation of a large number of carefully-selected word lists (along the lines of the ones I used to test my propositions above – only more thorough). Each word list is then applied to the target site’s search facility and the result-counts are summed. Since the word lists are chosen so that each represents a particular psychological characteristic, then a comparison of the popularity of each list should reveal the psychology of the site’s users.

Secondly, I propose an automated mechanism for applying the technique to a large number of web sites. This is so trivial I needn’t even bother with the mechanism here.

Finally I suggest a web site (the “Web Site Psychology Guide”) where the results are published. Surfers could then discover potential sites to visit or join by navigating through a hierarchical structure of psychological characteristics or could elect to undertake a conventional psychological profiling test themselves and as a result, automatically be matched with the sites where they are likely to find people they will get on with. Alternatively, parents could check out the type of people with whom their children are conversing by looking up their favourite chat rooms on the Web Site Psychology Guide.
-- dobtabulous, Nov 27 2003

Halfbakery Industrial Average http://www.halfbake...ndustrial_20Average
[galukalock]'s measure of HB idea quality [dobtabulous, Oct 04 2004, last modified Oct 05 2004]

Useless 1/2B Statistics Crawler http://www.halfbake...tatistics_20Crawler
[unabubba]'s idea for discovering HB trivia [dobtabulous, Oct 04 2004, last modified Oct 05 2004]

You mean, like this? http://www.halfbake...ychological_20types
But for any web site? [lintkeeper2, Oct 04 2004, last modified Oct 05 2004]

Vocabuladex http://www.halfbake...om/idea/Vocabuladex
related web site vocabulary idea (blatant self promotion) [krelnik, Oct 04 2004, last modified Oct 05 2004]

To paraphrase Groucho Marx, any site that has folk like me as members, I wouldn't sign up.
-- PeterSilly, Nov 27 2003


//Custard (450) vs Marmite (41)//

You should include Vegemite in this as well.
-- Detly, Nov 27 2003


[Delty] Vegemite comes in at a lowly (26) and therefore fails to alter the dynamic of the marmite/custard coefficient to any significant extent. Your point is well made though and an example of the improved thoroughness I was talking about when constructing the word lists.
-- dobtabulous, Nov 28 2003


Thanks for the link [lintkeeper2]. It’s not quite the same thing. The linked idea is about identifying types of roles on the HB and recognising how each of us performs those roles in differing proportions. I am looking at something which would assess a web site in terms of the characteristics of its users manifested in their contributed content. The analysis would be less in terms of their behavioural patterns and more in terms of the kind of things they like to talk about. For example, the roles like "ruler", "judge", “peacemaker” mentioned in the linked idea may well be found on the HB, or a site discussing gun laws, or on a site for keen fishermen but recognising this would not help me select a site at which I would feel at home. On the other hand, using my scheme, these sites would clearly be differentiated in that the HB would be seen to be populated by users preoccupied by subjects like sex, beer, toilets, custard etc and who frequently use words like "troll", "car” and “arse”. I suspect that the fisherman’s site might emerge with different dominant characteristics.
-- dobtabulous, Nov 28 2003


Positive - 419 : Negative - 470
Kirk: - 20 : Spock: - 18
Girlfriend - 210 : Sister - 192
Alpha - 89 : Omega - 20
Blood - 582 : Plasma - 101
Dog 1717 : Cat - 899
Wonderful - 599 : Pigeon - 102
Who gives a - 17 : rat's ass - 21
-- thumbwax, Nov 28 2003


how can you have a rat's ass without a - who gives a?
-- po, Nov 28 2003


[Arby] - thanks, I think that tells us all we need to know about your preoccupation!

[thumbwax] Nobody gives a rat's arse unless the word lists are chosen carefully. It's easy to find examples which reveal little, but If I noticed a site where the words "saviour", "revelation", "redemption" and "testify" outranked "britney", "madonna", "mariah", "whitney" etc by a considerable margin then I could assume that its subscribers were more interested in religion than female singers.

[po] - as in "JouLoundsbury's ideas aren't worth a..."?

[Rods Tiger] - that depends. You could start with limited objectives like trying to characterize sites based on factors like interest in food, aol-speak, legal-terminology, aggression, swearing (or whatever). If the approach proved successful you could expand your word lists. There is no obligation to assess every word, nor would words like "it" really tell you much.
-- dobtabulous, Nov 28 2003


Err...google?
-- DrBob, Nov 28 2003


Err.. is a search engine. [DrBob] - Look up the HB on Google and you can't tell anything about what it's for, who uses it, or anything about its users. You could use Google to search for a word like "God" and possibly even discover how many times that word appears on HB pages (embedded within a trillion other results) but to do the whole word-list/profiling thing would be very hard.
-- dobtabulous, Nov 28 2003


Searching for the name of a web-site wouldn't but stick the words that you're interested in into the search engine and back comes a ranked list of sites of interest. Which, although it doesn't fulfill the exact function of your idea, does fulfill it's purpose.
-- DrBob, Nov 28 2003


[DrBob] - I dispute that. If my (imaginary) 10-year old daughter frequented an (imaginary) web site called "DailyChatter" and I looked it up on the Web Site Psychology Guide to discover that its user-community were characterised as evil, sex-obsessed, aggressive, torturers and drilled down to discover that the evidence for this was the overwhelming use of words such as "satan", "anal", "barbed" (etc ... big list) how could I possibly gain the same perspective using Google? Part of the added-value of the site is in the selection of the word lists by those with expertise, and the repeated searching and summing already performed. Google might help me find sites associated with apples for example but it wouldn't differentiate a site for people interested in uploading photos of themselves inserting fruit into their bodily orifices from a site offering advice about growing top quality granny smiths.
-- dobtabulous, Nov 28 2003


Not so, dob. Next time that you're using google click on 'Advanced Search'. You can search for sites that don't have particular words as well as ones that do have them. Or a combination. So, in your example, you'd search for 'apples' but exclude sites that have, for example, 'sex' or ''fetish'.
-- DrBob, Nov 28 2003


Carefull now [UnaBubba] - you're skewing the results for the HB single-handedly! I'm sure some baker will supply the link you desire!
-- dobtabulous, Nov 28 2003


Not sure I quite understand your point [Rods Tiger]. The word lists would be large and drawn up by teams of experts which is one reason why they would add value as you would avoid the problem you describe.

As you say, it is possible that a site may contain many entries along the lines of "it is wrong to molest chimps" but would score highly against a word-list which was attempting to assess an interest in animal-molestation. However this goes back to my original point that it's not always what you say, but what you choose to talk about which is more revealing.

If you were meeting your girlfriend's parents for the first time and stated at the dinner table that "plunging your daughter's face into a vat of warm bat urine would be a bad thing" they would probably agree, but would have formed a view about you which would probably not be "my what accurate observations he makes".
-- dobtabulous, Nov 28 2003


The whole analysis is a flawed exercise in solipsistic navel-gazing. a) we do not need to have a measure of common identity with other contributors to get the most out of this or any other site; b) even if it were worth doing, your sample questions are inadequate, even inappropriate, to the task.

Fishbone.
-- DrCurry, Nov 28 2003


//The whole analysis is a flawed exercise in solipsistic navel-gazing//

Dictionary.com: "solipsism: The theory or view that the self is the only reality", "navel-gazing: Excessive introspection, self-absorption, or concentration on a single issue"

Are you suggesting this idea indicates I am obsessed with myself? Or do you mean "flawed" in the sense of "nothing to do with"? Sorry but I don't see this.
-- dobtabulous, Nov 28 2003


Agree with Rods, and agree with Curry, and would tack on "silly waste of time" also.
-- waugsqueke, Nov 28 2003


//Are you suggesting this idea indicates I am obsessed with myself?//

I read it as meaning that you were obsessed with your own navel. But then I'm not the spikiest stilletto in the wardrobe.
-- Fishrat, Nov 29 2003


In order to properly analyze the results, you'd have to compare the amount of instances those words were used by different members, not just the same ones repeatedly, and you'd have to count more than just the # of idea pages for which the word shows up, (which is the search engine numerical result shows, I believe), you'd also want to count the # of hits per page. Not difficult to program, but a huge sink of resources on the system I'd imagine.
-- RayfordSteele, Nov 30 2003


[waugsqueke] - I think your rider may be somewhat trivial - most of the internet would qualify... unless you meant comparatively. :P
-- Detly, Nov 30 2003


Thanks for the link [krelnik]. Slightly different but very relevant. Interesting that your idea got 14+ (no-) whereas mine is languishing at 0 with 7+ and 7-. I suspect people liked the straightforwardness of your vocabulary-measurement and objected to the psychobabble aspect of my assessment technique.
-- dobtabulous, Dec 01 2003



random, halfbakery