Author Topic: Question regarding stats  (Read 6230 times)

TRex

  • Eulexic
  • ***
  • Posts: 2038
  • ~50 miles from Chicago, in the Corn (maize) Belt
    • View Profile
Question regarding stats
« on: June 21, 2009, 12:35:15 AM »
If the information is not too much trouble to find, Alan, I'm curious about the following:
- how many common words are there?
- how many allowable words are there?
- how many nine-letter words are there?
- what is the typical percentage of users (registered or all) by country of origin?
- what is the typical percentage of users (registered or all) by browser type?
- what is the typical percentage of users (registered or all) by operating system?
- what is the typical percentage of users (registered or all) by screen resolution?

Just insatiably curious.

pat

  • Eulexic
  • ***
  • Posts: 3376
  • Rugby, England.
    • View Profile
Re: Question regarding stats
« Reply #1 on: June 21, 2009, 03:38:33 AM »
There's something I'm curious about too, Alan. You mentioned quite recently that there have been some puzzles whose 9-letter words no one got. What were some of them?

Alan W

  • Administrator
  • Eulexic
  • *****
  • Posts: 4961
  • Melbourne, Australia
    • View Profile
    • Email
Re: Question regarding stats
« Reply #2 on: June 22, 2009, 11:10:40 PM »
Yes, it seems you are intensely curious, TRex. Just lucky for you that I'm quite fascinated by statistics myself.

There are 34,706 common words, and 106,945 words in total. A lot of these words will probably never be part of a puzzle, simply because there is no suitable nine-letter word containing all their letters.

There are 5,210 nine-letter words that I consider to be familiar enough to generate puzzles from. However, some of these will never be used because they would allow too few words in a puzzle. (And some will only ever be used in "Your puzzles" because they make too many words for a daily puzzle.)

The following figures are for all "hits" on the site. Because of the way Chihuahua is implemented, each word played generates a hit, so the people who spend more time on the site are more strongly represented in the figures, although I doubt it would make much of a difference.

Browsers -

 IE6 24%
 IE7 35%
 IE8 10%
 Firefox 24%
 Safari 4%
 Opera 1.9%
 Google Chrome 1.5%

Operating systems -

 XP 64%
 Vista 23%
 Windows 2000 3%
 Mac OS X 8%
 Linux 0.7%

Countries -

See attached chart. Right-click on it and select "View image" to see it full size. (The Web statistics package I use infers these from IP addresses, which I believe is not totally reliable)

I do have some figures on screen resolution, which I'll dig out for a later post.

Pat, I'll answer your question in a later post,too, but if I said there were puzzles where nobody got the nine-letter word, I'm a fool. But I think what I might have said was that there were a few puzzles where nobody got a rosette. But I think that might be incorrect too - I think I'd been wrongly analysing the records. Anyhow, I'll come back to this question shortly.
« Last Edit: June 22, 2009, 11:35:34 PM by Alan W »
Alan Walker
Creator of Lexigame websites

pat

  • Eulexic
  • ***
  • Posts: 3376
  • Rugby, England.
    • View Profile
Re: Question regarding stats
« Reply #3 on: June 23, 2009, 12:51:48 AM »
Alan, you're right and I'm the fool. You did indeed say no one got a rosette, not no one got the 9-letter word. Silly me.  :D

Alan W

  • Administrator
  • Eulexic
  • *****
  • Posts: 4961
  • Melbourne, Australia
    • View Profile
    • Email
Re: Question regarding stats
« Reply #4 on: June 23, 2009, 11:22:08 AM »
As a footnote to the chart showing where people are from, and to reinforce my point about the possible unreliability of these figures, I hadn't noticed until I was writing my post that Australia is now shown as the biggest source of visitors to the Chihuahua site. Until a couple of months ago, it was always the US that appeared on top of the list.

In fact, the statistics show a dramatic shift in the national origins of Chi players, which I can't believe is genuine. It would appear that the US patronage fell by half from January to May, while Australian presence more than tripled! meanwhile, UK activity supposedly went up about sevenfold.

I'll look into this some more, but in the meantime I wouldn't make too much of any of those statistics.
Alan Walker
Creator of Lexigame websites

TRex

  • Eulexic
  • ***
  • Posts: 2038
  • ~50 miles from Chicago, in the Corn (maize) Belt
    • View Profile
Re: Question regarding stats
« Reply #5 on: June 24, 2009, 08:20:05 AM »
Thank you, Alan.

Yes, it seems you are intensely curious, TRex. Just lucky for you that I'm quite fascinated by statistics myself.
;D
FWIW, the MBTI type INTP (which I am) is supposed to be really into statistics as well as word games. Perhaps you're an INTP as well. ???

Since the browser and OS numbers are fairly close to total 'Net stats, the players here must not be too unusual. (Except maybe the hag!)   >:D

Linda

  • WordStar
  • ****
  • Posts: 7055
  • Cumbria, England
    • View Profile
Re: Question regarding stats
« Reply #6 on: June 24, 2009, 09:44:24 PM »
I truly have no idea what you are talking about, TRex, with your INTPs and MBT1 whatsits but the very mention of the word hag made me take notice!  There is more than one hag on the Forum, TRex, and a formidable bunch we make, too!!  >:D

TRex

  • Eulexic
  • ***
  • Posts: 2038
  • ~50 miles from Chicago, in the Corn (maize) Belt
    • View Profile
Re: Question regarding stats
« Reply #7 on: June 24, 2009, 09:59:03 PM »
I figured 'the hag' would elicit a reaction!  :laugh:

TRex

  • Eulexic
  • ***
  • Posts: 2038
  • ~50 miles from Chicago, in the Corn (maize) Belt
    • View Profile
Re: Question regarding stats
« Reply #8 on: June 24, 2009, 10:01:54 PM »
If anyone is interested, a Google search on MBTI (all letters) got about 1ΒΌ million hits. So if anyone wants to learn about it, there's plenty of material available.

Tom44

  • Paronomaniac
  • ******
  • Posts: 462
  • Pyrotechnics Live
    • View Profile
Re: Question regarding stats
« Reply #9 on: June 24, 2009, 11:37:18 PM »
Also, FWIW, as a psychologist I don't put much faith in the Myers-Briggs test even though its used around here a lot.  Any and every time you try to pigeonhole people you wind up with more mismatches than true descriptions.
Stevens Point, WI

birdy

  • Eulexic
  • ***
  • Posts: 3370
  • Brooklyn, NY
    • View Profile
Re: Question regarding stats
« Reply #10 on: June 25, 2009, 12:22:51 AM »
As one of those people who can't resist taking every personality test ever printed in a magazine (even Cosmopolitan on the rare occasions I see it), of course I had to take the abbreviated version of the MBTI test when I googled the term.

I can see why you don't particularly like it, Tom44.  I thought the choices were too arbitrary - there would be something I'd identify with in each opposing group.  Now if they listed all those choices in one column and asked me to pick the ones I would recognize in myself, and then sort it out, maybe it would have been more accurate.  Or maybe not, since in different situations I would choose diametrically opposing positions. 

In a training course at my job, we were asked to answer personality questions that would define our management style.  Many people seemed to have a very clearly defined style (which probably varied from laissez-faire to Stalinistic control freak) - I ended up with a mishmash that didn't fall into any of the choices.  Not sure what that said about me or the test!

TRex

  • Eulexic
  • ***
  • Posts: 2038
  • ~50 miles from Chicago, in the Corn (maize) Belt
    • View Profile
Re: Question regarding stats
« Reply #11 on: June 25, 2009, 01:08:55 AM »
Also, FWIW, as a psychologist I don't put much faith in the Myers-Briggs test even though its used around here a lot.  Any and every time you try to pigeonhole people you wind up with more mismatches than true descriptions.
Absolutely true. At best, IMO, it can offer some understanding into tendencies. Clearly, there are differences between introverts and extroverts and I don't have a problem with the idea that there are different axes which are important parts of a personality. But, like the 'choice' 'is it better to be just or compassionate', the test seems assume everyone is at one end of each of four axes. I suspect almost everyone is somewhere between the extremes for each axis.

ensiform

  • Paronomaniac
  • ******
  • Posts: 459
    • View Profile
Re: Question regarding stats
« Reply #12 on: June 25, 2009, 03:40:08 AM »
I know very little about the test, but it seems to me that question would be better phrased "Is it better to err on the side of justice or the side of compassion?" or some such, because as it is is almost implies "Is it better to be concerned with justice at the expense of compassion or vice versa?" which is, as you say, a false dichotomy.