h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Being quite new here, I still wonder ...
Newcomers have numbers around 130 000 right now, assuming that means so far 130 000 people have registered with H2G2 until now, further assuming that only 10 % "have come to stay", the others having lost interest very quickly. That makes 13 000 users. Further assuming that in order to be called a "regular user" you would have to spend around 20 min per week logged in to H2G2 (most users more I guess, but less than that wouldn't really make sense, so it should be a good, if careful average). 20 min are the 504th part of the week, so if you divide 13 000 / 500, you have 26 as the average number of participants online at any given time.
Oooops ..... I must have made a mistake ..... obviously, 42 is the correct answer.
However, there seem to be between 26 and 42 participants online at any given time, so both calculations and estimations seem to be not all bad. Still, they seem to me very careful ... a kind of worst case scenario, so to say. Anybody else who has thought about this and made calculations of their own?

Being even newer here I wondered the same, but I did not make such calculations (though generally I'm very likely to) but some observations: About 20% of the users onl is "new today" (12 of 59 right now) and at least another fifth is formed by people I already saw yesterday, so they seem to be (nearly) permanently online.

On the one hand this would explain why your calculation was off reality a bit but on the other hand it's rather sad because it would prove that quite a lot of people have lost their interest in h2g2!

well, there might be another reason: h2g2 is so time-consuming but interesting a lot of people maybe decide not to take the risk to check regularly
(btw 20 min a week is quite little!)

cya cyy!

well, the calculations were done both in a rough and in a cautious way. Of the new-today-crowd I'd expect 50 percent of them not coming back at all and in the long time only perhaps 10 percent to stay as regular participants (thus coming to 10 percent of 130 000). Concerning those 20 minutes, it was not an overall approximation of the mean time of all regular participants, but rather a definition of the minimum time a week any user has to spend online in order to be called a regular user. Less that that doesn't make sense. But quite a lot of the regulars will spend some 4 hours (some more, some less) online each day, making it 28 hours a week or 84 times these 20 minutes.

So, when you put a number of these "heavy users" into consideration it becomes obvious that for each of them the number of regular users is reduced by 83. So if you have 50 heavy users, you have 4150 less than the approximated 13 000, if you count 100 heavy users, you are left with 13 000 - 8300 = 4700 users being online the required minimum of 20 minutes. Minus the ones who are new today, but never come back, accounting perhaps for 50 (new users a day) * 7 (days in a week) = 350 (20 minute-shifts) you still have to subtract. Making it less than 4300.

Still, on the other hand I have noticed that very often there are more than 60 people online, so that again almost doubles the numbers.

When you do rough calculations such as this you must be aware that you will make a lot of mistakes because you have not all the details. On the other hand, I was really wondering how come that we have 130 000 participants and only 30 or 60 online at any given moment and the estimations I did just show that it is very much possible. Having a result of 13 000 users online each 20 min a week I was well aware of the fact that the real number was even less, because many are online quite a lot. With the above reasoning I have reduced them to 4250, but then with twice as much on the "who's online list" that makes 8500 again, which you can reduce with any number of people being online 2 hours each day and so on.

What I mean to say is that without knowing the details you won't find out the true number. 13 000 users is still much too high, I am aware of that and guessing from the list I'd say it's probable less than 1000 who are regulars. But that is an different calculation entirely.

Wir könnten ja eine Statistik erstellen um die Binomialverteilung aufzustellen und zu überprüfen wie genau unsere Schätzungen sind. Binomialverteilung müßt ich eh gerade lernen aber Statistiken mach ich jetzt lieber sinnvollere

(msg auf D weil ich keine Ahnung hab was Binomial- (od sonst irgend eine) Verteilung auf E heißt - weißt Du's?

auf jeden Fall sind zu viele Leute da wenn man viel zu tun hat und zu wenig die die Artikel schreiben, die man gerade sucht

cya,

Menschenfresser

Binominalverteilung? ..... hab ich vergessen, was das ist und Übersetzungen solcher Ausdrücke übersteigen meine Fähigkeiten bei weitem, ich bin froh, wenn ich den englischen Ausdruck höre und was damit anfangen kann.

Wozu brauchst du dnn die Binominalverteilung? Gab's da nicht einen Binominalkoeffizienten .... ich glaub ich muss das doch mal irgendwo gehabt haben ...hmmm. Was studierst du denn?

Und was für Artikel suchst du?

In 2 Wochen kann ich Dir (hoffentlich!) genau erklären was Binomialverteilung, -koeffizent und der gesamte Rest an mathematischem Grundwissen ist, weil ich noch gar nicht studiere

Ich hab einmal in 2 Wochen Matura

Und ich mach nix anderes als HHGG lesen und hab noch nix getan - äh, never mind

also das ganze brauch ich für meine Matura - und ich suche einen Artikel??

ciao
mf

Du suchst einen Artikel? Möchtest du lieber ein "der" oder ein "die"?

Matura hast du? Das klingt aber verdammt nach Österreich. Naja, ich drück die Daumen

Yup, ist es auch!


*lol* und DU meintest ja ich suche einen Artikel

thx
muß morgen (heute ;o) das h2g2 Spezialgebiet abgeben... wird noch länger die nacht *g*

baba!

wie? du machst h2g2 für die schule

und wir haben Jane Austen lesen müssen ((( Pride and Prejudice .... ich kann dir sagen .....puuhhh, das ist echt heavy stuff