It does matter what we do here?

UsersYes it matters. And this just because we are developers and we really try to improve our tool of choice.

Because some of you are asking about the polls if they are valid. The short answer is: Yes.

I humbly dare to say that here, with your help, we have more votes per issue than in any QC report ever And this implies, of course, that we have more votes per issue than at delphi.uservoice.com and of course more votes per issue than in any Delphi StackOverflow.com issue. …and when you think that in QC and at uservoice.com one developer can cast more than one vote then the things take other dimensions.

And then why I don’t publish the number of visits? It is a basic law of polls. Did you ever see in electoral polls published the actual number of voters? πŸ™‚ No. However, usually these polls are quite accurate, isn’t it? Well, I don’t want to enter now in details (anyway we’re in a Delphi blog) but perhaps is better to know that we have here a more representative sample (compared to the size of Delphi community) than any other electoral sample which a specialized company can afford (due of obvious reasons – the population of a country normally is way bigger than Delphi community).

But let share you some data about you πŸ™‚ (courtesy to tracker.icerocket.com which I have here):

Geographical distribution

Guys, you are (at the time of writing) from at least 107 countries and teritories! Yes, I wrote one hundred seven. The following list has 109 (nine) items, the 7th is (unknown). Also we have at least one Satellite Provider. See the list (sorted by rank) and the traditional pie at the end, this time snipped from the IceRocket’s Control Panel. Did I miss a country? Please let me know.

Countries
1 United States
2 Germany
3 United Kingdom
4 Netherlands
5 Italy
6 Russian Federation
7 (unknown)
8 Canada
9 Brazil
10 Spain
11 France
12 Australia
13 Poland
14 Ukraine
15 Sweden
16 Slovenia
17 Greece
18 Turkey
19 Belgium
20 Denmark
21 New Zealand
22 South Africa
23 Czech Republic
24 Japan
25 Norway
26 Hungary
27 Finland
28 Switzerland
29 Mexico
30 Indonesia
31 Bulgaria
32 Austria
33 Iran, Islamic Republic of
34 Argentina
35 Romania
36 Portugal
37 Croatia
38 Thailand
39 Malaysia
40 Korea, Republic of
41 Satellite Provider
42 Serbia
43 Taiwan
44 Colombia
45 Slovakia
46 Paraguay
47 Singapore
48 Belarus
49 Israel
50 Latvia
51 Hong Kong
52 India
53 Lithuania
54 Bosnia and Herzegovina
55 Vietnam
56 Syrian Arab Republic
57 Iceland
58 Estonia
59 Chile
60 Philippines
61 Kenya
62 Ireland
63 Saudi Arabia
64 Cyprus
65 Venezuela
66 Guatemala
67 Bolivia
68 Luxembourg
69 Kazakstan
70 Egypt
71 Albania
72 Moldova, Republic of
73 Pakistan
74 Macedonia
75 China
76 Bangladesh
77 Honduras
78 Puerto Rico
79 United Arab Emirates
80 Tunisia
81 Kyrgyzstan
82 Dominican Republic
83 Antigua and Barbuda
84 Reunion
85 Macau
86 Algeria
87 Mali
88 Malta
89 Azerbaijan
90 Costa Rica
91 Palestinian Territory
92 Uruguay
93 Cuba
94 Peru
95 Morocco
96 Andorra
97 Europe
98 Jamaica
99 Georgia
100 Faroe Islands
101 Guadeloupe
102 Benin
103 Jordan
104 El Salvador
105 Cote D’Ivoire
106 Asia/Pacific Region
107 Uganda
108 Ecuador
109 Nepal

Browsers

Browsers

…and

Operating Systems

Windows
Linux
Macintosh
(unknown)
iPhone
SymbianOS
iPod

…in the above order with Windows accounting 91% from the total.

But of course what the tracker cannot find is how old are you…

…and what Delphi version (SKU) do you use:

Vote. Developers matter.

18 thoughts on “It does matter what we do here?

  1. When one clinks a link in DelphiFeeds.com (highly recommended), the page is still wrapped by DelphiFeeds. Does this mess up you geolocation statistics?

  2. Thanks for mentioning tracker.icerocket.com so I can block it with Opera πŸ™‚
    As for polls here in Europe we say how many people voted.
    By the way nice pies

  3. I have to say I don’t know where you got the idea that the number of voters/people polled is routinely kept secret. I can’t think of any vote/poll where the participation level was not a part of the poll statistics in one form or another, other than in countries where the validity of the democratic process is, shall we say, somewhat questionable.

    In elections it’s called the “turn out”, in opinion polls it’s the “sample size”.

  4. Also, you forget that one of the things that gives election trackers an ability to derive statistically meaningful results from a small sample size is that the questions usually have two answers that are mutually exclusive. Thus, the standard deviation of the answers in the “true” underlying population is quite small. In most of you questions there are at least 4-5 possible answers. Unless you show your numbers it is impossible to determine the significance of any poll you have published.

  5. >And then why I don’t publish the number of visits? It is a basic law of polls.
    Sorry, that’s nonsense. Not showing the sample size is a sign of bad statistics or bad science.

    And it’s a basic fact of statistics that you’ll have to show the sample size (N). Means are not good enough, neither are means+ standard deviation, if you don’t give N. Potentially, I’d let you off with mean + standard error of the mean.

  6. All,

    We have votes in the range 480-520 / question with two questions at 850 and 1320 votes respectively. Note that I mention ‘votes’ not ‘choices’. So at the polls with multiple choices were more choices cast.

    It is quite obvious that it depends a lot about what is the question asked. And I cannot force anyone to answer. Hence there is no ‘sample size’. It depends of how important one thinks that the question is.

    On the other hand, we don’t have a flat community here. You (the visitors) are usually much more influent programmers than the ones who aren’t connected at all to the Delphi community. So, your vote weights much more compared with other’s in shaping the Delphi’s future on all planes: what features will be in, influence in buying decisions etc.

    That’s why I hesitated to publish the so called ‘sample size’. But if we look at the dry numbers we are in a very good shape compared to this.

    Also, as I said, just a part from the total number of visitors vote. Perhaps is better to have more guys casting their votes?

  7. Interesting results.

    I’ve got an idea for the next pool: “Do you think that a free version of Delphi is necessary for growth of the community?” or something like that πŸ˜‰

    Or: “Name two things that in your opinion will keep Delphi alive”.

    • hehehe… (Un)fortunately the results are somewhat predictable.
      For β€œName two things that in your opinion will keep Delphi alive”. I’d would opt for
      – spread the word
      – spread the product (lower the prices, give discounts etc.)

  8. “We have votes in the range 480-520 / question with two questions at 850 and 1320 votes respectively.”

    There.

    That wasn’t so difficult, was it?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s