Sunday, May 23, 2010
You probably have hundreds of passwords by now, for Facebook, Twitter, your social club, bank and who knows what else. You are told to use secure passwords for all of these, and you are also told to never use the same password on more than one site. Well, we all know this is just impossible to achieve.
I just started up a website http://www.passwordsquares.com/ to assist in solving this problem. If you think you can help me out, contact me.
Saturday, May 20, 2006
Whirlpool.net.au - The Statistics of Aura
Well, I finally decided - mainly because I like statistics and data analysis as well as studying language and human interaction to do some analysis of the new Aura system on the Broadband community site.
I also figured, that it would be my chance to show how screwed the system was.
Well in a word... I was wrong, the system is far from all pervasive and you really need to make an impact to alter it. Whilst doing the analysis, I came up with some interesting observations.
Out of the sample of over 30,000 user ids, the majority of users are just Plain Regulars or Participants.
|In the penalty box||1.57%|
|I'm new here, please be nice||3.88%|
|Whirlpool Forums Addict||0.87%|
In addition, the all pervasive Aura which has brought much critisicm (including from me), is a small speck of impact on the whole. The following demonstrates how few members are affected.
Notice that 95% of Whirlpool is entirely unnaffected by Aura, and 98% are not 'adversely' affected by Aura. Not also the larger slice that 'Low' has got due to the politically correct dropping of the 'Dark' and 'Dim' Auras.
The following further divides those members with Aura by the 'strength' of the vote. It seems that whilst it is easy to annoy a few people, it's a pure talent (a talent I have it seems), to annoy enough people to actually become pervasively anything. Notice also that the easy at which you move from Tentative, through Normal to Pervasive seems somewhat independent of the aura you are on.
I thought it would also be a bit of fun to see where the boorish brutes live, and where the enlightened few live. Well, it's official, Darwin is the home of the lowest over all Aura, and the city with the highest proportion of Low aura people. Canberra has the highest overall aura, but Perth boasts the highest proportion of very enlightened persons. This table includes only persons with a valid aura (not Indeterminate).
The final bit of statistics for now debunks the long held belief that people with high post counts, generate low auras. Well, I analysed the posts/day for all the persons with a valid aura (not indeterminite), and low and behold .... it isn't how much you say, it's what you say. The values are calculated based on the number of posts divided by the number of days from joining to last activity (the best I could come up with).
The following graph (added 27th September 2006), has a little bit of a disturbing future in its lines:
The information is a little busy, but fundamentally the red lines (and red axis) are numbers of members and the green lines are average durations of a members stay (before they leave for the long run). Now obviously, as the graph gets to the end, distinguishing between members that have left for good, and those away for a month is somewhat difficult. However, the over all trend is that Whirlpool is becoming less of a community and more of a news site (like news.com etc). The red lines demonstrate this significantly, whilst more and more people are joining, more and more of them are joining for very short durations and moving on. The rest of the graph, I'll leave for your digestion.
If there is any more information, or particular stats you would like to see, e-mail me by clicking on the contact me button up top, I will be adding to this page as time goes on (there is another 10 or more graphs/tables I have in the pipeline).
Sunday, March 27, 2005
Traffic vs Data Tutorial
Traffic vs Data Tutorial
Think of the internet as a road carrying cargo, vehicles are ‘packets’ and the cargo is your ‘data’. The entire group of vehicles and packets is ‘traffic’.
Some vehicles like the semi-trailer carry large amounts of data, with a small percentage overhead, others like the small truck carry little data with a larger percentage overhead. Yet others like the police car carry no data – and are there for special purposes.
Most traffic metering systems will measure the traffic that passes through the system, others (especially proxies and mirrors) will measure the cargo. This is why sometimes the two systems can have vastly different values.