Statistics?

[Gramphos]

Mark, you seem to be living to give us all kinds of statistics .
I'd be interested in knowing how many users start the week on each day, and a updated style statistic would be god too.

[Dauphin]

Graphs!
Charts!
Lists!

NOW! NOW! NOW!



Its sad I know, but I once considered making a chart of posts vs poster "rank" number (per the top 20... list). I wanted to see if there was an exponential decay curve.

[Gramphos]

Originally posted by Big Crunch
Its sad I know, but I once considered making a chart of posts vs poster "rank" number (per the top 20... list). I wanted to see if there was an exponential decay curve.

Sorry to disappoint you, but the best curve I could get (using top 200) was the one attached (I tried different curve-types.), and if it would have been true Ming would have had 22491 posts

[MarkG]

Re: Statistics?

Originally posted by Gramphos
I'd be interested in knowing how many users start the week on each day

didnt understand this....

and a updated style statistic would be god too.

you mean something like this?
<table width=100%><tr><th colspan="3">Styles Stats</td></tr><tr><td>Default</td><td>2871</td><td>87.58%</td></tr><tr><td>Default-Civ3</td><td>108</td><td>3.29%</td></tr><tr><td>Default-CTP2</td><td>51</td><td>1.56%</td></tr><tr><td>Default-Special</td><td>20</td><td>0.61%</td></tr><tr><td>Default-Special-Civ3</td><td>17</td><td>0.52%</td></tr><tr><td>Default-Special-CTP2</td><td>2</td><td>0.06%</td></tr><tr><td>Light</td><td>143</td><td>4.36%</td></tr><tr><td>Light-Civ3</td><td>25</td><td>0.76%</td></tr><tr><td>Light-CTP2</td><td>13</td><td>0.40%</td></tr><tr><td>Light-Special</td><td>8</td><td>0.24%</td></tr><tr><td>Light-Special-Civ3</td><td>15</td><td>0.46%</td></tr><tr><td>Light-Special-CTP2</td><td>5</td><td>0.15%</td></tr><tr><th>All Default</th><th>3069</th><th>93.62%</th></tr><tr><th>All Light</th><th>209</th><th>6.38%</th></tr><tr><th>All Default-Special</th><th>39</th><th>1.19%</th></tr><tr><th>All Light-Special</th><th>28</th><th>0.85%</th></tr></table>

[Zopperoni]

Wow...I thought a lot more people used Light-style...

[Gramphos]

Re: Re: Statistics?

Originally posted by MarkG
didnt understand this....

I wanted to know how many that starts the week on Mondays, and how many that starts the week on Sundays, and any other possible day.

you mean something like this?
&lt;...table...&gt;

Yes

[Dauphin]

Originally posted by Gramphos

Sorry to disappoint you, but the best curve I could get (using top 200) was the one attached (I tried different curve-types.),

You actually did this before I suggested it.

and if it would have been true Ming would have had 22491 posts

We have statistical proof that Ming does not post enough!

[Zopperoni]

Re: Re: Re: Statistics?

Originally posted by Gramphos
I wanted to know how many that starts the week on Mondays, and how many that starts the week on Sundays, and any other possible day.

...that would be an interesting statistic indeed...

[Gramphos]

From the graph we can also see that there are to many deities. all those dots in a very small bound.

[KrazyHorse]

That's not a stinking exponential decay curve!

You should have fitted to Ae^(Bx) (or AB^x) instead of Ax^B (A&B are the fitted constants).

[Gramphos]

Originally posted by KrazyHorse
That's not a stinking exponential decay curve!

You should have fitted to Ae^(Bx) (or AB^x) instead of Ax^B (A&B are the fitted constants).

That curve only hit a few posters in the midle.

[Dauphin]

Originally posted by KrazyHorse
That's not a stinking exponential decay curve!

You should have fitted to Ae^(Bx) (or AB^x) instead of Ax^B (A&B are the fitted constants).

Its going to be skewed anyway. The fact that we are ordering the posts in some way means it is not a simple decay. If you take a statistical distribution in a histogram it would be more revealing.

Perhaps "poster density" = An^xyⁿ

or something of that ilk, with N being the number of posts.