Some interesting statistics about Typophile

Top 30 thread creators in the history of Typophile (includes wiki pages):

| 490 | paul d hunt
| 378 | Miss Tiffany
| 349 | Jared Benson
| 261 | dan_reynolds
| 259 | hrant
| 228 | Joe Pemberton
| 152 | Stephen Coles
| 149 | sii
| 144 | Bald Condensed
| 137 | evan
| 130 | dezcom
| 121 | misalion
| 109 | kris
| 108 | pablohoney77
| 106 | Hildebrant
| 106 | union
| 105 | adriano
| 102 | Dan Weaver
| 100 | Miguel Sousa
| 95 | bjharvey
| 88 | julia
| 86 | Eric_West
| 85 | designalchemy
| 82 | Eben Sorkin
| 80 | signs79
| 77 | John Hudson
| 76 | James Grieshaber
| 75 | contohabc
| 72 | filip blazek
| 72 | Grant Hutchinson

Top 30 Commenters:

| 13923 | hrant
| 7852 | troubleman
| 6937 | tiffany
| 6699 | dezcom
| 5734 | coles
| 3436 | scoffer
| 3298 | pattyfab
| 3042 | tiro_j
| 2930 | billyb
| 2899 | marksim
| 2744 | GRG
| 2735 | bowfinpw
| 2526 | timd
| 2388 | sii
| 2215 | pilcrow
| 2144 | dan_reynolds
| 1943 | eben_sorkin
| 1658 | formlos
| 1505 | pemberton
| 1483 | tphinney
| 1454 | dgweaver53m
| 1441 | bjharvey
| 1309 | aluminum
| 1288 | Norbert Florendo
| 1255 | lexluthor
| 1175 | omine
| 1166 | cheshiredave
| 1150 | Linda Cunningham
| 1141 | marcox
| 1065 | grant

It's interesting how the top thread creators are different from the top commenters. Both are valuable in different ways. I'm interested to hear from you guys what you think makes for a good thread starter, and what makes for good comments. (btw I'm number 39 on the thread creator list :)


Top 30 Commenters:

| 13923 | hrant

Friggin spammer.


that's funny i'm in the creator's list twice, if you added my two totals together you get 598. I hope that's not all spam!
i wonder if i'd make the commenters' list if my two accounts were combined: maybe i don't want to know >^P (and i wonder what other folks who have 2 accounts would be on those lists?)

Interesting. Thanks for publishing these, Christian.

> I’m interested to hear from you guys what you think makes for a good thread starter, and what makes for good comments.

Hard to say. Perhaps you can tell how good/popular a thread/topic was by counting the total number of responses it got. Within that you might want to get the number of different users that were involved in the discussion. (You know, there are threads that look like ping-pong matches...)

> threads that look like ping pong matches

I think that looking at the diversity of posters is a good idea. For a while we've been thinking about ways to let some of the better threads float to the top. There is always the option of adding voting systems, etc, but it seems like queries that use intrinsic data that demonstrate interest are going to be better than prompting users for votes.

There may also be ways to look at number of reads, and how often a diversity of people return to a certain thread.

Paul: If you want I can merge your accounts, so that you get all the credit for your posts :)

For sure a pure comment count isn't a good indicator of the best threads. Case in point: the number 7 thread of all time by number of comments was the recent thread on why typophiles fight so much.

If you want I can merge your accounts, so that you get all the credit for your posts

actually, i'm not sure i want to be associated with some of my newbie postings. :P

Does this mean that Miss Tiffany and Tiffany are two separate people? I knew I felt split. :^/

I think a voting system would be good, on the one hand, but just like most everything else we discuss here it would be subjective. No?

At least my boss doesn't know about this site. :^/

The Miss Tiffany/tiffany disparity comes from how comments are stored in the database. When a comment is authored, the username is stamped in there to make future queries faster. If the user later changes their username, the comment names still show the old username, while the thread names are linked to the user database and change appropriately.

Here is a query that matches the comment count with the user ID. This list will be more accurate (and unite the schizophrenic pauls and tiffanies)

| 13933 | hrant
| 7854 | Bald Condensed
| 6945 | Miss Tiffany
| 6699 | dezcom
| 5734 | Stephen Coles
| 3438 | Nick Shinn
| 3298 | pattyfab
| 3045 | John Hudson
| 2931 | William Berkson
| 2902 | Mark Simonson
| 2744 | anonymous
| 2741 | bowfinpw
| 2526 | timd
| 2392 | sii
| 2219 | paul d hunt
| 2144 | dan_reynolds
| 1943 | Eben Sorkin
| 1658 | formlos
| 1505 | Joe Pemberton
| 1483 | Thomas Phinney
| 1454 | Dan Weaver
| 1441 | bjharvey
| 1311 | aluminum
| 1288 | Norbert Florendo
| 1255 | Lex Kominek
| 1175 | Eduardo Omine
| 1166 | cheshiredave
| 1153 | Linda Cunningham
| 1142 | marcox
| 1065 | Grant Hutchinson

Another interesting statistic: Max average comments per day for anyone who has been a member for over a week.

com/day name total comments total days of membership
5.5869 dezcom 6699 1199.05
5.4142 hrant 14034 2592.09
4.7083 Bald Condensed 7882 1674.05
3.7175 Linda Cunningham 1178 316.88
3.2295 Miss Tiffany 6963 2156.09
3.0682 litera 27 8.80
2.9928 jpad 766 255.95
2.9266 paul d hunt 2235 763.68
2.8530 pattyfab 3344 1172.10
2.5874 Stephen Coles 5734 2216.09
2.4254 Nick Shinn 3471 1431.09
1.9545 bowfinpw 2760 1412.09
1.8809 William Berkson 2940 1563.05
1.8791 John Hudson 3063 1630.05
1.8599 timd 2541 1366.17
1.7697 Norbert Florendo 1290 728.95
1.6888 Jan 159 94.15
1.6050 fontplayer 722 449.84
1.5785 Eben Sorkin 1946 1232.82
1.4507 Charlotta 89 61.35
1.4495 Mark Simonson 2918 2013.05
1.4292 anonymous 2744 1919.96
1.4263 Quincunx 381 267.13
1.3916 Lex Kominek 1266 909.74
1.3662 david hamuel 899 658.03
1.3612 Scott D 60 44.08
1.2841 sii 2419 1883.77
1.2815 cuttlefish 347 270.78
1.2457 Choz Cunningham 431 345.98
1.2265 Angel71 129 105.18

[ edit — updated statistics to reflect the fact that there are 24 hours in a day, not 12 ]

Tiffany: this should make your boss feel better if they ever find this site. 1.6 comments/day: that's easy during lunch ;-)

I feel human again. 1.6 comments per day seems healthy enough. ;^) Thanks Christian.

I think we can do much better than this, come on. These are Fox News stats.
Plus the "days of membership" can't be right, if you divide by 365.
BTW, are you counting the Lost Year?

> what you think makes for a good thread starter,
> and what makes for good comments.

One thing that might help is calculating "average comments caused per day": number of comments in threads one has started divided by number of days of membership.

BTW, I'm clearly weakening with age:
On Typo-L my record was 70 comments in one day.


BTW, more insightful than number of posts would
be number of bytes contributed. Is this possible?


BTW, more insightful than number of posts would
be number of bytes contributed. Is this possible?

lol. even more insightful would be the mean quality of users' postings. can you figure this out for us, christian? >^p

Stats are for quantity. The inference of quality from stats is whole other ballgame. The bytes-per-post number for example could be interpreted in ways that reflect upon the person - both positively and negatively - but in any case more meaningfully than the absence of a stat!

Looking at bytes instead of posts would rearrange the tables in
a big -and meaningful- way. As examples think of John Hudson,
and at the other extreme Chris Lozos.


BTW, generally the most interesting thing about stats is when they
factor in time. I would be interested to see how people's number of
posts, bytes per post, etc. have changed over time.


i'm just giving you a hard time, hrant.

Looking at bytes instead of posts would rearrange the tables in
a big -and meaningful- way. As examples think of John Hudson,
and at the other extreme Chris Lozos.

but wouldn't all chris' zany images count to a byte count? or would it just count the image tag? i'd think you would just want the text to count eh?

Ah, I hadn't thought of that! But we wouldn't want to not count the images at all.
Hmmm, maybe factor in the old "an image is worth a thousand words", counting
the average word as being 4 letters, but then what about the blank space? ;-)


You are right Hrant, my math was off by a factor of two (twelve hours in a day??). The actual number of posts/day are about double. Sorry, tiff. You can still tell people the first number if they ask ;)





Christian, this info is cool to see. I'm still wanting to see figures about the number of readers that don't post and the number of times threads get read, outside of the number of people who actually respond. (For every Typophile contributor there are ___ viewers.)

Joe, great stat ideas!


The actual number of posts/day are about double.

I kind of thought so -- I didn't sign up until the end of summer last year.... ;-)

even more insightful would be the mean quality of users’ postings.
and some have been very mean, indeed.

interesting lists...

Truly interesting stats.
Joe's suggestions regarding topic views and non-participating viewers would really be interesting.
At "peak forum" time of day, the relationship of "users" to "guests" are 1 to 30+.

(Personally I'm amazed that I'm on the stats since I've haven't been frequently posting for about 10 months now.)

I read this and thought, man I need to get out more! Actually my posts-per-day figure goes way up when I'm glued to my computer with work as I have been recently and goes way down when things are more balanced and I get to my studio (blissfully internet-free). Speaking for myself (and my own posts) I don't think quality and quantity are the same thing. There are people here I'd like to hear from more, others I'd like to hear from a lot less.

Also there's a real difference between a type ID thread and a thread that turns into a discussion. Number of comments is not a barometer, the recent enfant terrible thread runs to 5 pages and most of it is bickering among 3 people. I wish there were a way to filter the threads (or users) so you could actually read the content without being distracted by the flaming.

And of course, I'm wondering how all the Typophile technical outages factor in to the statistics.

I read this and thought, man I need to get out more
Lol! And I thought: thank goodness I'm not on it! Phewww (which certainly doesn't make my life more interesting than yours, Pat!)

Lies, damned lies and statistics. What do we need them for? To be honest I don't think it's relevant to know who are the most prolific posters. I think it's the subject of the thread that matters. How many Type ID, logo critique, technology related issues, type history etc, book design, advertising, Giggles' questions etc.

A User Filter is a good idea. And let the user know he/she's been filtered! I'm using my "own" filter at the moment and god knows it's not very reliable!

If a picture is worth 1000 words, how much are the common pictures of words we post? 1001? And a pdf, 1024?

If you wanted a handy dataset to float interesting posts, a "track this thread" button would be nice. Then the most tracked+commented or most increased tracked status in the x days could exist. This thread would likely lend itself to more technical stuff, but thats fine by me.

On a tangent, I'd love to see a subpage similar to the front one for each major forum. With some nice icon on it to remind me I'm down one step.

If there were lurker statistics I would be #1.

If there were lurker statistics I would be #1.
Not anymore! :)

Today I’m a member for 12 weeks and 6 days. How come it says 174.28 days above? Are my calculating capabilities leaving me? Showing up in statistics rather frightens me.

Relax, we are just trying to find out why's that Typophile attracts so many psychos. Go on Christian.

| 1658 | formlos

Top 18 Commenter, sure feels good. Nice. ;)
(Still, as I am a 'Typophile' for 5 years and 15 weeks by now, according to my profile page, I guess my posts/comments per day stats would be pretty lame.)


Dav, I miss your frequent posts! You used to be the go-to guy for IDs.

Ouch. Number 2 commenter doesn't count for the Type ID Board cleaning lady. I just work here, so get me out of those stats.

Ow, where are the days that I simply used to be the orange L-spike...

Sorry Yves!

But Dav was on a real tear for awhile. He was nailing them all, faster than I could refresh the page.

Thank you.
(I am still working on my Typophile comeback. ;)


This early-morning's "fornode" spam attack actually provides a means of
calculating another interesting stat: assuming the spam was distributed
totally randomly, it's possible to see what percentage of threads a member
has participated in, by dividing the number of the member's threads hit
by fornode by 200 (the number of threads spammed). I'm 39%.


> Top 30 Commenters:
> | 13923 | hrant
> Friggin spammer.

Some people just can't appreciate greatness (and apparently don't know the definition of spam).

When it comes to stats, "...there is hrant, and then there is everyone else."

Admittedly he is a little weak in the thread creation stat, but I'm betting that was just an oversight, and he will step up to the plate.
; )