Psst, hey lost_spook!

[eve11]

Guess who is the most well-rounded writer on Teaspoon?

By which I mean the entropy of the distribution of eras in your stories is as closest to uniform across all eras as anyone gets in the database. And you have a lot of stories too!

I shall have a true post up at some point but am off to a picnic now so pictures will have to wait.

Comments

[thisbluespirit]

Really!? Wow. I'm impressed with myself. So all those times I've claimed that I love all DW is proved true. :lol:

I thought I'd skewed myself to Three with all the Brig/Liz fics, and then there's my failure to write much New Who...

*basks in new-found... er... what would you call it?*

:-D

(I like your Teaspoon stats even more now. That's probably the only thing in my life I'll ever come out on top of.)

♥

[eve11]

Your most popular era is Third Doctor with 22% of your 90 stories (as of Feb 24, 2013) and 20% of your stories when weighted by word count. There are several authors who have a flatter peak than that (eg, their most popular era has a lower percentage), but they score lower in the uniformity metric because they write across fewer of the eras, or because their flatter peaks are also coupled with more extreme lows.

The measure of uniformity that I am using is called entropy: it reaches its maximum when a distribution is completely equal across all categories and its minimum when one category is the only category there is (think, eg, 15 Ten/Rose stories and that's it). It is calculated by taking -1 times the sum of each of the percentages times their natural logarithm.

There are only four authors in the database who have entries for every one of the 16 eras (1-11, Other Doctors, Other Era, Multi-Era, Torchwood and Sarah Jane Adventures): you, JJPOR, johne, and Gary Merchant.

But for example, comparing your entry with infiniteviking who also scored very high on well-roundedness, she has a slightly more even distribution across the Doctors but nothing for Torchwood or Sarah Jane, which will reduce her score.

Comparing your entry with jjpor who also scored high AND covered all 16 eras, he has a few smaller peaks than you (Four, Seven, Other) but it's also accompanied by a few more eras with eg, percentages less than 2% (First, Second, Fifth, Sixth, Ninth, Other, Sarah Jane), while you have the higher spike with Three, but have only 3 eras that have percentages less than 2% (Ninth, Other, Sarah Jane).

Here is the data. "StoryMode" and "StoryMax" are the category and value for your highest percentage. StoryEraEntropy is the entropy calculation. I have the data for stories weighted by word count as well but sticking just with counting each story equally regardless of length, this is what you get:

name                  infiniteviking vvj5         JJPOR        
NumStories            74             90           84           
First.Doctor          0.0147         0.0650       0.0189       
Second.Doctor         0.0868         0.0372       0.0130       
Third.Doctor          0.0756         0.2221       0.0408       
Fourth.Doctor         0.0895         0.0962       0.1840       
Fifth.Doctor          0.0783         0.0592       0.0160       
Sixth.Doctor          0.1053         0.0900       0.0189       
Seventh.Doctor        0.0373         0.0434       0.1457       
Eighth.Doctor         0.0508         0.0375       0.0114       
Ninth.Doctor          0.1301         0.0067       0.0031       
Tenth.Doctor          0.1093         0.0789       0.1106       
Eleventh.Doctor       0.0405         0.0418       0.0575       
Other.Doctors         0.1059         0.0167       0.0119       
Other.Era             0.0068         0.0815       0.1833       
Multi.Era             0.0692         0.0759       0.0977       
Torchwood             0.0000         0.0368       0.0788       
Sarah.Jane.Adventures 0.0000         0.0111       0.0083       
StoryEraEntropy       2.4894         2.5092       2.3030       
StoryMode             Ninth.Doctor   Third.Doctor Fourth.Doctor
StoryMax              0.1301         0.2221       0.1840       
NonZeroEra            14             16           16           

The highest possible entropy someone could achieve across 16 categories is the score you get for having 1/16 in each category, eg 2.77. You are at 90% of that with your score.

Now for 14 categories, the highest possible entropy you get is for having 1/14 in each category, eg 2.639. Discounting Torchwood and Sarah Jane, infiniteviking's score of 2.4894 is 94% of the 14-category maximum. Across just those eras, you would score an entropy of 2.406 which is only 91% of the maximum. But despite some Third Doctor preference, you gain points for having branched out into the spin-off categories.

It may be a bit difficult to interpret outright because of the flexibility with which authors may or may not use the "Multi-Era" or "Other Era" tag. But I think it's pretty well principled.

[dbskyler.livejournal.com]

I think the biggest issue with "Multi-Era" is when authors use it for a multi-Doctor story instead of assigning it to all the eras.

(And, um, *cough*, I just may have been on your list of authors who wrote in all 16 categories if I hadn't done that . . .)

[thisbluespirit]

BUT FOR THAT MY GLORY MIGHT HAVE BEEN YOURS.

;-p

[dbskyler.livejournal.com]

CURSES, FOILED BY YOUR MOST ASTUTE USE OF THE TAGGING SYSTEM!!!! Isn't that always the way?

Athough actually, I'd be interested in a separate "roundedness" measure that only looked at single-era stories -- for example, I wouldn't say I'm much of an Eighth Doctor writer just because I included him in my "Spotter's Guide" fic. Using a single-era definition, I'm sure you'd still be well ahead of me. ;p

I'd also be interested in seeing "roundedness" defined by the number of different characters written. eve11 has nothing better to do than answer all our different research questions, right?

[eve11]

for example, I wouldn't say I'm much of an Eighth Doctor writer just because I included him in my "Spotter's Guide" fic.

Well I did count stories with multiple eras differently: So that the counts for tags added up to the number of stories, if a story had K tags, then I added 1/K to each of the counts for the tags. So a Multi-Era story that was tagged as Multi-Era and with Doctors 1 through 11 eras would put a count of 1/12 on each of those tags. So if you wrote one Eleventh era story and one multi-era story with doctors 1 through 10, you would not count as completely uniform; your single era story would have more of a weight on Eleven.

The other way that I did it (which I didn't report these answers) was as above but also weighted by word count. So, if you wrote one Eleventh era story of 100 words, and a 1000-word story that had doctors 1 through 10, then your distribution would be uniform. But if you wrote a 150k word Eleventh era novel, and also had an Eight ficlet of 1000 words, you would be weighted much more toward Eleven.

I was also thinking about doing the distribution by character. That would pick up on your stories for example, that you labeled "Multi-Era" but had all of the doctors from 1 to 11. (which is also why you did not show up as having representation for all 16 eras as you rightly guessed: I thought you had meant that you labeled all the eras separately but didn't bother clicking "multi-era" but it was the other way around. My db says you have 26 stories as of the end of February this year, across 10 eras; looking at your stories, I think if I coded them with the eras suggested by the characters in the summaries, you would have 15 out of 16, missing only "Other Doctors" possibly).

The only problem is that because of how they were scraped from the website, I have a little bit of manual recoding to do for the character lists. There is a hard space limit on the webpage, and so for stories that list lots and lots of characters, sometimes the names got chopped off at the end. So I'd need to fix those and then figure out which characters to count (all of them?) in the calculations or not. But yeah, that also will be able to highlight & pick out the New Era authors who write lots of New Era Characters. Or eg, distinguish between the person who writes only Four and Sarah Jane vs. the person who writes Four and Sarah Jane, Harry, Leela, Romana and K-9. It will be sparser than the Eras and also there is likely a bit more uncertainty due to self-reporting. You have to pick an era but you don't have to list characters (do you?). And some people list every character, and some just the main characters, etc.

[dbskyler.livejournal.com]

Well I did count stories with multiple eras differently: So that the counts for tags added up to the number of stories, if a story had K tags, then I added 1/K to each of the counts for the tags.

Oooh, yes, very nice way of handling it!

And yes, I clicked only Multi-Era and not any of the other eras for my two multi-Doctor stories. Also, for one of them ("Fallacy Somewhere"), the character list was long, so instead of listing out all the companions who appear in it (11 of them), I just clicked "other characters." It's people like me who mess up archival data . . . ;)

And some people list every character, and some just the main characters, etc.

True! And I have come across authors who list characters just because they were mentioned in the fic, when the character otherwise has nothing to do with the fic at all.

My db says you have 26 stories as of the end of February this year, across 10 eras; looking at your stories, I think if I coded them with the eras suggested by the characters in the summaries, you would have 15 out of 16, missing only "Other Doctors" possibly

Not that I looked, except that I did . . . ; ) My fic "Fallacy Somewhere" has "Other Doctors" in it, but those Doctors only "appear" as snatches of dialogue because I needed thirteen Doctors (one per G&S operetta). So yes, I agree with you.

*makes plans to write a proper "Other Doctor" story so she can pretend to be a well-rounded author*

Anyway, really interesting stuff! I hope I haven't been too annoying . . . and I hope that when you eventually disseminate this more widely, you don't get bombarded with a bunch of people asking you to change the data based on their idiosyncratic situations!

[eve11]

Ha, well, measurements are almost always imperfect. What makes data analysis useful is that iterative loop between summarizing and understanding, that uncovers a story. It's the stories that lead to insights. Too many people think analysis ends with a number. That is actually where it begins, in my book :)

So I certainly welcome the input. I've been tooling around this data set for a little while and am quite enjoying it. I wish I had a better way of updating all of the data, because I've added a Third Doctor and Fourth Doctor story to my Teaspoon set since February but my percentages in the data are still 0s for them...

[thisbluespirit]

*smug*

Yes, I think I have written at least one story for every era - I'm not absolutely sure about Nine, though. I've written him in multi-era fics and in drabble collections, but I can't think of any others. My "Other Doctors" fic won't surprise anyone - I only have one of those because I wrote fic for David Collings's audio Doctor from Big Finish's Unbound series.

If you wrote Curse of Fatal Death fic or something, that could go in Other Doctors? That could be a thing. ;-D

Yes, I'm absolutely sure eve11 has nothing better to do than analyse Teaspoon! (Not analysing a teaspoon, though, that would be silly. Well, unless this was S&S, not DW.)

[eve11]

I only have one of those because I wrote fic for David Collings's audio Doctor from Big Finish's Unbound series.

Oooh, I really liked that story too. :D

I think Richard E Grant's WebCast 9 would count as "Other Doctors", as well as Fatal Death, any attempts to scoop Moffatt on the John Hurt character, any of the other Unbound audios, and perhaps Peter Cushing from the 60s movie(s?).

[thisbluespirit]

Aw, thanks. I really wish there were more FFF fics - it's so very twisted up and an interesting AU to explore.

And, yes. I think the number of other Doctors is quite large in total! There would also be the stage play Doctors (if anyone's ever written those) and no doubt Doctors from fan-made stuff.

[eve11]

ha, i got it. i think there might also be some strangeness with the eras labeled "oother" as well. I checked the numbers earlier & thought that the 15 counts were not missing due to multi era. There were only 4 authors who had 15 out of 16. I'll check it out when I'm at home :)

[thisbluespirit]

Ah, very interesting, thank you! I am surprised Gary Merchant didn't come out more highly - he seems to have lots of stories for every era, and Johne's got a lot too. If I'd have guessed, I'd have gone for one of them. And I thought JJ must be near in the stats, too - though, as you indicate here, he is much more faithful to Seven! I am an indiscriminate flirt of a writer, really. :lol:

Anyway, thanks - your stats are always interesting and it cheered me up a lot this morning!! :-D

[eve11]

The top 5 by entropy were you, followed by infiniteviking, Gary Merchant, Sceadu and Gabriel Chase. You were all pretty close. I'll have to check Gabriel Chase's #'s; he has 12 non-zero eras so must be pretty evenly distributed across them.

If I go by the Word Count entropy score, the top 5 are Crackers and Cheese (7 stories), golden orange (8 stories), DTA (13 stories), you (90 stories), and Arthur Dent (2 stories). But the difference is that those authors all have actually very few stories, but lots of multi-era stories that are well tagged. DTS is the only one other than you in that list who has more than 25000 total words. She has 213000, and you have 734000. Which is why for the ranking I went with entropy by story and not by word count.

[a-phoenixdragon.livejournal.com]

Oh wow!! That is massively cool!

*BOUNCES*

[(Anonymous)]

Congrats lost_spook!

I am fairly entertained by these statistics. As soon as I started reading I started trying to figure out what other authors might be in the AllSixteen club. I thought of the JJ and Johne straightaways, and, while I haven't read much of his fic, Gary_Merchant doesn't surprise me any because he does have A LOT. I was somewhat surprised that Dbskyler, AstroGirl, and a few others who regularly show up on Calufrax didn't make it.

[thisbluespirit]

Why, thanks, Nonny Mush, whoever you may be!

(And, yes, I'd have expected there to be more of us - but probably excluding spin-offs, there are - no doubt people like Calapine and Clocket Patch as well.)

[clocketpatch]

Lol, that was me, not realizing that I wasn't logged in (Calapine and Livii were the two Calufrax people who I couldn't remember but expected to see, if that even makes sense. They're on Eve11's big list below though)

[eve11]

Gary Merchant was right behind lost_spook in the entropy value by story, and I think he took a hit due to a high mode in "Multi-Era", which could be interpreted as less of a preference. So he and she are likely pretty comparable. :)

The distribution of eras (discounting characters, which as dbskyler points out above can lead to some skipping of characters) is:

   1    2    3    4    5    6    7    8    9   10   11   12   13   14   15   16 
1935  886  506  281  169   98   59   47   29   23   23   27   16   11    4    4 

I looked at the 35 authors with 13 or more out of 16 eras represented. Depending on the breakdown of who they wrote and when, even those with lots of eras can have (relatively) lower entropy, but yes, lots of familiar names show up. Here is them ranked by story entropy, but I also included their stats for the breakdown weighted by story (S) and word count (W):

(in next comment due to size)

[eve11]

I should say, above is the distribution of the number of eras reported per author, so eg, 1935 authors had only one era listed, 886 had 2, etc.

Other authors should go in this table likely, barring tagging inconsistencies/peculiarities as discussed.

name            Stories Words   SEntropy        WEntropy        SMode   WMode           SMax    WMax    NonZeroEra
vvj5            90      734366  2.5092          2.2802  Third.Doctor    Third.Doctor    0.2221  0.2041  16
infiniteviking  74      27849   2.4894          2.2367  Ninth.Doctor    Sixth.Doctor    0.1301  0.2865  14
Gary Merchant   199     513541  2.4565          2.1872  Multi.Era       Multi.Era       0.2203  0.3776  16
Sceadu          48      54299   2.4550          2.1961  Ninth.Doctor    Second.Doctor   0.1667  0.1991  13
ClocketPatch    53      217930  2.3731          2.1753  Tenth.Doctor    Multi.Era       0.2459  0.2146  14
neveralarch     48      185693  2.3466          2.1827  Other.Doctors   Third.Doctor    0.1944  0.1989  13
JJPOR           84      858410  2.3030          2.2296  Fourth.Doctor   Other.Era       0.1840  0.2341  16
johne           87      335942  2.2851          1.7511  Second.Doctor   Second.Doctor   0.3101  0.5194  16
Calapine        136     417863  2.2782          1.7267  Tenth.Doctor    Tenth.Doctor    0.2904  0.4748  15
Biichan         34      59498   2.2417          2.1272  Multi.Era       Multi.Era       0.2206  0.2755  14
drox            61      22425   2.2368          2.0397  Eighth.Doctor   Eighth.Doctor   0.2951  0.2733  13
King Rat        33      92258   2.2341          1.4757  Ninth.Doctor    Third.Doctor    0.1894  0.5831  13
Lyricwritesprose51      143737  2.1396          2.1091  Eleventh.Doctor Eleventh.Doctor 0.3595  0.3281  14
Paranoid Seat   25      62893   2.1309          2.0515  Tenth.Doctor    Multi.Era       0.2400  0.2711  13
amaresu         37      29266   2.1043          1.9723  Seventh.Doctor  Seventh.Doctor  0.2982  0.3515  13
were_lemur      83      12876   2.1002          1.8945  Tenth.Doctor    Tenth.Doctor    0.3775  0.4254  14
Derek Metaltron 16      55989   2.0667          2.2015  Tenth.Doctor    Tenth.Doctor    0.2562  0.2651  14
Spydurwebb      36      269173  2.0657          1.8776  Fourth.Doctor   Fourth.Doctor   0.2812  0.3340  13
aralias         52      277425  2.0168          1.8037  Tenth.Doctor    Eighth.Doctor   0.2676  0.4122  15
LilacFree       73      490817  1.9825          1.4067  Fifth.Doctor    Fifth.Doctor    0.3333  0.6260  13
Danny_B         64      452926  1.9357          1.5562  Multi.Era       Multi.Era       0.4622  0.6090  14
livii           95      152699  1.8717          1.6606  Multi.Era       Multi.Era       0.4281  0.5404  14
Daystar Searcher37      85682   1.8332          1.7215  Fourth.Doctor   Third.Doctor    0.3989  0.3552  14
TimeLord89      12      15652   1.8026          1.6856  Multi.Era       Multi.Era       0.4231  0.5075  13
bluedawn        11      244188  1.7963          1.7802  Ninth.Doctor    Ninth.Doctor    0.3613  0.3344  13
Library_Lover   17      19460   1.7929          1.9170  Tenth.Doctor    Tenth.Doctor    0.3922  0.4586  13
Marcus S Lazarus24      504193  1.7548          1.9727  Tenth.Doctor    Tenth.Doctor    0.4552  0.3825  14
enchantment     37      242401  1.6187          1.4924  Tenth.Doctor    Tenth.Doctor    0.4908  0.5220  13
finmagik        40      78898   1.5574          1.9583  Sixth.Doctor    Sixth.Doctor    0.5515  0.3829  13
schildkroet     126     533739  1.5555          0.9783  Tenth.Doctor    Tenth.Doctor    0.5368  0.7276  13
nostalgia       204     272946  1.4832          1.2743  Tenth.Doctor    Tenth.Doctor    0.4559  0.4962  15
missvmarigold   123     121661  1.1515          1.1441  Torchwood       Torchwood       0.7053  0.6899  13
betawho         150     330715  1.0841          1.4109  Eleventh.Doctor Eleventh.Doctor 0.7367  0.6013  13
rickmaniac101   108     214777  0.8127          0.8614  Tenth.Doctor    Tenth.Doctor    0.8167  0.7808  14
Vicky Smith     212     336872  0.6644          0.5282  Tenth.Doctor    Tenth.Doctor    0.8110  0.8152  15

(ETA: LJ is being stupid slow for me :( )

[thisbluespirit]

Ooo. *looks at info with wide-eyed interest*

:-)

[clocketpatch]

Shows how small Who fandom can be, while being at the same time so big, with the way it divides up between who writes what, ships what, etc. Very interesting stats to be sure.