June 5th, 2013 / 2:20 pm
Craft Notes

“Furthermore, the amateur poets were more likely to use more emotional words, both negative and positive.”

sciencebarrowY’all relax OK Australia finally figured out contemporary American poetry y’all. Post your scores, losers. > 0.5 is professional and < 0.5 is amateur. Be sure to read the, um, “methodology.” According to poetry journalist Patrick Gaughan, the highest so far (at 4.2) is WCW’s good old wheelbarrow coloring book, and noted anti-poetry skeptic/fiction engineer Jonathan Volk reports that “an untitled Jewel poem just got a 1.17.” Stay tuned for breaking developments. (Discovery credit goes to poetry scientist Anne Cecelia Holmes.)

Tags: , , , , , , ,


  1. Adam Robinson

      Mike Young’s “Giggle or Stop It” = 1.20885436893

  2. Adam Robinson

      Also, wow, nice tags

  3. Quincy Rhoads

      One of my poems scored higher than Frank O’Hara’s “The Day Lady Died.” BAM!

  4. Quincy Rhoads

      I scored 1.87656097561 on the unpublished poem I tried. Clearly you need to give me a book deal, Adam.

  5. Deborah Lisoway

      4.03246153846, bitches!

  6. Trey

      this is a fun game, but reading the methodology makes it clear that the creator or creators have not seen dead poets society

  7. Trey

      also according to the methodology, “warmth” is just slightly more difficult to define than “alimony”

  8. Adam Robinson
  9. Bobby Dixon

      I got a 0.296911877395

  10. Deborah Lisoway

      The algorithm used has a clear Australian bias. My husband entered this:

      G’Day Mate

      Dingos howl on Red Rock.
      Babies devoured, Fosters on tap.

      The wallabies fashioned
      after Paul Hogan’s mug.

      Something about Vegemite.

      That’s not a knife;
      this is a knife.

      and it got a 3.14328571429

  11. A D Jameson

      I submitted three of my unpublished poems and got the following scores:


      As I’ve long suspected, I’m a poetry amateur.

  12. mimi

      i entered the first thing that entered my mind, which was “Never Mind the Bollocks, Here’s the Sex Pistols”, and got a score of 3.354

      bully for me

  13. mimi

      publgeni gonna git mad hits brah

  14. mimi

      and then i entered “bully for me” and got a score of 1.82066666667

      so bully for bully for me

  15. mimi

      and then i entered “bully for bully for me” and got 2.074

      so bully for bully for bully for me!!!!!!!!!!!!

  16. A D Jameson

      “Bollocks for me,” don’t you mean?

  17. A D Jameson

      Tao Lin’s “Poems That Look Weird” got a -1.74397297297. Appropriate?

      I like that this system is accurate to more decimal places than Pitchfork’s rating system. I give it a 7.91358302145.

  18. mimi
  19. JosephYoung


  20. mimi

      cumm on, let’s see it, bitch

  21. A D Jameson

      Bobby, you are so clearly a 0.296911877396.

  22. Adam Robinson

      But that is a good poem.

  23. A D Jameson

      Can someone please summarize the device’s criteria/methodology? I glanced at the PDF but am feeling too lazy to really read it.

      I’ll understand if no one wants to go to the trouble. I imagine it’s looking for word/grammar variety, lots of polysyllabic words, and for the poem to be in conventional verse? Or something like that?

  24. Mark Cugini

      The Pitchfork review of “Trouble Will Find You” got a 0.34724009324

  25. Mark Cugini

      My above comment (with line breaks) got a 2.184.

  26. GlorgusTimeMachine

      I typed in a poem that read


      and got a 2.729

  27. A D Jameson


  28. Mark Cugini

      Your above comment got a -28.746 (and I’m going to downvote this comment)

  29. A D Jameson

      That’s actually pretty good, for one of my comments.

  30. Jeremy Hopkins


  31. Trey

      I also only glanced, so this is not fully informed. it was developed by running 100 relatively contemporary (20th century?) poems by famous people and 100 poems from some amateur writing website through the machine with their respective labels (amateur/professional). so the machine just associates whatever is present in the pro poems with pro poems and whatever is present in the amateur poems but *not* in the pro poems with amateur poems. there are a few notes on the differences, Mike pulled some for the tags of this post. one is “The study found that the successful poems had fewer syllables per word in their first lines and were more likely to have an initial line consisting of monosyllables”

  32. A D Jameson

      That is amazing and wonderful. Thank you!

  33. A D Jameson

      I just submitted the following poem (see below), which I wrote by copying text from here and adding line breaks. It received a score of 0.638166666667.

      I am so totally adding this poem to the collection I’ve been working on!

      Here is the poem:

      “Written language is relatively more complex than spoken language”

      Written language is relatively
      more complex than spoken language.

      Written language is grammatically
      more complex than spoken language.

      It has more subordinate clauses, more
      “that/to” complement clauses, more
      long sequences of prepositional phrases, more
      attributive adjectives and more
      passives than spoken language.

      Written texts are shorter and have longer,
      more complex words and phrases. They have
      more nominalisations, more noun based phrases,
      and more lexical variation. Written texts
      are lexically dense compared to spoken
      language — they have proportionately more
      lexical words than grammatical words.

      The following features are common
      in academic written texts:

      Subordinate clauses/embedding, Complement clauses,
      Sequences of prepositional phrases, Participles,
      Passive verbs, Lexical density, Lexical
      complexity, Nominalisation, Noun-based
      phrases, Modification of noun-phrases,
      Attributive adjectives.

  34. A D Jameson

      (Adding, a 0.638166666667 is a real step up for me after my earlier -0.0220048721072, -0.184882803943, and 0.189744680851, respectively. And to put it in further perspective, Frank O’Hara’s “Why I Am Not a Painter” scores a 0.280395939086. Suck it, Frank!)

  35. Bobby Dixon

      a low hanging 0.000000000006

      i felt really bad when i got that score

  36. A D Jameson

      Australian poems all score pi.

  37. Tracy Dimond

      A group of “non-writer” friends wrote this poem and it got a 3.099:

      The autumn leaves fall to dust
      I just want to rhyme
      Crunch crunch squeak crunch
      I once saw a gnome

  38. bartleby_taco

      sticking to my day job i guess. -3.33547368421 for all of my hard work:

      “art is subjective

      whatever you want to read into this poem is right

      i fucked your mom last night”

  39. Michael Coleman Dalvean

      One this worth mentioning is that the scale used in the paper (>.5, 0, <0). In other words, in the application a score above 0 means the poem has more in common with "professional" poems while a score below 0 means the poem has more in common with an "amateur" poem.
      Hope this helps. http://www.poetryassessor.com

  40. Michael Coleman Dalvean

      If it scores well it probably should be submitted somewhere for publication!

  41. Michael Coleman Dalvean

      The reason this scores well is that 1) It uses concrete language 2) It has no emotional words (love, fear etc) and 3) it has no psychological words (imagine, ruminate). These are characteristics of professional poetry. This does not mean that this poem is a professional poem; it simply means that it has more in common with the average professional poem than it has with the average amateur poem.
      Hope this helps.
      Michael Dalvean http://www.poetryassessor.com

  42. Michael Coleman Dalvean

      This poem is more like a professional poem than it is like an amateur poem. THis is because it is concrete and has few emotion or psychological terms.

  43. Adam Digged

      boethius’ meter 1 from consolation scored just a tad higher than one of my poems at 2.02270229008, which is surprising because it is filled with non-concrete, emotional and “psychological” words. muses and such.

      Carmina qui quondam studio florente peregi,
      flebilis heu maestos cogor inire modos.
      Ecce mihi lacerae dictant scribenda Camenae
      et ueris elegi fletibus ora rigant.
      Has saltem nullus potuit peruincere terror,
      ne nostrum comites prosequerentur iter.
      Gloria felicis olim uiridisque iuuentae,
      solantur maesti nunc mea fata senis.
      Uenit enim properata malis inopina senectus
      et dolor aetatem iussit inesse suam.
      Intempestiui funduntur uertice cani
      et tremit effeto corpore laxa cutis.
      Mors hominum felix, quae se nec dulcibus annis
      inserit et maestis saepe uocata uenit.
      Eheu, quam surda miseros auertitur aure
      et flentes oculos claudere saeua negat!
      Dum leuibus male fida bonis fortuna faueret
      paene caput tristis merserat hora meum;
      nunc quia fallacem mutauit nubila uultum
      protrahit ingratas impia uita moras.
      Quid me felicem totiens iactastis, amici?
      Qui cecidit, stabili non erat ille gradu.

  44. Michael Coleman Dalvean

      The system was calibrated on contemporary English poems so, although any given text will generate a score, the score only really makes sense if it is a contemporary English poem. It’s like a medical diagnosis system that has been calibrated on a particular population. Using such a system on a very different population will yield results that are not valid.
      Hope this helps.

  45. Michael Coleman Dalvean

      The highest score so far for a poem is 6 for The Image by Robert Hass

  46. columbusmatt

      Is this from the creators of the “processor”?

      (I played with this last night, got lots of shorter poems from 4s all the way up to a 9.934.)

  47. columbusmatt

      Find this interesting and useful, as I changed “tears” to “blood” in one poem (thinking “tears” might be considered too sappy/emotional) which did, imo, increase the power of the poem (and also raised its score from a high 4 to a 6.83733333333