Poemetrics, Part 1

Poetry and other works of art are famously open to opinion and interpretation. What I describe as poignant, you may declare repugnant. What I feel is hackneyed, you may find hilarious. What I cast off as atrocious, you may consider genius.

We can continue that dance all day and night, but neither of us will ever prove ourselves “right”.

At least not until one of us brings some data.

During the Madness! 2012 poetry tournament that I hosted here at TKT in March, 128 poems were posted for public evaluation (126 written by our tournament participants, plus one each by me and my alter-ego Charles Mund). Unlike, say, a museum, where passersby may take a moment to observe a piece, offer an approving nod or uncertain tilt of their head, and then shuffle off; during MMPoetry (that was our Twitter discussion hashtag and now-commonly accepted blogger abbreviation), observers were able to leave more permanent marks of their opinions before leaving each page. That is, they voted.

Some 12,000 votes were cast in all. With each vote, we have a documented opinion about something. Was it always as simple as someone reading two poems and saying “I like this one better” or “I could relate to this one more” or “my friend wrote this one” and then clicking through, or can we learn anything more specific about what exactly people liked about the poems that they voted for, in aggregate?

I decided to code all 128 MMPoetry poems on eleven variables that I felt lent themselves to reasonably objective evaluation. The variables, and possible resulting “values”, are:

  • Does the poem rhyme? Yes, no, or imperfectly so.
  • What was the structure of the poem? Distinct form (e.g., limerick, pantoum, triolet), typical structured verse, or unstructured.
  • Did the poem attempt to be funny? Yes or no.
  • Was the poem actually funny? LOL, smile, huh? (i.e., didn’t really get it), or wince (got it, but it was pretty bad).
  • How was the rhythm / meter? Perfect or close to it (leniency is granted for poems written on 36 hours notice), okay, or not-so-good.
  • What age was targeted? General audience (7-15yo but appropriate for anyone in a sense), Younger, or Older.
  • What was the subject? Animals, people, characters, or other.
  • From what perspective was the poem written? First, second, or third person.
  • Did the poem aim to deliver a punchline or otherwise strong ending? Yes or no.
  • Could the poem have been perceived to break any “rules” of the contest? Yes or no.
  • How central was the assigned word to the poem? Thematically important, key to a line or image, or used matter-of-factly/hidden.

From these questions, I am hereby announcing the creation of a new branch of applied statistics called Poemetrics — the quantitative study of poetry-related facts, trends, preferences, and whatever else we can think of that hopefully won’t take the fun out of writing/reading the stuff in the first place.

Now, before I reveal any findings, I want to know which attributes from the above list YOU think were more often associated with winning poems over the course of the tournament. Do you think funny beats serious? Do you think people care about meter? Do you think people like characters in poems? Do they value reading new poetic forms?

Leave your comments below!

Then, in a few days — by Friday 5/11 at the latest — we will review the world’s first ever poemetric analysis, and see if it turns up anything interesting or not.

Like it? Share it!

Like me? Subscribe to TKT!

  • http://www.katyaczaja.com/ Katya

    I’m going to guess that ‘attempted to be funny but failed’ is a negative while ‘actually funny’ will be a positive. I suspect that using an actual form will be a bonus… and having a punchline or twist at the end will be a strong positive…

    I can’t wait to see the results.

    • http://www.thinkkidthink.com Ed DeCaria

      Thanks for being the first to offer a guess, Katya. We’ll see how it turns out!

  • http://poetryonparade.com Janet F.

    Kind of like you say potato, I say potato. Poemetrics, I love it. More on my thoughts later, but for right now I go with rhythm, rhyme and punchline. From a statistical POV perhaps you might ask some others to rate the poems at some point and compare with your analysis? I also think the “idea”/ subject/ topic was also important, but how that teases out in an analysis will be interesting to find out. You are always up to something, Ed. Hooray.

    • http://www.thinkkidthink.com Ed DeCaria

      From a statistical POV perhaps you might ask some others to rate the poems at some point and compare with your analysis?

      I completely agree, and this will be addressed in Poemetrics, Part 3. We are all about to embark on an incredible adventure together …

  • http://www.MattForrest.com Matt Forrest Esenwine

    I think it’s an interesting task you’ve created for yourself, to statistically analyze these poems. I think these are mostly pretty reasonable variables to use. A few thoughts:

    Whether a poem was funny or not is only an appropriate factor to consider IF the poem was part of the ‘Humor’ category, yes? I’d say the broader question would be, ‘Did the poem elicit the intended response?” (whatever that response was, according to the rules)

    Also, I’m fine with free verse – unmetered and unrhyming as it is – but I feel that proper meter & rhyme should be factored in heavily if the poem was written in that style. Granted, 36 hours is not a lot of time to be tight with your meter, but in my opinion, there’s no excuse for a slant rhyme.

    Third, and this may be impossible to analyze, it would be interesting to somehow statistically learn how often other poetic devices were used – such as alliteration, internal rhyme, simile/metaphor. What purpose this would serve is anyone’s guess…but it might be fun, just for the sake of numbers!

    The contest was a great idea, Ed…keep up the good work!

    • http://www.thinkkidthink.com Ed DeCaria

      Yes, anything that did not attempt to be funny was “unrated” on the actual funniness variable.

      As for other poetic devices, I don’t think that these are impossible to analyze at all, but I wasn’t about to classify all 128 poems on those points. As above, hang tight for Poemetrics, Part 3 in a few weeks for more on my proposed approach to this.

      Thanks for contributing your thoughts, Matt.

  • Suz Blackaby

    This is SO HARD.
    Do you like ice cream or cake?
    Sunshine or shadow?
    Elves or fairies?
    Moms or Dads?
    Hardware or software?
    Lyrical, rhythmic, funny, thoughtful, poignant, free-form, form-fitting, or punchy?

    Seriously (I know, I know, this never happens) iIf you’re going with rhyme and rhythm, then it’s gotta scan. Gotta. With no cheating and no contortions. But those refinements often don’t come until you’ve been grooming something for weeks or months), not hours (or minutes).

    So once again, BRAVO, BRAVA!

    All brilliant. Data shmata.


    • http://www.thinkkidthink.com Ed DeCaria

      YES, this is and will be hard. You cannot completely quantify art. But it can be fun to try! Did you know that Jonathan LuCroy of the Milwaukee Brewers is the best catcher in baseball at framing pitches — i.e., taking a pitch that is actually a ball and manipulating his glove and body in such a way that the umpire behind him calls it a strike more often than he probably should? It’s as close as one can get to performance art without being on stage, and it has now been quantified — admittedly after 100+ years of increasingly innovative data collection and analysis techniques.

      So, with poemetrics, we’ll start with an isolated event referring to one person’s hurried classification scheme matched up against the indirect opinions of a small sample of undocumented voters that may or may not represent the general population. Because we have to start somewhere. Who knows, maybe someday we’ll read poetry with heart monitors and brain wave sensors and arm hair micro-rulers and we’ll know the difference between the health effect of reading a sad poem and a happy poem, or the half-life of enjoyment from reading a love poem the first time, to the second time, to the third time, etc.

      I’m kidding! Okay, I’m only kindof kidding. (But really, I’m not kidding.)

      I happen to completely agree with you on rhyme and rhythm. For me, it needs to be ultra-precise. Rhyme I think you can sometimes get away with small sins, but not nearly as much as musicians do in sung verse. But rhythm and/or meter have to be precise; one wrench and it is disruptive; multiple wrenches and it destroys a poem’s worth to me. It’s like taking a long, hot shower and every once in awhile someone dumps a full bucket of dirty, ice cold water on your head. Sure, when the shower is over, you’re mostly clean, and in between the bucket dumps the shower was nice, but the experience was cruel and you never want to be in that shower ever again.

  • http://www.quicksilverpartners.com Jeanne Poland

    Dear Ed:
    Thank you for the invitation to stimulation.
    My responses:
    1 “Funny” always beats “serious”, but irreverence never wins.
    2 Meter grips but free verse meanders and has a soft arrival.
    3 Every character gets shunned or adopted. None go ignored.
    4 As a visual and color person, I have to be directed to the new poetic forms, ie. excavate through sound, texture, smell, rhyme, twists and relevance. If those grabbed me, I’ll study structure to model it.
    Happy to join the literati,
    Jeanne Poland

    • http://www.thinkkidthink.com Ed DeCaria

      Thanks Jeanne. Should be fun to see how everyone’s opinions (predictions?) turn out.

  • http://gottabook.blogspot.com Greg Pincus

    You know I’m all for the analysis, Ed. I think your categories are relatively broadly objective (as in we could all agree)… though something being funny is subjective and consistent only within one person’s analysis (i.e. YOUR sense of humor). Also… what if something has humor in it but isn’t designed to have a punchline at the end? Or variations like that? Tricky, though worth looking at.

    I agree that the data tells us something. But I can’t tell you what it will tell you about what poem “wins” in a battle, as that carries with it too many other factors. Or put another way, if the same 600+ people who voted in the finale had voted for each battle, I guarantee you the results would have been different. If a different group of 600+ people, totally randomized, had shown up… the results would be different still!

    For me, meter and rhyme matter, no matter if you have 36 hours or not. If you choose to go that way, you have to get it right. (Of course, many lines in many poems in the contest can be read more than one way – a pause can make it scan whereas ignoring a pause makes it fall apart, etc. Tricky!).

    Still, what usually determined my winner, beyond the above things you mention, was simply how the poem impacted me – did it deliver what it was trying to deliver. That’s more than just landing a punchline, though it’s kinda like that (particularly in a funny poem). What’s unclear to me is if it’s possible to “deliver” if one messes up on the things you’re analyzing!

    Good times. Looking forward to more….

    • http://www.thinkkidthink.com Ed DeCaria

      Thanks for adding your POV, Greg!

      I completely agree on the “humor” thing — no doubt that my sense of humor differs from that of other readers. Again, more to come on how to gauge that in Poemetrics, Part 3 in a few weeks.

      Also agree on the sample population; I’ll get to that in a future post as well.

      Your last point was one that I simply could not include in my analysis — the IMPACT of a poem personally. For that, I would need to collect input from many people along with their votes. (And, for the record, I do think that the poetic details affect the overall impact.)

      Thanks again!

  • http://mainelywrite.blogspot.com Donna Smith

    My guess is that most found humor, a twist for interest, and rhythm to be the important factors in the poetry. Some more serious poems made the cut, but the needed the twist or rhythm. Those are my guesses.
    Loved your cold dirty water analogy! Very true.

    • http://www.thinkkidthink.com Ed DeCaria

      C’mon kids … keep these guesses coming!

  • Melinda Harvey

    Hi, Ed! Your brain intrigues me! I can’t wait for the release of your poem data. Regarding my two cents, I prefer a tightly rhymed poem. If it packs a funny twist or an emotional connection, all the better. I am not a huge fan of stats, but I think your poemetrics might alter that!

    • http://www.thinkkidthink.com Ed DeCaria

      I am not a huge fan of stats, but I think your poemetrics might alter that!

      This is an interesting comment, and I want to draw a parallel for you.

      Many people — including kids — also say “I’m not a huge fan of poetry”, but then when it is presented to them in a way that is familiar and comfortable, they naturally enjoy it.

      The same is true of statistics. I have published various studies at BaseballHQ.com and in the annual Baseball Forecaster book that took pretty intense statistics and presented them in a way that was easy for people that “hated statistics” to understand and use.

      So, I now find it especially fun (if a bit counterintuitive) to take TWO things (poetry and statistics), that on their own people may not be “huge fans of” and combine them in a way that makes people like and understand BOTH of those things even more.

  • http://poetryonparade.com Janet F.

    Interestingly you have referred to Poemetrics part 3 but nowhere to part 2. That leads to some interesting speculation as to what you, dear Ed, may have up your poetic sleeve for us in part 2. But honestly, when I am 80 and someone wants to wire me up to find out how I do with poems from TKT the debut album, I am not sure I will be able to contain my glee!!!

    • Melinda Harvey

      I love it, Janet….do we have to wait until you’re 80 though?!!

    • http://www.thinkkidthink.com Ed DeCaria

      Years ago, there was a funny South Park episode involving Underpants Gnomes, industrious creatures who mysteriously went around collecting underpants. When their motive was questioned, they explained as follows:

      Phase 1: Collect underpants
      Phase 2: ?
      Phase 3: Profit

      Here’s a picture of their “business model”: http://en.wikipedia.org/wiki/File:Gnomes_plan.png

  • http://thereisnosuchthingasagodforsakentown.blogspot.com Ruth

    You crack me up. If you figure it out, you could market it as a tool for teachers. Grading my students’ poems is always a big challenge for me.

    • http://www.thinkkidthink.com Ed DeCaria

      Now you are zeroing in on the end game, Ruth.

      This is where we are headed. Poemetrics (as we will discuss them in future weeks) can be used as a means to teach poetry, as a means to evaluate poems, as a means to measure/monitor the poetry market as a whole, as a means to assess the marketability of a poem/collection in today’s market, etc.


      It all starts with the right inputs. I can present my vision, provide some direct inputs (or automate the collection of certain types of inputs), and facilitate the effort, but to really make this work it will require participation from a much larger group of people, and a diverse one at that.

      But we’ll try to do it in a way that is FUN so that we don’t even have to think about it as “data collection” or “statistics” … just enjoying poetry in a new way.

  • http://poetryonparade.com Janet F.

    The underpants model, really, Ed, are we getting kinkier or thinkier??? Data and poetry…..gotta’ love it. You may have a wider impact than you ever imagined. Phase 2, now what might that be? Could you be looking for supergnome poetry helpers???? I got a Rassmussen poll phone call today. They wanted my input on the economy. I would have preferred talking about something much more fun, say….poetry!!!

  • http://readingyear.blogspot.com Mary Lee

    Wow. This comment thread started with potatoes and ended with underpants. Sometimes it’s fun to come late to the party.

    I’m looking forward to this day:

    “Who knows, maybe someday we’ll read poetry with heart monitors and brain wave sensors and arm hair micro-rulers and we’ll know the difference between the health effect of reading a sad poem and a happy poem, or the half-life of enjoyment from reading a love poem the first time, to the second time, to the third time, etc.”

    As for the factors that seem to be assigned to the winningest poems, I’m pretty sure that rhythm, rhyme, funny, and punchline will come out on top. Maybe you could survey the winning poets to see which of your factors they CONSCIOUSLY chose to use, and cross-check your back end (underpants) data with your front end (potato on a dinner plate?) data.