Tuesday, June 30, 2009

Homeric Epithets and Mercurial Forward Ron Artest

I'm not sure whether this is simple lazy sportswriting or the emergence of mythic "tradition" in sportswriting, but considering the pre-occupations of this blog, let's go for the mythic.

In writing about the NBA, writers often refer to Ron Artest as "mercurial forward Ron Artest." And when I say often, I mean that whenever you read or hear the phrase "mercurial forward" you should assume the name "Ron Artest" will follow. This morning when I google the phrase, "mercurial forward", I get 9,980 hits. When I google the phrase, "mercurial forward ron artest", I get 7,760 hits. Do the math. This morning, on the Internet, the phrase "mercurial forward" refers to Ron Artest about 80% of the time. The phrase is so ingrained in the minds of basketball writers that it nearly reaches the level of Homeric epithet. This is the current culture of basketball writing. What does it mean? Does this justify a mythic approach to basketball? I, of course, say yes. The players aren't just self-mythologizing Big Shoguns or League-promoted X-Men. They are epic heroes in the classical tradition because that is the language with which we, or at least basketball writers (who are an inevitable filter on the season unless you somehow manage to watch every single game), discuss them. What I am trying to say here is that, at the very least, thinking of Ron Artest in the terms of a character in epic poetry isn't just a plausible mode, but is actually a rather common mode, cliche even. So Ron Artest is a Homeric hero? Sure. Now, what other players in the league are discussed in such formulaic and grandiose modes?

Here is the new project: try to find the epithets by which players are consistently and predictively referred to. A special prize if anyone can find an Epithet Certainty Score for an NBA player(JUST NOW INVENTED) that's higher than Ron Artest's 0.775. This is an important first step in developing a useful taxonomy of basketball writing's neo-Homeric traditions and cliches.

Epithet Certainty Score: Divide the Google hits for "[epithet] [player]" by the Google hits for "[epithet]". Leave the results in the comments.

HINT: Surprisingly, "bovine Tim Duncan" doesn't seem to get any hits


  1. sorry buddy, i have a few issues with this formula.

    for my own clarification: the formula is "[descriptor and position] [player's name]" divided by "[descriptor and position]." this way you eliminate non-basketball uses of the descriptor. if you just use the "[descriptor] [player's name]"/"[descriptor]" format, your artest score is reduced to 2.2x10^-6.

    one problem i see with your formula is that all-stars aren't normally preceded by their position upon introduction. say, for example, you wanted to do something with Steve Nash. Google finds about 2,580,000 results for "steve nash" but only 14,700 for "point guard steve nash". i know the latter filters out things like the Steve Nash Foundation but you might also be losing articles where someone might give Nash an epithet. in the case of ron artest, your number did not include the 111 results that google found with "mercurial ron artest". granted, a bunch of these were from one article, but it does bring up concerns for finding other examples.

    it also fails to recognize shorthand for positions (people especially like to use PG for point guard for some reason). That's an easy fix, just add the results for "[epithet abbrev position player's name]" and "[epithet abbrev. position].

    the formula also doesn't account for specific versus general postitions. this doesn't really apply to ron artest because i only got one real return for "mercurial sf ron artest" and none for the full-hand phrase. he seems to be refered to as just a forward a lot. (perhaps that is his playing nature, though, since he plays in between forwards). it might apply to other players so, to be thorough, a search involving steve nash would need to also look for "[epithet guard steve nash]"/"[epithet guard]" as well as the point guard. it then needs to add both of the results for shorthands of PG and G. (nash is probably another bad example here because he is almost always referred to as by the specific position, PG).

    also, "guard" and "forward" have other uses apart from basketball. so epithet + position might also include things describing fantasy novels or something similar(in the case of guard) or soccer players. "evil guard" for example is worthless as a search even though Kobe totally is one (but he is rarely introduced by position anyways). i guess adding basketball to every search outside all quotation marks might clear this one up.

    or maybe i am just complicating things and just need to embrace the simplicity of your formula. this need to break it down probably stems from my feelings of inadequacy for not being able to come up with any examples of epithets.

  2. You aren't seeing the purpose. The purpose is to find unique strings of words that almost always refer to an NBA player. There is no need to include position. It just happens that "mercurial forward" works.

    Likewise, yep, these strings of words can refer to other players and completely unrelated things; which is why the denominator exists. "Evil" isn't a unique epithet because it's used so much, which is why it scores so low when paired with an NBA player. Epithets, to be effective need to be unique. Indeed, "mercurial Ron Artest" is a good case study, because "mercurial Rasheed Wallace", "mercurial Kobe Bryant", and "mercurial Gilbert Arenas are all worth more hits than lil Ronny.

    Unique epithets are rare. The formula enforces a strictness while allowing for weirder constructions without the rigidity of enforcing a position.

  3. I know I'm a bit late to the party on this one, but:

    Results 1 - 10 of about 3,020 for "troubled forward". (0.11 seconds)
    Results 1 - 10 of about 3,350 for "troubled forward qyntel woods". (0.14 seconds)

    (no, this doesn't make any sense to me either)

  4. Same thing happens with 'Future Rookie of the Year' and 'Future Rookie of the Year Tyreke Evans'