The Cox Proportional Hazard Model on Fire

April 12, 2012 7:45 PM

The Hunger Games Survival Analysis is kinda neat, but also an interesting use of statistics (a basic chi-square test and a more sophisticated survival analysis) in solving not-quite-real world problems. In this case, whether the Lottery in the Hunger Games was rigged and what factors predict the winner of the Games. This work joins a tradition of analyses of the world of Panem, such as whether the economy was realistic, the problems of teserrae inflation, and this (slightly problematic, as the first link points out) application of game theory.

For those who find the statistics, rather than the dynamics of Panem, intereting, there are some nice online resources, which can be combined with the Hunger Games dataset to get your feet wet in survival analysis, an important statistical technique in medicine and social science.

A reasonably painless introduction to the topic can be found in this Nature tutorial (you can skip the math if you want). If you want to start working with the data, start with the relatively straight-forward Kaplan-Meier survival function, which has a nice explanation (and online calculator) at Vassar. To actually run the model, you will need to enter the data into a hazard model regression calculator, which can be found online.

posted by blahblahblah (10 comments total)

Oh my, I won't have time to properly think about all this until July or so, but as a former high school English teacher who just finished learning survival analysis and uses Stata, I can vouch that this is relevant to my various interests. And goddamn, I love geeks and nerds.
posted by smirkette at 7:55 PM on April 12, 2012 [1 favorite]

The Hunger Games Survival Analysis


I survive by giving into and BELIEVING in the system, with my very soul, and plastering away any doubt behind about four feet of pink pancake makeup and a manic rigid grin.
posted by The Whelk at 8:35 PM on April 12, 2012 [5 favorites]

From the "realistic economy" link.

The injustice of the situation led eventually to African mass resistance and the overthrow of colonial rule. But in almost every case, the new elite simply started running the same extractive institutions for their own benefit.

Surely this undermines any pretension to economic realism the article may have. The new elites didn't "simply" start doing anything, they were encouraged and enabled or bullied and coerced to carry on running the extractive institutions largely for the same elite in whose benefit they had always been run, Western capitalists. When comparing science fiction to reality, it's useful to not be living in a fantasy world.
posted by howfar at 8:58 PM on April 12, 2012

I love statistics, but come on: The major finding here is that Suzanne Collins did a good job creating a fictional dataset that shows some significant differences between groups. Yes, that's because statistics measures deviations from randomness, and Collins *made up the data* as part of her novel's plot.
posted by yellowcandy at 9:11 PM on April 12, 2012 [4 favorites]

Is the story written from your point of view? No? Then I'm afraid it's not looking so good for you.
posted by ODiV at 9:42 PM on April 12, 2012 [6 favorites]

I didn't read the books, but based on the trailer they pick out the characters name from a bowl.

But I don't understand, if Panem is supposed to span North America, and there are only twelve districts, shouldn't they need more then a single bowl to do the lottery? Or do they pick the town randomly and then do a lottery there?

The Tesserae Inflation problem could be solved simply by making a the payout based on a proportion, rather then a fixed value.

The economic situation didn't really seem that realistic. If you have an unintentional famine/misallocation of resources, then I can see the situation being stable. But if people think the government is holding out on them, it seems like a rebellion is much more likely. You would need to keep a large contingent of well-fed people to oppress the populous. It's obviously been done in the past.
posted by delmoi at 9:58 PM on April 12, 2012

If you have an unintentional famine/misallocation of resources, then I can see the situation being stable. But if people think the government is holding out on them, it seems like a rebellion is much more likely. You would need to keep a large contingent of well-fed people to oppress the populous. It's obviously been done in the past.

Yeah that's...basically the plot.
posted by The Whelk at 10:08 PM on April 12, 2012 [5 favorites]

But I don't understand, if Panem is supposed to span North America, and there are only twelve districts, shouldn't they need more then a single bowl to do the lottery? Or do they pick the town randomly and then do a lottery there?

There's not many people in Panem. The reason why the Capitol is trying to avoid an uprising is that they are concerned that if there is further violent conflict, the population of Panem will drop below the level where it could replenish itself.
posted by His thoughts were red thoughts at 10:08 PM on April 12, 2012

And some districts being kept in constant near-starvation and thus beholden to the Capitol for food and supplies is well, look at the title.
posted by The Whelk at 10:12 PM on April 12, 2012

delmoi brings up something I keep thinking about -- District 12 is basically all of Appalachia, right? And the entire mining center of Panem? When you see all the kids standing at the reaping, that's definitely not all the kids 12-18 inclusive in District 12 -- it couldn't be. From the books I was picturing a much larger crowd.
posted by fiercecupcake at 11:08 AM on April 13, 2012 [1 favorite]