Stats in the Wild

So I was in Whole Foods the other day and they often have vendors in there giving out free samples of their products.  Everyone loves a free snack.  But this time I came across Chiro One, a chiropractic group in the Chicago area.  They were excited to talk to me until I started talking to them and the conversation quickly ended.  I finished my shopping (I was getting a ton of soup because it was cold and rainy and that’s what you do when it’s cold and rainy) and went to complain to a manager.  I asked is she knew what they were claiming they could treat and she was really nice and said she would go take a look at their display.  I’m really interested to see what they do (I’m going to follow-up of course) because Whole Foods seems to be pretty big into selling “supplements” and “natural cures”.  So maybe chiropractic fits in with their vision for their company.

So, I’ve had chiropractic on the brain for the last few days and I was talking to my wife about it last night.  She mentioned that she had recently heard a piece on NPR where they talked about how chiropractic is now actually recommended as a treatment for lower back pain based on a recent study.  So naturally I wanted to look into this more and I found the manuscript that the NPR piece was based on.  It’s a research article in the Journal of the American Medical Association (JAMA) entitled “Association of Spinal Manipulative Therapy With Clinical Benefit and Harm for Acute Low Back Pain: Systematic Review and Meta-analysis” by Paige et. al.  From the abstract the authors conclusions are as follows: 

Conclusions and Relevance  Among patients with acute low back pain, spinal manipulative therapy was associated with modest improvements in pain and function at up to 6 weeks, with transient minor musculoskeletal harms. However, heterogeneity in study results was large.

Basically, there conclusion is that there is a small benefit to spinal manipulative therapy (SMT), which is very specific and NOT the same thing as chiropractic as a whole.  Though SMT is most often administered by chiropractors.  They also note that “heterogeneity in study results was large”.  Basically, the study results are all over the place.  This article was published along with an editorial entitled “The Role of Spinal Manipulation in the Treatment of Low Back Pain” by Richard Deyo, which is also worth reading.

So what did Paige et al do in their systematic review?  They did something called a meta-analysis.  The idea of a meta-analysis is to collect study that are similar in concept and combine these results to get a better overall estimate of the effect.  You can think of a meta-analysis as being a “study of studies” (i know.  So meta.) Now meta-analysis is a legitimate statistical procedure, but as with all statistical procedures when you ask yourself the question “Can I apply this technique here?” the answer is always “it depends”.  And with meta-analysis it’s easy to go astray.  You have to make assumptions that the studies are similar enough to even be combined together, there is the potential for publication bias to be amplified by meta-analysis (i.e. The File Drawer Problem) (they do mention in the article “There was no evidence of publication bias in the overall pooled result, with a Begg rank correlation of 0.92 and an Eggar test P value of .58.”), and you have to assume that the individual studies themselves are done well.

Well….I have a few issues with this meta-analysis.  For instance, of the 26 studies included in the effectiveness analysis, the authors state:

None of the studies met the criterion for blinding of providers. Only 4 studies met the criterion for blinding of patients.

The gold standard for randomized clinical trials in double-blind placebo controlled studies.  Most of these studies didn’t bline the provider OR the patient.

The authors also say:

Two studies of SMT vs sham SMT reported non-statistically significant results.

“sham SMT” here is acting as the placebo, so only TWO of the studies considered here actually used a placebo.  These two quotes alone are enough to render this meta-analysis basically meaningless.  But I could go on and on about the weaknesses of this meta-analysis and the individual studies comprising it, but that blog post had already been written in much greater detail by Dr. Mark Crislip.  Dr. Crislip concludes:

Does SMT help for low back pain? Who knows. I have zero idea. Not that I could tell from the primary literature. Any effect of SMT is impossible to dissect out from all the other interventions in the studies. What little beneficial effect that is reported is small and likely from placebo and/or bias. And with a hodgepodge of interventions, it would appear to be a nonspecific effect, like one ape grooming another.

I’d like to point out one more thing about the meta-analysis before moving on to how this was reported by the media.  Several of the works cited in this meta analysis are from a journal called the Journal of Manipulative and Physiological Therapeutics.  This is a journal published by Mosby on behalf of the American Chiropractic Association (ACA).  I, for one, do not trust a journal published by the ACA.  Just to give you a taste of the kind of articles they have published in that journal over the years there is this:

Manner, Harold: Amygdalin, vitamin A and enzyme induced regression of murine mammary adenocarcinomas. Journal of Manipulative and Physiological Therapeutics, Dec. 1978.

That article, according to quackwatch.org, which is run by Stephen Barrett, M.D.:

Manner entered the public spotlight in 1977 by announcing at a National Health Federation seminar that he had cured cancer in mice with injections of laetrile, enzymes and vitamin A. The experiment was published in 1978 in a chiropractic journal [7]. What Manner actually did was digest the tumors by injecting digestive enzymes into them, which cannot cure metastatic cancer.

There is also this article from 2004 where a chiropractor treated a child with Attention Deficit / Hyperactivity Disorder (ADHD) in an article entitled “Cervical Kyphosis Is a Possible Link to Attention-Deficit/Hyperactivity Disorder” written by three chiropractors.  The conclusion of that manuscript is as follows:

This case study shows that spinal correction using the CBP approach may have effects much greater than relief of musculoskeletal conditions. Altered spinal biomechanics associated with abnormal posture clearly relate to significant neurological stress and malfunction. This is particularly evident when considering the effects on the brainstem and the autonomic nervous system. Thus, even in obscure cases with systemic, organic, or chemical dysfunction like ADHD, we suggest optimal spine equals optimal health.

This conclusion is based on ONE patient with no comparison to a placebo.  That’s not science, that an anecdote.

tl;dr: Chiropractors are still quacks.  Save your money and go get a back massage.

Cheers.

I originally made this plot for the 2012 Stat Geek Idol contest run by Team Rankings, and they are still cool.  The original article describing them can be found here. Below I’ve updated the plots for games through March 19th.  (For games that haven’t been played yet this year, I’ve assumed the higher seed wins for the sake of the plot.)

Cheers.

Cheers.

In 2010 CBS signed a deal with the NCAA for the rights to broadcast the NCAA tournament for 14 seasons from 2011-2024.  According to the NCAA, the deal is worth 10.8 BILLION dollars.

Let’s do some fun quick math, shall we:

There are 347 NCAA division 1 NCAA basketball teams and they each offer 13 scholarships.  That’s a total of 4,511 spots.  If you paid every single one of these players $50,000 per year, that would cost about 225 million dollars ($225,550,000).  Let’s assume salaries go up 3 percent a year for inflations, which would mean that in 2024 you are paying players around $73K per year.  The grand total over 14 years for this plan would cost about 3.8 BILLION dollars ($3,853,820,415).  That’s a ton of money.  BUT it’s dwarfed by how much money this deal is worth.  If you took this money out of this deal to pay the players, you’d still have almost $7 BILLION left over  ($6,946,179,585).  And keep in mind, this is just for the tournament.  That’s 67 games over three weeks in march.  There is more money than this for regular season TV contracts not to mention the money schools make for selling tickets and merchandise.  Merchandise, some of which, has players’ names on it!!!! The school can make money off of a player’s name, but the player can’t! (Well the player can, he just then can’t play in the NCAA).  Think about that.  Screw you NCAA.  Screw you.

The NCAA is a cartel that makes billions of dollars off of the back of labor that has absolutely no representation.  I don’t know why as a society we put up with this.  In literally every other area of life, if you are good enough at something to get paid for it, someone can pay you for it.  The exception is sports.  Why do we value amateurism in sports?  I have no idea.  But I’ll continue to call out the hypocrisy of the NCAA until they pay the players who actually make this stuff worth watching.  Coaches get to negotiate a salary and can come and go as they please.  Hell, some coaches get huge contract buy out to NOT COACH.  But players have basically no rights.  So until the NCAA changes it’s ways, I hope they get sued from every direction until they have no choice but to pay the players.

Cheers.

After 48 games we are in 97th place in this year’s March Machine Learning Madness.  I’m going to guess that we finish in 45th.  

Cheers.

In 82nd after 32 games.  In 101st after 40 games.  Not a great day for us.  Need something big tomorrow to get within striking distance.  I’m guessing we will finish in the top 50, but it looks like our lucky run of three top 4 finishes in a row is pretty much over.

Here are our predicted probabilities for Sunday’s games (with ridiculous precision):

• Louisville over Michigan: 0.7051029
• Kansas over Michigan St: 0.8346071
• UCLA over CIncinnati: 0.6074323
• Oregon over Rhode Island: 0.7488454
• Duke over South Carolina: 0.7873011
• Kentucky over Wichita State: 0.6272257
• North Carolina over Arkansas: 0.8434368
• Baylor over USC: 0.8055379

Cheers.

@statsbylopez and I finished the first day of March Machine Learning Madness in 140th.  After 32 games we are currently in 82nd.

Here are our predicted probabilities for today’s games (with ridiculous precision):

• Villanova over Wisconsin: 0.7490162
• Florida over Virginia: 0.5047416
• Gonzaga over Northwestern: 0.896185
• West Virginia over Notre Dame: 0.6114829
• Florida State over Xavier:  0.7793296
• Arizona over St. Mary’s (CA): 0.5626207
• Butler over Middle Tennessee State: 0.7420982
• Purdue over Iowa State: 0.5234515

Cheers.

Here is my bracket based entirely on my rankings using betting market data:

Midwest

Nothing too interesting here.  I’ve got all of the higher seeds winning all the games except for two first round upsets.  I like Oklahoma State over Michigan and Rhode Island over Creighton.

South

I’ve got three first round upsets in this bracket, the 9, 10 and 11 seeds.  That’s Seton Hall over Arkansas in one of the worst games of the first round, Kansas State over Cincinnati (if Wake Forest had beaten Kansas State I would have picked Wake Forest here; I really don’t like Cincinnati), and Wichita State over Dayton.  Wichita State being a 10 seed is the most head scratching seeding of the tournament in my mind.  They probably won’t be favored over Kentucky in the second round, but I’d guess that Kentucky won’t be favored by much more than 3 or 4 (Contrast that with North Carolina who would be favored over Seton Hall by about 12 or 13 points).

I’ve also got UCLA getting to the Elite 8 in this region over Kentucky, with North Carolina coming out of this bracket.

West

This is the most boring region in my bracket.  Straight chalk the whole way with only Florida State with a small upset over Arizona.  The Northwestern-Vanderbilt game features teams that have a combined 26 losses on the season with FIFTEEN of those belonging to Vanderbilt.

East

Another pretty boring region.  I’ve got no upsets in the first round here and the only small upset I have is Virginia over Florida to make it to the Sweet Sixteen.  I’ve got Villanova coming out of this region triumphing over Duke.

Final Four

I have Villanova over Gonzaga by 1 point to make it back to the finals and I have North Carolina, also getting back to the finals, over Kansas by 3.5 points.

National Champion

I think North Carolina gets back to the championship game, but this year they pull it off.

I’ll be posting some more brackets later tonight if I have time.

Cheers.