Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

New England won the SuperBowl XVI coin toss snapping a 14-game streak by NFC teams, which lead ~2 to 1 in tosses.

The graphic below shows the coin toss process realization (red) to SB XVI, superimposed on 100 pseudorandom i.i.d and unbiased binomial random walks (white) which displays the concentration of measure relative to the cone of possible outcomes (boundary: green dashed lines).

enter image description here

From a physics standpoint, there is no coin bias or NFC "kabbala" or telekinesis: each year there are distinct coins, different people toss the coins, and most important, it's not heads vs. tails that determines the outcome but how the tosses are called. In other words the toss process is fair and any determination to the contrary is "Gambler's Fallacy".

Nevertheless, we can assume unknown bias in the form of a uniform (ie, uniformative) prior distribution of bias in [0,1]. The sequence of Bayesian posteriors is computed via the binomial distribution because of the Cox consistency axiom as described in Sivia & Skilling Data Analysis.

Normalizing to probability distributions at each step, the sequence of posterior distributions is:

enter image description here

Where each curve is one of the posterior distributions and ones based on more data (ie, up to more recent SB's) are rendered in darker gray. The straight line corresponds to the posterior after SB I (ie, with one data point available).

For example the Maximum a posteriori estimator (MAP) after 46 tosses is ~.65 in favor of NFC. Ok that's driven by the data and the a-priori physical information "fades". This introduces a bias in this case.

My question however concerns the narrowing of the posterior distribution over time. This limitation of Bayesian analysis has been noted in Guyonnet and Ferson's "Bayesian methods in risk assessment".

In general, since the posterior distribution is obtained by multiplying the prior by the likelihood function, whenever these overlap only a little, the posterior will be very narrow, see G&F, p.12.

Is there a general methodology that can mitigate this? I don't have a particular solution in mind, but some way to complement the tendency to narrow down, some enveloping method that gives a conservative estimate to counter the bias? Or would any amendement to how the posterior is computed introduce inconsistencies with respect to the Laplace/Cox/Polya axioms?

If we can't address the limitations of the methods in the simplest possible stochastic process - coin tossing - how can we apply Bayesian methods to complex systems like climate forecasting, biomedical risk assessment etc where these tools are being currently used?

Since the posterior is derived from Bayes theorem, which in turn is based on the product rule: P(A|B)P(B)=P(AB)=P(B|A)P(B). I wonder if Bayes theorem can be balanced by looking at a (suitably normalized) sum of distributions? That may lead to an envelope or conservative posterior estimate?

share|improve this question

closed as off topic by cardinal, Henry T. Horton, Clive Newstead, 5PM, Ittay Weiss Feb 4 '13 at 0:30

Questions on Mathematics Stack Exchange are expected to relate to math within the scope defined by the community. Consider editing the question or leaving comments for improvement if you believe the question can be reworded to fit within the scope. Read more about reopening questions here.If this question can be reworded to fit the rules in the help center, please edit the question.

Is it possible that this question would be better answered on stats.SE? –  Asaf Karagila Feb 3 '13 at 21:26
Downvoters, it's customary to comment on rationale for vote. –  alancalvitti Feb 3 '13 at 21:39
(-3): (-1) for shameless self-promotion, (-1) for asking for a universal answer to a vaguely stated and, at this stage, somewhat ill-posed question and (-1) for the unnecessary and remarkably brash ignorance displayed in your first comment. Alas, unfortunately only one of them counts. –  cardinal Feb 3 '13 at 22:15
The best that can be currently made of your question so far is: Why does (insert favorite data-analysis paradigm here) actually use the data to arrive at an inference instead of ignoring it in place of my a priori assumption? In other words, you're accumulating evidence. Why shouldn't your posterior distribution get narrower! –  cardinal Feb 3 '13 at 22:19
@cardinal nailed it with his comment. If your prior really is that all possible probabilities for the NFC winning the toss (from $p=0$ to $p=1$) are equally likely, then the specific narrow posterior you've calculated is correct. Like many so-called Bayesian analyses, this one gets a silly answer because it starts with a silly prior. If you don't pay attention to that and instead try to modify the method to mitigate the posterior narrowing (equivalently: you ignore evidence), then you're just compounding the mistake. –  Greg Martin Feb 3 '13 at 22:23