Tom Brady and Trading
Let’s get a few things straight. I don’t follow football. I have not watched a game since Joe Theismann and the Redskins won the Super Bowl. I am neither a Patriots fanboy nor a hater of the franchise but having read most of the stories about “deflategate” scandal I am pretty confident that the Pats are guilty as sin.
For those of you not living in North America, let me explain what all the fuss is about. The New England Patriots -- the preeminent team in American football -- has been accused of cheating by systematically making their “football” softer which in turn allows the quarterback and other players on the team to have better control of the ball and therefore have far fewer turnovers to the opposition.
This is not a story about football but about cheating. There is very little doubt that the Patriots are objectively the best team in football, “deflategate” or not. But imagine if Lionel Messi coated the outside of his cleat with just a smidgen of some sticky substance that allowed him to have just a fraction of second more control over the ball. His natural talents would still make him the greatest scorer in the game today, but that tiny little tweak would give him an unfair advantage and if caught he should be suspended from the game. Just as Patriots should be suspended and their championship revoked even though they most likely would have won the title without cheating.
Again this is not about who is best, but who is cheating. To understand why cheating is so corrosive to all human activity let’s look at non-trivial example like building codes. Suppose you have builder who puts up high rise building in any urban area of the world and is generally recognized as producing decent housing stock, but decides to dilute his concrete by just 1% below code. This is not bad enough to cause any immediate structural problems for the building but it does allow the builder to win more contracts because he is able to save on materials.
Still think such “minor” cheating is no big deal? Still feel safe to have you two year old jump up and down on the floors of such a building?
Anyway back to the Pats. Why am I so confident that they are cheaters? Data and conditional probability.
First we have a lot of circumstantial evidence from the SMS texts of Patriots staff showing Tom Brady’s almost pathological obsession with the pressure of the football. You have mysterious “absence” of footballs just prior to kick off. You have a wide variety of statistics on the underinflation of the footballs -- which have been viciously criticized by pro Pats supporters as being inaccurate, but are frankly the least important aspect of the case
Secondly you have the much maligned study showing that Pats fumble the ball at rates so low as to make them almost superhuman athletes. And while the primary argument against that statistical analysis is that fumbles are not “random” in nature and therefore basic sample analysis does not apply -- that’s nonsense. Fumbles by their very definition are random and even if some players have a much higher predilection for losing the ball, the law of large numbers ( this was study done on many seasons of data and hundreds of games) applies to this case as it does to all other evidence based activities. Put another way, Ted Williams batting 400 for one season is an amazing feat of athletic performance, but Ted Williams batting 400 for ten seasons in a row ( the equivalent of Pats fumble accomplishments) is an achievement deserving of massive skepticism since it stands well outside the norm of the game.
Third and most damning of all is Tom Brady complete refusal to cooperate with the investigation. But not allowing the NFL to examine his phone and data records he has tipped his hand like a bad poker player. This isn’t a court of law, but the court of science where his silence speaks volumes as to his complicity based on the idea of conditional probability.
So what does Tom Brady have to do with trading? Quite a lot actually. Good trading is simply data observed and data analyzed. The VT system that we day trade hundreds of times each day is founded on the idea of conditional probability and we constantly analyse our myfxbook.com accounts to see if the underlying assumptions hold true.
Now am I CERTAIN that Brady is guilty? No. In social science it is almost impossible to prove with 100% accuracy that correlation is causation. That’s why when we trade we always have to be mindful of the possibility of the “mirage” ( that our analysis of the underlying causes was false and that market truth lies elsewhere). But ultimately the “ironclad” certainty that we enjoy in physical sciences is impossible to replicate in social sciences so we have to rely on the “preponderance of evidence” standard. In that case I will continue to believe that Pats are cheaters and that VT will churn out pips for us as it has done for the past four months running.