There is a widespread problem of results bias that I can’t bring myself to ignore, and it’s practically everywhere. In poker everyone claims to be aware of it but only in the most obvious examples are they being truthful. They understand that if they get their money in with the best hand and lose they likely didn’t make any mistake. They don’t alter their strategy when the situation comes up again, thus they are immune to results bias. The issue is that the end result of the hand isn’t the only result that might bias our perception of how we should our strategy. The cards you were dealt, the particular flop, turn and river cards, the decisions made at every street by both you and your opponent are all individual results that can bias our opinions. I’ll try to make a less esoteric example before I keep going.
Those of us who live in a big city will all have strong opinions on the fatstest route driving from point a to point b. The next time someone tells you to take the one lane alley behind the park or to stay in the right lane during a specific stretch of highway, ask how they came up with their method and really dissect it. Most likely you’ll hear how they’ve taken this route a thousand times and they always make it from door to door in under 8.3 minutes. Even more likely is that there was an unusual circumstance slowing down traffic along the logical route back when they decided on this and that’s why they can’t explain what makes their way the fastest. Meanwhile, someone traveling that path using mostly highway and avoiding stop lights, left turn lanes and pedestrian traffic is getting door to door in 8 minutes flat (impressive, i know). The point here is that a sound theoretical explanation needs to accompany your good results to separate your strategy from results bias.
As an observer this teaches me to be a skeptic when I see fantastic results of any kind. If a trader got rich off short selling the market in 2008, I don’t run to dump my life savings into his new hedge fund. I look for articles he’s written or lectures he’s given and I analyze his approach to investment. If he exhibits a clear edge in theoretical knowledge over the other guys, then I’ll give him fair credit. Before that, results mean next to nothing–especially in something so volatile.
On that note, this is part of what makes economic theory so vitally important to a government. You can’t just look at the years with economic growth, pick out our common strategies in those years and pray it still works. Even more, macroeconomic models are so complex that we can’t simply simulate the results of our entire market like data in a statistics class and keep running trials until the sample size is absurdly significant. There’s an immense amount of variance in our stock market and the results for a given quarter or year or even decade are so difficult to predict that we can do nothing better than spend tons of resources on building strong theoretical ideas.
It’s an extreme example of why theory matters but on a smaller scale the same is true in all cases. If you want to properly construct a trading strategy, determine the optimal decision in a poker hand, or just settle an argument over never taking lake shore drive into downtown at 6:30pm on a Tuesday, theory is everything.
2 Comments
really enjoyed this entry Krabby ^^
someone went to the U of C