Let me clear my mind

Yesterday was the first day I really hunkered down and started doing my research for fantasy football. It was at this point that I realized what all of the things that I like have in common. All of the things that I like involve using constrained problem solving and analytical thinking.

Take poker for example, everyone in the world thinks that when poker players are taking a long time staring at their opponent they are looking for the guy to scratch his head or look to the left and get some “tell”. The truth is that the majority of their brain power is solving a puzzle. They try to think about all of the hands their opponent could be holding and eliminating them based on their betting actions on all of the betting rounds. That is what makes Texas hold’em an interesting game. There are four betting rounds, all of which are different pieced to the puzzle. At this point a good poker player starts applying rough percentages on the odds the player has particular hands and makes a play that maximizes his EV (expected value). Maximizing one’s EV is the goal of poker and life. In poker as in life, decisions must be made, luck is heavily involved, information is vital, and the goal is to end off with the best chance of being best off.

Engineering has elements of EV, especially in the professional world, and certainly has constrained problem solving. Most engineering involves designing something of commercial value. There are often constraints and definitely analysis to determine how design changes effect efficiency in comparison to cost (the bottom line). This type of maximization is what interests me.

How is fantasy football like this in any way? Well, it depends on how you approach it. Some guys go to the draft with a magazine and a gut instinct and just pick someone near the top of the list that they like or think will improve. This is a less sophisticated version of what is common practice for evaluating talent in the NFL and many sports franchises. Billy Beane changed this for many baseball teams. His book, Moneyball, described how he was able to use economic statistical analysis to put together great baseball teams for the Oakland A’s in a small market with a small salary bankroll. Running a fantasy football is more like running a baseball team than it is like running a football team. In baseball, especially at the plate, players work as individuals and it is easy to access information on how their play affects the team in the W/L column. The same is true in fantasy football. A player’s value is strictly how many points they can bring to your team on a weekly basis. Basketball teams have started using this technique but football teams are behind the curve. If I were in the front office of a football team and someday I hopefully will, I would implement this type of analysis to determine the value of players.

My system for drafting a fantasy football team is first to estimate how well a player will do the next season. I try to implement “experts’” opinions, key changes, stats from previous seasons, and the trend of their stats from previous seasons. After estimating a point total for about 200 players, I use historical context of previous drafts to determine about what rank on my list for each position will be available after the draft (for example, We have 12 teams each team must start 1 qb, most teams draft 2 qb’s, some draft 3, maybe someone will draft 1. I think about 26 qb’s will be drafted. Seeing as some people will rank the talent in a different order I would guess that the 21st qb on my list will go undrafted. The 21st qb on my list then sets the benchmark). A player’s value in pure fantasy points is worthless. The top 10 qb’s will outscore most other players in the league. But that does not make them the most valuable; in fact it makes them quite the opposite. The fact that there are so many good qb’s makes them less valuable. The value of a player in my system is how much more he will score from the best substitute. If the best QB can score 100 more points then the 21st best QB he has a value of 100 pts. I rank all players by their point differential from the best substitute.

In the NFL this type of methodology could be used, the constraints are different, and it would be more difficult to analyze a player’s value, but I believe that it could be done. More complicated social problems have been analyzed using this econometrics method, so I believe it could be applied to the NFL.

Recently, I have found a new interest in pursuing engineering as a career. I read this article that talked about the cutting edge of technology that might be 15-30 years away. The world is radically changing at an exponential pace. By the time I die the world will probably be more different then it was between my birth and the birth of my great great grandfather. My main concern is the day to day work could be quite tedious, but to be an engineer at the front lines of these major changes would be very rewarding work. A path in engineering is still on the back burner but I am farther from ruling it out than I have been in years.

Tonight I will be back at the beer exchange. Brad is finally back from Connecticut and he had a good time. I think Danny did find a way to sell his ticket to the World Series. He won 12,000 dollars from a free tourney. You can’t complain about that, but I wanted to see him compete. I have earned 4 buy ins back from my poker slide so hopefully this will continue. Only two weeks left in Madison so I better make the most of it.