FA+
Shop
Summer Project
14 years ago
On June 7th at 7:18pm, my summer will officially begin. Between then and September 21, I won't have any homework or any classes to deal with for over 3 months. It's going to be amazing!
As of right now, I'm hoping to get this actuarial internship in Pepper Pike, just east of Cleveland. If that happens, all bets are off.
Another goal of this summer, with or without that internship, is studying for and passing Exam FM, which I will take in mid-August if all goes according to plan. By going according to plan, I mean that I will pass Exam P next Tuesday. This stupid exam has been running my life for the past 4 weeks when school and work haven't been. So, I better pass it or that will be hundreds of hours of studying wasted and $200 down the drain. Ah, the joys of actuarial science.
Now, if I don't get this internship, I have a backup plan that involves a lot of baseball umpiring and possibly a side job as well. But, those have been common for the past several summers are of no interest to you, the reader. The question I pose is: What else should I do with my time this summer?
In general, I'm usually bored during the summer. I get to see my friends more, I get to go outside more, but without school and Big Ten Network stuff to occupy two-thirds of my life, I have a hard time finding ways to keep my body and mind occupied. So, I was thinking about using the past 3 years of calculus, statistics, linear algebra (maybe not so much this one), and probability theory do some sports-related research.
So, I ask, what kind of project should I take on? I have a couple of ideas.
Team Rating Algorithm - Simplest idea. Come up with a way to rate teams based on their performance up to a specific point in the season, usually being that day. It would revolve around the thought process of which team's accomplishments would be the most difficult for an average team to accomplishment? This formula would be strongly based on probability theory.
Golfer Rating - Another simple idea, this one would involve less probability theory and more statistics. Taking a player's average score can help determine how well he plays the sport. However, course difficulty and comparisons against his fellow golfers may be more important.
Golfer Odds - This takes the Golfer Rating idea one step further. But looking at the consistency of a golfer's play, the effects of specific courses and how golfers age, it's possible to come up with better probability to win golf tournaments, then compare those odds with the Vegas lines to see which are better predictors.
Fantasy Football Rankings - I don't know if any of these could be profitable, but this idea has the most potential to be. Analyze the performance of veteran players and the draft status of younger players along with the effects of age, experience, and changes in coaching style to determine who to pick in fantasy football drafts this August (assuming there is even a season). Add in one more factor for auction drafts to know how much specific players are really worth.
Fantasy Baseball Rankings - Fantasy baseball isn't a lucrative as fantasy football, but it's a lot more involved for those who own teams. With baseball, you need to win more stat categories than your opponents, but which stats are the most predictable and which stats are the most important? How does a player's performance change over his time in the majors? And how much of a correlation is there because your basic baseball stats to advance sabermetrics? And again, add in a factor for auction drafts. Yes, I'm very fascinated by auction drafts.
Baseball Player Rankings - Very different from rating a batter or pitcher as a fantasy player is the idea of rating how they help their real life team. Granted, this idea has been done to death, but it's still an interesting idea. If you don't believe me, read Moneyball.
There are several more ways to apply mathematics, statistics, and probability theory to sports. The inexact science of scouting and recruiting is great example. Sadly, I don't have access to the data to do make a useful algorithm for various sports. I could also look at player rankings for football, basketball, or hockey. Look for correlations between college basketball success and NBA success, or by the same token, college football success and NFL success. There are so many different directions one can go with this. But alas, I'm only one man.
As of right now, I'm hoping to get this actuarial internship in Pepper Pike, just east of Cleveland. If that happens, all bets are off.
Another goal of this summer, with or without that internship, is studying for and passing Exam FM, which I will take in mid-August if all goes according to plan. By going according to plan, I mean that I will pass Exam P next Tuesday. This stupid exam has been running my life for the past 4 weeks when school and work haven't been. So, I better pass it or that will be hundreds of hours of studying wasted and $200 down the drain. Ah, the joys of actuarial science.
Now, if I don't get this internship, I have a backup plan that involves a lot of baseball umpiring and possibly a side job as well. But, those have been common for the past several summers are of no interest to you, the reader. The question I pose is: What else should I do with my time this summer?
In general, I'm usually bored during the summer. I get to see my friends more, I get to go outside more, but without school and Big Ten Network stuff to occupy two-thirds of my life, I have a hard time finding ways to keep my body and mind occupied. So, I was thinking about using the past 3 years of calculus, statistics, linear algebra (maybe not so much this one), and probability theory do some sports-related research.
So, I ask, what kind of project should I take on? I have a couple of ideas.
Team Rating Algorithm - Simplest idea. Come up with a way to rate teams based on their performance up to a specific point in the season, usually being that day. It would revolve around the thought process of which team's accomplishments would be the most difficult for an average team to accomplishment? This formula would be strongly based on probability theory.
Golfer Rating - Another simple idea, this one would involve less probability theory and more statistics. Taking a player's average score can help determine how well he plays the sport. However, course difficulty and comparisons against his fellow golfers may be more important.
Golfer Odds - This takes the Golfer Rating idea one step further. But looking at the consistency of a golfer's play, the effects of specific courses and how golfers age, it's possible to come up with better probability to win golf tournaments, then compare those odds with the Vegas lines to see which are better predictors.
Fantasy Football Rankings - I don't know if any of these could be profitable, but this idea has the most potential to be. Analyze the performance of veteran players and the draft status of younger players along with the effects of age, experience, and changes in coaching style to determine who to pick in fantasy football drafts this August (assuming there is even a season). Add in one more factor for auction drafts to know how much specific players are really worth.
Fantasy Baseball Rankings - Fantasy baseball isn't a lucrative as fantasy football, but it's a lot more involved for those who own teams. With baseball, you need to win more stat categories than your opponents, but which stats are the most predictable and which stats are the most important? How does a player's performance change over his time in the majors? And how much of a correlation is there because your basic baseball stats to advance sabermetrics? And again, add in a factor for auction drafts. Yes, I'm very fascinated by auction drafts.
Baseball Player Rankings - Very different from rating a batter or pitcher as a fantasy player is the idea of rating how they help their real life team. Granted, this idea has been done to death, but it's still an interesting idea. If you don't believe me, read Moneyball.
There are several more ways to apply mathematics, statistics, and probability theory to sports. The inexact science of scouting and recruiting is great example. Sadly, I don't have access to the data to do make a useful algorithm for various sports. I could also look at player rankings for football, basketball, or hockey. Look for correlations between college basketball success and NBA success, or by the same token, college football success and NFL success. There are so many different directions one can go with this. But alas, I'm only one man.
And ugghh... maths + sports :S It's quite scary for me, but you look to be comfy, so go with it :)