Sunday, January 26, 2014

Early Season Power Rankings - 2013 Update

Exposition universelle de 1900 - portraits des commissaires généraux-Charles Spearman
Charles Spearman
Nothing drives pageviews like portraits
of Victorian-era statisticians
In this post from last year, I took a look at the accuracy of various NFL power rankings, where I defined accuracy as the ability to predict future wins. In a follow up post, I added additional years to the study. Using the Spearman rank correlation coefficient as my measure of accuracy, I found that the betting market rankings that I publish here are the most accurate in terms of predicting future win-loss records. Surprisingly, the second most accurate ranking is what is known as the "Simple Ranking System", which outperformed both the statistical models and the "experts" at ESPN, despite being based solely on scoring margin.

The 2013 NFL Season

As promised, here is the 2013 season update. In order to shed more light on my methodology, I will breakdown the rankings and win-loss records at a team level below, so we can see exactly how right (and wrong) each ranking system was.

As a reminder, I am looking at the rankings after week 4 of the regular season, and seeing how those correlate to each team's win-loss record for weeks 5-16. The ranking systems I'm comparing are: the ESPN power rankings, Football Outsiders' DVOA Rankings, the Advanced NFL Stats team efficiency model, my betting market rankings, and the Simple Ranking System.

The $1 Billion Bracket - Part One

This is the first of a two part blog post on the probability of picking a perfect March Madness bracket. This first post will focus on what we can learn from the betting markets. The second post will take a more robust look at the problem using my college basketball rankings.

One......Billion Dollars

If you're looking for bigger upside in your March Madness bracket this year, Quicken Loans is offering a $1 Billion payout to anybody that can submit a "perfect" bracket. Quicken Loans founder and chairman Dan Gilbert is a wealthy man (he's also majority owner of the Cleveland Cavaliers), but a billion dollars is still a lot of money. So he has partnered with Warren Buffett (through Berkshire Hathaway) to provide the financial backing for the contest.

Buffett did not disclose how much they are receiving from Quicken in exchange for insuring this contest, saying only that the amount was less than what he would have wanted, but more than what Gilbert would have liked to pay.

Saturday, January 18, 2014

Win Probability Graphs for (nearly) all Australian Open Matches

Last year, I published win probability graphs for selected US Open matches, starting with Victoria Duval's upset of Samantha Stosur. For the remainder of the Open, I published a new graph on a daily basis, trying to highlight the most interesting and/or dramatic match of the day. For the mens' final, I was also able to provide a live win probability graph as the match was being played.

Thanks to my recent work on NBA Win Probability, I now have a more robust infrastructure for storing and sharing win probability data. So, it was easy enough to tweak my code to provide the same level of detail for tennis. The result:

Win Probability Graphs for the Australian Open

Sunday, January 12, 2014

New Feature: NBA Win Probability Added Totals by Player

DeAndre Jordan free throw
DeAndre Jordan attempts a free throw
The purpose of this post is to announce that Win Probability totals at the player level are now available for the 2013-2014 season. But before getting to that, I wanted to add some historical context to a recent improbable comeback by the Cleveland Cavaliers.

A Magical Collapse

A few weeks back, I began publishing win probability graphs and box scores for the 2013-2014 NBA season, based on my win probability model. Not even a week into publishing these, the Cavaliers did their best to break the model. Down by 9 points against the Magic with 1:03 to go in the fourth, my model pronounced the Cavaliers dead, their win probability flatlining at 0.0%. Somewhat embarrassingly for me, the Cavaliers defied those odds, and were able to force overtime, ultimately winning the game by a score of 87-81. See Yahoo Sports' Dan Devine's blog post for a more thorough breakdown of the Magic's collapse (with an assist from the inpredictable win probability model).

Wednesday, January 8, 2014

NFL Playoff Probabilities Available

NFL Playoff Probabilities are now available and will be updated daily. This is the same format I used last year.

Comparing my "fair" Superbowl odds to those offered by, the Broncos appear to be a bargain. I have the fair odds listed at +175, with 5Dimes offering +250.

I also project the probability, point spread, and over/under of each possible Superbowl matchup. These look roughly in line with the 5Dimes futures odds. The most likely matchup is Broncos/Seahawks, in which the Broncos would be a 1 point favorite (it would also feature the number one offense against the number one defense).

Saturday, January 4, 2014

The NFL Playoff Race on One Page

The Falcons Playoff Seed
Probabilities before Week 1
Think of this as an experiment in data art.

Image Link: Playoff seed probabilities for every team and every week of the 2013 NFL season.

Okay, maybe art is a bit of a stretch (I'm a color blind codemonkey, not a graphic designer).

As a reminder, blue=Seeds 1-6 (playoffs), red=Seeds 7-16. Bars are in ascending order from seed 16 to seed 1, with the height of each bar proportion to that team's probability of achieving that particular seed. The probabilities are generated from my team rankings and a 10,000 run simulation of the remainder of the season.

I debated whether to stack this vertically by week or team. Stacking by week puts each team in its own column, allowing one to easily see how each team's outlook evolved from week to week. The Broncos, by and large, are who we thought they were. The Falcons, on the other hand, were not (as their distribution shifts from blue to red, it's almost as if they're bleeding out).

Thursday, January 2, 2014

NFL Playoff Seed Probabilities - In Motion

In October of this year, I added playoff seed projections to my daily NFL ranking table, coverting each team's playoff chances into a bite-sized bar graph. As I did with my NBA seed probabilities, I can use Google's Motion Chart tool to show how those seed probabilities evolved over the course of the season.

Fans of the Panthers can see how their team's initially modest playoff hopes transformed into a two-seed and first round bye. Texans fans can relive their team's once high playoff expectations slowly dwindling and blinking out of existence, like so many dying stars.

Step 1: Select a team:
Step 2: Wait for chart to load
Step 3: Click Play button at bottom left of chart

Wednesday, January 1, 2014

Turnover Index: Annual Letter to Shareholders

To the Shareholders of Turnover Indices Incorporated:

2013 was a disappointing year for Turnover Indices Incorporated (TII). After posting moderate gains in its first year of existence, the fund suffered significant losses in Week 4 of the 2013 season. Gains were modestly positive in subsequent weeks, but not enough to offset that initial misstep. See below for a weekly summary of fund performance:

amount betprofitsending
4 2 0 $1,000 $268 (26.8%) ($268) $731
6 1 1 $731 $32 (4.5%) $29 $761
7 3 3 $761 $130 (17.1%) $118 $879
8 3 1 $879 $70 (8.1%) ($12) $867
9 7 2 $867 $166 (19.2%) ($66) $800
10 4 1.5 $800 $95 (11.9%) ($45) $755
11 2 0 $755 $14 (1.9%) ($14) $741
12 1 1 $741 $37 (5.1%) $34 $775
13 4 2 $775 $46 (5.9%) $21 $797
15 1 1 $797 $18 (2.3%) $16 $814
16 2 1 $814 $11 (1.5%) $2 $816

Top Games of the 2013 NBA Season (so far)

Win probability graphs have been updated with last night's New Years Eve games. To start off the New Year, I thought I would highlight some of the top (and bottom) games of the NBA season.

The Most Exciting Games

Each win probability graph has an associated "Excitement Index" (EI), which just sums up how far the win probability graph travelled over the course of the game. The most exciting game of the year, according to EI, was the December 2 triple overtime game between the Bulls and the Pelicans. The distance travelled by the win probability graph covered 16.9 "wins". The margin graph (shown below the win probability graph) shows that this was a game filled with multiple lead changes, particularly in the 4th quarter and overtime.

Ryan Anderson was the MVP of the game, amassing 85.9% in total Win Probability Added (WPA), in part due to two key three pointers, late in both the second and third overtimes.

Here are the top 5 games of the season, according to the Excitement Index: