Sunday, February 28, 2010

NBA Leader Martingale System 1

To start the research for the NBALMS, I will research the top 4 teams (by record), measure total wins, point differential, and duration of their two longest losing streaks, and record after 20 games (or 25% of the season).

Cleveland, 66-16, +8.9, 2 games, 2 games, 17-3.

Lakers, 65-17, +7.7, 2 games, 2 games, 17-3.

Boston, 62-20, +7.5, 4 games, 2 games, 18-2.

Orlando, 59-23, +6.7, 3 games, 2 games, 15-5.

Boston, 66-16, +10.3, 3 games, 2 games, 18-2.

Detroit, 59-23, +7.4, 3 games, 2 games, 14-6.

Lakers, 57-25, +7.3, 3 games, 3 games, 12-8.

New Orleans, 56-26, +5.3, 3 games, 3 games, 13-7.

San Antonio, 56-26, +4.8, 4 games, 3 games, 17-3.

I will go back as far as I can using ESPN's data, but so far, it looks like the longest losing streaks have been 4 games and three games.

The Martingale System revisited

The Martingale System is a betting system whereby the bettor doubles his bets until he wins. The theory being that the event being bet upon will eventually result in a win, gaining back all the money lost and plus one extra betting unit. The problems will the Martingale system are well documented. The first being that the event being bet upon is independent. This means that a long string of losses does mean that a win is "due up" and more losses are equally probable.

To read more about the Martingale system, click here.

However, the Martingale system could be used on something that is not independent, like the outcomes of sports game. My original research was with respect to the points spread, and unfortunately the results were inconclusive.

However, what if I combined the Martingale system with research done by David Berri and Wayne Winston? David Berri talked about the short supply of tall people. What he means is there aren't enough seven footers who are skilled for each team to have one. Therefore, the competitive balance in the NBA is the worst of all the major sports leagues. Good teams continue to have good teams for a run of years because there are simply not enough talented big men.

Wayne Winston said that an NBA players performance one season is most strongly correlated to his performance the past season; compared to players in other leagues (NFL, MLB, NHL), it is a greater predictor of future performance.

So while Brett Favre or Tom Brady may have a record setting season one season and mediocre one the next, an All-Star NBA player will perform at a high level more consistently year to year.
This research confirms my earlier hunch, that top NBA teams don't lose games to inferior teams because of chance, but because of effort. That a top NBA team (like the Cavs or Lakers this season) wins 75% of the games it plays and most of the losses are when the team is lethargic.

This leads to my version of the Martingale System, whereby bets of increasing amounts are to be bet on a single top NBA team to win a game. If the team loses, they will likely try harder the next night to "avenge" the loss; and two losses in a row brings lots of publicity and media harassment, whereby the top team tries even harder. And a top NBA team, trying hard, is going to win a lot of games.

The NBA team "has money" thereby breaking the "no memory" independent variable weakness of the Martingale System.

I shall now do some analysis of top NBA teams and the lengths of their losing streaks over the past few seasons to determine the length of losing streaks for the top two or three teams in the league.

This could provide me with evidence of a new system that could yield a net positive return over infinite time.



I finished reading Mathletics, by Wayne Winston, about a week ago. It is a much more technical book than Wages of Wins, which is itself a technical book. Wayne Winston does a lot of linear weights analysis to determine which sports have the most consistent performers, and which sports have the least consistent performers.

He found that basketball players performance are most dependent on how they performed last year. Football players, the least. And baseball, somewhere in between.

He obviously had many other findings, but the basketball section was the most interesting.

Saturday, February 6, 2010

Wages of Wins

I just finished reading Wages of Wins, by David Berri. The book was a birthday gift and after reading the other books I was given, it was time to crack open Wages of Wins. I've followed his blog for some time and really enjoy the posts and the insights he gives. It was time to read the book and I did.

I don't want to simply summarize the book or give chapter by chapter run down. Instead I'll list a few key points and also extrapolate a bit.

Scoring in the NBA is over valued with respect to paychecks and wins. High volume shooters who score hurt there team by using up so many shots and converting at a low percentage. Rebounding, defense, and assists are undervalue with respect to winning and pay.

This I already knew from his blogging. There is a heavy emphasis on regression analysis and multiple season data collection to come to conclusions. This makes sense. The more data, the more accurate the conclusions can be.

It also got me thinking about the proper place of anecdotal evidence. When I think of teams losing a lineman in football, especially a star lineman, and then losing, is there a way to properly study this? What about occurrences that are rare and so are essentially anecdotal? What about so-called black swans? This data do not necessarily lend themselves to the type of analysis that David Berri does. I noticed a difference between the types of analysis done; while I don't have the knowledge, database, or computing power to do the analysis he does, this blog can be more focused on anecdotal analysis and trying to distill low-frequency data analysis from the mountain of sports data being collected.

I don't have the means to do what David Berri does; but anyone can do what Nicolas Nassim Taleb does.


Are the Lakers focused enough?

The latest Laker debacle, a 126-113 defeat at home to the Denver Nuggets (playing without Carmelo Anthony) has made me want to write about the Lakers and there possible lack of focus.

Ron Artest wears shoes from a Chinese shoe company; he complains about foot pains and says that it feels like he is wearing lead boots and it has resulted in a lack of quickness on the court. Why is he doing this? The money. He is sponsored by them and has refused to switch to a better shoe.

Kobe Bryant has suffered many small injuries and has repeatedly refused to rest his body or seek treatment to fix them. He simply tries to play through the pain. But this will have a long-term cumulative effect on his game play and could manifest it self after a sustained period of stress....such as the playoffs.

They continue to take a band-aid approach to their point guard position, continuing to start Derek Fisher despite his defensive struggles and limited offensive creation. Rather than develop a younger talent and hope that he is ready for the playoffs, than are hoping that they won't need a good point guard to win the championship.

All these factors make me think the Lakers simply aren't as focused as they need to be to win a championship. Its not easy, after all.