Your verification ID is: guDlT7MCuIOFFHSbB3jPFN5QLaQ Big Computing: September 2011

Thursday, September 29, 2011

MLB end of the season 2011. Unusual? Momentum in Baseball? Tired?

So as I am trying to recover from lack of sleep because of last night's extra inning and rain delayed MLB games. I am starting to ask a few questions about the last day of the regular season itself.

First was the number of extra inning game unusual? The answer appears to be yes, but not unheard of. About 10% of all MLB baseball games go into extra innings so last night was not really that out of whack. I am still looking into overcoming a seven run deficit to win a game with less than two innings left which I do believe is an extremely rare event.

I also noticed that the both the Rays and the Orioles were the teams that came back late in the game and went on to win. So my knee jerk reaction was that momentum was a major factor in those teams winning the game, Momentum is huge in sports. Then I had a flash back moment and realized that momentum in sports is a popular truism, but a practical myth. Here is the link that talk about the lack of momentum in Baseball.

Either way praise is in order for the Tampa Bay Rays and condolences for the Boston Red Sox. The Sabermetrics guys at Tampa Bay do an amazing job of finding potential on a limited budget, and the Red Sox's analysts like Bill James and Tom Tippett showed them the way. Remember we never fail, but always see ways to improve our prediction

Wednesday, September 28, 2011

An Update on the Floop iPhone App for Polling and Surveys

Floop continues to very well with something in the area of 30,000 downloads so far which is amazing given that it was kind of a stealth launch. In addition to significant downloads and good user traction Floop is starting to get some press. I have posted some of the links below:

Tuesday, September 27, 2011

The American League Wild Card Race

So the Boston Red Sox and the Tampa Bay Rays are all tied up with one game to play in the regular season. I am a big Sabermetrics fan and the timing of this event with the recent release of Moneyball the movie is ironic, but I believe in the last week I have not looked at a single stat. Instead I watch night after night riding the joyful rollercoaster of emotion. I had forgotten this part of baseball. Dan ducquette once told me to remember to enjoy the game, and this week I have. After the regular season is over and the emotion drains away I will sit down and become the mathematician that I am and try to understand what happened here and why. However, until then I will watch the Red Sox And the Rays play their last games of the season purely for the enjoyment of it.

Monday, September 19, 2011

Strata NYC Jumpstart 2011 wrap up

Today was the first day of Strata NYC 2011 which is a five day Big Data conference in New York. The first day of talks are the so called Jumpstart talks that deal with the issues of Big Data at a very high level. The down side for me on the high level talks is they lose the granularity and coolness of the more detailed analytics talks that I go to. Nonetheless they were good talks given by some very influential thought leaders in the field of analytics and Big Data. Here are some of the highlights:

Nolan Goldberg gave the first talk and potential buzz kill session of the day.  Nolan is a lawyer and gave a sweeping overview of the development of the law with respect to big data and privacy. Basically the law and the technical development pace has resulted in a moving target. Solutions that protected people's privacy only a short time ago are no longer capable of doing that. This poses problems for figuring out what is ok, and determining if your solution or application might have a legal problem. Also as the US, individual states and the EU try to address this issue the is no reason to believe they will do this in sync or even come up with a similar end solution. Kind of makes me feel like I am flying without a net, but at least Nolan was honest about it.

Cathy O'Neil, Hiten Shaw and DJ Patil gave talks about various requirements to hire, utilize and build cultures that are successful with data scientists. I loved their approaches to organizations. I do not care what industry you are in if you empower your people in a positive collaborative environment you will succeed.  Simple ideas, but here are three people who followed their own advise and were successful with it.

Simon Wardley gave a talk on the development process of an idea from new to a commodity. Hopefully I can use his thoughts to avoid following my next business from boom to commodity and death. That ride down sucks so I hope Simon's ideas help.

Michael Nelson give great talk on transparency of your data. I liked Mike's take on this. I really enjoyed the idea that we should start by thinking we should share all data instead of we need to protect all data. We have been so paranoid about this issue for so long maybe by starting with the counter argument we will achieve the righ balance.

Talks dealt with the importance of reputation and credibility online. Turns out that is more important today then ever before and a good Data Scientist can put a value on it ( something like 20%).

Overall a good day, although I can not wait until we take a deeper drive.

Friday, September 16, 2011

Slides from the weeks R and Open Statistical Programming Meetups

There were two great meetups this week, and I want to share the slides with everyone.

The first slides are from the R for Beginners session at the Greater Boston Area R Users Group. I am so glad that a useR group finally did this, and that it was well attended. In fact, I think it ws double the usual attendance for this group, and there was a waiting list! Also good to know that they are planning another R for Beginnners session later in the year. Here is a link to the slides

The recently renamed NYC Open Statistical Programming Meetup which is run by the recently married and employed Drew Conway had a talk by Wes McKinney. Wes gave a thought excellent talk on the things he is working on in Python and some thought provoking opinions on both Python and R. That was a brave thing to do since most of those in attendance are primarily R users. I do not personally agree with Wes on some of those opinions, but his presentation has already resulted in a healthy discussion of the issue. Here is a link to the Slides

Thursday, September 15, 2011

R Users Group in Connectict (#rstats)

I am not sure if I am tired of driving to the R User meetups in NYC and Boston or if I want to prove that we can have a great RUG in Connecticut, but I went ahead and formed a group. I set it up on Meetup this morning. I have not set a date, topic or location yet, but as soon as we get enough people I will get that done as well.

Connecticut has a long history with the R language. We formed Revolution Computing ( now Revolution Analytics) in New Haven, and many of the people in New Haven who helped us get off the ground were avid R users and contributors long before that. In fact our first customer were also Connecticut based.

Please join up, and I look forward to seeing you at our first meeting.

Wednesday, September 14, 2011

Slides from the R for Beginners Meetup in Boston

Last night the Greater Boston Area R User Group had their monthly meeting which featured talks on R for Beginners. The waiting list to go to this talk was longer than those who got to attend! Obviously these kinds of talks are really needed and thanks to this group for doing one and already planning another. Josh has already put the slides up on the site. Here is the Link.

Monday, September 12, 2011

My Memories of 9/11 - Still Horrible after ten years

Every year I dread the anniversary of 9/11. There is not a day that I do not have a memory of that awful day. The anniversaries are even worse for me. All the anger, frustration and helplessness for failing to stop this evil thing from happening just overwhelms me.

The twin towers used to be a happy memory of my youth. I first became aware of them when WPIX Channel 11 in New York adopted them as there symbol in the 70s. For whatever reason I loved those commercials.

When I was in High School I got invited to a big New York Party at Windows on the World. The party was no so great, but I loved the Restaurant and the view! I had never been more than 12 stories above the ground in my life. Amazing!

I went to the towers whenever I was in the city. I loved the place. When I was in college my fraternity had a scavenger hunt that had us streak the top of the Tower. We got away uncaught by the Port Authority cops, and returned to Ithaca with a good story to tell.

After college and grad school I went to lower Manhattan less and less frequently. On September 10, 2001 I bought tickets to a show, and recalling pleasant memories of the towers in years past I made a final reservation at Windows on the World for Friday September 15, 2001. The next morning I watched the towers burn and fall at Eli's in Hamden with Miles Russell, a good friend. Some memories I wish I would forget.

Wednesday, September 7, 2011

Four Tips to get the most out of Floop

Floop is a new iPhone App that takes the mircoblogging of twitter to a whole new level.  Floop provides the social media user the ability to gauge the thoughts and opinions of the community dynamically with an opinion graph and text. I think that Floop is both a very entertaining application and a powerful social tool.  Here are the things I think you need to do to improve your Floop experience:

  1.  Follow Interesting People: Social Networks are built on associating with interesting and thought provoking people. Floop provides a method to find these people by looking for Floops that you like or by finding an individual's comment on a Floop that interests you. Also just like any social network the people you follow are likely to follow you back.
  2. Find interesting Floops and participate: This is not hard to do, and it is fun. The quality of your participation is probably more important than the volume of Floops you do. Comments like “yes or “no” add little value to the dialog, but quality comments improve the Floop, and get more people to follow you.
  3. Create interesting Floops: There is simply no better tool out there is get real time crowdsourced feedback so use it. Test the community’s opinion on your ideas or convictions. The more thought provoking the better.
  4.  Participate in the Floops of people who follow you. Floop's value is in the ability to query the community and for the community to voice its opinion on a topic.

Other than that just enjoy this great new social media application!