April 2012 Archives

10 Papers/Books You Should Read

am_turing.png2012 marks the 100th anniversary of the birth of Alan Turing.  Arguably he is the most important computer scientist ever, and the ACM A.M. Turing Centenary Celebration will be held June 15-16, 2012 in San Francisco to commemorate his birth.  This will be an amazing event.  Thirty four Turing Award winners will be attending.  The first Turing award was given for the year 1966 (46 years ago), so having 34 award winners in the same room is remarkable in its own right.  For those not in the field of computation, winning  the Turing Award is our industry's equivalent of winning a Nobel Prize. For those of you in the field, if you are are not desiring to attend this event, I just can't take you seriously.

In commemoration, I present my top 10 list of all time greatest papers/books written in the field of computation.  At least,the top 10 papers and books that I've read and that h
ave influenced me.  I consider this all most required reading for anyone claiming to be in technology.  In no particular order of importance:

10. "Hint for Computer System Design",  Butler Lampson.

9. "The Mythical Man Month," Frederick P. Brooks.


8. "Bayesianism and Causality, or, Why I Am Only a Half-Bayesian," Judea Pearl.
(In 2012, Turing would have been 100 years old.  So, I guess this is really the 99th anniversary and arguably the centenary of his birth is in 2013.  I find this "off by one error" to be deliciously ironic.)

It's Important to Have a Goal

"Would you tell me, please, which way I ought to go from here?"
"That depends a good deal on where you want to get to," said the Cat.
"I don't much care where--" said Alice.
"Then it doesn't matter which way you go," said the Cat.
"--so long as I get SOMEWHERE," Alice added as an explanation.
"Oh, you're sure to do that," said the Cat, "if you only walk long enough."

-- Alice's Adventure in Wonderland by Lewis Carroll
Moneyball: The Art of Winning an Unfair Game by Michael Lewis is a great story.  I read the book, saw the movie, and heard Michael lecture about it.  And, yet, at the end of each, I was left with the same question.  What was this story all about?  Or more precisely,

What was the goal of Billy Beane and the Oakland A's during the time period when Billy was the general manager?

Reading the book, watching the movie, hearing the lecture, and even searching Wikipedia doesn't make this clear.  Was it:

  • Win a World Series?
  • Win the most games?
  • Increase yearly profitability?
  • Increase the market value of the team?

You might say it's nitpicking.  I say these are pretty different goals.  And goals that might not be consistent.  And, from an objective standard, if the goal was to win a World Series during this time period, then Billy and the A's were unambiguous failures.

Sure, they were able to win a lot of games.  And, they were able to win a lot of games per dollar spent. Some argue this was the goal.  But, in reading, watching, and listening, this is not clear.

I have no idea about increased profitability or increased market value. Judging by the empty seats -- rather the tarp covered upper deck  -- at the Coliseum, I'm guessing, they failed at the latter potential goals as well.

Statistics:  A Sound Tactical Approach

In general, I'm a believer of the approach taken by Beane -- using math and statistics to make objective decisions. Which, lead to good results. Define metrics, test, measure, repeat.  If you throw in some intuition and "hunches," I think it  works even better.  Stuff like Bayesian Inference even allows for this.  And, going off the deep end, there's even hope for a science of causation to be even more effective.  Thank you, Judea Pearl.

However, in Moneyball, all the great math, statistics, and science seem to be for naught.  Why?  Because of the lack of clarity regarding a goal.  What were Beane/A's trying to accomplish?  I have not idea if they were successful.  Whether they chose the right metrics.  Or tests.  How could they without knowing what they were trying to accomplish?  (Is it possible that they did know and only I missed the point?)

In the words of Yogi Berra, misquoting Carroll:

"If you don't know where you are going, you will wind up somewhere else."

The good news is that, maybe "somewhere else" is somewhere good too.

What Does This Mean for Silicon Valley Startups?

It seems so obvious.  Having a goal is important. It's strategic.  However, sometimes we lose sight of this because we get overly wrapped up in the tactics.  You know -- why it's important to test and measure.  Run that A/B test.  Understand your customers.  And, the granddaddy of them all -- the importance of the pivot.  While these tactics are important, it seems like we should be laser focused on understanding the big goal.

Maybe you measure your company against lots of visitors, big revenue, or awesome user experience.  You might be "hitting it out of the park" on these metrics.  And, sometimes, you prioritize your actions against these metrics.  In the end, however, how do these numbers reflect the progress towards your goal?  Or worse, are you spending resources to achieve these metrics and losing sight of the big picture -- your end goal?

Of course, maybe I'm wrong.  You could wander around drunk on Burbn, stumble onto some pictures, and then exit for a billion dollars.  Then, it doesn't matter what your goal is or was. Without question, I'd say a billion dollars is success. 

And, then again, Michael Lewis didn't do so bad for himself either.

(Aside: Here's my naive opinion on what it takes to win the World Series:  You need to put the right arms on the mound to beat the best teams in a 5 and 7 game series.  Being able to beat a lot of mediocre teams won't get you there, even though you will win a lot of games.  Pitching is key.  Good pitching costs money.  You need the pitchers to win.  Thus, beating the best in the World Series costs money...Unless, as Bill Gerth reminds me, you were the 1975 Cincinati Reds.)