My Photo

Your email address:

Powered by FeedBlitz

November 2016

Sun Mon Tue Wed Thu Fri Sat
    1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30      
Blog powered by Typepad

Become a Fan

« The Data is the Query | Main | Master Data Management (MDM) vs. Sensemaking »

April 09, 2011


Feed You can follow this conversation by subscribing to the comment feed for this post.


I'm ecstatic to have found your blog. Actually, I skipped a few years of high school but ended up with a Ph.D. anyway. This means that in some areas of math I am superb and other areas a bit iffy, yet I did just fine as a statistician. This taught me that there is often more than one way to do things, whether it is with calculus, algebra, geometry or just pounding on it with data. It also gave me a healthy skepticism about some of the answers produced by our equations. Sometimes, as you point out, the math hides more than it reveals.

Marcel Blattner

Hi Jeff

Honestly I never understood such debates. It remembers me the discussions in physics where theoreticians fight against experimentalists whether the theory needs the experiment (observation) or vice versa. But it is so obvious that both need each other. I think Sir Arthur Conan Doyle was simply wrong with his statement: “It is a capital mistake to theorize before one has data”, because there may be situations where it is better to think (theorizing) about some "scenarios" based on different "data topologies" before having the actual data, because it could save time, money or even prevent some serious damages. And of course there are many situations where you simply need the data before you can do anything useful. So the statement from Sir Arthur Conan Doyle is only one part of the game.
In fact there are only "SOLUTIONS". Whether these solutions are approached by hard math or by some clever "on the fly" ideas, which might be pretty unconventional but functioning is not relevant at all. Use math where it is needed and use "non-math" approaches where it can help you to come up faster with a SOLUTION. Then solutions are the only thing which matter. Have a nice day!


I'm glad I came across your blog Jeff :).

I'm a college undergrad. And I totally agree. This is a matter of common sense, don't it? Pounding a problem with math while having incomplete data is just... stupid. It's common sense that if you don't have enough evidence to a "crime", then you go roam your beat, do stake out, ask informants, whatever.

My old professors (years ago) said that one needs to be really good in math to be working with computers. They were wrong. I'm working with computers, and I'm not good in math.

Your post reminded me of the Numbers TV series by the way :).

Great post, Jeff. Isn't what you are suggesting the essence of the scientific method. Propose a theory and then prove it. In your case, you propose the two records are the same and then gather data to support that theory.

What's changed is we are suddenly able to gather and understand more data in more different forms than ever before. Proving theories (within an acceptable statistical error margin) has never been easier.


Interesting post. What you are doing here is to create a hypothesis based on the data that you have. Isn't that biased? Wouldn't it be better to create a hypothesis and then gather data to check if its valid?
Just a thought :)

Account Deleted

brilliant, I thought I was alone on this one. Now I have one more data point :)


Hi Jeff, I just got your blog. And many more than a few post are great. Data beats math just remainds me that how both sides of our mind work together to get something cool.

The comments to this entry are closed.