LEGOS:
1. Square red pieces can be stacked to build towers.
2. Long skinny pieces can be used as roads.
3. Long skinny pieces on top of square red towers can build bridges.
4. Curved skinny pieces can make turning roads or bridges.
5. Tubes, motors, wheels, etc. can make pretty much anything you can imagine.
STATISTICS:
1. Univariate statistics can be used to clean errors out of your data.
2. Correlations of past data with current data can predict future events.
3. Linear Regression can combine many correlations to predict even better.
4. Logistic Regression can predict curved relationships.
5. Simultaneous equations, Non-Linear Regression, etc. can be applied to improve pretty much any business process you can imagine.
THOUGHT PROVOKING QUESTION
Which of the following best describes how is your company is organized?
A. We hire really creative builders who have access to all the pieces.
B. We hire creative people who have only ever seen the red square pieces to tell the builders what to build.
C. We hire builders who have a big sack of Legos with lots of different pieces, but they’re not too interested in building anything they haven’t seen before.
D. We use robots to automatically build “stuff” that we use as bridges. The bridges usually work, but sometimes we fall through.
E. We don’t have any Legos.
David Young
March 11, 2011
With permission from the original responder:
James Long, Senior Risk Economist at RenaissanceRe
Commented on another group the following:
David, I like this post, in particular the thought provoking question at the bottom. I’ve been in firms who didn’t realize the needed people who understood all types of blocks and that firm tended to define problems as a ‘red block’ problem and then hire ‘red block experts’ to work on these problems. It produced a very dysfunctional analytic community.
David Young
March 11, 2011
With permission from the original responder:
Pablo B, Entrepreneur, http://www.tweets4investors.com
Commented on another group the following:
Fun comparison! The most important piece is the imagination of the child playing with the Legos
David Young
March 11, 2011
With permission from the original responder:
Josh Franta, Entrepreneur – Finance, http://tradelink.googlecode.com
Commented on another group the following:
good metaphor
David Young
March 11, 2011
With permission from the original responder:
Vincent Granville
CEO – Business Analytics, BI, Web Mining, Ad Optimization, Fraud Detection, Scoring, Predictive Modeling, Web Crawlers
Commented on another group the following:
The A people usually develop the robots describe in D. If you don’t hire them (and they are very hard to hire), they will be your competitors.
To which I responded:
Well, I suppose my characterizing the robot built bridges as sometimes falling down showed a personal bias of mine since I’m more inclined to work on problems with a lot of financial leverage behind a single task. In that situation it normally pays to have someone think carefully about the problem, rather than sending it to the automated solution finder.
However, I’ll readily concede that there are some problems where a fully automated approach could be the best approach. Text Mining of complaint forms comes to mind as an example. The text could vary so much that any predefined solution would be difficult and the topics being complained about would change over time. In that kind of a scenario I think an auto-miner searching for high priority letters to be read could be the best approach combined with somebody reading a random selection of the other letters to catch what could slip through the miner and give feed back to the miner over time.
David Young
March 11, 2011
With permission from the original responder:
Mickael Bäckman, Method and Model Developer
Commented on another group the following:
Hi David.
I think an important variation is:
B2. We hire creative builders with stories about all the fine quality pieces that are available but consequently only give them square red pieces to build with because ‘that is the way everyone else does it’.
With permission from the original responder:
Joel Brodsky, Analytic, CE, and Text Mining Professional
Commented on another group the following:
As a long time statistical consultant and analytic strategist, I want to thank David for his thought provoking question. Based upon my professional experience as the leader of internal analytic teams and as an external consultant, I believe he really highlighted the key points: Is one really interested in advancing his/her competitive advantage (answer A), only interested in appearing to do so (answers B-D), or not interested at all (answer E).
Joel
David Young
March 11, 2011
With permission from the original responder:
Chenghui Cai, Quantitative Researcher at Cantor Fitzgerald and Founder of AITrading.org
Commented on another group the following:
Very nice, David 🙂 I personally like building robots that build up stuff with a higher probability of success than failure.
David Young
March 11, 2011
With permission from the original responder:
John Totten, President at Totten & Associates
Commented on another group the following:
Another plug for creativity – I have seen good people with wildly different tool kits start with the same data, follow two quite different analysis routes, and converge on the same results and recommendations.
David Young
March 11, 2011
With permission from the original responder:
Tom Hughes, Leader in online marketing, systems development, project management and product strategy
Commented on another group the following:
My company does advanced mean-reversion research and we help our customers use that research in their trading. We aspire to “A” (creative builders using all the pieces) and I think we often succeed; when we don’t succeed, we tend to fall back to “C” (builders with lots of pieces but sometimes only building what they’ve built before).
The metaphor works well for the “original research” dimension of the company, which is naturally the indispensable value-originating piece. There is also a big distribution side — we take the research and turn it into services (subscriptions, education, membership services) that benefit our customers. On that side, “A” is not so desirable, the customers put a value on repeatability and predictability. The services each have unique value-creation but should actually all work in similar ways as much as practical. In that dimension, we aspire to “C” and when we fail we tend to fall back into “B” (creative, but really only comfortable with the red square bits).
David Young
March 11, 2011
With permission from the original responder:
Michael Bigger, CEO, Bigger Capital, LLC
Commented on another group the following:
We hire artists who can build never before seen shapes in a trading sandbox.
David Young
March 11, 2011
With permission from the original responder:
John Meredith, Sr. Manager at Perot Systems
Commented on another group the following:
Personal Opinion of John Meredith. I love the analogy. We are just beginning to apply the power of the EMR. It has been my contention that if you build an EMR the associated applications must support workflow, make information accessable in meaningful way to clinicians, actionable through alerts, reportable so data can be easily extractable and analyzed. It is my belief that the recent legislation affecting healthcare automation is ultimately aimed at getting to the analytical components. With ICD 10’s, the data resulting from automation and sent to the CMS will give the CMS the ability to compare resource consumption for a given diagnosis and/or procedure from multiple sources. This then can easily lead to evidenced based national guidelines of care.
David Young
March 11, 2011
REPOSTED WITH PERMISSION OF THE CONTRIBUTOR
Group: Advanced Business Analytics, Data Mining and Predictive Modeling
Discussion: Statistical Modeling is like building with LEGOS
Enjoyed reading your blog! Gave me something to think about on how to improve my models. Thanks!
Posted by Marty Epstein
David Young
March 11, 2011
REPOSTED WITH PERMISSION OF THE CONTRIBUTOR
Group: Citi Global Alumni Network (Citi Network)
Discussion: Statistical Modeling is like building with LEGOS
Sidenote: Provided the independent variables don’t suffer severe autocorrelation, regression should be fine. A!
Posted by Jaime Adriano Jr
David Young
March 11, 2011
REPOSTED WITH PERMISSION OF THE CONTRIBUTOR
Group: Citi Global Alumni Network (Citi Network)
Discussion: Statistical Modeling is like building with LEGOS
Thank you David, very good thread!
Posted by Lloyd Malcolm Nguyen (Minh Linh Nguyen)