Statistics with Probability

Bogdan Soltoianu

Mr. Brooks tries to summarize in his articles the ways which the almost unlimited amount of data available to us, impacts our decision-making. This question becomes increasingly important as more and more technology becomes affordable and easy to master and use. Politics, business, sports bets, teaching –they are just some of the fields where data is profuse and used constantly to derive the most effective strategies, generate earnings and increase effectiveness. But at what point does guessing outcomes by looking at previous patterns stop being and aid and directs us on a wrong path?

According to Mr. Brooks, there might be unexpected advantages to using information to think about things, both big and small.  First and foremost, we might be able to eliminate emotions from our decision-making, when this is advisable. He uses an example from basketball, but it might this be extrapolated to stock-market investments, or any other fields where is might be very tempting to connect past performance with future prospects, when there is no evidence that they are related.  It would otherwise be quite difficult to persuade a hotheaded supporter or a novice investor that there is no such thing as a sure bet, if their favorite team or company keeps performing well.  Secondly, data can help us better understand the past, so that we can identify patterns when they occur again in the future and tailor our reactions to them. While this is not absolute, it can apply to human behavior, and the results can seem counter-intuitive, such as in his example about the frequent use of the noun “I” which denoted lack of confidence, rather and its presence, in an individual. Also, who would have thought that older writers tend to look at the future with more optimism then younger writers. Maybe if we correlate this with other data we might change our general perception of aging, and look forward tour next decade, rather than dread it.

Conversely, having an abundance of data at our disposal can create unexpected difficulties, because it must be correctly sorted, analyzed and must take context into account. This is especially relevant in the case of human social interaction, where values, perceptions and experience can be more relevant and hold more weight than efficiency. Ms. Brooks uses the example of a bank deciding to take the money-losing path, because it decides that, whatever the numbers, in the long run, perception is more important. It is interesting to notice that while this might seem like a good long-term decision, because it might increase the public’s confidence in the institution and ultimately attract more clients, the decision remains entirely subjective and risky.  Perhaps we would gain a better understanding of how much of a leap of faith this was, if we had access to the numbers he used to gauge the likelihood that Italy’s economy would recover, and at what rate.  Tradition is strong in Italy, but our CFO is American, so they must not have been that bad.

Furthermore, in the case of the examples on the meager impact of increasing the amount of ads in political campaigns, it would be useful to know, while at the same time difficult to determine, what would be the impact on the public’s perception of not running the extra ads. While on their own they might seem not to have had an expected of proportionate impact, in specific contexts, they might have helped a candidate remain in the race, or maintain a competitive advantage, by compensating and drawing attention from other potentially damaging issues. The story here is not all in the numbers.

I believe there are strong limitations to what we can do with data, when it comes to human interaction, and while we are getting better at designing machines that analyze it and produce expected outputs, we are still a long way from an absolute solution. More likely, predictions using data analysis will remain a tool rather than a substitute for human decision-making.  While computers can pick up verbal cues and help us derive relevant some relevant conclusion, language is complex and involves many levels and subtleties. A good example is the Watson computer build by IBM’s R&D department as an example of artificial intelligence and loaded with massive amounts of random data. While winning Jeopardy was indeed an impressive feat, it remains to be seen if this machine can offer practical improvements when it comes to everyday activities, beyond that of a more evolved encyclopedia.  An idea that was often discussed when Watson became Jeopardy champion was using it as a diagnostic tool, to replace a doctor.  Programmers felt confident that they could design it so that it can accurately analyze symptoms and produce correct diagnostics and courses of treatment.  Doctors were however quick to point out that far from being a matter of gathering data and matching it to a set of possible results, diagnosis is a complex process that is based on what a patient does and how he behaves and speaks, just as much as on what he says.  This is data that cannot be inputted.

To further elaborate on Mr. Brook’s assumption that the CEO of the bank made his decision to stay in Italy “based on his values”, we can take a look at the trends in the news media about the banks’ public relations apparatus. The overall trend in Europe and the United States is for banks tochange their image. Some of the strategies involved in changing public opinion range from offering better investment on returns, or improving customer service, a move strategized by Brian Moynihan, CEO of Bank of America. Brian McCarthy writes in the Reuters blog opinion page about such examples of European banks such as UBS, Barclays and Deutche Bank – that try to change public opinion on how banks conduct their business. “Big banks — at least in Europe — are putting on a new, highly branded, and more contrite face.” Their strategies come under names such as Project Transform which was devised by Barclays or 2015+, a plan devised by Deutche Bank.

According to the Charlotte Business Journal, Bank of America has changed its public relations account to Burson-Marsteller, a global PR and communications firm. This change was brought because “The shift comes as BofA is looking to improve its public image following a years-long recession and financial crisis that saw BofA accept taxpayer bailouts, experiment with unpopular new consumer fees and suffer through a disastrous foreclosure crisis that put the bank at odds with millions of homeowners.” To answer the question of whether data did or did not play an important role in the Italian banks’ decision to remain in Italy, we can look at current trends and deduce they are a part of the long term business strategy for the banks. According to William Cohen, bringing real changes to banks is to hold bankers “financially accountable, up to their entire net worth, when things go wrong”. Brooks was wrong; data did have an important role in the bank’s strategy. The field of public relations is heavily data-driven. Strategies are devised on polls and statistics. Ultimately the bank had probably more to lose by leaving the country than by weathering the storm.

Ryan McCarthy (February 20, 2013)How to do PR for Banks  Available: http://blogs.reuters.com/great-debate/2013/02/20/how-to-do-pr-for-banks/

Adam O’Daniel (January,29 2013) Bank of America will move PR account to Burson-Marestellar. Available:  http://www.bizjournals.com/charlotte/blog/bank_notes/2013/01/bank-of-america-will-move-pr-account.html

Bank of America CEO tells employees to give customers better service. (January 29, 2013) Chicago Tribune.

Available: http://articles.chicagotribune.com/2013-01-29/business/chi-bank-of-america-ceo-tells-employees-to-give-customers-better-service-20130129_1_bank-of-america-ceo-ceo-brian-moynihan-customer-service