Probability and Statistics: Gambling on Public Perception with Misleading Numbers

Hello and welcome to talking points. I’m dave kelly, director of advanced media production at Long Beach. Today our topic is probability and statistics. We’ll try and determine if it really is all about the numbers.

My guest today is Dr. Omer Benli. Dr. Benli is the associate dean in the college of business administration. welcome omer, and thank you for joining us on talking points. >> omer benli: thank you. glad to be here.

>> dave kelly: well, as we talk about this topic of probability and statistics, it’s hard to overstate the pervasive way in which numbers, statistics, and probability are used in everyday life. it’s everything from buying insurance premiums that are based on actuarial tables and actuarial science. and then we have during election season, we’re all inundated with statistics from both sides to try to sway our vote.

And we know that companies are using massive amounts of data to determine strategic direction and also marketing campaigns. and we know that the government is using copious amounts of data to regulate industry, to solve crime, and deter terrorism, and things of that nature. so given that this is the case, what can we do today as individual consumers of information? are we statistically literate enough or is there more to be done in that case? >>omer benli: that’s a very important point.

In my personal view, i don’t think the general public is well educated in this area or as much as they should be. it had been and recently with the advent of computers and computation that become so easy, and the collection of data is immense in these areas, and the conclusions that presented to public of some analysis based on this data can be misleading. i think the public, a general reader, general newspaper reader should be very much questioning how the data is collected based on what assumptions and the conclusions that whoever is stating the conclusions is based on.

So it’s a very, very responsible thing on the part of the public, i think, to be very much questioning this analysis that we are receiving. >> dave kelly: and today with the advent of the internet, the ability to capture data has grown dramatically over the years and it’s only going to become even more so in the future. and so we look at numbers maybe a little bit differently than we did 30 or 40 years ago, and certainly the numbers are used by all of industry now to try to determine what’s going to happen for future business opportunities.

Let’s take, for example, the insurance industry. they use actuarial tables and actuarial science to determine what kinds of premiums they’ll be charging their customers. explain the theory behind actuarial science and how that’s used to determine these premium rates. >> omer benli: well, this is a very good point. it is insurance as you well know is an essential part of modern day living and it has a long history that goes back.

And, basically, those tables or the guides how much to charge for certain things is just the spading the risk basically. it’s based on huge amounts of data for a particular subset of populations. depending on your age, for example, what is your depending on what you are being issued for is the chances that you need to collect from the insurance. so based on these huge data sets that the companies have the methods, statistical and probabilistic methods, are designed to have some forecast as to some prediction which is not always hundred percent because it is you are telling something to the future.

So based on these statistical approaches, those tables are generated depending on each subpopulation that is of interest. >> dave kelly: and you brought up a good point that although they try to predict the future as much as they can, probability does not give us an exact prediction of the future. what it does is it allows us to take a look at past data and then extrapolate from that what the most likely outcomes will be in the future. and another example of where this is used is in forensic science.

Now we’re going to have another show, we’re going to go into the whole aspect of forensic science in great detail. so for the viewers, you want to stay tuned to talking points because that show is coming up and we’ll get into all of the details about for forensic science. but for purposes of this topic, let’s just touch on it for a moment. forensic science, dna analysis, what it’s all about and sort of the broader sense? >> omer benli: this is very similar to the previous topic that we mentioned. basically, what they do is they gather data from the past of similar instances and they try to predict.

And forensic science traditionally been used in the courtrooms and presenting the evidence to support whatever the claim is in question. so they basically, again, they look at huge amounts of data that had been generated for that particular area and then they make a prediction or they have a likelihood of whatever the conclusion is. again, exactly, it is not hundred percent so sometimes the way that is also reported to public in the case of famous court cases, for example, can be misleading.

So probably when someone reads it in regular english it is very difficult to differentiate what is exactly meant by the expert’s opinion on say dna testing and the similar results. >> dave kelly: right, and we should talk about that for just a moment. if they talk about a dna match, we’ve all watched csi, the various programs about csi which are very popular on television that they talk about we have a dna match. well, that’s not exactly correct. it’s not 100%, because what the csi forensic analysis is really telling us is that the likelihood that this particular suspect committed the crime is very high based on the analysis of the dna structure that was exhibited in the evidence. however, there may be ten other people in the world that have similar dna structure profiles but since those ten people don’t live in the vicinity and this suspect does live in the vicinity at the time the crime is committed we’re going to call that a match.

>> Omer Benli: exactly. and that is why it is so important that someone, jury for example, would decide on the likelihood of this crime is being committed by this result. so all of these are decision aides to decision makers in the case of the jury or the judge to make a decision. so it is not a hundred percent certainty because we live in a probabilistic world and everything is based on likelihood. these sciences do not give an exact hundred percent accurate conclusions.

>> Dave Kelly: well, i think this topic came up in another show that we did where we said the science is digital but our lives are analogue. we’re analogue people in a digital world >> omer benli: exactly, that’s right. >> dave kelly: in terms of the numbers. let’s move to another topic which is gambling because gambling is all about probability. and for those that spend time in the casino they know that they usually walk out a loser but there’s a certain investment of time and entertainment value that goes along with that. so you assume a risk when you go in the casino.

And so people apparently assume that the entertainment value outweighs the risk of losing the money. >> omer benli: exactly. the thing is that as the odds that certain they say in gambling, again, is based on probabilistic odds and those odds are either formalized through some theory. and in some other cases it’s based on statistical analysis of past information. again, it’s very similar to the other sciences that we have been talking about is that looking at the past or the theory behind it how to predict something. this is how, for example, a gambler bets on however much he or she is betting on a particular [inaudible] based on the intuition or just the past experience.

But the fact that as you put it very well is the fact that the gambling is especially playing against the casinos, one should not expect to beat the real money casino sa. the idea here is that one should always remember the entertainment value for attending such events rather than trying to beat the casino or hoping to beat the casino. >> dave kelly: right, the house always wins and the main reason is because they have such deep pockets. >> omer benli: exactly.

>> Dave Kelly: we can’t hope to compete with what they have in terms of resources. >> omer benli: exactly, exactly. >> dave kelly: and when the california lottery came into existence back into the 1980’s, i remember they were talking about the statistics and the probability of winning the lottery. and at that time, i believe the statistic was 14 million to 1 meaning that if you bought a lottery ticket your chances were 14 million out of 1 that you would win. and as an analysis comparative to that, someone stated the fact that well the chances of you winning the lottery are about the same as being struck by lightning, not once but three times in three different locations.

So, in other words, you’d have to be struck by lightning three times in three different locations to have the same odds of winning the lottery which reminds me if you do survive three lightning strikes maybe you should buy a lottery ticket. >> omer benli: exactly. exactly, my point. >> dave kelly: but moving along here. i guess we can also use statistics to mislead people intentionally if we want to if we’re in marketing camp if we’re with a marketing campaign and we’re trying to sell a product, we can use a statistic that is technically true but not an accurate representation of what really exist in the real world. for example, you might see a diet weight loss product that says if you use our product you have three times greater chance of losing weight.

Well, if you look at the analysis comparing the other products, your initial opportunity to lose weight might be 1 out of 100, but with this product that has three times greater it’s now three times out of 100. so that’s three times greater but you still have a 97% chance of failure in this case. >> omer benli: absolutely.

And this is very important point especially in the business world. it is not as you said, it’s not technically incorrect what they have stated in that but it is it really opens the doors for misunderstanding by general public. now is this an ethical issue, is this proper to say that knowing that what an ordinary person will interpret that the result is a very important issue. for that point is very important in our business education, educating business professionals.

We do have programs, ethics across the curriculum in every course that there’s an ethical aspect, and we are trying to educate our students, the future business professionals is this an ethical statement or not. so this is something that is very important. if one says technically correct things but nobody really has like in that example that you mentioned is know who are going to look at the data.

In these commercials, the data where this conclusion comes from is not really given. if one has to find that data then the results are 97%. so this is something the public should really question and some other mechanism should be inherent in our advertising and other practices to have some sort of control over these things. >> dave kelly: we’re going to have to take a break right now. and stay with us. when we come back, we’ll continue this conversation about big business, big government, ethics and probability.