100 per cent improvement

I am not sure where I first heard the expression: "There are lies, damn lies, and statistics."  It could have been something dad relayed, or maybe it was a Statistics teacher I had at the University of Albany.  Regardless, the phrase has stuck with me.

It means that statistics can lie--more accurately, it means that people can lie with statistics. You can do that by sleight of hand because many people just don't get stats--even simple stats. Sometimes people don't even know they are lying with statistics. Often people lie with statistics as a conscious act.

So, let's take the title of this blog post.  100 per cent improvement.  Sounds good. One would like anything to improve 100 percent.  Quality of diet, air quality, income level. Whatever. However, taking the last item: income level as an example: if your employer was paying you 1 dollar per year that would be an abysmal salary. If he or she would say they raised your salary 100 percent, it might sound good, but it would only mean that now you were getting two dollars a year.  Still, you cannot buy a candy bar unless it is on sale.  Toothpaste entrepreneurs used to claim that brushing with a particular brand reduced cavities 100 per cent. Compared to what?  A group that never brushed teeth at all?  A group that brushed teeth with candy cane shards?

Well, I improved my March Madness prognoses 100 per cent on Friday. On Thursday the first group of 16 games were played when I predicted, as relayed in a prior post, a grand total of 2 games accurately.  2 out of 16.  12.5%.   Friday I was something else. Predicting 4 games accurately.  25%. A 100 percent improvement.  Clairvoyant, right?  A true pundit.

I know quite a bit about college basketball and yet, against the spread, I picked only 6 of the first 32 games correctly. It is true that two of those games I "predicted" with my heart and had I really used my limited wisdom on those games the tally for Thursday and Friday would have been 7 wins 24 losses and one push (tie).  On the money line I have picked 21 out of 32, but picking the money line during the first two rounds is not that difficult.  Consequently, 21 out of 32 is nothing to write home about.

Bottom line: soothsayer not. And the moral is, if you are going to Las Vegas to make money, think again.  This has not been a good year for me-and fortunately I have not wagered a dime--but even when I have done well, and even if I do well today, the chance of bringing home shekels is extremely slim.

With that as preface, here are my predictions for today's games.

1. Villanova take the 4 against Purdue
2. FSU give the 5 against Murray State
3. Kentucky -5 1/2 over Wofford (take the money and run)
4. MSU -10 over Minnesota
5. Michigan - 6.5 over Florida
6. Auburn give the 2 over Kansas (even though Auburn barely won in the first round)
7. Gonzaga -11.5 agains Baylor
8. Maryland +2 1/2 over LSU