Lies, Damned Lies and Statistics

I hold a PhD in statistics and perform reseach in clinical trials. I also review statistical sections of peer reviewed articles for publication. I can tell you one of the fun things about sports statistics is that we do not hold them to any real accountability. Therefore, these stats should be taken with a dose of enjoyment…but far from objective. If we really wanted to hold this up to a high standard, we would need somebody to decide the hypothesis they wanted to explore and then the methods (statistics) they were going to use. After agreement that these are the best statistics, the research would begin. In a perfect world, this would be a prospective study and we wouldn’t use past stats, just stats going from this point forward.

Sometimes I laugh when I think of all of the positive statistical posts on Orton on this site. I could be led to believe he is one of the greats of all time. But as long as we except that this is not objective statisitcs then it is fun to enjoy.

I always lean on the stats-heavy side when I post, but I know that they can be manipulated, if not kept in perspective, but I’ll defend my stats till the day I die.

I personally think the greatest value in statistics is that they can provide the observer with a way to identify patterns and trends — such as on 2nd & 4 with 6 yards to go, is the Broncos’ tendency to pass or throw & to which side?

Although, most any debate can be argued either way, not only those involving statistics.

For example, one person could say “Josh McDaniels is an arrogant and cocky dictator”, while another person could say “Josh McDaniels is a confident person who takes control of the locker room.” Both statements could be used, depending on your point of view. Bias is a fairly difficult thing to get rid of, statistically or in prose.

They’re also not the end-all, be-all of any debate.

I completely agree with this. Stats can’t be trusted, by themselves, but neither can opinions. Everything must have some type of context or reason, but in the end, it’s important for us all to be able to determine what constitutes a fair and balanced stat/argument and what is an unreliable, ridiculous lie.

during an eight year period where I really tried to “play” the stock market. I had no real knowledge of stats before that. What I found out was that if you didn’t use statistics while others were, you were sure to lose. However, your pocketbook would quickly inform you if your stats were bogus. When I later came into contact with advanced professional traders I found that each would use a different set of quants to make their decisions. They became very familiar with their statistical analysis sets, but would then interpret them to make their decisions. Experience and gut reactions was a big part of how they operated.

I’ve heard it posited that intuition is the response to data that you are not quite consciously aware of…yet. Later on you can find out what stirred the intuition and it can then be codified.

A book I found fascinating -" Against the Gods, the Remarkable Story of Risk", by Peter Bernstein. There was a time when the attempt to predict the future mathematically or otherwise was considered a sacrilege; the future and the knowledge of it’s unfolding was the province of God and God only. The insurance companies were the first to break this taboo.

Nice post – rec’d.

-a game so slow and excruciating boring that the spectators needed something to occupy their minds and help them kill the time in the constantly recurring and ever lasting intervals of no action. Following that, it has spilled over into all other sports. In most of these, stats were not needed but it was an easy way for announcers to both find something to talk about and enable them to sound informed.

Twenty years ago, it would never have occurred to anyone to count the number of punches thrown in a boxing match and to determine the number of those that landed. A judge who based his scoring on such math rather than "feel" of dominance would have been subjected to ridicule and likely banned from judging matches. Before the US got into soccer, it would never have occurred to anyone to count the numbers of shots on goal or the number of saves and corner kicks. Too many now consider these kinds of tabulations the "scientific" revelation of who performed the best.

The kind of stats provided by T J Johnson on this site is usually very enlightening. The kind typically provided by sports commentators and journalists tend to detract from the enjoyment of the game by basically suggesting quality is nothing more than a matter of quantifying performance components. You can’t determine which restaurant serves the best meal by comparing the prices andt sizes of the portions. You don’t find out who appreciates you the more by counting the number of words in their praise of you and you don’t determine which date was the best by comparing the number of smooches.

Not Twain, although he’s responsible for popularizing it.

http://en.wikipedia.org/wiki/Lies,_damned_lies,_and_statistics

The term was popularized in the United States by Mark Twain (among others), who attributed it to the 19th Century British Prime Minister Benjamin Disraeli (1804-1881): “There are three kinds of lies: lies, damned lies, and statistics.” However, the phrase is not found in any of Disraeli’s works and the earliest known appearances were years after his death. Other coiners have therefore been proposed. The most plausible, given current evidence, is Charles Wentworth Dilke (1843-1911).

Stats are rarely used objectively and when they are it still doesn’t mean crap about 60% of the time. Especially in something like football where emotions can play a huge role, John Elway is a good example of that because in many of his games you could look at the stat line, 26 of 39 314 yards 3 touchdowns and 1 interception, but from that you don’t get how he converted a 4th and 23 with 57 seconds left that took them to the 26 yard line where they Rich Karlis kicked the game winning field goal. That applies to every kind of statistical analysis you can put together is just that numbers will never be able to fully explain a game of football

Conclusions drawn by studying the stats are USUALLY more reliable than preconceived notions/hypotheses supported after the fact by statistics.

The latter tends to leave out contributing factors when they don’t support the original thought and statistics with so many variables – such as individual performance in a team sport generally DO have; …ache to be misinterpreted in the first place.

The latter is generally the rule in sports talk. (however writers attempt to disguise it as the former!)

I enjoy reading these statistical motivated viewpoints but while I try to be open minded and use them to filter out my own biases… In the end I trust my eyes and the eyes of those I respect, as much or more than the numbers.

With such a tough act to follow…KB’s QB ratings mathtacular…we get the stat-tacular post! Nothing but good, clean knowledge here people. The rest of SB Nation are just football fans, we’ve got brains as well.

Great post, and rec’d.

"Oh, people can come up with statistics to prove anything. 14% of people know that."
- Homer Simpson

“He uses statistics as a drunken man uses lamp-posts… for support rather than illumination.”
- Andrew Lang (1844-1912)