Shop for New York Giants gear at Fanatics.com

The wrong kind of stats

John Fennelly recently posted a bunch of stats over here and he prefaces the entire post with:

Here are some facts for you to digest….don’t read too much into them…..

I’m not going to trash Mr. Fennelly per se, but I’m going express my extreme distaste for the type of meaningless numbers presented not only in this post but so often in the sports media. These are stats that make you say “Huh, go figure”, take up all kinds of TV time in between commercials, millions watch and absorb them – but nevertheless they say NOTHING about the game to be analyzed at hand… in fact they only muddy the water.

Here’s an example… In the Tom Coughlin era, Fennelly says:

“The Giants are surprisingly 6-5 against the Cowboys: 3-3 in Dallas (including the 21-17 Divisional Playoff win in 2008); and 3-2 at the Meadowlands.  These numbers aren’t too bad considering this is a divisional rivalry, but if  you drill down further, you’ll find the Giants are only 2-5 vs Dallas when Tony Romo is at QB.  He’s starting Sunday night, by the way.”

So here we are supposed to be getting nervous and then there’s this nugget that shortly follows:

“…the Giants are 5-1 vs Dallas with Osi in the lineup.  As we all know, he’s back…..FYI Ware and Umenyiora were college teammates at Troy State…”

So we’re supposed to believe that the Giants have a winning AND losing record heading into Dallas Sunday. They have no shot to win because of Tony Romo, yet because Osi is going to be starting the Giants will most likely dominate. I HATE stats like this. And the sad part is these stats make their way onto certain pre-game shows and things get really off course, people lose focus and then everyone wonders how the Giants can win the NFC East when they were initially picked to finish last in the division.

Like past performances in the stock market, what happened in the past guarantees nothing for the future and depending on what stat you highlight, the information can certainly be misleading. The information I’m highlighting above is a perfect example — in the end it’s all a wash because it’s impossible to make heads or tails of any facet of either team using stats this way. There is no insight to the situation realative to the data points, no particular keys to the game are discussed, and so no intelligible information can possibly be gleaned without establishing that context. Questions brought to my mind when I see these numbers include:

  • Who was injured during those games that made a difference?
  • What time of year were they playing and who stood to gain the most?
  • What are some of the defining characteristics for each of these games that made a difference?
  • And most importantly, let’s discuss where the Giants and Cowboys are right now comparatively!

Stats are stats, you can prove and disprove anything with them and they do not take into account intangibles that are so important…. Tony Romo’s ankle may act up on Sunday and his passer rating plummets, Osi might blow his knee out (god forbid), the point is you don’t know what is going to happen that will make or break the game. It should go without saying that particulars for each game MATTER. But I still see so many of these meaningless stats used to determine the outcome of a game without accounting for the intangibles that make team sports so thrilling to follow. The law of probability and the randomness of “any given sunday” work together in that sense alone – the synergy of the quantitative and qualitative provide for an exciting game, and hopefully a thrilling outcome.

Analyze where the Giants are NOW and illustrate some keys to the upcoming game… like we did here:

Giants Roundtable Week 2: Dallas Cowboys

Otherwise the numbers are basically meaningless.


Enjoyed this post?
Subscribe to Giants Gab via RSS Feed or E-mail and receive daily news updates from us!

Submit to Digg  Stumble This Story  Share on Twitter  Post on Facebook  Post on MySpace  Add to del.icio.us  Bark It Up  Submit to Reddit  Fave on Technorati

2 Responses to “The wrong kind of stats”

  1. Andrew…you must learn to relax….these are things I put out there that I don’t think many people know….there are pros and cons to everything…..are these facts/stats meaningless? Most of them, but its my job to let the unknown be known….rgds
    JF

  2. andrew ilnicki says:

    I know John, with all due respect that comes across because you prefaced the post, much appreciated for the candidness there.

    My problem with stats is the larger issue I’m making about being able to prove or disprove anything with them. It all comes out in the wash… the exercise is thus empty. I like the debate and the insightful discussion, not the numbers that strip out context.

    best regards

    -ai

Leave a Reply