FanPost

Lies, Damned Lies, and Statistics

Berzerker did a damn fine job with this one, so let's move it on up. It's a pretty long post after the jump, but it's worth the read. - Chris

It’s great that Purple Faithful has started a rip-roaring debate on where to go with the #3 pick, but I have a couple of concerns about the logic of this argument. I started writing it up, but it got really long, and rather than bog down the Fanpost, I made a separate one. Follow the jump for a different slant on how to decide on using the #3 choice in the draft.

Firstly, teams gain personnel through free agency as well as through the draft. The data set would need to be standardized for free agency movements before you could comment with a high degree of confidence on the implication of draft choices.

Secondly, the sample size for Super Bowl wins is small in statistical terms, meaning that drawing conclusions on one specific set of variables is a difficult exercise in any case. To focus on draft selection without standardizing for other factors, for example turnovers, risks drawing false conclusions.

I think it’s better to look at winning in general, on the basis that teams that win more often are in a position to win the big games, and that this larger data set equalises the effects of “luck” on which team happens to win individual games. A few examples of the kind of “luck” I have in mind are:

  • If Welker makes the catch that he dropped, do the Pats win the Super Bowl?
  • If Ted Ginn Jr doesn’t get injured against the Taints*, does Kyle Williams fumble punts and let the Giants back into the game against the 49rs?
  • If Sterling Moore, an Undrafted Free Agent, cut by the Raiders and picked up by the Pats in October 2011, doesn’t bat the ball out of Lee Evens hands in the end zone, do the Ravens win the game and go to the Super Bowl?

The point here is not to say that some of the above didn’t involve great plays, but that these plays are so unusual and so decisive for the final result that they are essentially random and can’t really be correlated to other factors. At the top end of the game, a team has to be a winning team in general and then have a great game plan/execute better/get the “luck”/get the calls/all of the above, to win in these one-off games. As an example closer to home, with all that went on in the 2009 season NFCCG, if we hadn’t kept dropping the damn ball could we have won the game? Going further back, if we had re-signed Matt Birk for the 2009 season, would our line play have been good enough to get first seed and hold home field advantage?

As these are all discrete individual events, looking at overall winning, and what contributes to overall winning, is, IMHO, a better approach.

The website referenced by Purple Faithfull is one I regularly look at, and they have developed a set of their own sabremetric style stats for the NFL. At the end of the last season they did an article on which of these stats (and some of the traditional stats) were the best at predicting whether a team would win or lose a game:

http://www.coldhardfootballfacts.com/Articles/11_5326_Winners_and_losers%3A_beauty_runs_stat_deep.html

The quick summary of this article is that the accuracy of the various stats at predicting victory, in descending order, was:

1. Real Quarterback Rating (0.869 success rate)

2. Scorability-Bendability (0.850)

3. Passer Rating Differential (0.787)

4. Interceptions (0.781)

5. Real Passing Yards per Attempt (0.723)

6. Passing Yards Per Attempt (0.712)

7. Negative Pass Play Percentage (0.709)

8. Third Down Success (0.703)

9. Rush Yards (0.674)

10. Passing Yards (0.509)

11. Rushing Yards per Attempt (0.476) (also known as the “Minnesota Vikings Syndrome” apparently!)

Some of these stats are related only to how good team A’s offense is compared to team B’s defense, or how good team A’s defense is compared to team B’s offense. Others combine the relative performance of team A’s offense and defense against team B’s offense and defense. What these stats indicate as a set of general conclusions could be taken as:

  • To win consistently you need an elite QB (nothing startling in that);
  • Having a stud WR is less important than having a set of WRs which combine to perform above average as a group (usually referred to as the “shiny hood ornament” theory, as in, don’t get distracted by it – it pains me to say it, but Green Bay are a good example of this approach);
  • Overall line play is critical to many of these stats, but the focus is on the overall aspect, not being elite at specific positions. Essentially, if the line are all generally above average, they can compensate for each other collectively more effectively than if there are glaring weaknesses. This kind of supports Purple Faithful’s point that there’s no point being great at LT if you are dreadful elsewhere on the line, but I’ll come back to that again later;
  • Overall defensive play is critical to many of these stats, either directly as a part of the calculation, or indirectly by lowering a competitors stats and thus providing competitive advantage. In particular, defensive play against the pass is critical, defensive play against the run, not so much; and
  • Being a great running team is no great advantage to winning games. No matter how much we love AP, and we do, the reality is that a RB can’t consistently put a team on his back and win games.

So, taking account of the above, what to the Vikings need to do to become a winning team and be in a position to win big games? Three key things, all of which I’m sure will be agreed by everybody:

  • Have a good/great QB – Ponder will get the chance to show if it can be him, but right now, nobody knows for sure;
  • Get better on the o-line to improve performance in the passing game on the offensive side of the ball; and
  • Get better in our pass defense (as in, not trip over each other, miss tackles and generally behave like the keystone cops – it would be nice to see a DB, you know, compete for a ball or jump a route or something like that)

Turning to the draft, the key question to be asked at any selection point is whether selecting Player A will improve your overall performance more than selecting Player B, recognizing that you have several selections to increase your overall talent pool and that securing a good/great player from a limited field of DBs, say, and matching that with two good o-line players from a deeper field may be better for your team than selecting an elite o-line player from a deep field and missing out on any good DBs. None of this, of course, is helped by the crapshoot nature of whether picks work out or bust.

Taking that logic forward, if the Vikings feel that the o-line needs one player to stabilize it, and that there are some good DBs likely to be available later on, then taking Kalil is the right choice (National Football Post has him graded at 8.0, for what that’s worth).

If, however, the Vikings feel that the o-line is a mess, and needs multiple players, but that while there will be good options available later to fill those holes, there will be no good DBs left, then the Vikings need to take Claiborne (NFP has him graded at 8.5, slightly higher than Kalil).

I don’t agree with Purple Faithful’s analysis that taking Kalil would be the wrong choice because of draft history, but I do agree that taking him simply because he’s the best LT available is not logical if that is not going to be the best method of improving the overall ability of the team to both pass and defend the pass effectively.

It would be like deciding that although you are really hungry, and there’s a plate of bread rolls available which will be gone in 10 minutes, you’re going to spend the next 15 minutes enjoying a spoonful of caviar. It might make you feel like a big shot, and you might even enjoy it, but in 15 minutes you’re still going to be hungry and there’s going to be nothing left to eat.

So with that, how about a poll to find out what we think will be best approach to improving the team.

This FanPost was created by a registered user of The Daily Norseman, and does not necessarily reflect the views of the staff of the site. However, since this is a community, that view is no less important.

X
Log In Sign Up

forgot?
Log In Sign Up

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to SB Nation going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to SB Nation going forward.

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Daily Norseman

You must be a member of Daily Norseman to participate.

We have our own Community Guidelines at Daily Norseman. You should read them.

Join Daily Norseman

You must be a member of Daily Norseman to participate.

We have our own Community Guidelines at Daily Norseman. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker