tewall

Does anyone know the odds of making a 4th and 1? Specifically I'd be
interested in the Packers chances of making the 4th and 1 they had
against the Eagles, if anyone has any idea (support by data).

mpoconnor7

I have a book called The Hidden Game of Football, which has tables based on
play-by-play of the entire 1986 season broken down into downs and yards on Page
128. If I'm reading the chart correctly, on 4th and one teams in 1986 were
30.4 percent successful. By comparison, the chances of getting a first down on
3rd and 1 is 74.9 percent (but that also includes 4th down if the team doesn't
make it on 3rd down). I would have guessed it would have been around 50
percent, as virtually all 4th and 1 plays are running plays right up the middle
or QB sneaks.

Michael O'Connor - Modern Renaissance Man

"The likelihood of one individual being correct increases in a direct
proportion to the intensity with which others try to prove him wrong"
James Mason from the movie "Heaven Can Wait".

marty

See Easterbrook's TMQ column this week on nfl.com. He discusses that and has
links to the football outsiders site which tracks that data.

jerry donovan

There is a page at the following website that discusses this issue:

http://users.frii.com/davejen/nfl/nfl_4thdown.html

Jerry

heavyarms

Well, I don't know about overall in the NFL, but in that game, the Packers
were averaging 5.7 yards per carry. The odds of rushing on 4th and 1 and
getting the 1st down were pretty good.

In that game, Ahman Green had 25 carries, and only rushed for no yards or
negative yardage twice, once for no gain, and once for -1 yards. He only
had three carries over 10 yards, one for 26, 33, and 13. If you take the
two long runs out of his total yardage, he rushed 23 times for 97 yards, or
4.2 yards per carry. In obvious running situations (1st and 10 or less, 2nd
and 6 or less, or 3rd and 3 or less) the Packers rushed 26 times for 163
yards (6.3 yards per carry). The Packers longest runs (the three by Green
and one by Davenport for 12 yards) all came in obvious run situtations. In
Mike Sherman's defense, the only 4th down conversion the Packers attempted
was a 4th and 1 at the end of the 1st half, Ahman Green rushed for no gain.

magoo

I had to sit here for about a half an hour to figure out what is wrong
with the ****ysis on that page. I knew there had to be something wrong
with the conclusion since the author concludes you should go for it on
4th and 1 from your own 24 yard line. That can't be right.

The ****ysis breaks down because the author takes only1 change of
possession into account. If you go for it on 4th and 1, you're 65%
likely to make it. If you makit it, you might not score anyway. If you
don't make it, they might not score anyway. If you punt, they might
score anyway.

But what the author doesn't take into account is that if you punt, they
might not score and then you get the ball back and then you might score
anyway.

Of course, then, you might not score and you'd have to punt and they
might score anyway.


So you could go on and on like that but after a few iterations, the
probability that nobody would have scored is so low that it would be
negligable. But I don't believe it's negligable. after just 1 iteration
because that's the game of field position that a lot of coaches play so
successfully in the NFL.

However, the data on that page is useful in the case at hand, the
Packers vs Eagles game where Mike Sherman chose to punt instead of
going for it on 4th and 1 because the Packers weren't interested in
preserving field position at that point.

The only other thing I would say is that the Packers probably had a
better than 65% chance of making the 4th and 1. That's the league
average but the Packers are better than that even against the Eagles
defense (IMO).


I haven't plugged the numbers in yet but I'm pretty sure they'll show
that punting was a mistake.


--
Magoo
Packer Fan Discussion List: packfans@coollist.com

magoo

Can you post a URL? I can't find the page. Could be that it's because
I'm blind, and use a screen reader. Anmd it doesn't work very well with
the NFL web site.


--
Magoo
Packer Fan Discussion List: packfans@coollist.com

rich carreiro

Magoo <johnheim@nospam.tds.net> writes:

http://www.footballoutsiders.com

--
Rich Carreiro rlcarr@animato.arlington.ma.us

magoo

Thanks. That site is even worse than the NFL site. But I found this:

http://www.footballoutsiders.com/powerrush011204.php?view=message&mid=
131&showcomment=1&vmid=131

Which says that NFL teams converted 4th and 1 71% of the time (if they
ran the ball) this season.


--
Magoo
Packer Fan Discussion List: packfans@coollist.com

dave & janelle

(following up a little late)

I'm the author of the page referenced above. You may have a point about
iterating the value of a possession - but I think that's rolled up into the
'expected value'. My ****ysis uses the expected value of a possession to
derive its conclusions, and I currently think use of expected value removes
the need to iterate.

I think the weak spots in my ****ysis are:

1) I should use net punting average rather than 40 yards.
2) I should take field goals into account.

Still, in the first half of the game, I think teams should go for it on 4th
and 1 much more often than they should. According to my data, even going for
it on 4th and 1 from your own 10 isn't a terrible decision, even though the
outcome may be terrible. (the outcome also may not be terrible...)

For my model, the break-even point is the 24 yard line for 4th and 1. That
is, the expected value of going for it is about equal to the expected value
of punting - and both are negative because having a 4th down on your own 26
isn't a good thing, regardless what you decide to do.

My data might be wrong or my model might be wrong - but I think we should
reject its conclusions with data, not because they "can't be right".

---
Dave Boll
http://www.daveboll.com/