Dolphins' Performance in Last 3 Games Associated with 11.4 Regular Season Wins

[Shouright]

The following background information is required to make sense of this post:

http://www.advancednflstats.com/2007/07/what-makes-teams-win-part-1.html

Here are the data for the Dolphins' past three games:

	OFF YPA	DEF YPA	OFF INT RATE	DEF INT RATE	OFF FUM RATE	DEF FF RATE	OFF YPC	DEF YPC	PENALTY RATE
LAST 3 GAMES	7.43	6.73	1.8	2.5	1.96	0.0103	4.9	4.29	0.213
LEAGUE AVG	6.71	7.21	2.7	2.7	1.97	0.0147	4.15	4.16	0.416
LEAGUE SD	0.72	0.64	0.94	0.81	0.51	0.004144	0.46	0.4	0.0759
DOLPHINS Z	1	-0.75	-0.957446809	-0.24691358	-0.019607843	-1.061776062	1.630435	0.325	-2.674571805
COEFFICIENT	1.14	-0.92	-0.45	0.76	-0.33	0.42	0.46	-0.24	-0.39

When we work through the equation on this page:

http://www.advancednflstats.com/2007/07/what-makes-teams-win-3.html

...the result is 11.4 wins.

In other words, the Dolphins' performance over the last three games, if extrapolated to a 16-game season, is associated with 11.4 wins.

So, rather than being the beneficiaries of luck or some other random variable, it can be said in my opinion that they've been playing like a playoff team.

 

[DisturbedShifty]

Glad you're hear to tells these things.

 

[chriswarren876]

Did you just make a positive post... :stitches: I think my head just exploded

 

[phins_4_ever]

The following background information is required to make sense of this post:

http://www.advancednflstats.com/2007/07/what-makes-teams-win-part-1.html

Here are the data for the Dolphins' past three games:

	OFF YPA	DEF YPA	OFF INT RATE	DEF INT RATE	OFF FUM RATE	DEF FF RATE	OFF YPC	DEF YPC	PENALTY RATE
LAST 3 GAMES	7.43	6.73	1.8	2.5	1.96	0.0103	4.9	4.29	0.213
LEAGUE AVG	6.71	7.21	2.7	2.7	1.97	0.0147	4.15	4.16	0.416
LEAGUE SD	0.72	0.64	0.94	0.81	0.51	0.004144	0.46	0.4	0.0759
DOLPHINS Z	1	-0.75	-0.957446809	-0.24691358	-0.019607843	-1.061776062	1.630435	0.325	-2.674571805
COEFFICIENT	1.14	-0.92	-0.45	0.76	-0.33	0.42	0.46	-0.24	-0.39

When we work through the equation on this page:

http://www.advancednflstats.com/2007/07/what-makes-teams-win-3.html

...the result is 11.4 wins.

In other words, the Dolphins' performance over the last three games, if extrapolated to a 16-game season, is associated with 11.4 wins.

So, rather than being the beneficiaries of luck or some other random variable, it can be said in my opinion that they've been playing like a playoff team.

Are you under 'government' protection?
Where is that Thanks/No Thanks button for you?

 

[sampera]

Thanks, interesting stat. Good to hear that we're winning due to our own performance, NOT another team's "bad game".

If you did this with our entire season's information, would it show our actual win number?

 

[Shouright]

Thanks, interesting stat. Good to hear that we're winning due to our own performance, NOT another team's "bad game".

If you did this with our entire season's information, would it show our actual win number?

This actually doesn't rule out other teams' having bad games, however. What it tends to rule out is the impact of random variables, since the ones in the original post tend to be repeatable, or stable, over time. In other words, this tends to rule out "luck" or "the ball bouncing our way" so to speak.

It would be interesting to see what the full season data would show. I suspect the number they produced would be very close to our actual number of wins, since the variables in the original post account for over 80% of the variance in winning percentage, which is a huge number statistically speaking.

 

[greasyObnoxious]

we went 5-2 in our last 7 games. if you prorate that to a 16 game schedule you come up with 11.4 wins. all i needed was a calculator

 

[fleetwoodolfan]

The disclaimer at the end of the original post is excellent. Could not have said it better.

 

[FinfanInBuffalo]

we went 5-2 in our last 7 games. if you prorate that to a 16 game schedule you come up with 11.4 wins. all i needed was a calculator

How do you know your calculator isn't suffering from confirmation bias?

 

[Shouright]

How do you know your calculator isn't suffering from confirmation bias?

Your calculator wouldn't be susceptible to confirmation bias, but it would have a world of trouble controlling for random variance.

 

[dolpns13]

How do you have 4 tenths of a win?

 

[vt_dolfan]

The following background information is required to make sense of this post:

http://www.advancednflstats.com/2007/07/what-makes-teams-win-part-1.html

Here are the data for the Dolphins' past three games:

	OFF YPA	DEF YPA	OFF INT RATE	DEF INT RATE	OFF FUM RATE	DEF FF RATE	OFF YPC	DEF YPC	PENALTY RATE
LAST 3 GAMES	7.43	6.73	1.8	2.5	1.96	0.0103	4.9	4.29	0.213
LEAGUE AVG	6.71	7.21	2.7	2.7	1.97	0.0147	4.15	4.16	0.416
LEAGUE SD	0.72	0.64	0.94	0.81	0.51	0.004144	0.46	0.4	0.0759
DOLPHINS Z	1	-0.75	-0.957446809	-0.24691358	-0.019607843	-1.061776062	1.630435	0.325	-2.674571805
COEFFICIENT	1.14	-0.92	-0.45	0.76	-0.33	0.42	0.46	-0.24	-0.39

When we work through the equation on this page:

http://www.advancednflstats.com/2007/07/what-makes-teams-win-3.html

...the result is 11.4 wins.

In other words, the Dolphins' performance over the last three games, if extrapolated to a 16-game season, is associated with 11.4 wins.

So, rather than being the beneficiaries of luck or some other random variable, it can be said in my opinion that they've been playing like a playoff team.

That's pretty cool Shou. I wonder...does the numbers change drastically if we were to win the last two as well? I think its a good barometer of how we might be next season...

Sent from my SPH-L710 using Tapatalk

 

[Shouright]

That's pretty cool Shou. I wonder...does the numbers change drastically if we were to win the last two as well? I think its a good barometer of how we might be next season...

It depends on the team's performance in those two games. The finding wouldn't change if the performance didn't change substantially (for better or worse) with regard to at least one of the variables above.

In other words, the finding isn't a function of wins, per se, but of how the team performs with regard to the variables that strongly predict winning.

The Dolphins could lose the next two games due primarily to fumbles that bounce wildly around the field and get recovered by the opposing team (i.e., a random variable), yet still be performing in such a way with regard to the above variables (i.e., the more stable ones) as to be regarded as having the ability of an 11-win team.

 

[2413fanphins]

Love the italics, really hammers home the point. Without them the post may even be too cloudy to understand

 

[Shouright]

How do you have 4 tenths of a win?

Given similar performance, in 6 of every 10 seasons they'd likely have 11 wins, and in the other 4 of every 10 they'd likely have 12.