Graph of wins vs. white players in 2009 NFL season

[WASP]

-Edited by: WASP

 

[White Shogun]

What are you trying to prove or establish with this graph? That a team wins more games with more white players or that there is no real relationship between winning and the number of white players?

 

[devans]

I likeyour graph. It is quite helpful.
For example if you take the four teams with fewest white players, the total wins comes to 25
Then count up the total wins for the 4 teams with the most white players - it comes to 38!
I think, just by looking at the graph, if you did the top 6 and bottom 6 teams, the results would be even more uneven - but I don't have time right now.Edited by: devans

 

[jaxvid]

From the graph it clearly shows that the more white players you have on a team the more wins. We have known this for a while here of course. But it's nice to see it in graph form. Great job!!!!!

 

[icsept]

The correlation between number of wins and number of white players is 0.363 and the regression line is Wins = 2.366 + 0.365 * white players[/QUOTE]

Can you explain this in layman's terms?

Also, I think the playoff wins should be added to the graph for true accuracy.Edited by: icsept

 

[chris371]

Yeah, i always sucked at statistics., what does that correlation mean excactly?

 

[PhillyBirds]

Terrific graph, very good work. This is the type of statistical analysis that goes so far to argue a point. People can't debate numbers that are directly in front of their face. While it seems a little scattered in the middle, the correlation is noticeable, in my opinion.

I'm guessing that's the Colts and Packers as the far outliers? Edited by: PhillyBirds

 

[White Shogun]

Correlation and linearity

 

[TorontoArgos]

I noticed this year, compared to other years, white teams didn't do as well.

The past few years before this were probably much more certain in terms of correlation between white players and wins.

This year, some really black teams actually did well, like the Eagles and Saints. Much more so than in previous years where ALL the best teams were ALL the whitest teams. And all the blackest teams would always be the worst teams.

 

[WASP]

-Edited by: WASP

 

[dwid]

I noticed this year, compared to other years, white teams didn't do as well.

The past few years before this were probably much more certain in terms of correlation between white players and wins.

This year, some really black teams actually did well, like the Eagles and Saints. Much more so than in previous years where ALL the best teams were ALL the whitest teams. And all the blackest teams would always be the worst teams.

Saints have 17 White players, so not a "really black team" compared to the eagles who have only 10. While 17 isn't a big number it is around the middle section of this graph closer to the higher end than the lower end (the graph list them as 16 though) and an almost identical number to teams like the Vikings and Cowboys who most were rooting for and considered somewhat White friendly. They could lose the mostly sumo line and should have kept Hass but they aren't as black compared to the Haslett years. Eagles won against other teams that had a low number of Whites. Lowered playing field which is why some of these blacker teams will have success from time to time when they get to play other equally horrible teams.

Id say its a much better year than last Superbowl wise. Colts (22) vs Saints (17) compared to last year's Cardinals and Steelers. Edited by: dwid

 

[White Shogun]

You could also look at this relationship by breaking the data into categories.
There were 6 teams with 13 or fewer white players, and these teams had an average of 6.2 wins.
There were 9 teams with 14 or 15 white players, and these teams had an average of 7.9 wins.
There were 10 teams with 16 or 17 white players, and these teams had an average of 8.0 wins.
There were 7 teams with 18 or more white players, and these teams had an average of 9.7 wins.

Or, of the fifteen teams with 8 or fewer wins, 7 of them had fifteen or more white players. In other words, half of the losing teams had more whites than the other half.

Did the number of white players fluctuate throughout the year?

I admit I don't know much about statistics, and I can see from the explanation I've quoted here that adding just 5 more white players should = 3 more wins. I'm just not sure looking at the way the numbers are spread around the chart that it is really that convincing.

 

[chris371]

"The correlation between number of wins and number of white players is 0.363 "

Does this mean that an all white team would have a 36.3% higher chance of winning, or how can it be interpreted?

 

[devans]

I have a bit more time now, so...
If you take the SIX teams with fewest white players, the total wins comes to 37
Then count up the total wins for the6 teams with the most white players - it comes to 59.
A difference of 22 wins.

 

[white is right]

This also shows the clear scouting bias that exists in the NFL. As the blacker teams tend to be the youngest and most inexperienced teams. I noticed this trend 20 years ago with Bungles around the time they drafted James Francis.