Any suggestions for calculations for Math IA?

[maverick410]

Im doing my IA on the relationship between a professional sports teams salary and the teams total number of wins. I have collected dat from 3 sports leagues the NBA, MLB, and the NFL. For each team, i have the teams salary for a certain year and the teams total wins. I was wondering what kind of calculations can i do and also tips on organizing my data...any suggestions?

[Drake Glau]

Calculations: Linear Regression Line and a Pearson's Correlation Coefficient. If you have enough data (25 points) you can do a standard deviation as well. If you SOMEHOW have 100 points, you can do a chi-squared test too

I would display your data in a chart with columns X Y XY X^2 Y^2 and whatever else you need to do the linear regression and the pearson's. I forget exactly what you need but I had all the combinations of X and XY X^2 etc etc at the top and then summed them all in my last row for whichever column I needed sums for.

[Guest Positron]

BTW, by "salary", do you mean the amount of money the team makes or how much the players make? I think you mean the teams total income. Remember that the players' salaries depend on their contracts, which might be made years ago. That means, that players might make alot of money / only little money regardless of how the team is doing.

And of course the fact that usually teams in big cities, like New York, will make a lot of money anyways, even if they lose all the time. You should also consider the fact that not all games, and therefore wins, are equal. If the team wins one normal match, the team will get a lot less money than if it wins a "big game" at the end of the season.

[maverick410]

BTW, by "salary", do you mean the amount of money the team makes or how much the players make? I think you mean the teams total income. Remember that the players' salaries depend on their contracts, which might be made years ago. That means, that players might make alot of money / only little money regardless of how the team is doing.

And of course the fact that usually teams in big cities, like New York, will make a lot of money anyways, even if they lose all the time. You should also consider the fact that not all games, and therefore wins, are equal. If the team wins one normal match, the team will get a lot less money than if it wins a "big game" at the end of the season.

Im talking about a teams payroll...how much they give out to their players

[maverick410]

Calculations: Linear Regression Line and a Pearson's Correlation Coefficient. If you have enough data (25 points) you can do a standard deviation as well. If you SOMEHOW have 100 points, you can do a chi-squared test too

I would display your data in a chart with columns X Y XY X^2 Y^2 and whatever else you need to do the linear regression and the pearson's. I forget exactly what you need but I had all the combinations of X and XY X^2 etc etc at the top and then summed them all in my last row for whichever column I needed sums for.

Thx for the post but I'm still a bit confused as to how can organize the data....Would you be willing to take a look at my data to help me organize?

[Drake Glau]

Chart:

Teams Wins (X) | Teams Salary | X² | Y² | XY |

I believe that it was you need for a linear line and a pearson's. If you need another just tag it on.

At the bottom of the chart add the totals you need, so at the bottom of the X column, add a sum of X, same for y and any other totals you will need later.

I guessed on what your X and Y were by the way. This test could be done either way depending on if you think the # of wins is dependent on the salary or if the salary is dependent of the # of wins.

[Guest Positron]

BTW, by "salary", do you mean the amount of money the team makes or how much the players make? I think you mean the teams total income. Remember that the players' salaries depend on their contracts, which might be made years ago. That means, that players might make alot of money / only little money regardless of how the team is doing.

And of course the fact that usually teams in big cities, like New York, will make a lot of money anyways, even if they lose all the time. You should also consider the fact that not all games, and therefore wins, are equal. If the team wins one normal match, the team will get a lot less money than if it wins a "big game" at the end of the season.

Im talking about a teams payroll...how much they give out to their players

That's the problem I was trying to tell you. Even though you'll probably find some correlation, you can't prove the causality. The payroll depends on the contracts, which have been made before the season and therefore one can't cause the other. There is probably a stronger correlation between the payroll and last seasons wins. Of course at least a part of the players' salaries might depend on the teams performance, but it's rather pointless to speculate with this as the contracts aren't exactly something we can get our hands on to check this.

[maverick410]

Chart:

Teams Wins (X) | Teams Salary | X² | Y² | XY |

I believe that it was you need for a linear line and a pearson's. If you need another just tag it on.

At the bottom of the chart add the totals you need, so at the bottom of the X column, add a sum of X, same for y and any other totals you will need later.

I guessed on what your X and Y were by the way. This test could be done either way depending on if you think the # of wins is dependent on the salary or if the salary is dependent of the # of wins.

jee wiz thx mister