Upcoming Project

Although I have collected alot of SEC game data and I have put them into charts, I want to see if the correlation for each chart will increase or decrease depending on the amount of data put into the charts. It will be interesting to see what happens when I take out or put in data.

I also need to put the passing yards chart for an SEC game, sorry about the delay.

Points and Yard Difference

Although we have found the correlation for first downs, total yards, turnovers, and rushing yards, we still have three more charts to post. There are two charts we will post today, for points and yard difference. Here is the chart for offensive points scored ... I am expecting a big correlation from this one.

This chart, as I expected, shows a big correlation. Winning teams scored as many as 52 points, and as few as 24 points. Because its correlation is slightly smaller than rushing yards does not mean rushing yards are more important than points. By correlation it does, but I think points has causation, which means it really does cause victory in a game. The team that has more points wins the game, but the team that has more rushing yards doesn't win all the time. The team with more points wins all the time. Now, here is the chart for yard difference.

As you can see, the correlation for yard difference is small. In fact, one team got 120 fewer yards than its opponent, but still scored more points. Well, one more point! I am not surprised at the weak relationship between yards difference and points difference since the correlation for total yards was small as well.

My theory is that points is the only variable that has causation in a game's point spread, which means it is the only variable that can really determine who is the winner. For example, Team A can have more rushing yards, passing yards, and less turnovers, but if they socre only 7 points and Team B scores 9 points, who wins? Team B does. I also learned that total yard differnece doesn't have much of an impact on an SEC game.

Next Up: Passing Yards, how will the correlation for this one go?

Rushing Yards and Turnovers

Now that we have shown you the correlation of total yards and first downs with point spread in an SEC game, let's see how important rushing yards and turnovers are. Here is the rushing yards chart:

As you can see, the correlation for rushing yards (R-squared equals 0.37) is huge! I didn't know rushing yards were so important in an SEC game! Rushing yards has the biggest correlation so far, and its correlation is greater than first down's and total yard's correlations combined. However, here is the chart for turnovers.

Turnovers, as you might have noticed, has a negative correlation with the point spread. Makes sense. It just means that the more turnovers you commit, the smaller amount of points you will score. The equation line seems to indicate that two turnovers can cost you a game, but only one is survivable. Turnovers have more value than first downs, which I expected, but the R=squared is smaller than rushing yard's value, which surprised me.

From these charts, I learned that the more rushing yards you have in a game, the higher the point spread. I also learned that rushing yards are more important than turnovers, which for me at least, is hard to believe. Does it make sense? Maybe. But let's wait to see all the results before we conduct an anlysis. For the time being, remember that "Correlation is not causation." Maybe it is the point spread causing higher rushing attempts and yards.

Next Up: Points and Yard Difference, will there be another shocker?

Total Yards and First Downs

Now, it is time for the charts from total yards and first downs. To be honest with you, I was very surprised with the results from the charts. Behold, the charts!





In this first chart, each red and gray dot represents a game. On the bottom, or X, axis is the point spread between the winning team's score and the losing team's score. On the left, or Y, axis is the explanatory variable, which in this first chart is the number of first downs by the winning team. For example, the dot on the bottom represents a game where the winning team had only ten first downs but still won by a margin of one point.

The equation in the chart is y = 0.205x + 15.92. That means on average, even if x = 0, then y = 15.92, or that given everything else, the winning team makes about 16 first downs. Every additional five points equals another first down. Below that is an R-squared for the relationship between X and Y variables in the equation, which in this case equals 0.2133.





As you can see from the formula R squared in the charts, first downs had a bigger correlation than total yards (o.21 versus 0.04). In other words, first downs matter more than total yards in an SEC game! Wow, I thought the amount of yards accumulated in a game would matter the most. Also, the correlation for total yards is very small, total yards don't have a big impact in an SEC game at all. However, first downs have a much bigger correlation than total yards do, first downs mean a lot in SEC games!

To sum it all up, total yards don't have a significant impact in an SEC game, but first downs really change the game!

Next Up: Rushing yards and Turnovers, which one has the bigger correlation?

Victory Data

Here is the raw data for the 12 SEC games we looked at to find what was most important for margin of victory. The first thing is a list of twelve numbers which is the variable "Points Difference"


POINT DIFF
39
30
24
20
18
14
12
11
11
6
4
1


Then there are seven variables we considered, and are trying to determind which ones matter the most in explaining POINTS DIFF. These are ...


PS 1D TYd PYd RYd TO YdD
49 22 373 188 185 0 -25
51 22 475 210 265 1 154
30 16 243 96 147 0 -15
29 23 366 188 178 1 193
31 21 409 307 102 1 194
52 16 443 249 194 0 -54
26 29 458 310 148 2 249
41 21 334 205 129 1 10
31 19 358 216 142 0 35
27 16 353 215 138 3 -29
24 15 326 219 107 2 -33
31 10 325 185 140 1 -118



where PS is points scored, 1D is first downs, TYd is total yards, PYd is passing yards, RYd is rushing yards, TO is turnover, YdD is yards difference.

Sorry that the raw data looks so sloppy.

What causes victory in an SEC game?

An SEC game is full of action. There are smashing hits, big plays, momentum-turning interceptions, and best of all, touchdowns. However, my dad and I want to find what matters most in an SEC game -- or what leads to victory. Is it turnovers or rushing yards? First downs or total yards? Passing yards or points? At first, all of these seemed equally important. But my dad and I used something called correlation to find out which one mattered most.


First, let's describe what we mean by victory. We decided to use the difference in points as a measure of victory, which is Team A points minus Team B points. Each game has a winner, but if that team wins by 30 points, then you know it was a more solid victory.


Second, let's list all of the things that might lead to victory (or "bigger margin of" victory):

• Offensive Points
• First downs
• Rushing Yards
• Passings Yards
• Total Yards
• Total Yards difference (Team A minus B)
• Turnovers

Third, I had to actually collect the data. I decided to pick some good teams from the SEC and find out what happened in each statistical category after each game when they played one another last year. I picked Alabama, Florida, Georgia, LSU, Ole Miss, and Tennessee. These teams played a total of 12 games against one another, but not every team played every other. After I collected all the data from ESPN.com into an Excel spreadsheet, I saw what happened in each statistical category for each game.


Before I looked more closely at the numbers, I thought total yards difference would matter the most for victory. I was surprised to find out about the results, I hope you will too... (see my next post for the results!)