Yard Totals per year is absolutely meaningless. Why? Because it doesn't paint a clear picture. It would be like talking about how a car goes 500 miles on a full tank vs another car that goes 300 miles on a full tank. Without knowing how many gallons each tank holds, it is hard to tell what that number really means. Is the 500 mile car getting more out of each gallon or have they just put in a really big ass gas tank? This is why Mile per Gallon is a useful stat for comparison and miles the vehicle can go in a fill up is just a useless factoid.

So, if you want to keep Yards as a measuring stick for the offense, here are the steps necessary to make the stat less useless. The first step would be focusing it down to Yards per Game as the 2009 team played only 12 games, where the rest played 13. However, even at the per game level, it is still not a very good or helpful stat. This next step can actually turn a measure of yards into a useful stat. Say a team gains 400 yards in a game, does that sound like a good game for them? Well, what if it took team A 35 plays to get there and team B needed 55 plays? By focusing down another level to Yards per Play, you have something that matters. It measures something in a meaningful way by putting the results up against something that directly controls those results. However, some data cleaning up is needed to make this a better stat. You would need to remove the plays where the QB takes a knee to either end the half or end the game. Plays where the QB spikes the ball to stop the clock should also be removed so that this stat only measures plays where the offense actually intended on gaining yardage. That would give you a dataset and from that a stat that measures how many yards an offense gained per play when they were trying to gain yards. This is the miles per gallon stat compared to the miles you can go in a full tank that Yards per Year is.

Ok, now on to Touchdowns. Again, Touchdowns per year is useless, so how can we fix it? Again, the first step would be to focus in on TD per game since not all years presented have the same number of games in their seasons. But again, that isn't very helpful, and it can be better. So, we must find the right thing to measure. We found the right one for yards is per play designed to gain yards. The best way to measure Offensive TD's is to measure per offensive possession that wanted to score points. Cleaning this dataset would be removing possessions that ended the game with QB kneel downs and also before halftime possessions that only featured QB kneel downs. If the before halftime possession tried to make any actual offensive plays before kneeling down to end the clock, it needs to remain in the dataset as an Offensive TD could have been scored as it was in some way attempted to be scored. What you would then have is a percentage of offensive possessions that wanted to score that ended with a TD. If your goal is to use Offensive Touchdowns to compare offenses, then the stat needed for that is the percentage of offensive possessions that tried to score that resulted in a Touchdown.

What you have now are meaningful stats. One way to look at them is Yards / Number of Attempts to Gain Yards and the other looks like Touchdowns / Number of Attempts to Score Touchdowns.

Advanced stats take other factors into account such as the opponent, field position, and score. That is far too granular and time consuming to get into unless you worked for the team.