Wow. What a great idea. So, if I take the three biggest rushes against the Vikings this year and subtract them from the overall total this is what I get.
29 A.Abdullah left end to 50 for 29 yards
26 I.Crowell right guard for 26 yards,
19 M.Campanaro right end to BLT 44 for 19 yards
74 yards. 732-74=658 658/212 = 3.1 yards per carry. Good for moving up to 2nd in total yardage and at 3.1 YPC, tied for first.
I should do this for passing yards as well. Tell me, is it limited to 3 or can I use 5?
Well, this right here demonstrates the lack of understanding between median and mean (average).
If one running back runs for 100 yards on 1 run and 9 others run for zero on one run, then the median is 10 ypc for running backs and the mean is zero.
Especially because the Vikes haven't given up any unusual big runs which would be outliers to the data set, those runs really don't change anything much nor do they establish a narrative.
The point in saying about specifically three runs out of 200+ is that just over 1% of the runs account for 20% of the total yardage. 3 runs!
So the mean is what it is. We average 118 per game. The median is 111.
However, the point is that if we understand that in football because so much is played at a high level that the smallest things can manifest as the difference, then things like scheme adaptation, weekly improvement, player health and game day execution all factor in.
Now, if someone is going to analyze the Vikings Run defense, the improvements have more to do with 2nd level contain. It's much more about not letting small runs become the 10-20 yard runs.
With the Rams, it's about gap responsibility and assignment execution because, as the data shows, other than that the Rams have been pretty solid.
Here's how you can tell there are outliers. If the median and mean aren't close, there's your sign. Currently the median and mean for the Rams Run D has a delta of 7 ypg. That's kind of a lot. When those 3 runs are removed as outliers, the median is 94 ypg and the mean is 95 ypg. That shows a coherent data set.
The point is that if the Rams are consistently at 95 ypg with no big runs and they want to improve their Run D, they have to take certain steps. If they are consistently playing at 95 ypg and on 3 defensive plays, there have been breakdowns, be they schematic or execution, then that's a different fix.
Moreover, on any given play, the median is MUCH MORE likely an indication of what will happen. If a team routinely doesn't give up yards and about 1% of the time gives up a big run, then that means opponents are relying on not only having that 1% happen for them, but also that the play in question will manifest in points.
It's not just about manipulating numbers. It's about understanding what they mean in context.
Hope this helps.
