Introduction

The purpose of this analysis is to show how efficient rushers (QBs and running backs) were in the in the NFL during the 2025 season. This analysis only considers rushers that had at least 5 attempts.

Efficiency here is defined as how many more yards did the rusher produce than they were expected relative to the rest of league’s average rush yards based on the same set of conditions. The conditions being how many yards are needed for a first down. Yards expected to produce is calculated as the average of the number of yards produced versus the number of yards needed to gain a first down.

The analytics uses regression modeling based on plays that were intended to be running plays (i.e. not plays that are broken down pass plays that turn into run plays).

This first graph shows how many yards were rushed (the y axis) relative to how many yards that were needed to go to get a first down (the x axis). You will note that there is a heavy emphasis on data points at the 10 yards to go point. That is because that there are a lot of runs on the first down.

All rushes

This next graph shows the same data, but broken up by where the rusher went with a regression line. The only point of the regression line is to show the general relationship between how many yards were rushed versus how many were needed to get a first down.

The lines show you that as the number of yards needed increases, the rusher will will typically rush more (aside from when the rushers go to the edge, which has a slight decline and rushers should probably avoid rushing at the the edge). This shouldn’t be surprising as defenses will typically soften as yards needed increases.

All rushes with regression

Average rush yards given a number of needed yards

The next graph shows you the average numbers of yards gained for a given number of yards needed. This more clearly shows the positive relationship that was discussed in the previous section. And is also the basis for calculating was an average expected number of yards given a certain amount of yards needed.

The calculation the data in the graph above was used to calculate a metric which is named RYOE (rush yards over expected). The data / graph above give us what we would expect a given rusher to run given a certain number of yards needed. RYOE is calculated as the number of yards rushed versus the average number of yards expected. Total RYOE is the sum of of the number of yards over expected in the season. RYOE_PER is the average number yards above expected per play.

The table below ranks players by run yards over expected per play using the data discussed above. Viewing this data we can see that T.Bigsby is the most productive rusher on designed run play (he has the highest RYOE_PER in the league). But we must consider that he only has 8 rushing attempts. If we move the filter to 25 rush attempts at least, you will see that J.Taylor has been the most productive rusher. He has 225 yards of RYOE total for the year through 9 weeks.

Top 10 leading rushers

The following graph shows that there was little influence on rush yards based on if a team was winning or losing and by how many points. If anything they tend to rush less when they have a lead, which is somewhat perplexing since you would assume the goal win your winning would be to run out the clock.

The next chart shows you rushing distribution by team and down. This gives you an understand how of teams are using their rush plays on a given down situation.

The graphs below show you the average rushing yardage per down by team. There is outlier behavior in the fourth down area because there are few fourth down rush attempts.

Below is the table mentioned above, but organizes rushing by average rushing yards by down. It can be filtered based on team, down, or average rush yards. It’s base filtering is based on average rush yards regardless of down. It can be sorted by any column.