Hey everyone! Back in late March I wrote a media article where I compared the defensive and offensive performance of all previous Ultimus champions in the ISFL as to answer the old question that permeates a lot of discussions in several collective sports: “Does Defense win Championships?” for the ISFL landscape. My conclusion was that the answer was a slight yes, since championship winning teams were slightly better on defense than on offense. As someone who has played as a mostly defensive player on a variety of sports growing up, I understand, however, that it’s very difficult to put the art of defending into numbers. And resuming it to just points conceded, as my article had limited itself to, made me curious of the details of the defensive statistics of each team. So that led me to write this article: a more in-depth analysis of the defensive statistics of each ISFL Ultimus champion.
I decided to do this by gathering all team-based defensive stats available to me on the ISFL index and put them on a table for comparison. As you can see below, I expanded my previous table of Ultimus champions to include 6 stats: sacks, interception + pass deflections, forced fumbles + fumbles recovered, passing yards per game, running yards per game, and total turnovers (INTs + FR). I also avoided color coding the advanced stats for this table, as I believe it was visually polluting.
Again, I think it’s really important to view all league seasons separated into 3 main eras: the early era (S1-S15, which featured 8 teams and a 4 team playoffs, with S1 as an exception, that had 6 teams), the modern era (S16-S21, which featured 10 teams and an 8 team playoffs), and the contemporary era (S22-S27, which featured 12 teams until in seasons 22-24, and now features 14 teams, maintaining the 8 team playoff structure). As such I looked at the importance of each stat over every era and see how it has changed over time and, most importantly, calculated the weighted rankings: the absolute ranking of a certain stat divided by the amount of teams on the league during a season, as noted in the table below.
There are a lot of numbers all over your screen right now, so I’ll make them more understandable. First of all, lets look at the importance of each stat. A high number means that a stat has more teams with bad rankings on this stat, so it doesn’t determine a champion as much as low numbers. With this in mind and considering all 27 seasons, each stat is ranked in this order, from most important to least:
1. Rushing Yards allowed per game
2. Total Yards allowed per game
3. Sacks per game
4. INTs + PDs per game
5. Turnovers per game
6. Passing Yards allowed per game
7. Forced Fumbles + Fumbles Recovered per game
That is determined by the weighted ratings of all season, since it takes into consideration how each ranking has become more valuable every season (In S1, for example, a team had to beat 5 other teams in order to have the best defense. In S27, a team had to beat 13 other teams, which is considerably harder). The importance of considering weighted ratings is also valid for analyzing how certain aspects of the defensive game have changed in significance over the seasons, which is what the third table shows below. In it, I calculated in the percentile in which teams finished, on average, for each stat over the three periods of the ISFL.
Some very interesting conclusions can be taken from here. The biggest outlier personally has been how defending the running game has become very important in recent times after a drop off in between seasons 16 and 21. With the improvement of offense over the last few seasons (as championship teams are 16% more likely to be good offensively than defensively, as my last article shows), stopping the running game in key situations has become of utmost importance (all of the last 4 Ultimus champions have been mediocre to bad defensively overall stats wise, but have been dominant on the ground, registering the least Running Yards allowed per game in the league).
And the second biggest outlier would be how it would seem by the decline in stats percentile that teams have gotten worse on defense over the past 5 seasons. But that is far from the truth. I believe that has to do with the league’s size. When the league had 8 teams, it was considerably easier for teams to form dynasties (such as the ones established by the Outlaws and Otters who won all the first 6 titles of the ISFL) and superteams who were dominant in both sides of the ball and in all aspects. Now, with almost twice as many teams in the league, superteams are extremely unlikely. It is very difficult for a team to be great on both sides of the ball and almost impossible to be great in all aspects of defense. This means that teams are great not because they are dominant, but because they are good when they need to be and can close out games or stop important late 4th quarter drives. Most teams don’t have good pass rush, good secondary players, are good at stopping the run, and can get a lot of fumble. They can be two, or maybe even three of these, but not all.
It should be noted, also, that there is a possibility to write an even more in-depth analysis. Some stats, however, were either not readily available (such as TPE per position and total defensive TPE) or would take me a long time to write up (such as Yards per Drive, Yards per Rushing attempt, expected points conceded, and other advanced stats that, whilst readily available for the average NFL fan, aren’t so much here). I do plan to eventually do some kind of advanced stats-based look for a team or multiple teams in the future based on one season, where I could go over every single play and take notes on them. For now, however, I believe this is as further down the rabbit hole as I can go.
I will also be using these stats sometime over the next few days to make an article in which I analyze the best and worse defenses to ever win it all from an analytical standpoint, which I will link here when it is uploaded, so keep an eye on that!
I decided to do this by gathering all team-based defensive stats available to me on the ISFL index and put them on a table for comparison. As you can see below, I expanded my previous table of Ultimus champions to include 6 stats: sacks, interception + pass deflections, forced fumbles + fumbles recovered, passing yards per game, running yards per game, and total turnovers (INTs + FR). I also avoided color coding the advanced stats for this table, as I believe it was visually polluting.
Again, I think it’s really important to view all league seasons separated into 3 main eras: the early era (S1-S15, which featured 8 teams and a 4 team playoffs, with S1 as an exception, that had 6 teams), the modern era (S16-S21, which featured 10 teams and an 8 team playoffs), and the contemporary era (S22-S27, which featured 12 teams until in seasons 22-24, and now features 14 teams, maintaining the 8 team playoff structure). As such I looked at the importance of each stat over every era and see how it has changed over time and, most importantly, calculated the weighted rankings: the absolute ranking of a certain stat divided by the amount of teams on the league during a season, as noted in the table below.
There are a lot of numbers all over your screen right now, so I’ll make them more understandable. First of all, lets look at the importance of each stat. A high number means that a stat has more teams with bad rankings on this stat, so it doesn’t determine a champion as much as low numbers. With this in mind and considering all 27 seasons, each stat is ranked in this order, from most important to least:
1. Rushing Yards allowed per game
2. Total Yards allowed per game
3. Sacks per game
4. INTs + PDs per game
5. Turnovers per game
6. Passing Yards allowed per game
7. Forced Fumbles + Fumbles Recovered per game
That is determined by the weighted ratings of all season, since it takes into consideration how each ranking has become more valuable every season (In S1, for example, a team had to beat 5 other teams in order to have the best defense. In S27, a team had to beat 13 other teams, which is considerably harder). The importance of considering weighted ratings is also valid for analyzing how certain aspects of the defensive game have changed in significance over the seasons, which is what the third table shows below. In it, I calculated in the percentile in which teams finished, on average, for each stat over the three periods of the ISFL.
Some very interesting conclusions can be taken from here. The biggest outlier personally has been how defending the running game has become very important in recent times after a drop off in between seasons 16 and 21. With the improvement of offense over the last few seasons (as championship teams are 16% more likely to be good offensively than defensively, as my last article shows), stopping the running game in key situations has become of utmost importance (all of the last 4 Ultimus champions have been mediocre to bad defensively overall stats wise, but have been dominant on the ground, registering the least Running Yards allowed per game in the league).
And the second biggest outlier would be how it would seem by the decline in stats percentile that teams have gotten worse on defense over the past 5 seasons. But that is far from the truth. I believe that has to do with the league’s size. When the league had 8 teams, it was considerably easier for teams to form dynasties (such as the ones established by the Outlaws and Otters who won all the first 6 titles of the ISFL) and superteams who were dominant in both sides of the ball and in all aspects. Now, with almost twice as many teams in the league, superteams are extremely unlikely. It is very difficult for a team to be great on both sides of the ball and almost impossible to be great in all aspects of defense. This means that teams are great not because they are dominant, but because they are good when they need to be and can close out games or stop important late 4th quarter drives. Most teams don’t have good pass rush, good secondary players, are good at stopping the run, and can get a lot of fumble. They can be two, or maybe even three of these, but not all.
It should be noted, also, that there is a possibility to write an even more in-depth analysis. Some stats, however, were either not readily available (such as TPE per position and total defensive TPE) or would take me a long time to write up (such as Yards per Drive, Yards per Rushing attempt, expected points conceded, and other advanced stats that, whilst readily available for the average NFL fan, aren’t so much here). I do plan to eventually do some kind of advanced stats-based look for a team or multiple teams in the future based on one season, where I could go over every single play and take notes on them. For now, however, I believe this is as further down the rabbit hole as I can go.
I will also be using these stats sometime over the next few days to make an article in which I analyze the best and worse defenses to ever win it all from an analytical standpoint, which I will link here when it is uploaded, so keep an eye on that!