I know that I comment a lot about opponent strength, but I believe in it deeply.... And as a result, I can't help myself... . I wanted to put together a post about this because I do think it is important to have this concept in our collective minds during this season. By all accounts, the schedule strength this season will be much harder than any previous season. I think schedule strength will shape our season as much or more than any other factor. Even if we have improved (which I think we have) it will likely look a lot "harder" than even last year.
I don't think this concept should be a controversial opinion. Vegas tells us every week that opponent strength absolutely matters. All of my years of watching football tell me that it is obvious to me that opponent strength matters. I am a college football fan too. This concept is even more evident in college football because there is no salary cap and there is a huge differential in talent. There is no salary cap in college. Teams can hoard talent. Every Power 5 football team looks like they could go to the College Football Championship in their 1st 3 games when they play teams like Idaho, Kent St, and UTEP. I also do a lot of college football prospect film watching. I have learned over the years that opponent strength and supporting talent for a player makes a huge difference in how a player looks. I don't spend a lot of time watching tape when a player plays against inferior talent. In those cases, I just look for bad plays. Good plays are expected, but if you make a lot of bad plays against non-NFL caliber talent then it is a concern. An above average player that plays against lesser talented opponents and is also surrounded by tons of talent will look like a future HOFer. But this concept of opponent strength also applies to the NFL. While it is not as obvious as it is in college football. It is still important.
To me the concept of opponent strength mattering just makes logical sense. It is a major part of my cornerstone philosophies of sports...
Player success = player talent + player effort + player IQ + player preparation + player decision-making + player technique + supporting talent + scheme + scheme fit + opportunity - (opponent talent + opponent effort + opponent IQ + opponent preparation + opponent decision-making + opponent technique + supporting talent + scheme + scheme fit)
QB is the most important position in football, but it is one of the most dependent positions in football
Individual stats are Team stats... and Matchup stats
This is especially true in football. Football is the most complex and inter-dependent sport on the planet. It is rarely if ever just about one person. There are so many things that impact the success of an individual or a team.
However, there is an enormous paradox that is created in sports fandom. Stats rule the landscape of sports. They have been used since the beginning of time for our debates. They are used to crown MVPs and HOFers. They are used for Fantasy Football. We love stats. I love stats! They are at the core of our fandom. But as powerful and cool stats are, almost all stats at the core are wildly flawed. They assign stats to an individual player, but they ignore all of the things that we all have known since we started watching sports. They ignore that behind the player that gets the stats there are coaches, a scheme, and 10 other players that helped them get the stats. And they also ignore that there are 11 players on the other side of the ball that are trying to stop that player from getting stats.
The Paradox of Stats and Success:
Stats are great and can help us understand/measure what we are seeing, but they also can mislead us. So, that is where the paradox starts. We all (myself included) fall into Stat Bias. Stats seem very concrete and they give us comfort. They are backed by math and science and they are usually based on things that actually happened. While all of those things are good things and all stats have some level of truth in them. Stats tell a story, not the whole story. All stats are limited to the data that is their input, the context that they were taken, the correctness of the algorithm that generated them, the person created the algorithm, and the person interpreting the stat.
But the lure of stats draws us in. They give us great confidence for our sports opinions. We can make claims of this player is great or this team is great. That is a big part of why we are all here.
Another bias in sports is what I call Success Bias. We are all guilty of this. Success is the reason that we watch sports. It is the reason that they play sports. We need a winner and a loser. We need to see if one player can overcome the competition of another. We need to see the great highlights of a great play. When a player or team wins, it creates Success Bias. Success Bias changes our perspective on what we just saw. Success washes away some of the concerns for that team/player prior to the "win". The reverse is also true. Success bias is a real thing. As an example, look at these past two weeks. The Bills lost a close game to a very good defensive team and Josh made some questionable risky decisions/plays. The Bills and Josh were the talk of local and national media. Everything that was concerning with Josh and the offense during that game and even before the game became enormous concerns in our "glass case of emotion" (thanks for the analogy Hal184!). In week #2, we play a lesser team and we win handily but Josh also made some risky decisions, but they turned out OK and we went on to win. As a result, this past week we were in a "glass case of tranquility". The local and national media proclaimed, "Josh bounced back and the Bills are back in it".
These biases are real. But it does not mean that they tell us the whole story. Most stats do not account for supporting talent, scheme, and opposing talent and scheme. In that respect, they are all flawed. Success does not always mean greatness. If the opponent was weak, then success is expected. It really does not tell us if the player/team was great or not.
These biases also fold back into what we see on the field. The great retort to stats is that "stats lie". That is partially correct most of the time. I think it is more correct to say that stats tell a "limited truth". The follow up retort to "stats lie" is usually that "my eyes don't lie". That is also not entirely correct. Our "eyes" can (and often do) deceive us by the same biases that make stats flawed. If we watch a game and a player makes a great play, many will say, "wow! He is great". But to know if he is great you have to account for the opponent, the opponent scheme, the play call, and all of the other supporting talent that enabled the great play. Sometimes the player is actually great and made it happen all by himself, but many times the success of the play was a product of many things on the field and behind the scenes.
If we can't trust stats and we can't trust our eyes, then what can we trust?
This paradox creates distrust in all the things that we use to digest sports. But I believe that we can find goodness in both stats and our eyes if we consider the limitations of what we are reading or seeing. That is what I want to hopefully prove in this post. But for the sake of time and to save some digital pixels on this screen, I want to focus this post on the impact of Opponent Strength on stats and what we see on the field. There have been several recent narratives that have dragged this up again. Hopefully, I will be able to show that opponent strength does matter in these discussions.
Delta between the strength of a team vs its Opponent:
I have found that Pts/Drive is a really good indicator of a team's ability and success, so I chose this metric to highlight the impact of opponent strength. In this chart, I took the delta between a defense's ave pts/drive and their opposing offense's ave pts/drive to show the talent gap between a team and its opponents. If I plot this talent gap against the resulting offense's actual pts/drive you can see there is an incredibly strong correlation. Actually, it is incredibly linear.
For me, that is all that I need to see to prove if opponent strength matters. But obviously there is more to the discussion than just this chart. This data shows the averages over a whole season , so some of the game to game variation is smoothed out.
But Let's go a bit deeper
Here is a chart of a defense's Ave Pts/Drive relates to the actual Pts/Drive scored against that defense for each game.
As you can see, the better the defense, the more likely that the opposing offense will score less (note the trendlines). While there is variance around the trendline from game to game, overall there is a pretty strong correlation. And even though the slope of the trendline may not look significant in the graph, it can be. The right end of the trendline is at 2.5 pts/drive and the left hand is about 1.2 pts/drive. There are usually 10-12 drives in a game so a difference of 1.3 pts/drive ends up being roughly 13.0 to 15.6 pts a game. That is significant.
Also, it is interesting to see that the slope of the trendline for the Bills' actual Net Pass Yds/Att matches the slope for the whole league. That also shows that the impact of defenses on the Bills' offense is in line with the league. The trend of opponent strength is meaningful. The good news (and not surprising news) is that the Bills are above the NFL average by about a .5 Pts/Drive which translates to roughly 5-6 pts per game.
2021 vs 2022 "It Looked Harder in 2022":
So, now that I have hopefully shown you that opponent strength matters. I want to show some Bills-specific examples of this impact. Last year, the prevailing opinion was that it "looked harder" and it was "less consistent" for the offense in 2022 than it was in 2021. I don't think that is debatable. All of those things were true. But as I said in the preamble, what you see on the field can look good or bad based on many factors. Opponent strength is one of those factors. Let's take a look at the opponent strength between the two years.
2021 vs 2022 Metrics (Bills Offense's Opponents):
2021 vs 2022 Metrics (Bills Defense's Opponents):
While the 2022 Bills did not play a murderer's row of opponents, the 2021 Bills played a historically easy schedule. The defenses that the Bills played in 2022 we much better at containing teams wrt yards, but not at points. Maybe that was the reason that it "looked hard", but the team was still able to put up points. I know that these are not perfect metrics, but it is hard to ignore the stark difference between the opponent strength in 2021 compared to 2022.
Now I know that the Bills ended the year with two historically great playoff games (and they were), but the Bills' offense was not going against top defenses. The Pats were very good in raw stats, but they were aided by the same easy schedule that the Bills had in 2021.
Pats Wild Card Game:
KC Divisional Game:
Now. Don't get me wrong. I can see the pitch forks are already out. The Pats and KC playoff games in 2021 were epic and wonderful and all of the other adjectives that you want to give them. And the Bills were great in both of those games and they should get credit for a great performance. Josh put the team on his back in those games and he delivered. But if we are being real, the opponent strength should be considered. KC was not a good defensive team in 2021. I know we are fans, but part of good assessment is to try and be unbiased and consider all of the factors. Opponent Strength is one of those factors.
Listen, I am not pointing this out to trash on the 2021 Bills. The Bills are my favorite team. Trashing the team is not my gig. But I am passionate about trying to be realistic in my assessment and expectations. And part of that is my passion for accounting and measuring opponent strength in my metrics. Like I said they had a great run and they had a great chance against one of the best offenses in the league. Could the Bills have gone on and won it all? Sure. Once you are in the playoffs, anything can happen especially with a talent like Josh Allen at the helm. It was a missed opportunity for sure.
But I am posting this post because I wanted to hopefully show how important it is to consider opponent strength. It matters for sure. My hope is that this post will convince you to be curious of opponent strength, to be wary of stats that don't consider opponent strength, to consider opponent strength when comparing teams from year to year, to consider opponent strength when assessing a bad loss (like against the Jets in Week #1), to consider opponent strength when assessing a big win (like this past week), and to consider opponent strength when watching the game with your eyes.
As Ted Lasso said in his epic dart scene against Rupert when he quoted Walt Whitman... "Be curious, not judgmental".
I think we are going to need this mindset this year. On paper, the 2023 opponents are significantly stronger than in 2022. There may be some really ugly games this year, but if they are ugly wins, then that is all that we need.
2021 vs 2022 vs 2023 Metrics (Bills Offense's Opponents):
2021 vs 2022 vs 2023 Metrics (Bills Defense's Opponents):
It is going to be tough sledding this season! But I can't wait to see if they are up for the challenge and win a lot of ugly games! If we make the playoffs, I think this team has as good or better chance than we have ever had. I think we can if everyone pulls on the same rope and there are incremental improvements across the whole team. But just keep in mind that that Opponent Strength will Matter this season.
That's enough from me. What do you think? What do you see in the data?