A look at who is good at throwing the Deep ball and who isn’t from 2006-2021.

Before the Denver Broncos traded for Russell Wilson, I had heard that he was good at throwing the deep ball, but I had never really dived into it. How good is he? What is “good” in terms of deep passing, for that matter, how do you define “deep”?

My readers know that I write about questions that I want to answer. This article will be no different. First let’s look at how “deep” is defined.

Pro-football-reference.com defines a deep pass as one that lands 15 yards or more from the LOS. SIS.com sets that threshold at 20 yards. NFL NextGen Stats also uses 20 yards. Since stathead.com is a part of PFR and they have the most searchable data (and I have a sub there), I will use them for this study and their definition of deep.

The pass distance and direction data for every play only goes back to 2006. So I searched every regular season play from 2006 to 2021 to see which QBs were and are good at throwing deep and which QBs are not. But I also wanted to see what and how the league does as a whole on deep throws.

If you parse it another way it looks like this...

So using 15 yards from the LOS as deep, about 20% of all passes each year are deep. Oddly enough, the 16.6 percent value for deep passes in 2021 was the lowest value in the whole time frame. 2021 had more shorter throws than other recent years. On average about twelve percent of all completions are deep, but thirty percent of all TDs are on deep passes. This is not to be confused with short throws that turn into long TDs because of YAC. The most interesting thing to me was the 45 percent of all interceptions are on deep throws. This is interesting partly because PFR only has two throw distances in their play synopsis - short and deep - where short is any throw that lands 15 yards or less from the LOS. If I had access to SIS’ or NextGen’s full data set, I’d love to look at INTs by their distances with short, medium and long categories (see below).

During this time span, league-wide deep completion percentage was as low as 36.1 percent in 2017 and as high as 43.8 percent in 2020. Completion percentage in the NFL overall has gradually increased from its low of 59.8 in 2006 to a high of 65.2 in 2020. Overall TD:INT ratio has also been gradually increased over time peaking at 2.2:1.0 in 2020. It would appear that many things peaked in 2020.

If you prefer I have a plot below of some of this deep passing data, and it is interesting to see that inflection points coincided to the entrance of the three best active deep passers into the league.

In 2012 the league as a whole had a big change in deep TD:INT ratio - it jumped from 1.03 to 1.20 year over year. Deep TD:INT ratio has stayed at or above 1.10:1.00 ever since. 2012 was Russell Wilson’s rookie year. In 2018 there was a big jump in deep completion percentage year over year. That was Patrick Mahomes first season as a starter in the NFL. In 2020, there was a spike in both deep completion percentage and deep TD:INT ratio. That was Justin Herbert’s rookie year. In terms of career numbers the three best active deep passing QBs are

1. Russell Wilson
2. Patrick Mahomes
3. Justin Herbert

The only current AFC West QB who is not in top 5 in history (2006-2021) in deep passing is Derek Carr and he is 17th (which you can see below).

This is a good time to show all the career numbers from 2006-2021 keeping in mind that some of these QBs had careers that started prior to 2006. There were 112 QBs who had a minimum of 100 deep throws. Many of the guys, who are good deep passers, have 100 or more deep throws each year. The data below is sorted by passer rating

I was surprised to find the Jeff Garcia was so good on deep passes. If this data set gets expanded back to 1994, it would be interesting to see how early Peyton Manning fared and how late-career John Elway fared on deep balls. That would allow us to see the early and middle years of Garcia’s career.

Many of the best deep passers in the game entered the league in the past ten years. Only Garcia, Tony Romo and Drew Brees from the top 10 above have careers that began prior to 2006. Romo first started a game in 2006 though, so this time span covers 99 percent of his career passing attempts.

It would appear that QBs have gotten better at deep passing, but before we go too far down that path, we need to think about the effect the pass interference (and the lack of that call) has had on this data. The way in which pass interference has been called has changed fairly dramatically during this time and while that might not have a direct effect on whether a ball is caught or not (generally if the ball is caught and DPI is called, the penalty is declined) it has definitely had an indirect effect on how defensive backs defend deep passes. In my opinion, the way that DPI has been called in the past ten years has made it easier for QBs to complete deep passes because defensive backs are keenly aware of touching the receiver at all before the pass gets there. They also have to be careful to not hit the receiver too hard after the catch or they could get called for unnecessary roughness.

In the table above I color-coded two columns: passer rating (PR) and deep pass percentage. To be top 10 in deep passer rating you have to have a PR of 105 or better. To be in the “green” in deep pass percentage you have to be in the low 20’s in terms of percent of all of your passes that are deep. The average deep pass percentage for this group was 19.8 percent. Very few passers were below 16 percent, but many of those names that didn’t surprise me based upon what I know of those QBs: Daniel Jones, Alex Smith, Jimmy Garoppolo and Joey Harrington. I found it interesting that Herbert is really good at deep passes, but doesn’t throw them very often. Daniel Jones and Gardner Minshew are similar as was Jeff Garcia. In fact the only QB in the top 10 in deep PR who is above average in deep pass frequency is Russell Wilson - another reason he is the daddy of the deep ball.

Of these 112 QBs, there was a great deal of clustering around the average with fairly few guys near the margins for deep pass frequency. Just focus on the distribution below since you can’t read the names below each bar (or if you can you have really good eyes or a massive monitor/TV upon which you are reading this).

Only five of the 112 throw/threw deep 25 percent of the time or more: Vince Young, Deshone Kizer, JaMarcus Russell, Brandon Weeden and Tim Tebow. Weeden and Tebow both threw deep on one of three of their passes during their careers. Interestingly enough, Tebow was the best of these five with a passer rating of 66.0 on deep throws (that was 66th of 112). Kizer and Russell were 110 and 111 of 112 in deep passer rating. Only Brady Quinn was worse among qualifiers.

So we have discussed the guys who do it frequently and are good at it (Wilson, DeShaun Watson and Aaron Rodgers), the guys who don’t do it frequently but are good at it (Daniel Jones, Justin Herbert and Dak Prescott) and the guys who did it frequently and are mostly terrible at it. The other quadrant is guys who are bad at it and don’t do it very often, you can see that in the chart below. Brock Osweiler and Joey Harrington fit into the last category (quadrant)

The vast majority of QBs were clustered near the X axis (average for frequency), but to the left of the Y-axis (poor in proficiency).

I also looked just at last year. Looking at what receivers got deep targets helps us understand something - to be a good deep passer you need receivers who can catch the deep ball. Justin Jefferson led all receivers with 44 deep targets in 2021, but Tyler Lockett was tied with two others for second with 41. Wilson’s other deep threat in 2021, DK Metcalf, had 37 deep targets. The vast majority of those throws were from Wilson.

Despite missing three games, Wilson managed to throw 98 deep passes in 2021. 51 passers threw ten or more including the other Seattle starting QB, Geno Smith, who threw 13. For the Broncos in 2021 Teddy Bridgewater and Drew Lock combined to throw 93 deep passes (67 and 26 respectively).

Sorting by passer rating, Geno Smith actually had the highest deep PR in the league at 138.6 last season (admittedly on only 13 throws). Russell Wilson was second with a deep PR of 112.0. Smith only threw deep on 13.7 percent of his passes in 2021 while Wilson threw deep on 24.5 percent of his. Herbert was 4th and Mahomes was 15th last season in deep PR. Tom Brady throw the most deep passes in 2021 with 136. His deep PR of 98.3 was 14th. The QB with the highest percentage of deep attempts in 2021 was PJ Walker with 25.8 percent. He was, however, 50th of 51 in terms of deep PR. The quadrant view is below. Notice that Russell Wilson is in a space in the upper right by himself.

If we shift to SIS data, which uses 20 yards from the LOS as the threshold, we find that only 40 passers had the minimum 10 attempts. SIS has a stat that is similar to passer rating called IQR (the formula is proprietary), but in terms of deep IQR, the best in the league in 2021 was Tyrod Taylor with a value of 110. I found this quite intriguing because, according to PFR, Taylor was not a good deep passer in 2021. His PR on attempts 15 yard or longer was 51.9. It was also interesting to see him with a deep IQR of 110 considering his deep completion percentage of 33 percent in 2021 was also fairly low (average for the top 40 QBs was 38.9% on deep throws according to SIS). According to PFR he was 4 of 19 with 1 TD and 1 INT on deep throws. According to SIS he was 4 of 12 on deep throws (they don’t show TD and INTs). Oddly enough none of the stat sites I use NFL.com, ESPN.com, SIS.com and PFF.com has throw distance data for QBs (at least not free to access). This site shows completions by distance, but it’s yards gained on play, not throw distance and they don’t have attempts (or TDs or INTs).

By inference for Taylor, all seven on his deep incompletions were in the 15-20 yard range. If you have the time, you can tease out the data from the game-by-game throw charts (available for most players). Looking at Taylors six starts from 2021, he indeed only completed four deep throws (20 yd or more) all year. The first three were in week one vs JAX and the fourth was a TD throw in week 12 against the Jets. The reason why Taylor’s PR and IQR are so different is that (I think) two of his interceptions were in the 15-20 yard from LOS range. Oddly, the Nextgen stat data is hit or miss since his throw chart from week 12 fails to show him throwing an INT even though he did throw one in that game. So it’s possible that SIS only has Taylor as only throwing one deep INT in 2021 and they may discount that INT if they deem it “not his fault”. Who knows? They don’t tell us how they calculate IQR. So let’s get back to Russ.

By SIS data Russell Wilson was 12th in terms of deep IQR among qualifiers in 2021. NFL NextGen Stats mostly agrees. They only show the top 10, but they have Russell Wilson as the 8th best deep passer in 2021. They use “passing score” instead of passer rating or IQR. You can read about it here, but it basically is a much more balanced evaluation of passer prowess than passer rating but it does not take into account running ability like ESPN’s QBR - making it much more useful for comparing deep passer ability. The best deep passers in the league in 2021 had passing scores of 99 on deep throws (20 yards or more from LOS). Josh Allen and Matt Ryan were tied for first at 99. Wilson’s 93 placed him eighth. Aaron Rodgers, who many fans would think is the best deep passer in the league, was 10th in their ranking for 2021 with a deep passing score of 91.