## Remember when 300 yard passing games were for losers? Part II

Back in December 2009, Jason Lisk wrote about a recent trend in the NFL: quarterbacks throwing for 300 passing yards and actually winning. Jason wondered whether that was something fluky, or a sign of the shifting nature of the NFL. With the benefit of hindsight, I think the answer is…. well, I think it’s pretty clear.

Including playoffs, quarterbacks who threw for 300+ yards in a game during the 2009 season won an incredible 63.3% of games. And that mark remains the highest in modern history. Over the last five years (2010 to 2014), quarterbacks have won 52% of games when cracking that mark; during the decade of the ’90s, quarterbacks won 53% of their games when throwing for 300+ yards.

Of course, the likelihood of a quarterback throwing for 300+ yards has increased significantly. Over the last four years, quarterbacks have thrown for 300+ yards in 25% of all games, an enormous increase relative to most of NFL history. The graph below shows both pieces of information: in blue, and measured against the left Y-Axis, shows winning percentage by year when a quarterback throws for 300+ yards; in red, and against the right Y-Axis, is the percentage of all games where a quarterback hit the 300+ yard mark:

The 2009 season was certainly an outlier; other than that, there doesn’t seem to be much of an increase over time (at least since the mid-’80s) in winning percentage for 300+ yard passers. The table below shows the information from the chart above but in data form:

Year300 passWinsWin Perc300+ Rate
197013753.8%3.6%
1971114.540.9%3%
19728450%2.2%
1973104.545%2.7%
1974114.540.9%3%
197517952.9%4.7%
1976151066.7%3.8%
19776466.7%1.5%
197817741.2%3.8%
1979462145.7%10.3%
19805827.547.4%12.9%
19816123.538.5%13.6%
1982432046.5%17.1%
1983692637.7%15.4%
1984753952%16.7%
1985663147%14.7%
19866331.550%14.1%
1987603151.7%14.3%
19885629.552.7%12.5%
1989693550.7%15.4%
1990472961.7%10.5%
1991572950.9%12.7%
1992452146.7%10%
1993412151.2%9.2%
1994693753.6%15.4%
1995844148.8%17.5%
1996502958%10.4%
1997472042.6%9.8%
1998562951.8%11.7%
1999754560%15.1%
2000683652.9%13.7%
2001743851.4%14.9%
20028447.556.5%16.4%
2003673349.3%13.1%
2004873843.7%17%
2005663248.5%12.9%
2006683754.4%13.3%
2007834351.8%16.2%
20088148.559.9%15.8%
20091096963.3%21.3%
2010994949.5%19.3%
20111297155%25.2%
20121336347.4%26%
20131246250%24.2%
201412974.557.8%25.2%

Update: As requested by Nick in the comments, here’s a graph showing the win percentage of the team that threw for more gross passing yards in each game.

{ 20 comments }

better stat to explore: % of games where passing leader was winner

• Thanks, Nick. Good idea: I’ve updated the post.

• Richie

It’s interesting, but I don’t know if it’s better. It answers a slightly different question.

One is basically: Is a heavy pass-based offense a good way to win games?

Your question is: Does the more effective passing team usually win?

I think its addressing the same basic question, right?

• Richie

It’s similar, but not the same.

In the extreme, if a team rushes for 240 yards and passes for 135 yards and beats a team that only passed for 120 yards, I don’t think that does much to tell us about the ability of a passing offense to win a game.

Is this total team passing yards per game? Or individual? And are sacks included here?

• In the last graph, it’s total gross team passing yards — so not the individual and excluding sack data. In the rest of the post, it’s individual passers and also not including sack data.

The thought that I had was that a QB could throw for a portion of the 300 yards, get hurt (or benched), and his replacement could get the team total passing yards over 300. Not sure if this would affect the overall win% results significantly, just an idea.

Looks like passing is more volatile now – maybe we have a lot more “good” big-yardage passing games and a lot more “garbage” big yardage passing games

• It makes sense to me that passing for 300+ wouldn’t indicate losing as much as it used to. 300 yard games use to be an outlier, but they are much more common now. In 1970, the average team gained 161 net yards per game. It went down as far as 141 in 1973 before reaching an all time high of 237 last year. As it becomes increasingly favorable to pass, why not pass?

Some more interesting tidbits since 1970:

In 1970, 57.3% of offensive yards came from passing; last year, 68.0% came from passing (the highest portion in history).

https://pbs.twimg.com/media/CMnlGUYWUAY4XfB.png

In 1970, 51.9% of first downs were on passes; last year, 61.5% of first downs came from passes (and 29% from runs, with about 10% from penalties).

In 1970, 37.3% of points came from passing touchdowns, and 24.9% came from rushing touchdowns. Last year, those numbers were 42.5% and 18.6%, respectively.

I’d like to factor out garbage passyards.

• If you can find play by play from the 70s, have at it dude.

• I’m still surprised, though, that the 300-yard rate hasn’t changed much over the last 30 years. Thoughts?

• Richie

Yeah, I do find this surprising. It seems like everybody has gone pass happy. Maybe this is a residue of adding 4 NFL teams over the past 20 years. I assume the raw number of 300+ passing games has gone up a little over the last 30 years, but the percentage hasn’t changed much.

• Going through my offenses of NFL history series, I noticed that the 70s and 90s look like outliers in an otherwise general upward trend in offensive production. I think the 300 yard rate falls into that same shape. If you were to look at is as a line graph, I think the trend would show up visibly. I can’t do it on my phone right now, but I think I can visualize the table as a graph. I might be wrong though.

• Roger Kirk

Maybe this is too obvious to mention, but are you excluding games where both QBs passed for 300+ yards? Such games, which are more common in recent years, would tend to drag the overall average win % towards 50%.

• Yeah, I considered doing that, but opted against it (out of laziness). It would drag the overall win % towards 50, but I figured the effect probably wouldn’t be too big. Something to consider, tho.

• Josh Sanford

This is TOTALLY unrelated…but you have normalized sack seasons yet, where you adjust for league-wide sack rates and number of drop backs, etc. I was asking because it seems like these Justin Houston-type seasons would shrink when you compare them to Gastineau and LT. (If my question doesn’t make sense, what I am talking about is the same analysis that you recently did to interception seasons.)