The Factors: Rain, ball security, and efficiency

Based on the research I performed a week ago, quarterbacks are already at a disadvantage when it rains because their coaches tend to run the ball more in those conditions. But it stands to reason that rainfall does more than hurt a quarterback’s volume. The reason coaches would change their called ratios of passes to runs in the first place is because they anticipate the rain will either decrease the effectiveness of passes, increase the effectiveness of runs, or both, which would create a new equilibrium for the two. I can think of several potential reasons for those changes to test, most of which deal with the slickness of a wet ball.

In the passing game, there are actually two sides of the exchange that a slick ball could impact. The first is for the quarterback, who needs a secure grip of the football as part of his mechanics to make a strong and accurate throw. The second is for the receiver, who needs friction between his hands and the ball to make the catch. If a wet ball compromises either or both of those ends of an attempted pass, then one would expect to see that loss reflected in a decrease in completion percentage. When I ran weighted average comparisons of completion percentage in both light and moderate rain compared to no rain, I saw that expected trend. The heavier the rain, the lower the completion percentage.

Meanwhile, in both light and moderate rain, passing yards per attempt decline by about a third of a yard. That is a similar loss in passing efficiency as what quarterbacks see when they leave home to play on the road. Obviously, completion percentage is a driver of yards per attempt, so some decrease makes sense. But I found it odd that the two YPA numbers were basically the same when moderate rain created a bigger drop in completion percentage than light rain. I think the explanation for that dissonance can be found in the receiver side of the exchange.

As expected, receivers catch fewer balls when it rains. About half of that decrease comes from increased drops, which leaves the other half to come predominantly from less accurate passes. Those rates are pretty similar in both light and moderate rain, but notice how receiver yards per target is much higher in moderate rain than light rain. It has to be said that my sample size of pass attempts in moderate rain is much smaller than it is for light rain, so maybe the difference is overstated. But I think it’s more likely there is a real reason for the difference, which is an increase in the difficulty of the tackle.

Even in light rain, a football will become slick. It may take a bit longer than it would in moderate rain, but it will still happen, and I suspect it will still impact the majority of throws a quarterback makes in a rainy game. I suspect the difference of light and heavy rain is more acutely felt in players’ footings. Wet turf will hurt the performance of both offensive and defensive players, but I doubt it will do so equally. Offensive players can use their judgments on the slickness of the turf and on the proximity of defenders to them to inform their decisions on when and how to make cuts. But when a ball-carrier makes a cut, a defender has no choice. He has to attempt to make the corresponding cut if he wants to make the tackle. I expect that becomes very difficult in moderate rain, and I think that neatly explains why receivers see an increase in their yards after the catch in moderate rain versus light rain. That in turn explains why receiver yards per target is higher in moderate rain and why quarterback yards per attempt is not lower in moderate rain than it is in light rain when completion percentage is lower.

One final factor I wanted to test is fumble rates. I tested these for running backs specifically because they have distinct chances for fumbles in each of their attempts. That’s just simpler than it is for quarterbacks, who can fumble an exchange with the center, fumble as a result of contact with a pass-rusher, or fumble on a pump fake. It’s also simpler than it is for receivers, who must contend with the ephemeral boundary between a catch-and-fumble and an incomplete pass. But I suspect every holder of the football will see a similar impact on their chances for fumbles from rain as running backs do.

Another table, another confirmation of expectations. Running backs fumble a bit more in light rain than they do without rain, and they fumble much more in moderate rain.

While I was testing backs, I also ran weighted average comparisons for their yards per carry and avoided tackles per carry, and those results provide more evidence for the greater impact of moderate rain on the defensive player. Both yards per carry and avoided tackles per carry increase with rain.

Because of the Thanksgiving holiday, I’m writing this before the Thursday games but it will be published after. That makes for awkward timing in spinning the research forward to the games this week, since Sunday’s offerings are still four days away, which is murky timing for weather forecasting. As things stand now, the one game this weekend that looks to have a chance for rain is the late Sunday contest between the Broncos and Raiders in Oakland. That is a shame for Amari Cooper — he doesn’t need any added difficulties in catching the ball. Also be mindful of that possibility for Derek Carr, Michael Crabtree, C.J. Anderson, and Marshawn Lynch. If past coaches are any indication, it could be a Beast Mode kind of weekend.