The Investigation Of Fantasy Football QB Performance In High Winds
From the casual player to that guy at the office with eight different teams, we all hesitate when we hear the following words: “High winds expected during the game.” Weather forecasts of this type seem to have the same effect on all players due to the unpredictability of how the game will play out. Even from a qualitative standpoint, our expectations change from the foreseeable to the unpredictable. If there’s rain, we can mentally project a more conservative offensive approach. Snow adds another variable with the reduced temperatures, but yet again, owners can understand decreased performances. High winds throw a wrench into everyone’s lineups.
It was Week 5 of 2016, and DeSean Jackson was set to burn Ravens LCB Shareece Wright that had been torched the previous three weeks. As a contrarian upside play, I used him in a few lineups and was completely dismayed to see this 3-35-0 stat line at the game’s conclusion. Hindsight analysis produced two red flags: an injury to Wright kept him out of the game… and 16 mph winds were forecasted. Any podcast or article reviewing the game acknowledged that the wind could have made a difference, but couldn’t quantify its significance. A 2012 study conducted by Brian Burke showed a steady decline in average Adjusted Yards per Attempt (AYPA) as the wind speed increased for all quarterbacks.
However, averages describe the central value of a population, which is great for detailing trends. But as a fantasy owner, I don’t really care about all 32 quarterbacks in the population. I care about the one I drafted or the one I picked up off the waiver wire based on the matchup I spent hours researching. This study looks to decompose the trend to visualize how each quarterback approaches the game under less-than-ideal weather conditions. Projecting a range of outcomes is a large component of evaluating fantasy football matchups. Like poker, past experiences and historical context help establish that range.
The recent Super Bowl gave me my first two test cases in Matt Ryan and Tom Brady. Regular season stats were collected from Pro Football Reference for all games started. Games played in a dome or a stadium with a retractable roof (and the roof was closed) were assumed to have no influence from the wind.
|0-4 mph||5-9 mph||10-14 mph||15-19 mph||20+ mph|
|Tom Brady – 234 starts||59||81||56||27||11|
|Matt Ryan – 142 starts||103||24||12||3||0|
Brady’s longevity and home stadium make him the prototypical subject with double digit samples for each speed range. Ryan’s lived the ‘dome life’ with the advantage of a dome at home and a division opponent (Saints) making his experience a bit lopsided. In either case, the following metrics were collected and averaged for comparison:
- Adjusted Yards Per Attempt (AYPA)
- Number of Pass Attempts
The 2012 study was used as a baseline comparison to the averaged AYPA values for each quarterback. Number of attempts and completion percentage were used for direct comparisons with ’0-4 mph’ as the baseline. The calculated differences were used to highlight the impact of increased wind speeds. Factors that were not considered in this phase, but will in future iterations:
- Impact of combined weather factors (e.g. 9 mph winds plus rain)
- Strength of opponent
- Offensive motivation (e.g. Brady starting Week 16, but not finishing the game)
Looking at the individual average AYPA for Matt Ryan and Tom Brady, they both perform above average when compared to the original study. Subsets of data up to 2012 were formed to create an even comparison with a similar conclusion. For the first three brackets, Matt Ryan showed at worst an 8.25% decrease in AYPA while Brady actually jumped in AYPA by nearly 19%. In games with winds above 15 mph, both quarterbacks seem to take a step back. When compared to ideal (no wind) conditions, Tom Brady and Matt Ryan’s average AYPA decrease by 11% and 8.25%, respectively. To not be blinded by average values, distribution curves were generated for each subset to ensure the averages were not heavily influenced by a smaller set of samples.
So they’re throwing shorter passes, OK. But in fantasy, volume is king. If the wind picks up, do I have to tilt watching my QB hand the ball off three straight plays? For Tom Terrific & Matty Ice, not at all.
Matt Ryan’s APG, while taking an early 0.95% drop, actually outperforms the baseline for the higher wind speed brackets at an increase of 2.95% and 1.04%. While he’s never played in a game with a recorded wind speed of greater than 20 mph, this lends credence to the idea that there is not a massive offensive deviation when the wind picks up for the Falcons. Tom Brady? Just short of a gale tornado touching down, the recent Super Bowl champ would still be slinging the rock. He outperforms his baseline in all categories by an average of 4.51%. Think about that for a second. The wind speed is increasing and Tom Brady is throwing more. In a 2010 game against Chicago, the recorded wind speed was 30 mph and he threw the ball 40 times with a 10.21 AYPA. While you’re not benching Tom Brady for anyone, this information should help cement your decision to leave him in during these types of conditions.
Each matchup presents its own set of questions an owner must consider prior to roster lock. This study looks to eliminate the fear of wind by adding historical context to the debate on its impact. Early indications paint the picture of weather actually aiding your fantasy day. Remember the Week 10 Packers vs. Redskins primetime game? Pictures of flags almost flying off their poles, trash blowing everywhere, Twitter ablaze with hot takes, and an estimated 23 mph wind speed at kickoff. Cousins’ stat-line: 30 attempts, 375 yards, and 3 touchdowns with a 14.5 AYPA. Rodgers’ stat-line: 41 attempts, 351 yards, and 3 touchdowns with a 10.02 AYPA. There’s still a ton of work to be done, as mentioned before, to normalize the data with iterations coming as updates are made. However, it has been seen that the wind is not as devastating to fantasy as previously noted.
To discuss this and other fantasy football topics, feel free to follow me (and engage!) on Twitter @frolow625