Jared and I met in Calculus, and boy it’s been a great match. Our common love for pragmatism, efficiency, beach houses, and great food have served us well. I am thrilled to collaborate together on a series, Data Crunch that looks at the data behind long held pieces of popular conceptions. Can Peyton Manning throw in the cold? Does the color of tie really influence a Presidential debate? Visit us every Monday to see what we’ll look at next! Suggestions for future posts? Suggest above!
For full disclosure, our family bleeds blue and orange. And what a great year to be a Bronco’s fan. As the playoffs have approached, we have spent more than a moment fretting about — How will Peyton play in the cold? Will it be warm in New York? And I’m not the only one. If you ask Google, “Can Peyt-” it fills in the rest. Can Peyton Manning play in the cold.
Peyton Manning began his career in Indianapolis, playing at home in a dome at home for 13 years. He has been dogged by the story that he can’t play outside in the cold. So we set to looking at the data to see how closely we need to monitor the weather in East Rutherford, NY in the weeks leading up to Super Bowl XLVIII.
We took Manning’s win-loss record, measured statistics from every game, game day temperature and ran a logistic regression. [for those racking their brain back to college statistics — since the important outcome of each game will be one of two things, win or loss, you can use logistics regression to identify the important factors that determine whether it will be a win or a loss, and evaluate how much each factor influences the final game result. The resulting table shows you which factors you might want to look at more closely.]
Regression analysis indicated air temperature was a significant factor, so we took a closer look at game time temperature, results pictured below.
Peyton’s Win/Loss Percentages, broken down into three temperature ranges
(10°-40°, 40°-70° and 70°-100°)
Well, data doesn’t lie and this does not look good for January play.
*We decided to label anything colder than Denver “cold.” This makes us nervous, as we wouldn’t say Denver is warm, but we are using “cold” to mean colder than Manning typically plays in.
Let’s see how this can be explained.
Click below to see what may be behind this.
Two possible explanations for Peyton’s affected record in the cold:
Cold weather teams are are more likely to be legacy, experienced programs with better records and histories, and the increase in losses is just a reflection of the high caliber teams that are located in cold locations. Unfortunately, no. This scatter plot shows home teams sorted by citys’ average temperature in January (horizontal), and number Super Bowl appearances (vertical). Can you spot Green Bay, Steelers, New England, Miami and Cowboys? Looks pretty random to me. Coming from a cold weather city doesn’t appear to increase your shot at a Super Bowl Appearance.
*if a team plays in a dome, their adjusted temperature was recorded as 68°.
It’s a real drag to be in the Patriot’s division. Now we are getting somewhere. Here is Peyton’s win/loss record playing the Patriots in Foxborough, MA and at home. Yikes.
The fact Peyton struggles so dramatically in Foxborough, MA deeply affects his cold weather record. Since he played in a dome for so many years, he has relatively few outside “cold” games to evaluate. (he has played in only 24 “cold” games out of 260 career games). Almost half of his cold games, ten out of the 24 cold games — 42%– have been in New England. So is it the cold, or the Patriots he struggles against, hard to tell.
So what does that mean? I don’t know! Other than he has a rough time playing the Patriots on the road, yes. Are we very happy to have home field advantage should we play New England in a few weeks, oh yes. Pleased to see the forecast is a balmy 50° for Denver this weekend? For sure.
Official summary. He can do it. Sometimes. But pray for warmth, and the Colts to take care of New England on Saturday.
What do you think? Weigh in!
huge thanks to the kids helping with data entry, Jared performing analysis using JMP(R) and creating the graphics
data gathered from nfl.com, almanac.com, and SI for Kids First and 10, greatest book ever purchased for boys of all ages.