To listen, click on the right pointing triangle. A written version is available below.
San Francisco will face Kansas City in the Super Bowl, a game with many intriguing story lines.
Kansas City coach Andy Reid has won almost 62% of games in a 21 year NFL career, but he doesn’t have a championship ring. Can he get it done?
San Francisco coach Kyle Shanahan is the new kid on his block as a third year head coach. Can he join his dad Mike as a Super Bowl winning coach?
In the Super Bowl preview, I’ll break down match ups and identify the important factor that could swing this game. I’ll end with my best prediction based on analytics usually saved for members of The Power Rank.
San Francisco offense vs Kansas City defense
How will San Francisco’s offense approach this game?
To get an idea, let’s look at their play calling on 1st down in the first half. Late in the game, the scoreboard dictates play calling (teams facing a deficit throw the ball to score points, teams with the lead run the ball take time off the clock). Without these factors early in the game, we get a sense for the tendencies for an offense.
On 1st down in the first half, San Francisco runs on 54% of plays, 10th most in the NFL. The Niners showed this tendency in the NFC conference championship game. They ran the ball with great success against Green Bay in a convincing win.
Despite their tendencies, San Francisco is not that efficient at running the ball. To measure this, I’ll use success rate. For an offense, success means gaining the following fraction of necessary yards for a 1st down:
- 50% on 1st down
- 70% on 2nd down
- all on 3rd, 4th down
When I take success rate on run plays and adjust for opponent, San Francisco ranks 20th in the NFL. One factor in this lack of efficiency could be the tendency to run on 1st down in the first half. In this situation, NFL defenses usually prepare to stop the run.
However, San Francisco could have success with the run in the Super Bowl. Kansas City has the worst rush defense in the NFL in my adjusted success rate. To explain this lack of success, let’s look at snap count data for players from Pro Football Focus (PFF).
Kansas city plays a 4-3 defense, which implies four defensive linemen and three linebackers on the typical play. Two of these linemen are interior linemen, the large players that clog up the middle so teams can’t run the ball.
According to the snap count data on Pro Football Focus, interior defensive linemen take up 1.7 positions on average out of 11 players on Kansas City’s defense. This is less than the two interior linemen expected for a 4-3 defense.
This suggests that Kansas City’s does not focus as much on stopping the run. They’d rather stop the pass, in general a good strategy. During the 2019 NFL season, teams gained 6.3 yards per pass attempt compared to 4.3 yards per carry.
Instead of interior linemen, Kansas City puts an extra player in the secondary. The Chiefs average 5.0 players in the secondary, more than the 4 players listed as starters.
This extra player in the secondary helps their pass defense. In my adjusted success rate on passing plays, Kansas City ranks 12th in the NFL. In addition, they excel at preventing big plays. To see this, I look at yards per pass attempt, which includes negative plays for sacks. After opponent adjustments, Kansas City ranks 3rd in the NFL.
Kansas City’s success in preventing big plays comes from the safety position. Tyrann Mathieu and Juan Thornhill have the two best coverage grades according to the PFF scouting data. They have significantly higher grades than Kansas City’s top 3 cornerbacks.
Note: Thornhill is injured and will not play in the Super Bowl. Daniel Sorensen will take his place with a worse coverage grade, although still higher than the 3 cornerbacks.
Despite these strong numbers, Kansas City will face an excellent San Francisco pass offense. Based on adjusted success rate, San Francisco has the 6th best pass offense.
Kansas City offense vs San Francisco defense
In contrast to San Francisco, Kansas City likes to throw the ball. On 1st down in the first half, they throw on 65% of plays, by far the highest rate in the NFL.
Andy Reid runs a modern NFL offense. On 1st down in the first half, defenses usually prepare to stop the run. Instead of running, Kansas City has opted to throw with QB Patrick Mahomes.
When Kansas City does throw the ball against San Francisco’s defense, it is strength on strength. By my adjusted success rate, Kansas City is 5th in the NFL. They rank even better in my adjusted yards per pass attempt (1st).
Despite the strength of these numbers, Kansas City could have been even better. QB Patrick Mahomes missed two games with an injury, and top receivers Tyreek Hill and Sammy Watkins also missed games this year.
Kansas City will face an excellent San Francisco pass defense. By either success rate or yards per pass attempt (both adjusted for schedule), the Niners are the best in the NFL this year.
Based on the snap count data, San Francisco is even more aggressive about stopping the pass than Kansas City. The Niners also play a 4-3 defense. But the PFF data reveals they play 1.5 interior defensive linemen on average, even less than the 1.7 of Kansas City.
While Kansas City tends to put an extra player in the secondary, San Francisco puts another edge rusher on the field. They average 2.5 edge rushers (usually defensive ends, and does not include linebackers). For comparison, Kansas City averages 2.24 edge rushers.
These extra edge rushers have worked for San Francisco as they excel at getting pressure on the QB. When I look sack rate (sacks divided by pass attempts) adjusted for schedule, the Niners rank second in the NFL. Kansas City will need to handle this pass rush to throw the ball deep.
San Francisco is not as good at run defense. By my adjusted success rate, they rank 20th in the NFL. While Andy Reid tends to throw on offense, I’ll be interested to see whether he chooses to run more against San Francisco based on their personnel choices.
In addition, Kansas City has been good at running the ball. By my adjusted success rate, they rank 6th in the NFL.
Will either team have an edge in turnovers? In this section, I’ll focus on interceptions.
To get a preliminary estimate on interceptions, let’s look at the career interception rate for the starting QBs. Based on this data, it looks like Kansas City will have the edge.
Kansas City’s Patrick Mahomes has a 1.5% career pick rate on 1,241 pass attempts, which includes playoff games. For perspective, the NFL average was 2.3% during the 2019 regular season. San Francisco’s Jimmy Garappolo has 2.5% interception rate on 837 pass attempts through this season’s two playoff games.
It’s difficult to make a definitive statement about interceptions due to the small sample size of attempts with these young QBs. It would be different if this game had veterans like Drew Brees and Tom Brady.
However, there’s another factor that suggests that Kansas City has an edge in interceptions: head coach Andy Reid. He got his first NFL head coaching job in 1999 with Philadelphia, and he came over to Kansas city in 2013. In 21 seasons as a head coach, his offenses have had an above NFL average pick rate only three times.
These 3 seasons with an above average pick rate all occurred in Philadelphia. During his 7 years in Kansas city, his teams have had below average NFL interception rates each season.
There is luck associated with these numbers. The Chiefs came closest to the NFL average in 2018, the first year Patrick Mahomes started at QB. He threw 12 interceptions that season.
However, according to charting of Sports Info Solutions, defenders dropped 10 interceptions that year. If those defenders had held on to some of those interceptions, Kansas City would have been worse than NFL average in pick rate.
In general, Andy Reid has always had offenses that protect the football in the passing game. I don’t know how he coaches his players to achieve this goal, but the data suggests a long term trend.
Sports books usually have a Super Bowl prop bet for how many interceptions each quarterback will throw. This number is set 0.5 interceptions. The under implies the quarterback will throw no picks.
Based on inception rate and projected number of attempts, you can work out the math for the chance a quarterback will not throw an interception. For Patrick Mahomes, I’ll use his 1.5% interception rate and assume that he throws ball 35 times, his average this season. This gives a 59% chance that he will not throw an interception, or that he goes under 0.5 interceptions.
For Jimmy Garappolo, I’ll use his 2.5% career interception rate. Since San Francisco tends to run more, he projects to throw fewer pass attempts. This season, he has averaged 30 pass attempts. This gives a 47% chance that he will throw no interceptions.
For members of my site, I put together a prediction based on data from the current season, market rankings based on the median closing spread and preseason expectations. These numbers favor Kansas City by 1.6 points.
I’m very happy with this prediction. Before I ran the numbers, I thought Kansas City should be the favorite but not more than by 2.5 points. It should be a fantastic game as long as both teams play up to their ability.
Nik Wassiliew says
I think all the analytical pontification is great…going back..looking at an entire body of seasonal work as well as trends. However,each coach and staff understands the other teams strengths,trends and weaknesses..true? It will be the team..that best functions…OUTSIDE those Norms..that will win. I truly respect analytics,though as AI has yet the ability for Intuitive REASONING..Analytics cannot yet account for Coaching CREATIVITY within a Single game. I know you guys abhor short term trends,even though I feel they give you a better ‘current’ look for speculation..lolol for an “thats All folks” Last game. It will be a Great First Half..though..lol thank you for your time~nik
jack powell says
high scoring or not?
Dr. Ed Feng says
I think there are a range of outcomes. Kansas City is explosive, and their defense can be bad. So there’s definitely the possibility of a high scoring game.
AR Massey says
What is significant in basketball of the second number in your data, (points per possession) in basketball? Is high negatives, like -13.5, a positive for handicapping?
AR Massey says
What the significance of a high Negative number as the points per possession in your basketball data?
Dr. Ed Feng says
On what page are you referring to?