John Dorsey's Cleveland Browns making use of analytics after all - 03/22/18 06:47 PM
John Dorsey's Cleveland Browns making use of analytics after all
By Cynthia Frelund
NFL Network Analytics Expert
http://www.nfl.com/news/story/0ap3000000...ytics-after-all
The Cleveland Browns' new general manager, John Dorsey, might be teaching us more about analytics than he planned.
A proud traditionalist and self-avowed "football guy," Dorsey inherited 12 picks in the 2018 NFL Draft -- a parting gift from fired executive VP of football operations Sashi Brown, whose tenure included the Browns' controversial hiring of renowned analytics-minded former baseball executive Paul DePodesta -- when he took the job at the end of last season. Earlier this month, Dorsey relinquished a few of the mid- and late-round picks from the stockpile accumulated under Brown in order to acquire veterans who will presumably be immediate on-field contributors in 2018.
Here's where analytics -- real analytics -- come into play. Contrary to popular opinion, analytics are not about nerds trying to ruin sports. They're not a means for one person (or group) working in a silo to prescribe outcomes for others. In the NFL specifically, analytics are not defined by hoarding draft picks. The best definition of analytics I have found is this: They're about the practice of creating and refining processes to make decisions of the highest possible quality using historical, contextual and situational data.
Context and situation are especially key in the NFL, magnified by the relatively small number of games and strict spending floors and ceilings. On the field, these are things like: how close the game actually was (did the team lose on a missed field goal or by 20 points?); if it was, say, Deshaun Watson's first start or fifth; and -- this is a biggie -- if there were injured or missing players. One off-the-field example: Has this free agent played in a system like ours before, making for a potentially shorter learning curve?
In short, take lots of data, organize it so everyone who makes decisions can meaningfully use it, then work together to create the most overall value. Value can include not over-drafting a player or acquiring additional draft picks, but there is no way to define NFL value without including on-field results. And the search for value is pointless when the coaching staff and front office aren't on the same page.
Using both Hue Jackson's and Dorsey's past league experiences as reference, the number and speed of trades and free-agent acquisitions this offseason suggests they are executing a unified plan. Here's how the Browns' moves show us a cohesive front-office and coaching definition of value, which starts to reveal Cleveland's potential 2018 strategy on the field ...
Snag some proven playmakers in high-need areas
Say what you will about Tyrod Taylor, but there's absolutely no question the Browns have upgraded at the all-important QB1 spot.
With seven years of pro film on Taylor -- including three full starting campaigns -- the quarterback's production and style are far more known quantities than those of the other veteran options in his price range. The Browns gave up a third-round pick and will spend $16 million this season for a signal-caller who posted the NFL's lowest interception rate last year (1.0 percent; he threw just four picks) and logged just 21 total turnovers in the past three years (the fewest in the NFL among quarterbacks with at least six starts in each season). Last season, Browns rookie DeShone Kizer, who started 15 games, threw a league-high 22 interceptions.
Taylor has also averaged 5.6 yards per rush since becoming a starter in 2015, with no fewer than 427 rushing yards in each season over that span. Not to mention, 14 additional scores on the ground. Digging a little deeper shows Taylor ranked in the top 10 among all QBs in rushing average on first and second down during those seasons. This is important, as a team is more likely to call designed quarterback runs on those downs. Defenses have to respect Taylor's ability to extend plays, as he threw eight touchdown passes (against three picks) last season when holding the ball for 2.6 or more seconds. Under the same circumstances, Kizer threw two TD passes and 12 interceptions in 2017.
While Taylor hasn't cracked the top 26 in terms of passing attempts per game in any of the past three years, his results in critical throwing situations are strong positive indicators. Taylor ranked in the top six last season in both completion percentage and passer rating when under pressure. And when he was forced to scramble and make a play on third down, his accompanying completion percentage led the NFL.
"Polarizing" is a word often used to describe Tyrod Taylor, because he's taken the third-most sacks (124) over the past three seasons combined, and -- most importantly -- he is inconsistent in the red zone. Taylor threw half of his four INTs last season in the red zone, while ranking 25th out of 31 qualifying quarterbacks in yards per attempt (2.9) within this crucial area of the field.
Using the quarterback rubric I created by enlisting 22 NFL-level coaches to help define performance clusters (there are only five: elite, above average, average, below average and well below average), Taylor's production in each of the past three years has rated between QB12 and QB16. Not a bad return for the 65th overall pick, which would have been the Browns' sixth selection in next month's draft. They still have five picks in the top 64, with four occurring in the top 35. The value this year's Browns receive from Taylor projects to far exceed a single third-round pick.
The opportunity for Taylor in 2018 was sweetened through the Browns' acquisition of Jarvis Landry from Miami for a pair of Day 3 picks. The 25-year-old wideout caught a whopping 81 percent of his red-zone targets last season, resulting in nine touchdowns (most among WRs in this area). He also had only one red-zone drop.
Let's look over Landry's four years of overall NFL experience, with an additional focus on the past two. Landry has ranked in the top four in yards generated from the slot and in the top five in broken tackles over the past two seasons. Last year, he had 60 receptions that earned first downs, which is more than three times the number any single Browns wide receiver earned. For a team that ranked 31st in time of possession last season, moving the chains is an attribute Cleveland must improve upon this season.
Further, by trading for the franchise-tagged Landry two weeks ago, the Browns avoided uncertainty in the WR free agency market by merely giving up a fourth-round pick in 2018 and a seventh-rounder in 2019. They didn't burn any of their high-value selections. And this is key when it comes to Cleveland's most vital task of this entire offseason ...
Formulate a long-term plan at the game's most important position
Taylor and Landry not only provide the potential to win this year, but also create the space for Cleveland to optimize its draft/free agency strategy. Given that the Browns possess pick Nos. 1, 4, 33 and 35, they are in the unique position of being able to draft an instant-impact player like Penn State running back Saquon Barkley AND a highly touted QB prospect to develop behind Taylor.
Based on the on-field strengths Taylor brings to the team, a logical fit at quarterback -- in terms of someone Hue Jackson could cultivate for maximum upside potential -- could be reigning Heisman Trophy winner Baker Mayfield. The Oklahoma product checks in at 6 feet and 5/8 inches tall (or 6005, as scouts term it) and 215 pounds. That's approximately the same size as Taylor, who is 6 feet and 3/4 inches tall (or 6006) and 215 pounds. Mayfield also averaged 6.9 rushing attempts per game last year (averaging 3.2 yards per attempt) and had the best passer rating under pressure in college football (111.6, according to Pro Football Focus). The strategy Cleveland implements for Taylor has natural extensions to Mayfield -- based on their similarities in stature and certain attributes -- and would give the Browns the time to develop Mayfield from a college-spread QB to an NFL-caliber signal-caller.
The analytically defined value here lies in past data showing that developing a quarterback in a system (and with a playbook) optimized for his best attributes leads to greater NFL success as measured by wins (SEE: Jimmy Garoppolo). Broadly speaking, the data works out like this: In critical areas that are correlated with greater win percentages -- limiting turnovers, producing on third down and in the red zone -- QBs who get to add complexity to the playbook as their experience level increases win more often, while QBs who have to be more productive immediately do not win as often. Yes, the teams that immediately start rookie quarterbacks are often "worse" overall, but fit and play calling impact execution and results (SEE: Dak Prescott, Jared Goff). It also just makes sense.
Don't stop wheeling and dealing
With structure around the quarterback position, the Browns were able to extract value from a QB who didn't fit their plans by trading away Kizer in order to acquire corner Damarious Randall. Additionally, parting ways with nose tackle Danny Shelton (and a 2018 fifth-round pick) for a third-round pick in 2019 shows that the Browns' current strategy is not all about "spending" draft equity, but about a shared definition of current and future value.
Between signing potential impact free agents like RB Carlos Hyde and DE Chris Smith, tendering wide receiver Josh Gordon and creating a plan to soften the blow of 10-time Pro Bowl LT Joe Thomas' retirement (Cleveland signed tackles Chris Hubbard and Donald Stephenson), the Browns are also using their league-leading cap space to round out their roster. An additional consequence of addressing so many positions with free agents is the creation of uncertainty about whom they will be targeting with the first and fourth overall picks. This generates value if other teams believe Cleveland is targeting a player they wish to draft, allowing the Browns to be in a position to trade down and earn even more draft picks.
Given the pick trades we've already seen -- notably, the Jets jumping up to No. 3 and the Bills vaulting to 12 with the potential to move up further -- Cleveland's decisions will shape the entire draft. It's a safe bet we're not done re-shuffling the Round 1 deck.
Win today ... and tomorrow
I would never say Dorsey and Jackson aren't "football guys." That said, I believe they're using the resources amassed by past "analytics" -- in terms of both draft equity and cap space -- in a way that improves their potential to win now and, so far, follows some of the most important data-driven insights that have led to sustained success for teams in this league.
Collaboration between the front office and coaching staff -- with all this capital -- could be a one-year fluke. Or it could be a sign that analytics have indeed changed things in Berea ...
By Cynthia Frelund
NFL Network Analytics Expert
http://www.nfl.com/news/story/0ap3000000...ytics-after-all
The Cleveland Browns' new general manager, John Dorsey, might be teaching us more about analytics than he planned.
A proud traditionalist and self-avowed "football guy," Dorsey inherited 12 picks in the 2018 NFL Draft -- a parting gift from fired executive VP of football operations Sashi Brown, whose tenure included the Browns' controversial hiring of renowned analytics-minded former baseball executive Paul DePodesta -- when he took the job at the end of last season. Earlier this month, Dorsey relinquished a few of the mid- and late-round picks from the stockpile accumulated under Brown in order to acquire veterans who will presumably be immediate on-field contributors in 2018.
Here's where analytics -- real analytics -- come into play. Contrary to popular opinion, analytics are not about nerds trying to ruin sports. They're not a means for one person (or group) working in a silo to prescribe outcomes for others. In the NFL specifically, analytics are not defined by hoarding draft picks. The best definition of analytics I have found is this: They're about the practice of creating and refining processes to make decisions of the highest possible quality using historical, contextual and situational data.
Context and situation are especially key in the NFL, magnified by the relatively small number of games and strict spending floors and ceilings. On the field, these are things like: how close the game actually was (did the team lose on a missed field goal or by 20 points?); if it was, say, Deshaun Watson's first start or fifth; and -- this is a biggie -- if there were injured or missing players. One off-the-field example: Has this free agent played in a system like ours before, making for a potentially shorter learning curve?
In short, take lots of data, organize it so everyone who makes decisions can meaningfully use it, then work together to create the most overall value. Value can include not over-drafting a player or acquiring additional draft picks, but there is no way to define NFL value without including on-field results. And the search for value is pointless when the coaching staff and front office aren't on the same page.
Using both Hue Jackson's and Dorsey's past league experiences as reference, the number and speed of trades and free-agent acquisitions this offseason suggests they are executing a unified plan. Here's how the Browns' moves show us a cohesive front-office and coaching definition of value, which starts to reveal Cleveland's potential 2018 strategy on the field ...
Snag some proven playmakers in high-need areas
Say what you will about Tyrod Taylor, but there's absolutely no question the Browns have upgraded at the all-important QB1 spot.
With seven years of pro film on Taylor -- including three full starting campaigns -- the quarterback's production and style are far more known quantities than those of the other veteran options in his price range. The Browns gave up a third-round pick and will spend $16 million this season for a signal-caller who posted the NFL's lowest interception rate last year (1.0 percent; he threw just four picks) and logged just 21 total turnovers in the past three years (the fewest in the NFL among quarterbacks with at least six starts in each season). Last season, Browns rookie DeShone Kizer, who started 15 games, threw a league-high 22 interceptions.
Taylor has also averaged 5.6 yards per rush since becoming a starter in 2015, with no fewer than 427 rushing yards in each season over that span. Not to mention, 14 additional scores on the ground. Digging a little deeper shows Taylor ranked in the top 10 among all QBs in rushing average on first and second down during those seasons. This is important, as a team is more likely to call designed quarterback runs on those downs. Defenses have to respect Taylor's ability to extend plays, as he threw eight touchdown passes (against three picks) last season when holding the ball for 2.6 or more seconds. Under the same circumstances, Kizer threw two TD passes and 12 interceptions in 2017.
While Taylor hasn't cracked the top 26 in terms of passing attempts per game in any of the past three years, his results in critical throwing situations are strong positive indicators. Taylor ranked in the top six last season in both completion percentage and passer rating when under pressure. And when he was forced to scramble and make a play on third down, his accompanying completion percentage led the NFL.
"Polarizing" is a word often used to describe Tyrod Taylor, because he's taken the third-most sacks (124) over the past three seasons combined, and -- most importantly -- he is inconsistent in the red zone. Taylor threw half of his four INTs last season in the red zone, while ranking 25th out of 31 qualifying quarterbacks in yards per attempt (2.9) within this crucial area of the field.
Using the quarterback rubric I created by enlisting 22 NFL-level coaches to help define performance clusters (there are only five: elite, above average, average, below average and well below average), Taylor's production in each of the past three years has rated between QB12 and QB16. Not a bad return for the 65th overall pick, which would have been the Browns' sixth selection in next month's draft. They still have five picks in the top 64, with four occurring in the top 35. The value this year's Browns receive from Taylor projects to far exceed a single third-round pick.
The opportunity for Taylor in 2018 was sweetened through the Browns' acquisition of Jarvis Landry from Miami for a pair of Day 3 picks. The 25-year-old wideout caught a whopping 81 percent of his red-zone targets last season, resulting in nine touchdowns (most among WRs in this area). He also had only one red-zone drop.
Let's look over Landry's four years of overall NFL experience, with an additional focus on the past two. Landry has ranked in the top four in yards generated from the slot and in the top five in broken tackles over the past two seasons. Last year, he had 60 receptions that earned first downs, which is more than three times the number any single Browns wide receiver earned. For a team that ranked 31st in time of possession last season, moving the chains is an attribute Cleveland must improve upon this season.
Further, by trading for the franchise-tagged Landry two weeks ago, the Browns avoided uncertainty in the WR free agency market by merely giving up a fourth-round pick in 2018 and a seventh-rounder in 2019. They didn't burn any of their high-value selections. And this is key when it comes to Cleveland's most vital task of this entire offseason ...
Formulate a long-term plan at the game's most important position
Taylor and Landry not only provide the potential to win this year, but also create the space for Cleveland to optimize its draft/free agency strategy. Given that the Browns possess pick Nos. 1, 4, 33 and 35, they are in the unique position of being able to draft an instant-impact player like Penn State running back Saquon Barkley AND a highly touted QB prospect to develop behind Taylor.
Based on the on-field strengths Taylor brings to the team, a logical fit at quarterback -- in terms of someone Hue Jackson could cultivate for maximum upside potential -- could be reigning Heisman Trophy winner Baker Mayfield. The Oklahoma product checks in at 6 feet and 5/8 inches tall (or 6005, as scouts term it) and 215 pounds. That's approximately the same size as Taylor, who is 6 feet and 3/4 inches tall (or 6006) and 215 pounds. Mayfield also averaged 6.9 rushing attempts per game last year (averaging 3.2 yards per attempt) and had the best passer rating under pressure in college football (111.6, according to Pro Football Focus). The strategy Cleveland implements for Taylor has natural extensions to Mayfield -- based on their similarities in stature and certain attributes -- and would give the Browns the time to develop Mayfield from a college-spread QB to an NFL-caliber signal-caller.
The analytically defined value here lies in past data showing that developing a quarterback in a system (and with a playbook) optimized for his best attributes leads to greater NFL success as measured by wins (SEE: Jimmy Garoppolo). Broadly speaking, the data works out like this: In critical areas that are correlated with greater win percentages -- limiting turnovers, producing on third down and in the red zone -- QBs who get to add complexity to the playbook as their experience level increases win more often, while QBs who have to be more productive immediately do not win as often. Yes, the teams that immediately start rookie quarterbacks are often "worse" overall, but fit and play calling impact execution and results (SEE: Dak Prescott, Jared Goff). It also just makes sense.
Don't stop wheeling and dealing
With structure around the quarterback position, the Browns were able to extract value from a QB who didn't fit their plans by trading away Kizer in order to acquire corner Damarious Randall. Additionally, parting ways with nose tackle Danny Shelton (and a 2018 fifth-round pick) for a third-round pick in 2019 shows that the Browns' current strategy is not all about "spending" draft equity, but about a shared definition of current and future value.
Between signing potential impact free agents like RB Carlos Hyde and DE Chris Smith, tendering wide receiver Josh Gordon and creating a plan to soften the blow of 10-time Pro Bowl LT Joe Thomas' retirement (Cleveland signed tackles Chris Hubbard and Donald Stephenson), the Browns are also using their league-leading cap space to round out their roster. An additional consequence of addressing so many positions with free agents is the creation of uncertainty about whom they will be targeting with the first and fourth overall picks. This generates value if other teams believe Cleveland is targeting a player they wish to draft, allowing the Browns to be in a position to trade down and earn even more draft picks.
Given the pick trades we've already seen -- notably, the Jets jumping up to No. 3 and the Bills vaulting to 12 with the potential to move up further -- Cleveland's decisions will shape the entire draft. It's a safe bet we're not done re-shuffling the Round 1 deck.
Win today ... and tomorrow
I would never say Dorsey and Jackson aren't "football guys." That said, I believe they're using the resources amassed by past "analytics" -- in terms of both draft equity and cap space -- in a way that improves their potential to win now and, so far, follows some of the most important data-driven insights that have led to sustained success for teams in this league.
Collaboration between the front office and coaching staff -- with all this capital -- could be a one-year fluke. Or it could be a sign that analytics have indeed changed things in Berea ...