Category Archives: machine learning

Analyzing the Draft no Draft of NFL Quarterbacks on 2018 Depth Chart

Analyzing the Draft No Draft of NFL Quarterbacks 2018 Depth Chart. Below you will see a YES or NO next to most of the players names. If there is a YES, our machine learning NFL quarterback system would have drafted the QB – if there is a NO, our system would not have drafted the QB. Remember, our machine learning predictions of NFL QB success is based solely on college statistics. We are in the process of running analysis on the ones missing a yes or no and will update shortly. Learn more about our machine learning quarterback NFL draft analysis system which is based on college stats of Super Bowl winner quarterbacks and college stats of high draft pick busts. Learn more about our Machine Learning Franchise Quarterback draft system

We affectionately call our system “Beast or Bust”

Team Quarterbacks

ARI
Sam Bradford YES
Josh Rosen NO
Mike Glennon YES

ATL
Matt Ryan YES
Matt Schaub YES

BAL
Joe Flacco YES
Lamar Jackson NO
Ryan Mallett NO
Robert Griffin III YES

BUF
A.J. McCarron YES
Josh Allen NO
Nathan Peterman YES

CAR
Cam Newton YES
Garrett Gilbert NO

CHI
Mitchell Trubisky YES
Chase Daniel YES
Tyler Bray NO

CIN
Andy Dalton YES
Matt Barkley

CLE
Tyrod Taylor NO
Baker Mayfield YES
Drew Stanton

DAL
Dak Prescott YES
Cooper Rush
Mike White

DEN
Case Keenum YES
Paxton Lynch YES

DET
Matthew Stafford NO
Matt Cassel
Jake Rudock YES

GB
Aaron Rodgers NO* only 2 years college data
DeShone Kizer YES
Brett Hundley

HOU
Deshaun Watson YES
Brandon Weeden YES
Joe Webb

IND
Andrew Luck YES
Jacoby Brissett NO
Brad Kaaya YES

JAC
Blake Bortles YES
Cody Kessler YES
Tanner Lee NO

KC
Patrick Mahomes YES
Chad Henne

LAC
Philip Rivers YES
Cardale Jones YES
Geno Smith

LAR
Jared Goff YES
Sean Mannion YES

MIA
Ryan Tannehill NO
Brock Osweiler NO
Bryce Petty YES

MIN
Kirk Cousins YES
Trevor Siemian NO

NE
Tom Brady YES
Brian Hoyer NO
Danny Etling

NO
Drew Brees YES
Tom Savage

NYG
Eli Manning YES
Davis Webb NO
Kyle Lauletta

NYJ
Josh McCown NO
Sam Darnold YES
Teddy Bridgewater YES

OAK
Derek Carr YES
Conner Cook NO
EJ Manuel

PHI
Carson Wentz YES
Nick Foles YES
Nate Sudfield NO

PIT
Ben Roethlisberger YES
Mason Rudolph YES
Landry Jones
Joshua Dobbs NO

SEA
Russell Wilson YES
Austin Davis

SF
Jimmy Garoppolo YES
C.J. Beathard NO

TB
Jameis Winston (sus) YES
Ryan Fitzpatrick NO
Ryan Griffin NO

TEN
Marcus Mariota YES
Blaine Gabbert NO
Luke Falk YES

WAS
Alex Smith YES
Colt McCoy
Kevin Hogan YES

Visualizing College Statistics of Super Bowl Winning QBs vs. QB Draft Busts

Understanding how college statistics of quarterbacks translate to NFL success, is very important to understanding the draft day value of quarterbacks. The first chart below visualizes the COLLEGE completion percentages of Super Bowl winning quarterbacks like P. Manning, Aikman, Rodgers, E Manning, Roethlisberger, Brees, Elway and Brady. They are represented as the BLUE histogram. The orange histogram represents the COLLEGE completion percentages of famous high round draft picks QB busts like Young, Harrington, Quinn, Russel, Ponder, Carter and Plummer.

Blue = QB SB winners
Orange = QB Busts

Looking at the histograms, it is apparent that the Super Bowl winners’ college completion percentage is clearly shifted to the right which indicates that this is a good indicator that college completion pct is important and should be above 60%.

In comparison, let’s look at height as a predictor – the histograms below represent the height of the same QBs and it is apparent for the histogram with these Abs height was not predictive like college completion percentage.

Now lets look at the ratio of interceptions to touchdowns – clearly the sweet spot is around 40% and trying to stay less than 57% would be a good goal.

Interceptions / Touchdowns

The final histogram below compares the rush attempts per college game. From this chart, Super Bowl winning quarterbacks tend not to be runners whether by choice or by chase.

Rush Attempts per Game

We think the above charts provide great insight into how college statistics coupled with machine learning can help predict NFL success on draft day.

NFL Super Bowl Winning Quarterbacks

A complete list of Super Bowl winning quarterbacks

Super Bowl 1. Bart Starr (MVP), 2 TDs
Super Bowl 2. Bart Starr (MVP), 1 TD
Super Bowl 3. Joe Namath (MVP), 0 TDs
Super Bowl 4. Len Dawson (MVP), 1 TD
Super Bowl 5. John Unitas (Chuck Howley), 1 TD
Super Bowl 6. Roger Staubach (MVP), 2 TDs
Super Bowl 7. Bob Griese (Jake Scott), 1 TD
Super Bowl 8. Bob Griese (Larry Csonka), 0 TDs
Super Bowl 9. Terry Bradshaw (Franco Harris), 1 TD
Super Bowl 10. Terry Bradshaw (Lynn Swann), 2 TDs
Super Bowl 11. Ken Stabler (Fred Biletnikoff), 1 TD
Super Bowl 12. Roger Staubach (Harvey Martin & Randy White), 1 TDs
Super Bowl 13. Terry Bradshaw (MVP), 4 TDs
Super Bowl 14. Terry Bradshaw (MVP), 2 TDs
Super Bowl 15. Jim Plunkett (MVP), 3 TDs
Super Bowl 16. Joe Montana (MVP), 1 TD
Super Bowl 17. Joe Theismann (John Riggins), 2 TDs,
Super Bowl 18. Jim Plunkett (Marcus Allen), 1 TD
Super Bowl 19. Joe Montana (MVP), 3 TDs
Super Bowl 20. Jim McMahon (Richard Dent), 0 TDs
Super Bowl 21. Phil Simms (MVP), 3 TDs
Super Bowl 22. Doug Williams (MVP), 4 TDs
Super Bowl 23. Joe Montana (Jerry Rice), 2 TDs
Super Bowl 24. Joe Montana (MVP), 5 TDs
Super Bowl 25. Jeff Hostetler (Ottis Anderson), 1 TD
Super Bowl 26. Mark Rypien (MVP), 2 TDs
Super Bowl 27. Troy Aikman (MVP), 4 TDs
Super Bowl 28. Troy Aikman (Emmitt Smith), O TDs
Super Bowl 29. Steve Young (MVP), 6 TDs
Super Bowl 30. Troy Aikman (Larry Brown), 1 TD
Super Bowl 31. Brett Favre (Desmond Howard), 2 TDs
Super Bowl 32. John Elway (Terrell Davis), 0 TDs
Super Bowl 33. John Elway (MVP), 1 TD
Super Bowl 34. Kurt Warner (MVP), 2 TDs
Super Bowl 35. Trent Dilfer (Ray Lewis), 1 TD
Super Bowl 36. Tom Brady (MVP), 1 TD
Super Bowl 37. Brad Johnson (Dexter Jackson), 2 TDs
Super Bowl 38. Tom Brady (MVP), 3 TDs
Super Bowl 39. Tom Brady (Deion Branch), 2 TDs
Super Bowl 40. Ben Roethlisberger (Hines Ward), 0 TDs
Super Bowl 41. Peyton Manning (MVP), 1 TD
Super Bowl 42. Eli Manning (MVP), 2 TDs
Super Bowl 43: Ben Roethlisberger (Santonio Holmes), 1 TD
Super Bowl 44: Drew Brees (MVP), 2 TDs
Super Bowl 45: Aaron Rogers (MVP), 3TDs
Super Bowl 46: Eli Manning (MVP), 1 TD
Super Bowl 47: Joe Flacco (MVP), 3TDs
Super Bowl 48: Russell Wilson (Malcolm Smith), 2TDs
Super Bowl 49: Tom Brady (MVP), 4TDs
Super Bowl 50: Peyton Manning (Von Miller), 0TDs
Super Bowl 51: Tom Brady (MVP), 2TDs
Super Bowl 52: Nick Foles (MVP)

Predicting NFL Team Wins for 2018 using Machine Learning

Predicting NFL Team Wins for 2018 using Machine Learning

We have developed and applied a machine learning system to predict how many wins each NFL team will have for 2018-2019 season. We think the top 8 teams by wins this year will be:

Philadelphia with 11 wins
Baltimore with 10 wins
Minnesota with 10 wins
Detroit with 10 wins
LA Rams with 10 wins
Arizona with 9 wins
Jacksonville with 9 wins
New England with 9 wins

Below is the entire list of machine learning predicted wins for each NFL team for the 2018 – 2019 NFL season using machine learning.

TEAM target
Philadelphia 11.0
Baltimore 10.0
Minnesota 10.0
Detroit 10.0
LA Rams 10.0
Arizona 9.0
Jacksonville 9.0
New England 9.0
New Orleans 9.0
Dallas 9.0
Pittsburgh 9.0
Seattle 9.0
Carolina 9.0
Atlanta 9.0
Oakland 8.0
San Francisco 8.0
Tennessee 8.0
LA Chargers 8.0
Kansas City 8.0
Houston 8.0
Green Bay 8.0
Denver 8.0
Cincinnati 8.0
Washington 8.0
Miami 7.0
NY Jets 7.0
Cleveland 7.0
Chicago 7.0
Tampa Bay 7.0
Indianapolis 6.0
Buffalo 6.0
NY Giants 5.0

Below is the Vegas over/under totals for team wins for 2018

NFL over-under win totals

New England Patriots, 11

Green Bay Packers, 10

Minnesota Vikings, 10

Philadelphia Eagles, 10

Pittsburgh Steelers, 10

Atlanta Falcons, 9

Carolina Panthers, 9

Jacksonville Jaguars, 9

L.A. Chargers, 9

L.A. Rams, 9

New Orleans Saints, 9

Baltimore Ravens, 8

Dallas Cowboys, 8

Detroit Lions, 8

Houston Texans, 8

Kansas City Chiefs, 8

Oakland Raiders, 8

San Francisco 49ers, 8

Seattle Seahawks, 8

Tennessee Titans, 8

Denver Broncos, 7

Washington, 7

Buffalo Bills, 6

Chicago Bears, 6

Cincinnati Bengals, 6

Indianapolis Colts, 6

Miami Dolphins, 6

N.Y. Giants, 6

N.Y. Jets, 6

Tampa Bay Buccaneers, 6

Arizona Cardinals, 5

Cleveland Browns, 5

Machine learning predicts NFL success for quarterback Jimmy Garoppolo

The Niners announced Thursday they have signed Jimmy Garoppolo to a five-year contract extension through the 2022 season. Sources confirmed to ESPN the deal is for $137.5 million, including $86.4 million guaranteed in the first three years, making it the largest three-year total in NFL history.

We have analyzed Garoppolo’s college stats with out machine learning NFL qb draft prediction system and he would have been a “YES” draft with a 97% probability.

He went 5-0 after being traded to the 49ers last season. However, he through 7 TD and 5 INTS which will not get it done.

We will watch with great interest this year to see how he progresses.

Learn more about our machine learning quarterback prediction system

2018 NFL draft qb machine learning predictions

2018 NFL qb draft prospects Kiper vs. Machine Learning

Below is a comparison of 2018 NFL QB prospects Mel Kiper VS. a machine learning system. We watch no video and only use machine learning and college stats to project if a player should be drafted or not. More specifically our Machine Learning NFL quarterback system predicts if a quarterback will be a franchise type quarterback in the NFL based on the players’ college stats and is trained with superstar and bust draft pick stats to learn how to predict when to draft and not. The list simply reflects a yes or no for a draft or not and the probability that they belong on that list.

Here is our list and then below is Mel Kiper’s list for the QBs for the 2018 draft.

QB data read successfully!

name predict prob
2 Mayfield-Baker [YES] 0.999965
3 Falk-Luke [YES] 0.999244
4 Rudolph-Mason [YES] 0.993547
1 Darnold-Sam [YES] 0.954596

name predict prob
6 Lee-Tanner [NO] 0.996788
2 Allen-Josh [NO] 0.932838
3 Jackson-Lamar [NO] 0.836334
1 Rosen-Josh [NO] 0.632428
4 Litton-Chase [NO] 0.548834
5 Benke-Kurt [NO] 0.541672

Interesting Todd McShay and Mel Kiper are really high on Josh Rosen and also think Josh Allen should go maybe in the first round, but our system would not recommend them to be drafted.

Here is Mel Kiper’s top 10 QBs…
Quarterbacks

1. Josh Rosen, UCLA
2. Sam Darnold, USC
3. Josh Allen, Wyoming
4. Baker Mayfield, Oklahoma
5. Drew Lock, Missouri
6. Mason Rudolph, Oklahoma State
7. Lamar Jackson, Louisville
8. Mike White, Western Kentucky
9. Luke Falk, Washington State
10. Kurt Benkert, Virginia

Of Kiper’s top 3 our system would not draft 2 and of his top ten, our system would not draft 4 of the 10 – we have not analyzed Lock or White so 4 of the 8 meaning 50% of his top 8 picks would be avoided by our system. see Kipers 2018 QB draft prospects list

machine learning nfl draft qb prediction system

Machine Learning predicts Nick Foles Super Bowl success

Our machine learning system predicted that Nick Foles would be a franchise starting quarterback and would likely win a Super Bowl and he did.

THIS PREDICTION WAS MADE USING MACHINE LEARNING AND ONLY NICK FOLES COLLEGE STATS.

See our machine learning results below of many quarterbacks you may know – these predictions are based only on the their college statistics on the day they were drafted.

Nick Foles was almost an 80% chance to be star quarterback based on our machine learning system and his college stats.

Machine learning NFL QB draft analysis

Here is a link to read more about our machine learning NFL qb draft prediction system

2018 NFL draft quarterbacks analyzed with machine learning

We have analyzed the 2018 NFL draft quarterback prospects with our machine learning system. Our machine learning system predicts the probability that a draft pick will become a franchise Quarterback in the NFL. We have analyzed the last 20 years of Super Bowl winning quarterbacks and use this data applying machine learning to predict the probability of the 2018 draft quarterback prospects will be franchise quarterbacks and classifies them as a yes draft and a no draft and the probability they fit on that list. Machine Learning NFL quarterback system Here is the list…

QB data read successfully!

name predict prob
2 Mayfield-Baker [YES] 0.999965
3 Falk-Luke [YES] 0.999244
4 Rudolph-Mason [YES] 0.993547
1 Darnold-Sam [YES] 0.954596

name predict prob
6 Lee-Tanner [NO] 0.996788
2 Allen-Josh [NO] 0.932838
3 Jackson-Lamar [NO] 0.836334
1 Rosen-Josh [NO] 0.632428
4 Litton-Chase [NO] 0.548834
5 Benke-Kurt [NO] 0.541672

Interesting Todd McShay and Mel Kiper are really high on Josh Rosen and also think Josh Allen should go maybe in the first round, but our system would not recommend them to be drafted.

Here is Mel Kiper’s top 10 QBs…
Quarterbacks

1. Josh Rosen, UCLA
2. Sam Darnold, USC
3. Josh Allen, Wyoming
4. Baker Mayfield, Oklahoma
5. Drew Lock, Missouri
6. Mason Rudolph, Oklahoma State
7. Lamar Jackson, Louisville
8. Mike White, Western Kentucky
9. Luke Falk, Washington State
10. Kurt Benkert, Virginia

Of Kiper’s top 3 our system would not draft 2 and of his top ten, our system would not draft 4 of the 10 – we have not analyzed Lock or White so 4 of the 8 meaning 50% of his top 8 picks would be avoided by our system. see the list

2017 NFL Draft Eagles Best Value Quarterback machine learning

Our machine learning NFL franchise quarterback prediction system picked Jerod Evans as the #3 QB in the 2017 draft behind Mitch Trubisky and Deshaun Watson and just ahead of Patrick Mahommes. Despite this, Jerod Evans went undrafted. However, he was picked up last month by the Eagles. So now the Eagles have Wentz and Evans – both high “Yes Drafts” at Quarterback. It will be interesting to watch Evans.

Predicting the 2018 Super Bowl Winner with Machine Learning

Predicting the 2018 Super Bowl winner.

Last year the Patriots were are our short list made in May and they won the Super Bowl.

Predicting the Superbowl winner in December is hard, predicting the Super Bowl winner in June should be near impossible. However, using Machine learning and some simple statistics, narrowing down the Superbowl winner may not be as difficult as it appears. So let’s get started…

Looking back at the last 10 Superbowls, 80% of the time, the Superbowl winner was ranked in the top 10 (actually top 8) in total defensive yards allowed per game for the prior season. (Less is better) There were two Superbowl winners that were ranked 22nd (2006-07 Colts) and 23rd (2009-10 Saints) in Yards allowed. However, we will go with the 80% and stay in the top 10 defenses from 2016 listed below.

Houston Texans
Arizona Cardinals
Minnesota Vikings
Denver Broncos
Seattle Seahawks
Jacksonville Jaguars
Baltimore Ravens
New England Patriots
Los Angeles Rams
New York Giants

Now the next statistic we will use to rank the teams is the quarterback “yes draft” “no draft” from our machine learning NFL franchise quarterback prediction system.

Of the last 10 Superbowl winners 90% had a quarterback that was marked as a yes draft. So we will look only at teams that have a “yes draft” starting quarterback.

Houston Drafted Deshaun Watson from Clemson and he is a “yes draft” and was our systems second pick behind Mitch Trubisky. Despite being a rookie QB, the Texans are on the list.

The Arizona Cardinals have quarterback Carson Palmer. Palmer is a “yes draft” and we leave them on the list with a shot at the Super Bowl.

It is unclear if the Minnesota Viking will have Teddy Bridgewater back this year. He is a yes draft with our machine learning franchise quarterback system – until we can confirm his return, the Vikings are off the list as Bradford is a “no draft”.

Denver has Trevor Simeon as their quarterback – he is a “no draft” and the Broncos are off the list.

Seattle has Russel Wilson and he is a “yes draft” we will leave Seattle on the Superbowl winner possible list.

The Jacksonville Jaguars have Blake Bortles and he is a “yes draft” – Jacksonville is on the list.

The Baltimore Ravens Quarterback Joe Flacco is a “yes draft” and the Ravens are on the list.

Tom Brady is a “yes draft” and his resume speaks for itself. The Pats are on the list.

The Rams have Jared Goff and he is a “yes draft” – assuming Goff starts the Rams are on the list.

Finally, the Giants have Eli Manning at QB and he is definitely a yes draft as he makes up part of our model for selecting a franchise quarterback.

So this is our shortlist for the Super Bowl contenders for 2018 in June – before training camp even starts.

Texans
Cardinals
Seahawks
Jaguars
Ravens
Patriots
Rams
Giants

2017 NFL DRAFT Quarterbacks Drafted First Round

Interestingly, in the 1st round of the 2017 NFL draft, there were 3 quarterbacks taken. The Bears traded up to number and took Trubisky. Our machine learning franchise Quarterback system had him ranked as the highest rated quarterback. The Chiefs also traded up to 10th to take Patrick Mahomes II – Mahomes was the number 4 rated quarterback prospect according to our machine learning NFL quarterback draft projections for 2017. The 3rd quarterback taken was Deshaun Watson. Watson was ranked as the number three prospect by our machine learning NFL franchise quarterback system.

Here is a link to our 2017 NFL quarterback predictions using machine learning

Read about how our machine learning system predicts quarterback success based only on college statistics.

2017 NFL Draft Machine Learning predicts top running backs

Our primary focus and goal is to use machine-learning technology to predict elite NFL running backs based solely on their college statistics. Not good running backs, elite running backs. To this end we have analyzed numerous running backs over the years and training a machine learning model based on the college stats of these star running backs. Below is the 2017 NFL draft top running back prospects predicted with our machine learning system. We would pick form the yes list and avoid the “no” list. Here is our analysis for the basis of our Machine Learning NFL Running Back Analysis

<

2017 NFL DRAFT – QUARTERBACKS ANALYZED WITH MACHINE LEARNING

We have analyzed the 2017 NFL draft quarterback prospects with our machine learning system. Our machine learning system predicts the probability that a draft pick will become a franchise Quarterback in the NFL. We have analyzed the last 20 years of Super Bowl winning quarterbacks and use this data applying machine learning to predict the probability of the 2017 draft quarterback prospects will be franchise quarterbacks and classifies them as a yes draft and a no draft and the probability they fit on that list. Machine Learning NFL quarterback system Here is the list…

Machine Learning Analyze NFL Quarterback Clustering College Statistics

The problem sometimes with explaining machine learning outcomes to NFL general managers, scouts and coaches, regarding NFL quarterbacks, is that the results are black boxed and it is hard to understand why the computer selects certain players over others. This article is going to try and put some visualization to players to demonstrate how some great quarterbacks compare to some less than great quarterbacks based on college statistics.

It is pretty easy to come up with a list of great quarterbacks that have played in the NFL, but it is hard to compare the college stats of older generations in today’s more aerial based game. The list below will be comprised of some of the more prolific passers with successful NFL stats and some of the projected quarterback draft picks who did not pan out in the NFL.

By looking into these stats, we hope this helps you see what our NFL Quarterback Machine Learning System sees when it analyzes players’ college stats. The machine learning clustering will relate the quarterbacks in the clusters based solely on college statistics.

Good

Peyton Manning
Tom Brady
Dan Marino
Aaron Rodgers
Dak Prescott
Drew Brees
Eli Manning
Andrew Luck
Philip Rivers
Ben Roethlisberger

Not So Good

Brady Quinn
Jake Plummer
Vince Young
Tim Couch
Joey Harrington
JaMarcus Russel
Matt Leinart
Trent Dilfer
David Carr
Mark Sanchez
Christian Ponder

Here is the graphic representation of the college statistics and there relation to all the players in both lists.

click image to see more detail

unknown-1

Here is the textual version of the graph showing the 5 clusters.

Cluster 1: manning|peyton, carr|david, leinart|matt, brees|drew, roethlisberger|ben, manning|eli, rivers|phillip, luck|andrew, prescott|dak
Cluster 2: brady|tom, aikman|troy, young|vince, rodgers|arron, sanchez|mark
Cluster 3: marino|dan
Cluster 4: ponder|christian, dilfer|trent, harringotn|joey, russell|jamarcus
Cluster 5: quinn|brady

We think that this provides great insight regarding how college statistics can predict NFL success. Cluster 1 is particularly interesting how the names (except Carr) are very elite. Yet cluster 4 and 5 are not. Cluster 2 shows some very elite quarterbacks as well.

Drafting Winning NFL Quarterback using Machine Learning

If an NFL team could predict, with a high probability, that a quarterback would have a greater than 55% winning percentage by his third season, using only college statistics, would that be something valuable to most NFL teams during the NFL draft?

We think the answer is absolutely a yes and the good news is we have developed such a system using machine learning.

And you read that right, our system ONLY USES COLLEGE STATS to predict the NFL performance of these quarterbacks.

Below is our analysis using our Machine Learning NFL Draft Quarterbacks Prediction System

For our analysis, we started with a list of active NFL quarterbacks for the 2016-2017 NFL season. We then limited the list to only those quarterbacks with 13 or more career starts. Using this list, we ran each player through our machine learning NFL quarterback draft system. Here is the list.

Player GS W L T Win Pct
Tom Brady 234 182 52 0 0.778
Drew Brees 231 131 100 0 0.567
Eli Manning 198 107 91 0 0.54
Ben Roethlisberger 183 123 60 0 0.672
Philip Rivers 175 97 78 0 0.554
Carson Palmer 173 88 84 1 0.512
Matt Ryan 141 84 57 0 0.596
Jay Cutler 139 68 71 0 0.489
Joe Flacco 137 83 54 0 0.606
Alex Smith 135 78 56 1 0.581
Aaron Rodgers 134 89 45 0 0.664
Tony Romo 127 78 49 0 0.614
Ryan Fitzpatrick 115 45 69 1 0.396
Matthew Stafford 108 51 57 0 0.472
Matt Schaub 92 47 45 0 0.511
Andy Dalton 91 54 35 2 0.604
Cam Newton 91 51 39 1 0.566
Russell Wilson 79 55 23 1 0.703
Matt Cassel 79 35 44 0 0.443
Ryan Tannehill 77 37 40 0 0.481
Sam Bradford 77 31 45 1 0.409
Mark Sanchez 72 37 35 0 0.514
Andrew Luck 69 42 27 0 0.609
Josh McCown 60 18 42 0 0.3
Colin Kaepernick 57 28 29 0 0.491
Chad Henne 53 18 35 0 0.34
Derek Carr 47 22 25 0 0.468
Derek Anderson 47 20 27 0 0.426
Blake Bortles 44 11 33 0 0.25
Kirk Cousins 40 19 20 1 0.487
Blaine Gabbert 40 9 31 0 0.225
Robert Griffin III 39 15 24 0 0.385
Nick Foles 36 20 16 0 0.556
Christian Ponder 36 14 21 1 0.403
Shaun Hill 35 17 18 0 0.486
Brian Hoyer 31 16 15 0 0.516
Jameis Winston 31 14 17 0 0.452
Geno Smith 30 12 18 0 0.4
Teddy Bridgewater 28 17 11 0 0.607
Tyrod Taylor 28 14 14 0 0.5
Matt Moore 27 15 12 0 0.556
Marcus Mariota 27 11 16 0 0.407
Colt McCoy 25 7 18 0 0.28
Brandon Weeden 25 6 19 0 0.24
Case Keenum 24 9 15 0 0.375
Brock Osweiler 21 13 8 0 0.619
Kellen Clemens 21 8 13 0 0.381
Bruce Gradkowski 20 6 14 0 0.3
Mike Glennon 18 5 13 0 0.278
E.J. Manuel 16 6 10 0 0.375
Dak Prescott 15 13 2 0 0.867
Carson Wentz 15 6 9 0 0.4
Drew Stanton 13 8 5 0 0.615
Trevor Siemian 13 7 6 0 0.538

There were a total of 54 NFL Quarterbacks active in 2016-2017 that had thirteen or more starts.  Of those, 36 would have been “yes” drafts by our Machine Learning system and 18 would have been no drafts.

To better visualize how the “yes” drafts quarterbacks and no draft quarterbacks compare, we graphed them in the chart at the bottom of this page.  The no drafts are RED and the yes drafts are GREEN.

If you look at the chart, plotted along with the red and green dots, you will see a red vertical line.  This line represents the 51% win mark.  Any player to the right has 51% or greater career wins.

The horizontal green line represents 50 starts which is right at 3 complete regular season. So above the green line is 50 or more starts

We think the most compelling relationship exposed by this graph is that only 4 of the 18 no draft players have more than 55% of the wins regardless of how far along they are in there career.  This means that if you drafted a Quarterback that our system predicted was a no draft, you would have only a 22% chance of drafting a quarterback that would have a 55% or better win percent at any time in his career.  Conversely, that is an 78% chance that you would draft a quarterback with a less than 55% win percent probability.  We should note, that the one red dot in the top right quadrant is Aaron Rodgers and he came out of college after 2 years – so our system did not have enough data to rank him and defaulted to a No draft. Removing Rodgers, the no drafts win percentages are very pathetic.

It should be noted that all of the NO draft (red dot) quarterbacks were in fact drafted by a team and were in fact active at some point in 2016 – 2017.

Now looking at the “yes” drafts on the chart, there are 21 of 36 that have a winning percentage above 55% which is a more than a 58% probability.  So by avoiding the “no” drafts and sticking with the “yes” drafts, an NFL team would have over a 58% probability of choosing a quarterback that would have a at least a 55% win percentage.

What is really interesting is looking at the chart where all the dots are above the 50 starts green line, you can see there is a clear dichotomy here between the red and green dots.

Here is the breakdown once the “yes” draft quarterbacks and the no draft quarterbacks reach 50 starts.   There are 17 quarterbacks from the list that reached 50 starts of those, 15 had win percentages of at least 55% – that is a whopping 88%.

There are 9 “no” draft quarterbacks that achieved 50 starts, of those 9, only 3 had a winning percentage of at least 55%.  That means that only 33% of the no draft predicted quarterbacks achieved a 55% or higher winning percentage.

screen-shot-2016-12-29-at-1-06-05-pm

If an NFL team wants to draft a college quarterback in the NFL draft that is more likely to have a 55% winning percentage, they should ask us to help them using our Machine Learning NFL Draft Winning Quarterbacks Prediction System

Machine Learning Predicts Quality at Quarterback for Cleveland Browns 2016

For a team that has had a revolving door at the quarterback position for many years, we find it fascinating that the Browns have 3 of 4 quarterbacks that our NFL franchise quarterback machine learning system rates as “yes” draft. They usually have no drafts at the helm which causes the revolving door. Read more about our machine learning NFL franchise quaterback system.

Robert Griffin III is a yes draft, Josh McCown is a “no” and Cody Kessler and Kevin Hogan are both “yes” drafts. Obviously RGIII has shown he is injury prone, but we think our system takes that into account and we think he will have a break out year. And although Kessler didn’t look great in preseason, we think he will develop. And the addition of Kevin Hogan we think was a good choice for the Browns.

We think the additon of Paul De Podesta is the driving force behind these moves to acquire quarterbacks with a higher probability to perform at a high level.

We will watch the Browns QBs closely.

Machine Learning Predicts These 2016 NFL Teams Will Struggle at Quarterback

Using our Machine Learning NFL Franchise Quarterback prediction system for the 2016-2017 season, we project that these teams will struggle at the quarterback position. Our system predicts the career of a college quarterback in the NFL based soley on college statistics. Our system labels a quarterback “yes” draft and “no” draft. The “no draft” designation indicates that a quarterback will NOT likely be a franchise quarterback and that they will put up less than stellar numbers. It also is an indicator if a team will make it to the playoffs. Only 2 of the 12 quarterbacks that went to the playoffs last year (2015-2016) were “no” draft quarterbacks.

San Francisco 49ers – Not only has Kaepernick drawn unwanted attention to the 49ers, the 49ers have a true quarterback dilemma. They have Blaine Gabbert and Colin Kaepernick both of these quarterbacks are no draft according to our machine learning NFL franchise QB system. (kaepernick|colin, [‘no’], 90.90% – gabbert|blaine, [‘no’], 69.37%)

Buffalo Bills – The Bills look like they will start Tyrod Taylor (taylor|tyrod, [‘no’], 90.56%). He is a no draft according to our machine learning system. The good news in Buffalo is they have Cardale Jones (jones|cardale, [‘yes’], 97%) who is a yes draft and will some day be the starter.

Minnesota Vikings – The Vikings lost Teddy Bridgewater (bridgewater|teddy, [‘yes’], 92.25%) to a knee injury in practice. The replacment, for now is Shaun Hill a 36 year old veteran. Shaun Hill is a NO and the Vikings will suffer through this season we fear. (hill|shaun, [‘no’], 92.25%). A new development, Vikings picked up Sam Bradford – also a no draft. (bradford|sam, [‘no’], 51.92%) As we said, we fear the Vikings will suffer through.

Houston Texans – The Texans paid a lot of money for Brock Osweiler who was benched last year by the Broncos. He is a no draft and the Texans will have problems generating offense. (osweiler|brock, [‘no’], 80.78%)

Miami Dolphins – Ryan Tannehill is a No draft and the Dolphins will suffer another ho-hum season with him as quarterback. Good news for the dolphins, Brandon Doughty is a yes draft and he looked great in pre-season – for a rookie. Doughty was our best value in the draft last year as he was highly rated by our system and taken as the last quarterback overall.

Tennessee Titans – Not only is Mariotta young and inexperienced, Our machine learning NFL franchise quarterback prediction system thinks he is a no draft. (mariotta|marcus, [‘no’], 71.51%).

NY Jets – AHHHH the Jets, the season would not be complete without mentioning the Jets as they will continue to experience quarterback issues. Fitzpatrick is a no draft. (fitzpatrick|ryan, [‘no’], 73.80%).

Detroit Lions – Although Stafford puts up some impressive numbers, he can’t seem to get the lions into the playoffs consistently. Our system thinks he is a “no” and we predict the woes for the Lions will continue. (stafford|matthew, [‘no’], 94.08%)

2016 NFL draft running backs analyzed with machine learning

The following is a list of the 2016 running backs drafted in the order they were taken. After the name of the team that took them, we compare the draft picks with out Machine Learning running back analysis system to see what teams took “yes” players and which teams took “no” players. The players are rated yes or no if they would have been drafted and the probability the system thinks they fit in that category.

  • Ezekiel Elliot –  Cowboys – Yes – 99.862 %
  • Derrick Henry – Titans – Yes – 99.91 %
  • Howard Jordan – Bears – Yes – 99.992 %
  • Devontae Booker – Broncos – Yes – 99.994%
  • Kenneth Dixon – Ravens – Yes – 99.79%
  • Paul Perkins – Giants – Yes – 99.98%
  • CJ Prosise – Seahawks – No – 97.16%
  • Alex Collins – Seahawks – Yes – 99.991%
  • Kenyan Drake – Dolphins – No – 99.28%
  • Jonathan Williams – Bills – No – 97.98%
  • Tyler Ervin – Texans – No – 98.44%
  • Kelvin Taylor – 49ers – No – 98.90%
  • Keith Marshall – Redskins – No – 98.79%
  • Daniel Lasco – Saints – No – 98.30%
  • DeAndre Washington – Yes – 99.993%
  • Andy Janovich – Broncos – No – 97.38%
  • Wendell Smallwood – Eagles – No – 97.16%
  • Dwayne Washington – Lions – No – 99.22%
  • Dan Vitale – Bucs – No – 98.68%
  • Derek Watt – Chargers – No – 96.89%
  • Keenan Reynolds – Ravens – Yes – 99.18%
  • Zac Brooks – Seahawks – No – 97.79%
  • Darius Jackson – Cowboys – No – 98.15%

Machine Learning and NFL running backs draft analysis

NFL Running Back Analysis with Machine Learning

The following is a list of the early round (mostly 1st and 2nd) draft picks, at running back, drafted between 2009 and 2014. Our machine learning running back analysis system was applied to determine what picks were “yes draft” players and which were “no draft” players based only on college stats. The players are rated yes or no if they would have been drafted and the probability the system thinks they fit in that category.

In testing, our system was 86% accurate in finding productive NFL running backs.

*** THIS ANALYSIS IS BASED SOLEY ON COLLEGE STATISTICS OF THE PLAYERS BELOW.

If you picked “yes draft” backs for 2009-2014, you would have had an 86% chance of selecting a productive back.

If you picked “no draft” backs for 2009 – 2014, you would have had a 41% chance of selecting a productive back.

RB data read successfully!

name predict prob

    bernard|giovani [yes] 0.999922
      moreno|knowshon [yes] 0.999879
      wells|chris [yes] 0.999879
      mason|tre [yes] 0.999856
      bell|lavian [yes] 0.999843
      mccoy|lesean [yes] 0.999813
      hillman|ronnie [yes] 0.999795
      sanky|bishop [yes] 0.999777
      hyde|carlos [yes] 0.999777
      hill|jeremy [yes] 0.999757
      williams|ryan [yes] 0.999711
      martin|doug [yes] 0.999622
      james|lamichael [yes] 0.99962
      richardson|trent [yes] 0.999551
      mathews|ryan [yes] 0.999362
      ingram|mark [yes] 0.99917
      ball|montee [yes] 0.996881

    name predict prob

        spiller|cj [no] 0.997899
        sims|charles [no] 0.997508
        brown|donald [no] 0.995403
        lacy|eddie [no] 0.994026
        best|jahvid [no] 0.993481
        vereen|shane [no] 0.993481
        leshoure|mikel [no] 0.98901
        green|shonn [no] 0.988013
        wilson|david [no] 0.983055
        pead|isiah [no] 0.981617
        mccluster|dexter [no] 0.97806
        coffee|glen [no] 0.976095

      Below you can visual where the draft and no draft players stand relative to each other by year and “yes draft” – “no draft”.

      unknown-1

      unknown-2

      unknown-5

      unknown-6

      unknown-3

      unknown-7

      If you picked “yes draft” backs for 2009-2014, you would have had a 86% chance of selecting a productive back.

      If you picked “no draft” backs for 2009 – 2014, you would have had a 41% chance of selecting a productive back.

      We will soon release our analysis of the 2016 running backs that were drafted.