fantasy football     JOIN THE HUDDLE    
HOME ARTICLES NEWS DRAFT GUIDE REGULAR SEASON STATISTICS NFL TEAMS MESSAGE BOARDS

FANTASY FOOTBALL ARTICLES

NFL Power Rankings for Fantasy Football -- The 'Blvd. Way
Part 2: Predicting Touchdowns Against
Fritz Schlottman
August 17, 2005
Part 1 | Part 2

The pros on the Blvd come from various backgrounds. Some are ex bank executives, some have advanced degrees in statistics, and some are just wise guys that have been doing this for years. However, one common thread between them is how they think about football. This article goes through the process by which some of the best in the business think about NFL scoring and may provide the reader with some useful strategies for evaluating players.

Fantasy football is scored by awarding fantasy points for real points on the football field. Most fantasy points are awarded for offensive scoring: touchdown passes, touchdown runs, and touchdown receptions. There are points scored for other plays: safeties, field goals, defense, etc. but most points are scored when the player scores a touchdown for his team in some manner. Therefore there is a direct connection between team scoring and player scoring. Likewise there is a direct connection between teams giving up touchdowns and opposing players scoring.

“Who should I start?” is the oldest questions in fantasy football. This is actually a two-part question. The first part is, “Which team/player is more likely to score a touchdown?” and the second part is “Which team is more likely to give up the touchdown?”In Part I of this series I presented several statistics that were good indications of ability of teams to score or not score. This week, I’ll tackle the second part of the equation.

To identify why teams score points, I collected statistics from the 2004 season and spread them out on an Excel spreadsheet. I then ranked each of the teams in each category with the first ranking (number one) being given to the team with the best statistics in each category and the number 32 being given to the team who was last in that category. I then divided the 32 team database into two smaller databases: the TOP 10 database had all the statistics for the teams ranked 1-10 (with several ties) in touchdowns scored against and the BOTTOM 10 database had the 10 teams that opponents scored the most touchdowns against: The teams in each database were:

Team

TD Rank

Pittsburgh

1

Baltimore

2

Buffalo

3

NY Jets

4

Philadelphia

4

Washington

4

New England

7

Jacksonville

7

Denver

9

Tampa Bay

9

 

Team

TD Rank

Oakland

32

San Francisco

31

Kansas City

30

Tennessee

29

Dallas

28

Green Bay

27

Minnesota

26

Cleveland

25

New Orleans

24

St. Louis

23

 

The next task was to identify “strength”. I use the ‘Vegas term “strength” as a proxy or a kind of statistical shorthand to represent the likelihood that a variable is related to touchdowns scored. A statistic was “strong” if many of the TOP 10 teams in touchdowns scored were also in the TOP 10 ranking for this variable. Also, a statistic was “strong” if many of the BOTTOM 10 ranking teams were also among the BOTTOM 10 in this statistic. A statistic (variable) was strong if there were at least six teams in the TOP 10 touchdowns scored list that were also in the TOP 10 list for this statistic and at least 6 teams in the BOTTOM 10 list were also in the BOTTOM 10 list. Each of the variables making the cut was then ranked to determine which variables were the strongest predictor of touchdowns scored. Teams were then ranked in each of these variables to determine which were the “strongest” or “weakest” teams.

STATISTICS (VARIBLES) THAT DID NOT MAKE THE CUT

Some of the statistics that are commonly associated with good or bad touchdown scoring did not make the cut. These are the “Urban Legends” of NFL statistics. The statistics eliminated were Time of Possession, the Number of Plays an offense ran or were run against a defense, Penalties, Penalty Yardage, Giveaways, Takeaways, 3 rd Down %, and Turnover Margin.

Takeaways, Giveaways, and Turnover Margin were good predictors that teams give up touchdowns but were not good predictors that teams would stop opponents from reaching their end zone. This is the old “bend but don’t break” strategy where some teams will play passive defense rather than give up the big play (more later).

So what did make the cut?

#1 YARD PER PLAY (9 OF 10 IN TOP 10 AND 7 OF 10 IN BOTTOM 10)



Just as in Part I of this series, Yards per Play was the number one predictor of touchdowns against. Teams that give up the big play give up more touchdowns than do defenses that are able to keep opposing offensive players in front of them.

Jacksonville (18 th) was the lone exception in the TOP 10. It seams the Jaguars defense was able to overcome mistakes and keep their opponents from converting big plays into touchdowns.

Cleveland (12 th), St. Louis (18 th tie), and San Francisco (21 st) under performed. While they gave up less yards per play than many other teams (especially Cleveland), opposing teams found a way of getting into the end zone. Turnover differential explains this. St. Louis (32nd Turnover Margin), San Francisco (31 st Turnover Margin), and Cleveland (26 th Turnover Margin) consistently put their defenses in bad spots by turning the ball over. Bend but don’t break doesn’t work very well on a short field.

THE BEST
1. Buffalo
2. Washington
3. Pittsburgh
4. Tampa Bay
5. Baltimore

THE WORST
32. Kansas City
31. Tennessee
29. New Orleans
29. Minnesota
28. Green Bay

#2 YARDS PER GAME (9 OF 10 IN TOP 10 AND 7 OF 10 IN BOTTOM 10)

Just as was discussed in Part I of this series, teams that give up yardage also give up points. The bend but don’t break style of defense where defense give up yards between the 20’s and then stiffen to force a field goal doesn’t seem to pan out for either the best or worst scoring defenses.

In the TOP 10 Defense list the only team outside the top 10 group was Jacksonville (11 th). Not much to comment on here. In the BOTTOM 10, Cleveland (15 th), Dallas (16 th), and St. Louis (17 th) were underperformers. Turnovers were once again the reason for the poor performance as Cleveland was 26 th in Turnover Margin, Dallas was 28 th, and St. Louis was last in the NFL (32 nd).

THE BEST
1. Pittsburgh
2. Buffalo
3. Washington
4. Denver
5. Tampa Bay

THE WORST
32. New Orleans
31. Kansas City
30. Oakland
29. Minnesota
28. Tennessee

#3 FIRST DOWNS PER GAME (8 OF 10 IN TOP 10 AND 7 OF 10 IN BOTTOM 10)

This is getting monotonous isn’t it? Once again, as we saw in Part I of this series, first downs lead to scoring touchdowns. Not many teams can put together 80-yard touchdown drives, the squads that deny opponents’ first downs are not going to give up touchdowns either.

THE BEST
1. Denver
2. Pittsburgh
3. Washington
4. Buffalo
5. Tampa Bay

THE WORST
32. Oakland
31. Minnesota
30. New Orleans
29. Kansas City
28. San Francisco

DEFENSES TO AVOID-LOOK FOR BETTER OPTIONS

  1. Pittsburgh
  2. Buffalo
  3. Washington
  4. Denver
  5. Tampa Bay
  6. Baltimore

DEFENSES TO EXPLOIT-START ‘EM IF YOU GOT ‘EM

  1. Kansas City (tie)
  2. New Orleans (tie)
  3. Minnesota
  4. Oakland
  5. Tennessee

LOOK-OUT/WATCH-OUT TEAMS

BIGGEST UNDER-PERFORMERS (COULD BE GOOD UNITS BUT GIVE UP TOO MANY TD’S)

  1. Miami
  2. Detroit
  3. Tampa Bay
  4. Cleveland

BIGGEST OVER-PERFORMERS (LOOKOUT!-THESE TEAMS PREVENTED TOUCHDOWNS DESPITE POOR UNDERLYING NUMBERS)

  1. Jacksonville
  2. Arizona
  3. San Diego
  4. Indianapolis
  5. Houston