DFS NFL: The Subtle Science Of Predicting Field Goals
Some people do not like playing FanDuel NFL games because they find picking a kicker to be pointless. A lot of people feel picking a kicker is mostly a crapshoot, and get frustrated when a kicker goes off for 6 or 7 field goals and they didn’t have a piece. The truth is field goal kicking is one of the more predictable aspects of NFL daily fantasy. In this article, I’ll talk about the metrics you can use to help pick kicker on FanDuel.
I’ve written in previous articles about my thoughts on picking a kicker. For a lot of people, this may be a boring topic. For some reason, I find it fascinating. I think anytime numbers illuminate a truth about a sport, I get excited. The research I did on kicking was particularly rewarding because the factors I believed to be important in kicking ended up being very significant.
For a FG attempt to occur, three main things have to happen.
1. A team must get within a reasonable yardage to the end zone for a field goal to be plausible, normally this is around the 35 yard line or in.
2. A team must be in a position where they do not need a touchdown for a significant effect on a games outcome.
3. A team must fail to score a TD when they get into that area.
The first factor is tough to capture. You want an offense that will move the ball, but not so well that they are scoring a lot of touchdowns. It also helps to have a defense that constantly is stopping their opponents and giving them good field position. I felt like scoring margin (average margin of victory/loss) captured the first factor well, but also killed two birds with one stone capturing the second factor. A team will often employ a conservative strategy when winning and kick field goals on 4th down instead of going for it, but will have to go for TDs when down late in the game.
For the last factor, luckily stats are easily available that capture this for the most part. Red zone TD percentage tells you how often a TD is scored for a given team when they get 20 yards or closer to the goal line. It fails to include the 20-35 area, but still should do a good job finding teams that struggle to score TDs when within FG range.
I took those two factors and looked at how well they correlated with 2014 data. I looked at FG attempts instead of FG made, because I wanted to look at how field goal opportunities were coming about, not how quality a teams FG kicker was at making FGs.
Scoring margin had a strong correlation to FG attempts for 2014, coming it at .58. Red zone TD percentage was also somewhat robust, coming in at -.41 (the better a team is in the red zone, the less FGs they kick). Because both these signals were so orthogonal, fitting both of them to FG attempts was very robust, giving us an r-squared value of .634 and a correlation of about .8. In the graph below, you can see how well FG attempts follows the predictions from red zone TD percentage and scoring margin for 2014. As the predicted FG attempts increase, field goal attempts increase proportionately.
You probably notice one clear outlier, at 22 FG attempts and the lowest predictor value is the Oakland Raiders. They had the highest margin of loss, 12.4, but they also had the highest red zone TD percentage in the league at 71%. Because they were always losing and scored a ton of TDs in the red zone, we predicted about 15 FGs on the year for them. Ultimately, they were tied for second to last in FG attempts, but our formula predicted them to have 7 less field goal attempts than we projected.
What factors could we be missing here? Oakland does have a very good FG kicker in Sebastian Janikowski, which is certainly a factor that would have some influence on whether a team kicks a FG, punts, or goes for it on 4th down. But looking at the 2015 outliers, I get a sense that there is another factor this current model does not capture.
The predictor does not do as well on this data as the 2014 data, but the correlation coefficient of .623 on the 2015 set suggests it still is predictive. You can clearly see field goal attempts increase as the predictor value rises, but there is more error around the prediction. This should be expected with still only a little more than half the season in the books.
The team with the highest FG attempts this season is Tampa Bay with 29. They are the dot on the far right. They are far outpacing our prediction of 19. Why is this happening? One thing that this team has in common with the 2014 Raiders, the other outlier, is both teams started rookie QBs. Oakland had Derek Carr in 2014 and the Bucs have Jameis Winston this year. I think there’s a rookie QB factor that is hard to capture numerical but makes sense logically. Coaches are not going to push their rookies to do too much. They don’t want them to fail going for it on 4th down in an important part of the game and have them lose confidence. So both the Raiders and the Buccanneers have a lot more FG attempts then we would expect from scoring margin and red zone performance alone.
Using the 2014 data as a training set, I applied the FG prediction model to make predictions on a week to week basis. For projected scoring margin, I use Vegas spreads. For red zone TD percentage, I adjust the teams current season red zone TD percentage against their opponents red zone TD percentage allowed. You can see the predictions for this week and for the rest of the season on our Projected FGA tool.
The leader in projected FGA this week is Cincinnati. Not only do they have one of the highest projected margins of victory at 9, they are facing the top red zone defense in the league, the St. Louis Rams. Mike Nugent hasn’t been a great kicker this year, and his leg is not strong (his long this year is 47 yards), so I think he’s a solid kicker this week, but not the best. The best kicker this week is Steven Hauschka, who has only missed one FG the entire season and has a huge leg. Dallas has had an average red zone defense this season, but Seattle’s awful red zone offense makes them highly projected for FGAs.
New Orleans has by far the worst FGA projection at 1.23. They’re only small underdogs, but they are also one of the best red zone offenses in the league, facing one of the worst red zone defenses. San Diego is in the same boat, although their red zone offense is not quite as good as New Orleans. Interestingly, both these offenses have great matchups this week in fantasy. The fact that they are expected to score so little points with FGs definitely gives their QB and skill position players a slight bump for me, since more of their scoring should come from TDs.
Look for projected FGAs to be updated each week, and remember to consider the skill of a teams kicker when considering kickers to choose. I’d recommend using ProFootballFocus’ FG evaluations if you have a membership, but you can also look at the kickers NFL stats page on NFL.com to get a solid feel for kicker quality.
On a side note, this tool is great for streaming kickers in season long. In my league, I managed to pick up Mason Crosby. Not only does Green Bay have a 2.14 FGA projection, Crosby is one of the best kickers in the league and has a big leg.
Good luck this week for daily fantasy. Make sure to get in some lineups on FanDuel. Hopefully using our FGA predictions will pay off in a big way.
View all posts by Daniel Steinberg