Kick the Kicker: Remix Part Two

Jacob Feldman

After a few years away from the topic, I’ve decided to come back to my idea of kicking the kicker. If you missed part one of this reiteration, make sure you go back and check it out. I talked about the types of people who want to keep kickers in dynasty leagues and some potential flaws in their logic. I also went through the three requirements I feel must be met for a player/position to be a viable dynasty asset. In case you missed those, here they are in very brief summary.

  • Must have a measurable statistic
  • Must be consistent and/or predictable from one week to the next
  • Must be consistent and/or predictable from one season to the next (this is dynasty after all!)

Kickers meet the first criteria, but I think they fail miserably on the other two. Let’s take a look at some numbers while I discuss why I feel that way. For the future parts, I’m using standard PPR scoring for other positions. For kickers, I’m using 3 points for a field goal under 40 yards, 4 for a kick from 40-49 yards, 5 for something 50 yards or longer, 1 point for an extra point, and -1 for a missed extra point or field goals within 40 yards.

Week to Week Consistency

[am4show have=’g1;’ guest_error=’sub_message’ user_error=’sub_message’ ]

To help me examine week to week consistency, I decided to use some basic statistics. I started by finding the mean (or average) score for a player and then their standard deviation. For those unfamiliar with statistics, standard deviation is basically the average distance a data point in the set is away from the mean. Data sets with a small standard deviation have their values much closer together (more consistent week to week in this case) while data sets with large standard deviations are much more spread out (less consistent). To balance out the inequities in scoring between the various positions, I then calculated the relative standard deviation, which is simply the standard deviation divided by the mean to give up a ratio or a percentage. This tells us what percentage variation we can expect from one week to the next.

For kickers I took the top 15 scoring kickers from the 2016 season using the scoring stated above. The kickers coming off of the better offenses were a little bit more consistent, which was expected. This is largely because their teams get into the red zone more often, so the kicker is getting some kind of points in most cases. However, even being the most consistent of the kickers didn’t mean they were consistent. Even the best of the best in terms of week to week scoring consistency, Graham Gano, had a relative standard deviation of 41 percent. The worst from the sample kickers was the Chiefs’ Cairo Santos with a relative standard deviation of 74 percent. The entire sample had a mean relative standard deviation of 52 percent.

Now just giving you numbers without anything to compare it to isn’t very helpful, especially since I’m trying to convince you that kickers are more inconsistent than the rest of the positions. I did a similar exercise with the rest of the offensive positions and here is what I found for their mean relative standard deviations.

sdev

The way the skill positions broke down really shouldn’t surprise anyone. We all know quarterbacks are by far the most consistent position week in and week out. I was a little surprised at how close the other three positions were though. I was expecting receiver to be a fair amount more consistent than the running backs and tight ends, but the differences were relatively small between the three positions. I was also a little surprised that the most consistent player outside of the quarterback position (where Carson Palmer lead the pack at 20%) was Mark Ingram with 24%. That means he very seldom had great weeks, but he also didn’t kill your team with a 3 point week.

The majority of the inconsistency in the three positions came from huge games popping up. For example, a player who averages 16 points per week has a game where they score 36 points. Just that one week can have a pretty significant impact on their standard deviation. As I thought about this a little bit more, it occurred to me that this makes sense since at the running back, receiver, and tight end positions one huge play can be double digit points and possibly even come really close to a player’s weekly average in just that one play. Think about an 80 yard catch and run. We are talking about 15 points right there at one point for ten yards, one point for the reception and six for a touchdown. This isn’t true of the quarterback or kicker positions where even the biggest play isn’t quite as impactful on their weekly totals meaning they should be more consistent, which is true for the quarterbacks. However, the kickers are obviously the least consistent according to the numbers, which seems to go against the logic and the overall trend.

Week to Week Predictability

There was a second part to this week to week benchmark, predictability. I mentioned in the first part of this article if someone is inconsistent, but they are predictable in that nature, then they can still be a very valuable piece of your team. I used the example of a middle tier running back that produces against weak defenses but gets stuffed by good ones. You will know which weeks you can start them and which weeks they need to sit, and you can work with that knowledge.

The general way we predict production from week to week is based on the strength of the opposing team. In order to check the predictability, I took the strengths of the defenses they were playing each week and laid that over the top of their scores to see if there was a correlation. The quarterbacks, running backs, and wide receivers all had very strong correlations between the two. Some individuals had a more significant correlation than others, but there was a general trend. Meaning they scored most of their best scores against weaker defenses and most of their lower scores were against better defenses, which is exactly what you would expect. There were of course outliers, but the overall trend was there. The correlation for tight ends wasn’t nearly as strong, but it did still seem to exist.

Kickers were a completely different story. There really wasn’t any kind of correlation or true predictability on the part of the kickers. They didn’t score more against bad defenses, maybe because they ended up with extra points instead of field goals. They weren’t less against good defenses. It was by in large relatively random. Did their coach happen to give them a shot at a 50 yarder that week? Was the wind favorable on that day? Did the team go for two instead of kicking extra points? Kickers didn’t seem to be predictable when looking at the defensive strength like the other positions.

Part Two Conclusions

Knowing who to start and when to start them is a key foundation of fantasy football, and it is what the idea of week to week consistency and predictability is based upon. There will always be exceptions, but for the most part if we pay attention and notice the trends, we can pick the right players. We can do this with our quarterbacks, running backs, wide receivers, and tight ends. However, the kickers seem to fail in this regard.

This probably has something to do with the counter intuitive nature of the kicker’s role on the team. Think about it for a moment. If you want a kicker to score the maximum number of points possible, you want field goals. Not just field goals, but long field goals if you use distance scoring. This means you need the offense to move the ball across midfield, but then you need it to break down. Meaning you want the entire offense to be both good and terrible in the same drive, over and over. Therein lies the paradox which makes kicker consistency non-existent. By the time you mix in all of the other things like weather, coaches decisions on when to kick the field goal or not, the increasing trend of trying more two point conversions, and many other factors, this isn’t going to change anytime soon.

If you accept the premise that consistency and predictability from week to week is important in fantasy football, then you need to count this as a strike against the kicker position. They are easily the most inconsistent position and the ability to predict them from one week to the next, at least in any way similar to the way you predict production for other players, doesn’t seem to be there. This means the production of your kicker from one week to the next is more or less random, which amounts to little more than just the luck of a dice roll.

In the third and final part of this year’s edition I’ll be taking a look at the year to year consistency to see if the top performing players from the kicker position continue from one year to the next. If so, it means you might be able to gain an overall advantage on the league by acquiring a top kicker, even if the week to week production is a coin flip. If not, it will be yet another reason to Kick the Kicker.

[/am4show]

jacob feldman