Positional Tiers and Value

Tiers in Fantasy Football

As you being your journey in Fantasy Football, you’ll often hear people discussing tiers, i.e RB1, RB2, QB1, QB2, QB3 and so on. Intuitively these tiers are groupings of players by performance, so a Quarter Back that falls in the QB1 tier will be more valuable than a QB2 Quarter Back.

The definition of a given tier is arbitrary and varies from person to person. Some people might consider to top 12 RB to be RB1, while others only the top 6-8. There are other, non quantifiable decisions that often go into these tiers, such as coaching changes, usage changes, health issues and so on. While the definitions will vary, the concepts does not - the higher the tier a player is in, the more valuable he is.

As mentioned above, going into the new season, there are a lot of factors affecting tiers. For the sake of this analysis, however, we will define our tier purely based on last years performance. So our tiers will not necessarily be an indicator of next year’s tiers. That is to say that someone who we identify as a QB1 with or analyses, will not necessarily be a QB1 going into the next season, but that is besides the point.

We are establishing tiers purely to demonstrate positional value. We will look at which position has the biggest variance across their tiers, and in doing so, will gain some insight as to what the most valuable positions are.

Why Tiers?

You may be thinking to yourself that there is a more straightforward way to evaluate position value. Why don’t we just look at the average Point Per Game by position to establish which the most important positions are? If that’s a question you’re having, that is an absolutely valid one to have at this point. However, purely looking at PPG per position will not tell the whole story, and the reasons we’re looking at Tiers is to undercover the true value. So for now, bare with me. At the end of this report, we will tie everything together and see how Tiers can give us more insight into positional value than purely PPG by position.

Defining Tiers

We will keep our tiers indicators simple. We will make use of the average point per game, and the variance of points per game. Simply put, we will look at how good players are on any given game, and how consistent they are throughout the season. We will begin by analyzing the data for the Quarter Back position for the year 2018.

There are a couple things to note about this graph. First you will observe that both axes are on a scale of -2 to +2. That is because we have normalized the data. This allowed a cleaner read and puts the data on a similar scale. Essentially, the farther right (more positive) a player is, the higher the average PPG for that player is, and the farther left (more negative) the lower the average PPG is. Similarly, the higher up a player is, the less consistent they are, the lower down, the more consistent they are.

Let’s take a closer look at a few player. A perfect example of a player you don’t want, is Josh Rosen on the far bottom left. That position means that he has a a low PPG (significatnly lower than the average, which on the chart is at the 0 line) and that he is consistent on a game to game basis. So, bad… and consistently bad. Now look at Ryan Fitzpatrick on the top right. Above average PPG, but extremely inconsistent. Meaning that having him on your team is a roller coaster - boom or bust as they say.

Lastly, let’s look at this years MVP, and league winner for many, Patrick Mahomes on the middle right. He has a significantly higher PPG than the average, and is as consistent as the average QB in the league (close to the 0 horizontal line.)

Clustering the Scatter Plot

The natural next step to this plot, is identifying “clusters”, or simply put, groups within this plot that are very similar. We could intuitively try to group some of the high performers with average consistency, like Patrick Mahomes and Matt Ryan together, and do the same forabove average performers with low consistency (Fitzpatrick, Trubisky, Brees.) We would then have to find every one of these clusters and group players manually. While this is doable and intuitively makes sense, there is another way around it. There are certain algorigthms that we can use that can help us identify the optimal number of cluster, and what those clusters are.

In this analysis, we will make use of one of these algorithms, called a K-means. As mentioned, the K-means algorithm will identify how many clusters there are in our data, and build those clusters, grouping players with similar PPG and Standard Deviation. See below for the results.

Now, our algorithm has clustered players together, based on how similarly they perform. Let’s take a quick look at the summary for each cluster.

##   Cluster  Tier.PPG   Tier.SD Tier
## 1       1 18.961034  7.356099    3
## 2       2 13.678667  6.667779    6
## 3       3  8.651429  4.267737    8
## 4       4 16.589730  5.808217    4
## 5       5 22.601562  7.843288    1
## 6       6 14.008571  4.517360    5
## 7       7 19.228000 10.912020    2
## 8       8 13.202632  8.948080    7

The result above is how we define our tiers. Looking at the first three columns, we can see the “clusters” that our algorithm has identified, and we can look at the average PPG and Deviation of all the players within each cluster. Consequently, we can identify the cluster with the highest PPG as the Tier 1, or QB1. From the above result, we have identified cluster 5 as our highest performing cluster (unsurprisingly where Mahomes sits) and hence Tier 1. The next highest will be QB2, and so on. Note that every player within a certain cluster has a fairly similar consistency to every other player within the cluster.

Running Back, Wide Receiver and Tight End Clusters and Tiers

Now that we have established how to Tier our QBs, let’s apply the same principles to the other relevant Fantasy position. Below are the cluster plot for Running Backs, Wide Recievers and Tight Ends.

Normal Distributions

For the next part of this analysis, we will make use of normal distributions to interprete the position value in Fantasy Football. If you don’t know what those are, here’s a quick explanation - feel free to skip ahead if you are familiar with the concept.

A normal distribution, shown below, is simply a way to summarize a data set, looking at the average and the variance in that data set. Let us only focus on two aspects of the normal distribution plot: the peak and its breadth

In the above example, we can see two very different distributions.Let’s assume that they represent the Fantasy Points of two players, on a game to game basis. The peak (highest point) represent the average Point per Game. So the Red Player has a lower PPG compared to the Blue player.

The breadth of the plot, is an indication of variance (or consistency.) The narrower the plot, the more consistent the player is, and similarly, the broader, the less consistent. So while Blue Player averages more PPG, they are significantly less consistent that Red Player and hence less reliable. Note that the height inversly correlates to the width - a wide distribution will be shorter than a narrow distribution.

Positional Value in Fantasy Football

Quarter Back Analysis

Now that we have established our Tiers and brushed up on Normal Distributions, let’s see what insights we can gain from the positional tiers. Let’s first take a look at the QB Tiers.

In the grid below, we’ll look at the QB Tiers in comparison to each other (1st Graph), as well singling out QB1 vs the other tiers to highlight the differences.

So, what can we observe by comparing those distributions? For starters, in the “All QB Tier” chart, we can clearly observe that QB2 and QB3 are very similar to each other and not at all that different from a QB1, and that the biggest discrepancy is between QB1 and QB4. We can also simply focus the difference between the average PPG compared to QB1 for each other tier, obtaining:

Essentially, we observe that QB2 average about 18% less than the QB1 PPG, and are less consistent. Again, we see that QB3 are very close to QB2, in that they average roughly 19% less than the QB1 PPG, but are slightly more consistent. We do see a big falloff in PPG when it comes down to QB3, and the higher consistency of QB4 only means that, we are dealing with the Josh Rosen kind of QB - Bad, and reliably so.

This, however, does not tell us much. Is 18% between QB1 and QB2 significant, is it a lot, or very little? So, to put a context to this data, let’s compare across the other positions.

All Positional Value Fantasy Football

First, let’s look at what the normal distribution for all tiers across all positions look like.

The first thing that pops out is the substantial jump between RB1 and any other RB Tiers. We can observe a huge difference in PPG by just looking at where the peaks are located on the x-axis. The patterns we observe for WR and TE are similar to each other. We can see that the fdifferences between the first two tiers are fairly similar, and that the next two are a lot more removed.

To showcase these in a better way, let’s look a a bar plot for the previous grid.

Now we can really put into context the differences that we had talked about previously. Is 20% difference in PPG between QB1 and QB2 substantial? The answer is clearly no. Looking at the RB position, we can clearly see that there is a massive jump between RB1 to RB2 - the biggets jump across all position.

While WR only sees a 27% jump, we can see that the difference is almost double around WR3 and an even steeper cut off at WR4. We observe similar jumps at TE as well. In terms of variance, we can see that QB2 are less consistent than QB1, but then again, have a relatively similar PPG. Contrast that with RB2, WR2 and TE2 - we can observe much lower PPG, but more consistent.

Point Difference Quantification

All together, it means that there is a much smaller gap across tiers at the QB position than there is across other positions. So it is essential to have a RB1, WR1 and TE1 ahead of getting a QB1, for the simple reason that a QB2/3 will perform very similarly, but the same cannot be said about second tier RB, WR and TE. This is clearly demonstrated in the raw point differential shown below - which is simply the previous grid with the raw points instead of the percentages.

Another way to quantify the point differential, is looking at a draft hypothetical. Let’s look at a draft choice you can make in the 4th round and 7th round. Here are two options you can make:

1. Draft a QB1 in Round 4 and RB4 in Round 7
2. Draft a RB2 in Round 4 and QB3 in Round 7

Let’s use the average PPG and the deviation for each of these positional tiers to calculate the average, minimum and maximum point range for the positions drafted. Here are those ranges:

## [1] "Quarter Back Tiers"

##   Tier Deviation  Average  Minimum  Maximum
## 1    3  7.356099 18.96103 11.60494 26.31713
## 4    4  5.808217 16.58973 10.78151 22.39795
## 5    1  7.843288 22.60156 14.75827 30.44485
## 7    2 10.912020 19.22800  8.31598 30.14002

## [1] "Running Back Tiers"

##   Tier Deviation   Average    Minimum  Maximum
## 2    4  6.851612  7.404016  0.5524045 14.25563
## 4    2  8.275759 11.689362  3.4136024 19.96512
## 5    3  4.852324  8.067485  3.2151608 12.91981
## 6    1  9.162231 19.573214 10.4109829 28.73545

Using that information, we can calculate a range for each of the scenarios above:

##    Scenario  Average      Min      Max
## 1 QB1 + RB4 30.00558 15.31068 44.70048
## 2 RB2 + QB3 30.65040 15.01854 46.28225

## [1] "Differential between scenario 2 and scenario 1: "

##     Average        Min      Max
## 2 0.6448176 -0.2921412 1.581776

So the difference here is not big. We see a slightly higher average for the second scenario and higher ceiling, despite a lower floor.

The big difference, however, comes from potential of jumping tiers. For example, let’s look at players like Aaron Jones and Chris Carson. Currently two RB2s but with the potential of jumping tiers to RB1 next season. That is ultimately why you pick a RB2 in those rounds, as opposed to a QB1 - for the upside that your RB may have, if they jump tier. If they don’t, as we’ve seen above, you break even.

Let’s now assume that in Scenario 2, the RB2 picked in the 4th jumps tiers and becomes an RB1. Likewise, let’s assume that the RB4 in Scenario one, jumps a tier, and becomes an RB3. Here’s the point differential:

##    Scenario  Average      Min      Max
## 1 QB1 + RB3 30.66905 17.97344 43.36466
## 2 RB1 + QB3 38.53425 22.01592 55.05258

## [1] "Differential between scenario 2 and scenario 1: "

##    Average      Min      Max
## 2 7.865202 4.042483 11.68792

We observe roughly about an increase in 8 PPG on average, and a higher ceiling and floor for Scenario 2. So we can clearly see the huge incremental gain of the RB2 with RB1 potential, which also highlights the very insignificant difference between the QB1 and QB3. That is to say that if the difference in QB3 and QB1 was as big as that of RB2 and RB1, then we would have see little to no gain in the last scenario.

We in fact observe something very similar across all the positions.

## [1] "Scenario 1 and 2 (No Potential): QB1 + RB4/WR4/TE4 or QB3 + RB2/WR2/TE2"

##      Average    Minimum  Maximum
## RB 0.6448176 -0.2921412 1.581776
## WR 1.6996669 -2.1990858 5.598420
## TE 0.9742349 -2.0325684 3.981038

## [1] "Scenario 1 and 2 (With Potential): QB1 + RB3/WR3/TE3 or QB3 + RB1/WR1/TE1"

##     Average    Minimum   Maximum
## RB 7.865202  4.0424831 11.687920
## WR 1.420693 -1.2404319  4.081817
## TE 3.656950 -0.4689832  7.782883

In the no potential scenario, where we evaluate purely the tiers without any jump in potentials, we still see a small positive benefit to picking a Tier 2 RB/WR/TE (scenario 2.) We also observe a higher ceiling across all positions for scenario 2, with however a lower floor.

In the case where the Tier 2 RB/WR/TE picked in Scenario 2 jump to Tier 1, and the Tier 4 RB/WR/TE in Scenario 1 jump to Tier 3, we see similar results with even bigger gains and higher ceilings. Note that aside from the RB position, Scenario 1 retains the highest floor.

Conclusion

We have now evaluated the positional values of QB, RB, WR and TE in Fantasy football. Despite averaging more PPG, we have seen why QB are not as valuable as it may seem, and the importance of gettting a RB1, WR1 and TE1 ahead of a QB1. We have also observed the benefits of betting on the potentials of RB2, WR2 and TE2 over QB1 while drafting.

Lastly we have also covered the steep drop off between Running Back Tiers, which makes it the most valuable position in Fantasy Football.

Next up in this series, we will talk about when should QB in fact be drafted, on a case to case basis. Essentally, we will atttemps to validate the ADP of certain QBs and evaluate other alternatives. We will also take a look at the paper thin TE class and when to draft one of the top tier TE.