Introducing NFL Correlation Values

I was reading Jonathan Bales’ new fantasy football book (look for it soon!) and loved a part he had on correlations among DFS positions. The great minds over at Rotoviz created a huge correlations chart, which Jonathan then used to break down the positions best correlated. The goal of DFS and especially tournaments, of course, is to maximize the upside of your lineup – correlating positions is a great way to do that.

If you don’t know what I’m talking about, positions that are correlated are tied together (the exact amount can be seen in the Rotoviz chart with 1 being a perfect correlation). If a wide receiver goes off for three touchdowns, it’s highly likely that his quarterback is a big part of that. If you paired those two together, you’re getting points on both sides – from the passing and receiving side of things.

Anyway, the biggest correlations that Rotoviz found and Jonathan talked about are as follows:

WR2 and QB on the same team (0.39)

WR3 and QB on the same team (0.35)

QB and K on the same team (0.35)

K and D on the same team (0.35)

WR1 and QB on the same team (0.34)

This article is going to introduce the idea of “correlation values” that I’m going to make into a weekly article this upcoming NFL season. The basis is to see the value of pairing those two positions in either your DraftKings or FanDuel lineup. The idea is to try to get “free” upside — if you know a QB and his kicker are correlated positions and you’re trying to decide between kickers, taking the one paired with your QB doesn’t cost any extra salary necessarily, but it does add upside to your lineup.

Essentially what I’m doing is looking at the salaries of a position and comparing it to the salary of the correlated position – for example, if you’re taking Eli Manning (the ninth-highest salaried QB on FanDuel in Week 1) and are looking for a kicker, you can see that the Giants kicker, Josh Brown, is the 28th-highest salaried kicker (minimum salary). If Eli is the QB of your lineup, you’re not only getting the kicker at a discount, but you’re also getting the one correlated with Eli’s production. If Eli goes off, there’s a good chance Brown will hit value as well.

DraftKings: Quarterback and Wide Receiver Correlation Values

Let’s dive into one, as I think seeing the chart will make things easier to understand than my explanation above. Here’s the chart for DraftKings (sorted alphabetically), showing the correlations between QBs and their wide receivers.

correlation week 1 5
 

The way I would read this chart is from right to left. In other words, if you have a team’s QB, what’s the correlation value of taking one of his wide receivers? Or put another way — if you have Drew Brees in your lineup, what’s the value of taking his WR1 (Marques Colston)? Pretty high, as we see: Colston is a very cheap first wide receiver.

Quick note: I know, Brandin Cooks. So, right now I’m going off FootballGuys.com’s depth charts, and I’ll obviously update these correlation charts for Week 1 as we get through the preseason and know the depth charts better. For now, I wanted to go off a list for consistency, but it’ll change close to the season.

Also, we can invert the chart to see the opposite — the value of taking the QB if you have the wide receiver. But for the sake of space, I’m not going to post the inverted chart for receivers. Just invert the colors in your mind – the heavy reds will be heavy greens.

Of course, this chart is the most subject to context – although Tom Brady will have a cheap WR1 most weeks, we know that Rob Gronkowski is essentially his top receiver. I think it could be even more useful as the inverse chart – if you think DeAndre Hopkins is going to go off Week 1, you can get cheap correlation value by taking Brian Hoyer, if he ends up starting.

Among the high-priced wide receivers, you can see some correlation value with the Matt Ryan – Julio Jones and Eli Manning – Odell Beckham Jr. combinations. Both are elite wide receivers and the price of their QBs makes it so that in tournaments, if you think either one is going to explode for multiple touchdowns and lots of DK points, pairing them with their cheaper QB is an almost-must.

The highest correlation value among wide receivers, although they’re all pretty high, is with WR2s. The best values in Week 1 in this regard is Cam Newton and Devin Funchess (almost min-salary) and Russell Wilson and Jermaine Kearse. Of course, the addition of Jimmy Graham in Seattle is important context, but if you expect Newton or Wilson to have big games in the air, it’s fairly likely that those WR2s will hit or exceed value.

FanDuel: Quarterback and Kicker Correlation Value

correlation week 1 1
 

For this chart, you’re probably saying, “Alright, what’s the big deal? The best QBs are all at the top.” Correct. And I think this is informative – this won’t be the case when we look at the correlations among other positions for one specific reason: the pricing of kickers. The difference between the highest-priced QB and the lowest one ($3,400) is drastically more than the difference between the highest and lowest kicker ($700). For a correlation like a QB and his first wide receiver, you’re sometimes going to have to pay up a lot to get that correlation (think Aaron Rodgers and Jordy Nelson). However, for a quarterback and kicker, paying up for Rodgers’ kicker only costs you $600 above minimum salary and barely above the mean salary among kickers.

For Week 1, two of the better values are Josh Brown of the Giants and Graham Gano of the Panthers, $4,500 and $4,600 respectively. If you’re rolling with either Eli Manning or Cam Newton in a tournament lineup, I think it makes a lot of sense to get essentially free – opportunity-cost free, I mean, as you have to spend at least $4,500 for a kicker anyway so it might as well provide upside – correlation value in your lineup.

FanDuel: Defense and Kicker Correlation Value

correlation week 1 2
 

This is obviously going to be less drastic as both positions – kickers and defenses – are both generally priced close to the minimum salary. However, you can still see a big difference between some teams, like the Ravens, Broncos, and Falcons. Again, the way I would interpret this chart is from the first column to the second column. In other words, what is the value of taking the Ravens defense if I already have their kicker? Again, we can invert it to see the opposite – what’s the value of taking the Panthers’ kicker (Gano again) if I already have their defense?

correlation week 1 3
 

The reason I didn’t do this for the QB-K chart above is that I seriously doubt that anyone would choose their QB based on getting correlation with their kicker. Kicker and defense are generally the last pieces to the lineup puzzle and remaining salary is usually the biggest factor in those choices. Again, that’s how I would approach this article – if you’re unsure of a position (maybe it’s your kicker), I’d check to see if it’s possibly worthwhile to pair your QB and K or D/ST and K and get that correlation upside boost.

FanDuel: Quarterback and Wide Receiver Correlation Values

correlation week 1 4
 

We’re going to see some of the same values as the DraftKings chart, but sometimes more or less pronounced. A lot of the pricing among the sites are similar – Rodgers is the highest QB on both sites – but minor differences can lead to a lot of predictive value, as you’ll see with our Bargain Rating this upcoming NFL season.

Again, this is more of an introduction article and the values will change over the next month as we learn more about depth charts and expected volume. However, I think it could be a cool article to check during the season, especially with the kicker and defense values – those are so hard to pick anyway, and any time you can get essentially free upside in your lineup, it’s incredibly valuable.

I was reading Jonathan Bales’ new fantasy football book (look for it soon!) and loved a part he had on correlations among DFS positions. The great minds over at Rotoviz created a huge correlations chart, which Jonathan then used to break down the positions best correlated. The goal of DFS and especially tournaments, of course, is to maximize the upside of your lineup – correlating positions is a great way to do that.

If you don’t know what I’m talking about, positions that are correlated are tied together (the exact amount can be seen in the Rotoviz chart with 1 being a perfect correlation). If a wide receiver goes off for three touchdowns, it’s highly likely that his quarterback is a big part of that. If you paired those two together, you’re getting points on both sides – from the passing and receiving side of things.

Anyway, the biggest correlations that Rotoviz found and Jonathan talked about are as follows:

WR2 and QB on the same team (0.39)

WR3 and QB on the same team (0.35)

QB and K on the same team (0.35)

K and D on the same team (0.35)

WR1 and QB on the same team (0.34)

This article is going to introduce the idea of “correlation values” that I’m going to make into a weekly article this upcoming NFL season. The basis is to see the value of pairing those two positions in either your DraftKings or FanDuel lineup. The idea is to try to get “free” upside — if you know a QB and his kicker are correlated positions and you’re trying to decide between kickers, taking the one paired with your QB doesn’t cost any extra salary necessarily, but it does add upside to your lineup.

Essentially what I’m doing is looking at the salaries of a position and comparing it to the salary of the correlated position – for example, if you’re taking Eli Manning (the ninth-highest salaried QB on FanDuel in Week 1) and are looking for a kicker, you can see that the Giants kicker, Josh Brown, is the 28th-highest salaried kicker (minimum salary). If Eli is the QB of your lineup, you’re not only getting the kicker at a discount, but you’re also getting the one correlated with Eli’s production. If Eli goes off, there’s a good chance Brown will hit value as well.

DraftKings: Quarterback and Wide Receiver Correlation Values

Let’s dive into one, as I think seeing the chart will make things easier to understand than my explanation above. Here’s the chart for DraftKings (sorted alphabetically), showing the correlations between QBs and their wide receivers.

correlation week 1 5
 

The way I would read this chart is from right to left. In other words, if you have a team’s QB, what’s the correlation value of taking one of his wide receivers? Or put another way — if you have Drew Brees in your lineup, what’s the value of taking his WR1 (Marques Colston)? Pretty high, as we see: Colston is a very cheap first wide receiver.

Quick note: I know, Brandin Cooks. So, right now I’m going off FootballGuys.com’s depth charts, and I’ll obviously update these correlation charts for Week 1 as we get through the preseason and know the depth charts better. For now, I wanted to go off a list for consistency, but it’ll change close to the season.

Also, we can invert the chart to see the opposite — the value of taking the QB if you have the wide receiver. But for the sake of space, I’m not going to post the inverted chart for receivers. Just invert the colors in your mind – the heavy reds will be heavy greens.

Of course, this chart is the most subject to context – although Tom Brady will have a cheap WR1 most weeks, we know that Rob Gronkowski is essentially his top receiver. I think it could be even more useful as the inverse chart – if you think DeAndre Hopkins is going to go off Week 1, you can get cheap correlation value by taking Brian Hoyer, if he ends up starting.

Among the high-priced wide receivers, you can see some correlation value with the Matt Ryan – Julio Jones and Eli Manning – Odell Beckham Jr. combinations. Both are elite wide receivers and the price of their QBs makes it so that in tournaments, if you think either one is going to explode for multiple touchdowns and lots of DK points, pairing them with their cheaper QB is an almost-must.

The highest correlation value among wide receivers, although they’re all pretty high, is with WR2s. The best values in Week 1 in this regard is Cam Newton and Devin Funchess (almost min-salary) and Russell Wilson and Jermaine Kearse. Of course, the addition of Jimmy Graham in Seattle is important context, but if you expect Newton or Wilson to have big games in the air, it’s fairly likely that those WR2s will hit or exceed value.

FanDuel: Quarterback and Kicker Correlation Value

correlation week 1 1
 

For this chart, you’re probably saying, “Alright, what’s the big deal? The best QBs are all at the top.” Correct. And I think this is informative – this won’t be the case when we look at the correlations among other positions for one specific reason: the pricing of kickers. The difference between the highest-priced QB and the lowest one ($3,400) is drastically more than the difference between the highest and lowest kicker ($700). For a correlation like a QB and his first wide receiver, you’re sometimes going to have to pay up a lot to get that correlation (think Aaron Rodgers and Jordy Nelson). However, for a quarterback and kicker, paying up for Rodgers’ kicker only costs you $600 above minimum salary and barely above the mean salary among kickers.

For Week 1, two of the better values are Josh Brown of the Giants and Graham Gano of the Panthers, $4,500 and $4,600 respectively. If you’re rolling with either Eli Manning or Cam Newton in a tournament lineup, I think it makes a lot of sense to get essentially free – opportunity-cost free, I mean, as you have to spend at least $4,500 for a kicker anyway so it might as well provide upside – correlation value in your lineup.

FanDuel: Defense and Kicker Correlation Value

correlation week 1 2
 

This is obviously going to be less drastic as both positions – kickers and defenses – are both generally priced close to the minimum salary. However, you can still see a big difference between some teams, like the Ravens, Broncos, and Falcons. Again, the way I would interpret this chart is from the first column to the second column. In other words, what is the value of taking the Ravens defense if I already have their kicker? Again, we can invert it to see the opposite – what’s the value of taking the Panthers’ kicker (Gano again) if I already have their defense?

correlation week 1 3
 

The reason I didn’t do this for the QB-K chart above is that I seriously doubt that anyone would choose their QB based on getting correlation with their kicker. Kicker and defense are generally the last pieces to the lineup puzzle and remaining salary is usually the biggest factor in those choices. Again, that’s how I would approach this article – if you’re unsure of a position (maybe it’s your kicker), I’d check to see if it’s possibly worthwhile to pair your QB and K or D/ST and K and get that correlation upside boost.

FanDuel: Quarterback and Wide Receiver Correlation Values

correlation week 1 4
 

We’re going to see some of the same values as the DraftKings chart, but sometimes more or less pronounced. A lot of the pricing among the sites are similar – Rodgers is the highest QB on both sites – but minor differences can lead to a lot of predictive value, as you’ll see with our Bargain Rating this upcoming NFL season.

Again, this is more of an introduction article and the values will change over the next month as we learn more about depth charts and expected volume. However, I think it could be a cool article to check during the season, especially with the kicker and defense values – those are so hard to pick anyway, and any time you can get essentially free upside in your lineup, it’s incredibly valuable.