NFL
How Well Did High-Ownership Players Perform for Daily Fantasy Football in 2019?
DFS seemed to be a bit sharper in 2019 than it has been in years past. But how often did the chalk fail, and what conditions most often led to flops?

When Christian Kirk blew up in Week 10, it felt like all hope for DFS was lost.

Despite racking up a whopping eight yards in his previous game, Kirk was on 27.56% of rosters in the FanDuel Sunday Million. That means when he hung 34.8 FanDuel points on the board with a salary of $5,700, if you didn't have him, your lineups were dead in the water.

In the past, a bad performance like the one Kirk had in Week 9 would have kept him on the fringes of DFS until he had a big game we could point-chase. But with analysis and data being more broadly available than in the past, plenty of people rightly recognized that he was a great play in a potential shootout against the Tampa Bay Buccaneers' porous secondary.

Daily fantasy football is harder now than it was in the past, and this point seems to be beyond dispute. Thanks to the wealth of analysis out there, good plays are being pinpointed more often, and there are fewer non-viable lineups we can exploit. That's going to naturally make cashing -- much less hitting a top-end score -- even more difficult.

That doesn't mean the edge is completely gone.

Although there were certainly the Kirk examples from this past year, there were also plenty of instances in which the chalk didn't hit. Not every decision the public made was a good one. Those misses not only prove to us that DFS is still beatable, but they can give us a template for what to look for in identifying future whiffs, as well.

That's what we're going to try to do here today. We're going to look back at what popular players in the Sunday Million did during 2019, what commonalities we can find in the misses, and what it means for 2020. If we can pinpoint areas where the public is missing, it should allow us to continue to profit despite a game that is tightening things up rapidly.

DFS is harder now than it used to be, and we shouldn't expect box-score numbers to chase people away the way they did in the past. But what can we learn from looking at popular plays from 2019? Let's check it out.

Positional Hit Rates

Before we dive into the anatomy of a bust at each position, we should look at how often each position hit this year so we know where we need to deviate from the chalk on a weekly basis.

Throughout this piece, we'll be referencing the "popular" plays at each position on a weekly basis. We'll define that as the three most popular quarterbacks, tight ends, and defenses on a given slate. We'll select the top six running backs from each week along with the top eight receivers (given that roster requirements are different at those two positions than the others). Whenever we reference the chalk, that's the group we're referring to.

Additionally, in order to quantify what a "bust" looks like, we need some sort of baseline. This baseline will vary by position because the scoring distribution is different at defense than it is at tight end.

As such, we'll give each position a baseline value, with "value" being how many points were scored per $1,000 worth of salary. The baseline was calculated by looking at the average value of all players at that position who were on at least 1.0% of rosters in the FanDuel Sunday Million for a given week. This helps weed out players without a role and those who were injured who would drag that bar down. Here's the baseline value for each position this past year.

Position Baseline Value
Quarterback 2.48
Running Back 1.91
Wide Receiver 1.65
Tight End 1.49
Defense 1.88


The translation of this is that if a running back provided 1.92 FanDuel points for every $1,000 of salary, he would have paid off in that given week. This is how we'll be defining the "hits" and "busts" at each position.

Clearly, these numbers aren't very big, which means -- in theory -- it should be pretty easy for players to hit their baseline value. That wasn't the case. Additionally, merely hitting value doesn't mean that player was a good investment, so we should be a bit more wary of popular players than even the numbers here would indicate.

Even with a low baseline and abundant information, there were still a lot of busts at every position this past year. Here's how often popular players at each position hit their baseline value in 2019.

Position Percentage to Hit Value
Quarterback 58.82%
Running Back 55.88%
Wide Receiver 47.06%
Tight End 56.86%
Defense 54.90%


No position hit baseline value even 60% of the time, and wide receivers did so less than half the time. Football's a volatile sport, so we're never going to be devoid of misses, but to say that we're suddenly perfect at making these selections would be misleading.

What's even more interesting is that these numbers were fully in line with what we've seen in past years. This is the fourth time we've done this review of popular players in the FanDuel Sunday Million, so we can look back at those to see if we spot a trend. If hit rates were higher this year than in the past, then it would back up the claim that things were getting even tougher. That's just not how it has played out in reality.

The hit rate on defense this year tied the highest it has been in our four years. But it was second-highest for quarterback and third-highest at running back, wide receiver, and tight end. The chalk didn't hit more often this year than it has in the past. Rather, this seemed to be just another year.

That doesn't mean the narrative that DFS is getting tougher is false. Roster construction has gotten better, and there are fewer completely irrelevant lineups per contest. But there are still plenty of whiffs to exploit at each position.

Based on that chart, an easy route for finding an edge in DFS is being superbly skeptical of popular wide receivers. Even though you have the examples like the aforementioned Kirk one this year, only 47.06% of popular receivers this season returned even 1.65x value. If you up that number to at least 2x value, it drops to just 38.24% of popular wide receivers.

Hitting Value Thresholds At Least 1x At Least 2x At Least 3x At Least 4x
Running Back 84.31% 51.96% 24.51% 5.88%
Wide Receiver 74.26% 38.24% 17.65% 6.62%
Tight End 82.35% 35.29% 13.73% 0.00%


In other words, if you were to blindly back off of every popular wide receiver for the entire year, you'd avoid a disappointing investment about 61.76% of the time. Of course, the odds the alternative you select busts are also pretty high, so fading the chalk by itself is not a winning strategy. Still, there are clear incentives to deviate at wide receiver.

That same chart can be used as a justification for being skeptical of popular tight ends, as well. Although they did hit baseline value at a higher rate than every position except for quarterback, it rarely burned you if you skipped over a popular tight end. Basically, they hit their baseline value often because their baseline value was almost non-existent. A $6,000 tight end would need 8.9 FanDuel points to hit the baseline, and you generally need more than that to have an elite lineup.

This is not meant to be taken as a sweeping statement that you must avoid all popular wide receivers and tight ends. As you'll see in a bit, there are certain conditions under which even those positions can thrive. But if you are even the slightest bit wary of a tight end or wideout who projects to be popular, history says you've got good reason to look elsewhere.

One thing the Kirk example does show is that DFS players have gotten better about box-score scouting. They're not just selecting players based on who had a good outing the previous week.

There was still a good amount of point-chasing, though, as is to be expected. The table below shows how players performed when they were popular versus how many points they scored the week before they were popular. It also shows the percentage of players who scored more points the week they were popular than they did the week before.

FanDuel Points When Popular Previous Week Percentage to Improve
Quarterback 22.3 24.0 39.58%
Running Back 16.7 15.2 56.25%
Wide Receiver 13.2 16.1 37.50%
Tight End 11.1 12.5 39.58%
Defense 10.6 8.9 64.58%


At defense and running back, we generally did a good job of avoiding the shiny objects and picking objectively good plays. Everywhere else, point-chasing was the name of the game.

Not only did we still sprint after big outputs this year, we actually did it worse than last year. In 2018, wide receiver was the only position where players improved when popular less than 45% of the time. This year, wide receiver, quarterback, and tight end all failed to even hit 40%. If we had a weakness as DFS players this year, it was failing to be skeptical of players coming off big performances.

Just as with how we don't need to avoid all popular players, though, we also don't need to avoid all players coming off a big game. That's especially true if the player experienced a role change which led to the big game and that role change is not fully accounted for in their next-week salary. There are also some other conditions under which we should be willing to trust popular plays in DFS.

What are those conditions? Let's dig into that now, starting with the quarterbacks.

Quarterback

In order to have an idea of how to handle a popular player in DFS, you first must know who will be popular. Because of that, for each position, we'll also go through what the average popular player at that position looked like in 2019. Here's the breakdown of the popular quarterbacks, again looking at just the three who wound up on the most Sunday Million rosters each week.

Popular Quarterbacks Averages
Home 49.02%
Total 48.2
Spread -6.4
FanDuel Points 22.25
Previous Week 24.02
Wind Speed 5.6
Value 2.73
Baseline Value 2.48
Percentage to Hit Value 58.82%


It may be a bit surprising to see that more than half of all popular quarterbacks were guys who were on the road. But as we saw in our breakdown of perfect lineups from this year, road quarterbacks had a ton of big games. That may not have been a misstep by the public, even if it was a bit surprising.

One way we can identify conditions that are conducive to a bust is by comparing what a "bust" looked like compared to the "hits." If there's a major difference between the two groups, it might give us a signal about spots we'll want to avoid.

That's what this table shows. It compares the average conditions for a popular quarterback who managed to exceed the baseline value versus the conditions for a popular quarterback who did not.

Popular QBs Hits Busts
Average Salary $8,147 $8,095
Home 53.33% 42.86%
Total 48.3 48.1
Spread -7.2 -5.3
Wind Speed 5.3 6.0
Previous Point Total 23.5 25.1


The salaries were relatively even, so we don't need to back off of a popular quarterback based on that alone. There were some differences, though.

The big one is that 53.33% of the quarterbacks who exceeded value were at home compared to only 42.86% of those who busted. The hits were also more heavily favored than those who fell short.

In the investigation of perfect lineups, we saw the value of targeting close games, and quarterbacks who were on the road seemed to show a lot of upside. The data above does not dispute those facts. But if your sole purpose is to avoid a bust, we might want to be wary of the conditions that led to the most upside.

The other split between the hits and the busts is what they did the previous week. Those who busted averaged 25.1 FanDuel points the week before while those who were hits averaged 23.5. In other words, point-chasing often led us down a bad path. When we rostered a quarterback because of his game that upcoming weekend, things turned out a lot better.

This gives us somewhat of a blueprint for what risky chalk at quarterback looks like. They're more likely to be on the road with a tighter spread, and a big game the previous week could just further inflate the public's interest. We don't have to avoid quarterbacks who fit in those buckets, but we should at least give additional thought to looking elsewhere as the bust potential seems meatier.

The wind gap between the hits and busts isn't all that large at 0.7 miles per hour, but wind did still seem to signal when a potential bad output was on the horizon. Only 44.44% of the popular quarterbacks playing in winds of 11-plus miles per hour managed to hit the baseline value.

Wind Speeds Percentage to Hit Value
0 to 10 mph 61.90%
11-Plus mph 44.44%


A quarterback in blustery conditions can still pay off. But of the nine popular quarterbacks playing in wind speeds of 11 miles per hour or higher, only two returned more than 2.60x value, a mark hit at a 54.76% clip by quarterbacks playing in lower winds. Suboptimal weather is another reason to potentially deviate from where the public is siding.

One thing to keep in mind when considering quarterbacks is that popularity here tends to be much flatter than other positions. Across 2019, a quarterback was never on more than 30% of rosters in the Sunday Million, and only seven quarterbacks topped the 20% mark. This means the upsides of fading what is popular there are smaller than what they are at other positions. Still, we should have a solid template now for what leads to risky chalk, allowing us to decide how we want to handle those quarterbacks we do expect to be popular.

Running Back

As you saw earlier, the chalk at running back was pretty good this year. They hit at a high rate, and they often provided respectably high-end scores. There's still some things we can learn here, though.

First, here's what your average popular running back looked like in 2019.

Popular Running Backs Averages
Home 53.92%
Total 45.6
Spread -3.6
FanDuel Points 16.7
Previous Week 15.2
Wind Speed 5.9
Value 2.21
Baseline Value 1.91
Percentage to Hit Value 55.88%


We were more likely to get our running backs at home than we were our quarterbacks. That's interesting given that popular road running backs actually hit their baseline value at a higher rate than those at home.

Popular Running Backs Percentage to Hit Value
Home 52.73%
Away 59.57%


This does not mean that running backs at home are more likely to bust than those on the road. It's moreso to say we don't need to view a popular running back who is playing on the road as being a trap.

One of the more surprising pieces of data to come out of the perfect lineup piece was that heavily favored running backs didn't wind up making the cut all that often. The hypothesis was that they would get fewer targets later in games, putting a bit of a lid on their upside.

That would impact the ceiling of these players and keep them out of perfect lineups. But in theory, you wouldn't think that would prohibit them from hitting our baseline value. The data here runs counter to that thought.

Of our 102 popular running backs, 22 came from games where their teams were favored by 10 or more points. Only 45.45% of that sample wound up hitting value.

Spread Percentage to Hit Value
Favored by 10 or More 45.45%
Favored by 5 to 9.5 57.14%
Favored by Less Than 5 63.33%
Underdogs by Less Than 5 56.25%
Underdogs by 5 or More 53.85%


From a hit-rate perspective, the heavily favored running backs were the worst in our sample. It's easy to understand why people would want to target a running back on a team that is a massive favorite as it would seemingly lead to good late-game volume, but it wasn't a successful formula this year.

Once we look at these backs on a more micro level, you can start to see why these picks flopped. A lot of the trap plays were guys who don't catch many passes.

Tevin Coleman failed twice while popular as a heavy favorite, and he doesn't catch many passes. Neither do Raheem Mostert, Josh Jacobs, Marlon Mack, and Sony Michel, all of whom made this list. The only truly one-dimensional back who paid off while popular and heavily favored was Derrick Henry in Week 17.

When a team is heavily favored with a rush-first back, people are going to target them, thinking that the increased volume late in the game will make that player more likely to pay off. That very well could happen. But as we've seen both here and with the perfect lineups, you need to go nuts as a rusher to pay off without catching passes, and players in this mold fail regularly. If you get the sense that a back like this is trending toward being popular, that might be a sign for you to jump off the train and look elsewhere.

Another key parallel between what we saw with the perfect lineups and what we can see here is the value of viable lower-cost backs. There were only 12 running backs who were popular with a salary lower than $6,000, but eight of them (66.67%) wound up paying off their salary.

When looking at the perfect lineups, we saw that most of the successful options in this range were players who experienced role changes and caught passes. All of the cheaper backs who came through when popular fit this mold, as well. So, if you can find cheap volume that includes work in the passing game, that's a spot where you should feel comfortable swallowing the chalk.

Everything at this position revolves around catching passes. If players don't do it, and they project to be popular, we should proceed with caution. If the potential chalk does get targets, we can feel pretty safe about investing. It may be over-simplifying to view things this way, but multiple sources of data signal that it's a key in identifying both the boons and the banes of the position.

Wide Receiver

Now we get to the black sheep of the positional groups. Wide receivers fail -- a lot -- meaning identifying the chalk should be a key aspect of our process. Here's what the average popular receiver looked like this year.

Popular Wide Receivers Averages
Home 58.09%
Total 47.4
Spread -2.8
FanDuel Points 13.2
Previous Week 16.1
Wind Speed 5.7
Value 1.89
Baseline Value 1.65
Percentage to Hit Value 47.06%


We talked about the point-chasing angle before, but it's worth repeating quickly here: if a wide receiver is coming off a big game, think twice before investing. If they pass the sniff test, feel free to fire away. Just be sure to give them extra thought first.

When you look at the wide receivers who hit versus those who busted, there's a pretty major gap in the spread between the two. Those who hit were favored by an average of 3.3 points, and those who busted were favored by 2.4 points. That could lead you to think that we want to avoid wide receivers who aren't heavily favored. It's a bit more broad than that.

The main reason that the spread was smaller for the busts was that popular underdog receivers had a really bad hit rate relative to other options.

Spread Percentage to Hit Value
Favored by 10 or More 52.63%
Favored by 5 to 9.5 48.15%
Favored by Less Than 5 51.06%
Underdogs by Less Than 5 39.39%
Underdogs by 5 or More 40.00%


It didn't matter if you were favored by 2 or by 12. The hit rate there was about the same. It was moreso about being wary of those on teams that weren't favored at all.

Of our 136 popular wide receivers, 43 (31.62%) entered the game on teams that were underdogs. Those receivers averaged 16.0 FanDuel points the previous week but dropped all the way to 11.9 the week they were popular, and only 17 of them exceeded baseline value. They were largely bad investments.

Now, with that said, there were some big-time hits in this group, including Kirk, whose massive day came as a 5.5-point underdog. There is some upside in a team that's an underdog if the game shoots out, and as mentioned in the perfect lineup piece, underdog receivers are solid if the spread is tight. They're just not safe.

This means we face a difficult dilemma whenever a wide receiver who is an underdog projects to be popular. We know that the odds they bust are high, but we also know the upside is there for a day like Kirk had in Week 10. It doesn't make deciding whether or not to swallow that chalk very easy.

When we focus on just the popular underdog wide receivers, we do see two differences between those who hit and those who busted. Those who hit had salaries that were about $200 cheaper on average, and the totals for their games was more than a point higher.

If we look at just wide receivers who fit these criteria (underdogs, salary under $7,000, in a game with a total of 50 or higher), our list is narrowed to just nine players. But, of those nine, five hit their baseline, and three of them scored 21 or more points (including Kirk). We don't want to draw too many conclusions from such a small sample, but it does seem players in this group are at least less likely to crash and burn than other underdogs.

The total conversation is relevant for the group as a whole, too. Popular wide receivers in games with a total lower than 45 hit value just 40.63% of the time compared to 49.04% for those in projected higher-scoring games. Popular wideouts still busted more often than other positions when the game was high-scoring, but it was at least less gruesome than for the lower-scoring games.

Overall, this is a position where you largely want to avoid someone who projects to be popular because it is a volatile spot. However, we should be especially skeptical if the wide receiver is on a team that's an underdog or in a game with a lower total. Favored wide receivers in games with a total of 45 or higher hit 52.86% of the time, which is at least better than what you get in other splits, even if it does still leave plenty to be desired.

Tight End

As mentioned before, tight end is a tough position to diagnose. A lot of players there managed to hit the baseline value, but the baseline value was so low that this didn't necessarily mean the chalk was worth investing in. It's going to make this conversation a bit more nuanced than with the others.

First, here's what the average popular tight end looked like this year.

Popular Tight Ends Averages
Home 54.90%
Total 47.7
Spread -1.3
FanDuel Points 11.1
Previous Week 12.5
Wind Speed 6.6
Value 1.76
Baseline Value 1.49
Percentage to Hit Value 58.86%


On average, a popular tight end produced 18.12% more value than the baseline at the position. That was actually the second-best mark, trailing only defense and special teams. So maybe popular tight ends weren't actually the issue; it's moreso that the position as a whole is diseased.

One interesting note for tight end is that the popular ones were almost always spendy. Of the 51 popular tight ends this year, only 11 carried a salary below $6,000. But when the public did decide to trust a cheaper tight end, they tended to invest wisely.

Salary Percentage to Hit Value
$7,000 or Higher 50.00%
$6,000 to $6,900 52.94%
Lower Than $6,000 72.73%


Again, the sample for that bottom group is only 11 players. But having 8 of those 11 hit value is interesting.

If you want a bit of a larger sample, we can increase that salary cutoff to $6,500, which gives us 22 at that number or higher and 29 below it. There, the cheaper guys still take the cake, though the hit rate does take a slight step back.

Salary Percentage to Hit Value
$6,500 or Higher 50.00%
Lower Than $6,500 62.07%


Not only did the expensive tight ends fail to hit value, but they also just didn't score a lot of points straight up.

Of the 22 popular tight ends with a salary of $6,500 or higher, only one managed to score more than 19 FanDuel points. That was Travis Kelce at 20.2 in Week 2. There were five cheaper tight ends who topped that mark while popular this past season.

It seems pretty likely that this was just negative variance for the guys in this group. Zach Ertz and George Kittle both had sort of weird seasons from a fantasy perspective, and Darren Waller's usage varied wildly based on who was active around him. So we could very well see the more expensive tight ends rebound next year. Still, it's important to keep in mind that even the studs at this position can be risky investments when popular.

A big overlap between wide receiver and tight end is that tight ends in projected lower-scoring games were more likely to bust than those with higher totals on their side.

Total Percentage to Hit Value
50 or Higher 61.54%
45 to 49.5 57.69%
Lower than 45 50.00%


The sample for popular tight ends with a total lower than 45 was only 12, but there were 26 tight ends with a total between 45 and 49.5. That's a decent number, and the hit rate there was solid. So we can add a low total to the list of things that makes a chalky tight end higher-risk.

Let's combine those two points together. The best chalk at tight end seems to be a cheaper guy in a game with a respectable total. What's the hit rate when we check both of those boxes?

Overall, there were 21 popular tight ends this year who had a salary lower than $6,500 in a game with a total of at least 45 points. Of those 21, a whopping 14 (66.67%) hit their baseline value, which is one of the better splits for any position in 2019. Four of the seven popular tight ends who provided at least 3.0x value fit into this bucket, so as far as swallowing chalk goes, it doesn't get a whole lot better than that.

Even with the lack of top-end performances, tight end doesn't seem to be a major trap position for popular players. This is especially true if that player is cheaper and in a high-total game, so going the popular route there isn't necessarily a bad idea. We should, though, be a bit wary of tight ends in lower-scoring games and those who come with prohibitive salaries. There, the upside seems a bit more limited, allowing us to potentially turn our attention elsewhere.

Defense and Special Teams

As mentioned in the section on tight ends, no position provided more relative value when popular than defense and special teams. We were generally pretty good with this position in 2019. Here's what a popular defense looked like.

Popular DSTs Averages
Home 56.86%
Total 42.5
Spread -7.8
FanDuel Points 10.6
Previous Week 8.9
Wind Speed 7.7
Value 2.38
Baseline Value 1.88
Percentage to Hit Value 54.90%


In general, popular defenses were solid this year, so we don't need to be as wary of this position as we are others. But let's divide things up a bit more to see if there are any potential traps.

One of those traps was mid-priced defenses. There were 18 different popular defenses with a salary of $5,000 or higher this year, and there were 12 with a salary below $4,000. Both those groups performed well when popular. The mid-range defenses weren't so lucky.

Salary Percentage to Hit Value
$5,000 or Higher 61.11%
$4,000 to $4,900 42.86%
Lower than $4,000 66.67%


It's pretty easy to concoct a story for why this would be the case.

The expensive defenses are largely expensive for a reason. They're probably high-quality defenses that are in favorable spots. It makes sense why they would come through often.

For the cheaper defenses, it's easier for them to pay off. Value is based on salary, and when you have a lower salary, the threshold for coming through is lower. That's the advantage they have over the mid-tier options.

The struggles for that middle tier were even apparent when those defenses were in spots we search for when looking for a defense for DFS. Seven of the defenses with a salary between $4,000 and $4,900 were double-digit favorites, but only three of those seven hit baseline value. Only 4 of 11 defenses favored by 7.5 or more points paid off.

This could very well mean that the poor performance by mid-salaried defenses was just variance, and that's the most likely explanation. We don't have to avoid a popular defense in this range, especially if they fit with our process. Still, there's a narrative explanation for why this would happen, so it might not be a bad idea to take an extended look at defenses in this range when they project to be popular.

Looking elsewhere, we see something similar with defenses as what we saw with quarterbacks: hunting for upside is a different pursuit than looking for players who will simply "pay off."

In the perfect lineup piece, we saw that it was decently common for defenses from games with high totals to wind up in the perfect lineup. It wasn't as common as other positions, but it still happened.

The reasoning was that shootouts involve more drop backs than other games, and more drop backs means more chances for sacks, picks, and touchdowns. When you want a big score, that's what you need.

When you're instead focused on floor, then we see the lower-scoring games take on a more favorable glow. Of the 13 popular defenses in games with a total of 45 or higher, only 3 hit baseline value.

Total Percentage to Hit Value
45 or Higher 23.08%
40 to 44.5 75.00%
Lower Than 40 50.00%


This should not be taken to mean that we should avoid all defenses in games with high totals; the perfect lineup piece showed that was not true. Instead, it means we have pivot opportunities if one of those defenses projects to be popular.

Defense is a volatile position where the tournament-winning scores are going to come via teams that score touchdowns. We can increase our odds of getting those big outputs by targeting defenses that will face a bunch of drop backs, but predicting who will score the touchdowns is still a tough endeavor.

This is why pivoting at the position is still a good idea, even if it is one of the positions we are better at predicting. The games that give us big upside are also the ones that can lead to a lot of flops. As a result, we should still look at defenses in games with higher totals due to the expectation of an increased number of drop backs, but we should simultaneously look to jump ship if one of those defenses projects to be popular. The two lines of thought are not mutually exclusive.

To wrap it all up, popular defenses are not bad investments, and if you love a defense, you can use it even if you know it'll be popular. That's especially true if it's a super cheap defense, as we saw. But mid-range defenses and those in projected higher-scoring games are still capable of flopping, which presents us with an opportunity to pivot. As long as these defenses fit your process well, you can still go there, but even at a largely predictable position, there are still advantages to hunting for a more overlooked unit.

Related News

Which Quarterback in the 2020 NFL Draft Is Statistically Superior?

Jim Sannes  --  Feb 13th, 2020

Analyzing Trends in FanDuel's 2019 Perfect Daily Fantasy NFL Lineups

Jim Sannes  --  Feb 13th, 2020

Fantasy Football: The Late-Round Podcast, Sell Candidates

JJ Zachariason  --  Feb 13th, 2020