When I was a kid, there were only two things that could get me riled up. The first was the assertion that dragons weren't real, which was an outright lie the government pedaled so that they could keep the majestic beasts all to themselves. Don't think I can't see through your deception.

The second was nap time. Naps were evil, and anybody who tried to push them upon me was simply doing the devil's work.

It wasn't anything about the nap itself that bothered me. It was what I could have been doing if I wasn't being forced to sleep. Every second with my head on the pillow was one where I could have had an adventure with my Spider-Man action figures. What if this was the play session where we finally took down the mob of rogue Barbies? If I were to miss out on that, the entirety of my life's work would be for naught.

Fast forward to now, and while I have grown to appreciate the blissful experience that is a long nap, the fear of missing out is still there. Except instead of worrying about dismantling a network of plastic dolls, I'm trying to avoid missing out on Julio Jones' next 200-yard game. I'm going to guess that I'm not alone in that fear.

We know that NFL players are volatile assets for daily fantasy football. We've seen the ills of using a highly-owned running back who gets game-scripted to death in the third quarter and sinks all of our lineups. But we've also seen the opposite end of the spectrum, jumping up and down as that chalky tight end hits paydirt for the second time. This brings about conflicting emotions as to whether or not we should use players when we know they'll be popular.

The logic behind using those players makes sense. If Odell Beckham is on 35% of all rosters, and he blows up with 150 yards and 2 touchdowns, every lineup that doesn't have him is pretty much dead in the water. It's onto the next week, assuming your bankroll can take the hit.

But using this logic also assumes the odds that the player goes off are 100%. But what if Beckham's odds of a hitting on a given week were just 75%? What if they were 50%? Maybe even lower? Then that's going to play a major role in our decision-making.

Knowing those odds in advance is no small task, and if we did know them, we'd clearly make better decisions on a week-to-week basis. While we can never have a definitive answer here, looking back at what has happened in the past can at least give us a blueprint for assessing these situations.

Thankfully, we've got the historic side covered. Each week here on numberFire, premium members can see ownership totals on FanDuel for the previous week's main slate. Using those, we can take each player who was popular -- we'll get into what is considered popular -- throughout the year, see how often those players met expectations, and have a rough estimate for the odds that a high-ownership player will hit going forward. If the odds are high that they'll hit, we can feel better about plugging them in even if the public is doing the same. If not, then our fears of missing out would seem misplaced, and we'd be wise to look elsewhere.

Today, we're going to look back at the 2017 season to see what we can learn from the performance of high-ownership players in NFL DFS. We'll be basing this on FanDuel's Sunday Million contests, which are a tournament format, the game type in which considering ownership is more important. We can also dig a bit further to see under which conditions players were more likely to bust or thrive. We'll do a full position-by-position breakdown in a second, but first, let's just start with some general takeaways.

Overall Takeaways

Before we get into the analysis, let's lay out a few of the parameters that will be key throughout our discussion.

Because FanDuel's rosters ask you select two running backs, three wide receivers, and one each at quarterback, tight end, defense, and kicker, what qualifies as "popular" at each position will be a bit different. As such, we'll be looking at the performance of the five most popular running backs, eight most popular wide receivers, and top three choices at each of the other positions each week to judge the quality of these choices. This roughly works out to include most of the players who carried at least 10% ownership in a given week.

We also need a baseline expectation to determine whether or not a high-ownership selection was successful or not. But this is also something that will differ from position to position because quarterbacks tend to have different scoring distributions than any other position.

As such, a "baseline" expectation for value was set at each position. "Value" in this instance refers to how many FanDuel points were scored for each $1,000 in salary. For example, if Tom Brady scores 18 FanDuel points at a salary of $9,000, he would be worth 2x value in that week.

To determine a baseline value, we looked at all players at each position who carried at least 1% ownership in a given week and found the average value for those players. This weeds out injured players and those without relevant roles and lets us know what an "average" player at that position would be expected to fetch you. Here's the baseline value used for each of the positions.

Clearly, the bar here isn't very high. You generally want players to hit somewhere between 2x and 3x value at running back and wide receiver to have a shot at cashing, so even if a player hit baseline value for his position, you weren't necessarily giddy about the performance. Still, not every position was able to hit that with regularity.

Here's a look at how often the popular players at each position hit their baseline value. Again, expectations weren't high, but there was a whole lotta chalk busting lineups this year.

High-ownership wide receivers are not your friends.

Even though they had the second-lowest baseline average to hit, they still managed to get there less often than any other position. We can expand on why later on, but that was a bloody position this year.

Based on the chart above, tight ends appear to be DFS darlings. However, that's not necessarily the case. It's important to remember that they needed just 1.33x value to "hit," and that's much lower than any other position. Additionally, they had the lowest hit rate when we conducted this exercise last year, reaching baseline value just 39.22% of the time. Quarterbacks performed well both years, so their mark appears to be a bit more legitimate.

What the table above does do is show us the hit rate at each position. What it doesn't do is alleviate our previously mentioned concerns about the "fear of missing out." We want to know what the odds are of a player just obliterating value when he's not on our rosters. So let's tackle that here.

Instead of comparing to baseline values, this chart shows how often each position hit various thresholds. It is important to keep in mind that hitting 3x value is far different for a quarterback than a wide receiver due to the scoring differences at each position, but this should show how many true blow-up games occurred with chalky players.

If you truly had a fear of missing out, it shouldn't have been on a wide receiver; it should have been at defense or running back.

These were the only two positions at which high-ownership players hit at least 4x value more than 2% of the time. There were just six total instances of popular players hitting at least 5x value, and half of those were from defenses and special teams. Their hit rate -- as seen above -- also wasn't relatively terrible, so you can do worse than going with the crowd at that position.

Wide receivers are a different discussion. There was only one instance of a wide receiver exceeding 3.75x value, and that was Julio Jones in his infamous blow-up against the Tampa Bay Buccaneers. Outside of that, you never truly got scorched if you faded a highly-owned player at the position.

The chart above is also important in tempering expectations at tight end. Even though they wound up hitting their baseline value at a high rate, not a single highly-owned tight end exceeded 2.99x value. Only 37.25% of the tight ends topped 2x value, which equates to 12 FanDuel points for a tight end priced at $6,000.

To make things even worse, most of the positive performances at tight end were by a single player. Five of the 10 best outings from a value perspective came from Rob Gronkowski; no other player cracked the top 10 twice. As mentioned in our piece analyzing 2017's perfect FanDuel lineups, if you weren't paying up for Gronk or maybe Travis Kelce, you had better have been saving all you could.

Before we move onto the position-by-position breakdown, it's important to look at the role that point-chasing plays in our decision-making, and it can rear its head to varying degrees at each position. To (at least partially) quantify this, we looked at how many points each chalky player scored the week they were popular compared to what they did the week before that. Obviously, this means we'll be excluding Week 1 at every position, narrowing the sample a bit. But here's the rate at each position for 2017, showing how often they scored more points when popular than they did the previous week.

If you used a wide receiver at high ownership, there was a 60% chance he would score fewer points than he did the previous week. It was even worse for kickers and tight ends. We need to do whatever we can to avoid chasing points.

Even the positions that did well were still relatively underwhelming. It was a coinflip whether quarterbacks and running backs would outperform their previous outings, and only defenses exceeded 50%. If we're going to swallow chalk on a player, we need to make sure we're doing so because of his usage, a change in his role, or his matchup rather than how many points he stumbled into the previous week.

Overall, this paints a pretty grim picture for using players who will be popular. But can we find exceptions to that? Let's dig a bit deeper into each position to see if we can learn any additional information.

Quarterback

As you saw before, quarterbacks were not the objects of our ire in 2017. They were pretty solid for the most part. But there is still plenty we can learn in looking at those chalky assets.

First of all, here's a breakdown of what a highly-owned quarterback looked like in 2017. This can give us a bit of a blueprint for what forces a player into popularity. Again, this is for the top three quarterbacks in ownership each week for the FanDuel Sunday Million. "Previous Score" is how many FanDuel points that player scored the week before he was popular.

Before we dig into salaries and such, one piece that should stick out based on our perfect lineup analysis is the wind speed. The average wind speed for quarterbacks in our perfect lineups was 2.59 miles per hour; it was 5.18 for the chalky quarterbacks. Although that's still lower than the league-wide average for all games of 5.77 miles per hour, it's still exactly double what we saw among perfect quarterbacks.

This tells us that the public is not accounting enough for wind speeds when making their decisions. The maximum wind speed for any quarterback in a perfect lineup was 10 miles per hour, and 21.57% of all chalky quarterbacks played in games with wind speeds higher than that mark. We can gain a leg up on the field by focusing on quarterbacks who won't face the ills of wind in a given week.

Let's stick with comparing our chalky quarterbacks to those who wound up in perfect lineups when discussing salaries. There's going to be an inherent bias toward cheaper quarterbacks when looking at perfect lineups because they're all hitting the positive side of variance, and we don't have to deal with the unpredictability of the non-elite assets. Still, it's pretty clear that we should be making a more concerted effort to save salary here when possible.

Among popular quarterbacks, 62.75% had a salary of $8,000 or more. That pricing tier represented just 35.29% of the quarterbacks in perfect lineups. Although it's not a terrible idea to pay up at the position, we should likely be funneling more of our ownership into the middle tier of pricing.

The question then becomes whether or not we were good at predicting when those cheaper assets would go off. So let's divide our quarterbacks into two groups: those priced $8,000 and higher and those who were cheaper. What were the hit rates (again, hitting a baseline value of 2.20x salary) for each group? Just let the numbers do the talking.

This would be the perfect place for Wee-Bey dot gif.

Turns out we're pretty darn good at predicting when mid-priced quarterbacks will go off. So why don't we spend down more often?

Thankfully, this gives us a pretty solid blueprint of how we should handle quarterbacks whom we expect to be popular. If they're a higher-priced asset, then we really don't need to go there, but we also don't need to avoid them. If they're closer to that middle range in pricing, we should feel pretty solid about plugging them in. There were only 19 quarterbacks at $7,900 or lower on our list, but 15 of them hit at least baseline expectation. They also produced five of the seven best performances among chalky quarterbacks in terms of just raw FanDuel points. That's a massive hit rate, so paying down at quarterback seems to be our ideal strategy for tournaments.

Before we close up shop at this position, we should quickly touch on some of the other data above. The hit rate among underdog quarterbacks was just 50.00% compared to 68.29% for those who were favored. It went all the way up to 76.92% for quarterbacks favored by 10 or more points. We should be super wary of guys on teams that aren't expected to win, and quarterbacks on teams that are heavily favored should not worry us much at all.

Finally, it didn't seem to matter much where the game was played. The hit rate among quarterbacks on the road was 63.16% compared to 65.63% at home, which is a negligible difference. A quarterback playing on the road shouldn't exclude him from consideration as long as the other conditions are met.

If we're paying down at quarterback, we need to spend up somewhere. It seems like a good chunk of that salary should be sliding over to the running backs.

Running Back

Running backs are essentially the opposite of quarterbacks when we're comparing them to the perfect lineups. While we decided the public was spending too much on the signal callers, they should have been allocating more salary to their ball-toting ground crusaders.

Here's the same breakdown as we had with quarterbacks to give you the over-arching picture of what made a highly-owned back in 2017.

Again, wind speed, y'all. There is no logical reason that we should be playing quarterbacks in higher average wind speeds than running backs. This should be an important factor in your weekly process.

Back to the salaries, which is a big part of what we can control each week. We can dictate how much of our cap we're allotting to each position, meaning that if we spot a deficiency, we can adjust for it pretty easily. Here's the breakdown of the hit rates for running backs in each pricing tier out of a full sample of 85 high-ownership backs.

If we do the same as with quarterbacks and break them into two groups, exactly half of the running backs priced at $8,000 or higher hit value compared to 40.54% of the cheaper guys. This is the opposite of what we saw at quarterback, and it should influence the way we build our rosters.

Even though it takes a bit more juice for a high-priced running back to hit value than a cheap one, they still were able to do so at a higher rate than the cheaper backs. And while even the higher-priced guys weren't necessarily locks to perform well, they weren't abysmal, either. This means we should have a dump truck of skepticism when a cheap running back is projected to come at high ownership, and we can potentially just bite the bullet and ride a guy if he's an established stud.

This isn't to say that we always need to ignore cheaper backs. If there's an injury pushing one of those players into a workhorse role, then we'd be foolish not to go there. But more often than not, it does seem as if running back -- at least based on 2017 data -- is a spot where we want to pay up.

In last year's look at ownership, we saw that running backs on the road tended to be far less safe than their contemporaries at home. While the 2017 data still favored those at home, the gap was considerably smaller.

Those numbers were 73.33% and 40.00%, respectively, last year. Because the gap was so cavernous in that season, we should likely still be more willing to fade popular running backs who are on the road. This does, however, at least make that less of a hard-and-fast rule than it was based on previous data.

We saw the same thing with over/unders as 48.89% of the backs in games with an over/under of 45 points managed to hit baseline value compared to 42.50% of those in projected lower-scoring games. Yes, there was a difference, but it wasn't overly large. The biggest split seemed to be between the expensive running backs and the cheaper ones, making that our best takeaway from the data.

Based on the first two positions, it seems pretty clear that we want to be particularly skeptical of high-priced quarterbacks and low-priced running backs. At wide receiver, we need to be skeptical of everybody.

Wide Receiver

If you'll remember from our opening section, only 44.85% of wide receivers managed to hit their baseline expectation, the lowest at any position. That's going to make most of the data here look pretty bloody. But can we find any exceptions to this rule?

Let's first take a look at the data on highly-owned wide receivers as a whole to see what we're working with.

The average chalky wide receiver scored almost three fewer points when popular than he did the previous week. Point-chasing be a curse.

But we've already established that, so let's move on from the fact that wide receivers are just bad plays when they're projected to be popular. We want to see if we can find any exceptions to this rule, allowing us to still go here in the right scenarios. That would seem to be the most actionable approach here.

As such, we're going to break the 137 chalky wide receivers into various categories to see if those who fell into one area managed to hit at a higher rate than others. Here's that breakdown. For the salary, over/under, and wind speed, the median value in that category was used as the breaking point for dividing up the receivers.

Although this doesn't make wide receivers look much better, it does at least provide us with some answers. Let's go through the takeaways here.

First, receivers tended to be more reliable when they were outside of the upper tiers of pricing. Again, this is similar to what we saw in analyzing perfect lineups, where the most common pricing range for perfect wide receivers was between $6,000 and $7,400. By this point, that shouldn't be a surprise.

Second, having a high over/under didn't necessarily help at all. This goes counter to what we found last year with the chalky wide receivers, and it's not necessarily in line with what you would expect anecdotally. It's fair to take that data with a grain of salt.

The reverse is true, though, for the spreads. Wide receivers on teams that were favored hit baseline value more often than the underdogs in both years. Because this data has been true for both seasons, we should likely feel better about using a popular wideout when his team is projected to win.

All of this isn't to say that we should seek out wide receivers who fit into the groups above. Rather, it's to say we should steadfastly avoid those who do not. If there's a wide receiver we believe will be popular who is expensive or on a team that is an underdog, we should feel fairly comfortable in avoiding that guy. Frankly, all chalky wide receivers appear to be DFS death traps, but it's especially so for guys who fit into those two groups.

Tight End

Tight ends were our odd ducks from the beginning. Even though they hit baseline value often, they never really blew the top off of value. That's going to complicate things in this discussion, but let's try to see what we can decipher.

The hit rate was high, but the average value generated by a chalky tight end was 1.53x. That equates to barely netting double-digit points from the average-priced chalky player at $6,649. They're really not as warm and cuddly as they appear at first glance even if they are outperforming their positional peers.

So, which pricing ranges were the worst culprits here? Let's break it down as we have with the other positions, dividing the position into three tiers of pricing. It's worth noting that the upper rung is exclusively Gronk and Kelce.

Even the two best assets at the position were hard to trust. Sure, they had to obtain a higher score to reach that value threshold, but it's also a bigger pain when they fail to get there due to the salary you're paying for them.

It is worth noting that Gronkowski still managed to land in the perfect lineup three times, and Kelce was there once. So it's not always an awful idea to plunge into that range. But the more fruitful zone -- again, as seen in the perfect lineups -- was the lower rung.

Not only did they manage to hit baseline value most often, but they also had the highest hit rate for 2x value and the best average value of the group.

Even though the middle range hit baseline value more often than the upper echelons, they also had fewer truly impactful games. Only 4 of 18 chalky tight ends priced between $6,000 and $7,400 managed to hit 2x value, which is fully unfulfilling. Sure, there were some misses in the other categories, but for the most part, you needed to be at one extreme or the other.

Going forward, this means we should likely be skeptical of all tight ends who are projected to be popular, but this is especially true for the mid-priced assets. If they don't have a big enough market share to be a high-priced guy and aren't cheap enough to save us money, then those players are a pretty tough sell. The most fruitful tier -- both among popular tight ends and in the perfect lineups -- was the cheaper range, and we should be acting accordingly with our future lineups.

Outside of salary, another piece of data that appeared to be helpful was the over/under. The median over/under for chalky tight ends was 46.5 points. The tight ends in games at that mark or higher averaged 1.69x value; those in projected lower-scoring games produced an average value of just 1.33x. If a tight end is projected to be popular in a game that could be short on touchdowns, we absolutely need to look elsewhere.

The other thing we need to discuss with tight ends is wind. For some reason, we were obsessed with using tight ends in blustery games this year with an average wind speed of 7.16 among the popular options. Again, the league-wide average was just 5.77 miles per hour, and the average wind speeds for tight ends in perfect lineups was 5.00. If quarterbacks perform worse when the winds are higher, it makes sense that their tight ends would, too, and we need to do a better job of considering this when filling out lineups.

Defense and Special Teams

In the opening section, it certainly seemed as if highly-owned defenses and special teams were viable simply because they had a decent shot of blowing up. So, we don't have to jump ship here if we think a team is going to be popular. But there's still more we can learn from this data.

First, here's the breakdown we've been providing at each of the other positions.

Getting an average of 2.22x value when the baseline value was 1.89x really isn't too bad. So this position is far less daunting than running back, wide receiver, and tight end. It gets even a bit more interesting once you break it down by splits.

Let's do the same action as we did at wide receiver to look at the hit rates of chalky defenses in various situations. We won't be looking at the defenses that were favored or underdogs because only one defense was popular as an underdog. Once again, the split marks for salary, over/under, and wind speeds were set based on the medians of the data set.

I, personally, have always been of the mind that defenses at home are the preferred options because they're more likely to put the opposing team in negative game flow, and they get the advantage of a raucous bunch of fans. That wasn't the case here.

There were only 15 chalky defenses that were on the road, but 11 of them were able to exceed the position's baseline value (and all 11 were worth at least 2x value, as well). The success was even more pronounced if we looked at just teams that were favored by at least nine points where seven of nine such road units hit baseline value. The only two that didn't were both the Pittsburgh Steelers, a team notorious for playing down to its competition.

This data doesn't mean that we should actively target defenses on the road. However, it does show that we don't need to bail on a potentially chalky defense just because it's playing outside of its home stadium. If we believe that road defense has a path to a big day, we should be willing to plug them in.

From a pricing perspective, it also seems as if we shouldn't be afraid of paying up at this position. Even though it takes a higher output for a more expensive defense to hit value, they still managed to do so almost 60% of the time. Based on what we've seen at other positions, that's fully respectable.

It was the cheaper defenses that made things a bit iffy. If we lower the bar to $4,500, there were just six defenses that qualified as being popular; four of them failed to meet baseline value. The same was true for 8 of 11 defenses priced at $4,600 or lower, and this included some defenses that were pretty heavily favored. We don't necessarily need to bail if a cheaper defense is projected to be highly owned, but we do need to view them with an extra bit of skepticism.

Finally, it's worthwhile to note that the success rate did not vary much based on the over/under. In fact, it was the opposite of what you may have expected. It's just one year of data, but this does make it seem as if the Vegas line is more important than the over/under for defenses. This anecdotally makes sense given the fantasy scoring tied to sacks, interceptions, and other products of negative game script compared to what you earn based on the total points allowed for the game. Going forward, we should be sure to put a higher emphasis on finding a defense that is heavily favored over one playing in a potential low-scoring game.

Kicker

If you go back to our chart about how often each position hit each value range, you'll see that kickers managed to hit between 3x and 3.99x value 19.61% of the time, the second-highest mark of any position. They also had the third-lowest rate of finishing with less than 1x value. Clearly, we're not too bad at predicting this position. What more can we learn from the data?

Despite the solid hit rates at high ownership, we were still a bit point-chasey here with the average marks decreasing almost four points when popular compared to the previous week. It wasn't detrimental, but we do need to keep this in mind when evaluating kickers.

The other thing that stands out in that chart is the wind speed. Although 4.51 miles per hour isn't excessive, it's still higher than the 3.53 mile-per-hour average we saw among kickers in perfect lineups. Accounting for wind isn't the most important lesson we can learn here, but it is something we need to have on our weekly checklist at multiple positions, especially quarterback and kicker.

To get some more takeaways, here's the hit rate based on various splits among high-ownership kickers.

One split you will not see there is the kickers who were favored versus those who were underdogs. There were just five popular kickers this year who were on teams that were underdogs; none of the five hit baseline value. Don't use underdog kickers.

The table above does make it seem as if over/under is more important than what we saw with the defenses, which is a solid takeaway. Even some of the kickers in games with lower over/unders had respectable implied team totals (they were favored by an average of 6.77 points), but they were still more erratic in hitting value. As a result, we may need to be more selective at kicker when it comes to Vegas' info than what we saw at defense.

As a bit of compensation, though, the cheaper kickers in our sample were much more reliable than the cheap defenses in the previous section. They didn't hit as often as the more expensive kickers, but a 50.00% hit rate on non-elite assets is at least respectable. Based on this, if we're forced to choose between paying up at either defense or kicker, we should be favoring the defenses more often than not.

That said, we can't just go picking low-priced kickers at random if we want to fill out a successful lineup. These cheaper assets that hit baseline value were still in games with an average over/under of 45.32 points on teams that were favored by 6.27 points, putting their implied team totals at 25.80. If you can't find a cheap kicker who checks the boxes at the position, then you're likely going to have to suck it up and find the salary to get out of that pricing range. And if it seems like a kicker is going to be popular while failing to meet these criteria, we can feel safe about bailing without the fear of missing out.