Finding good quarterback plays when playing daily fantasy football on FanDuel and DraftKings

One of the most common themes in our considerable reader email each football season is, "how do I find a good QB/RB/WR/TE for double ups/big tournaments/heads up games." Which is to say, "how do I play daily fantasy football?" Well, that might be an oversimplification. But the feeling behind it is sincere - daily fantasy football is one of the most complex of all of the DFS contests because analyzing each position is an endeavor entirely unto itself. While there is some overlap, trying to figure out a good QB AND a good RB - to say nothing of a solid tight end or defense - requires analyzing all different kinds of data. This year, we're going to take selections out of our free NFL eBook and take an even deeper dive into how we go about picking players at each position for each kind of contest.

Today, we're going to get started at the top - Quarterback. We hope you enjoy!

In case you didn’t notice, on Saturday we released our brand new NFL DFS suite of tools including full projections, optimal NFL lineups and Player Lab which includes filters to help you create NFL lineups for FanDuel and DraftKings. It’s a robust set of tools; well beyond anything we’ve ever had and something we feel will help our users *ahem* tackle the NFL season.

Using tools like these for the NFL season is crucial and you can grab a free 2 WEEK trial to test out yourself.

FINDING GOOD DAILY FANTASY QUARTERBACKS

Finding a great daily fantasy quarterback can mean a lot of different things, because there is a huge difference in what our goals are across different contest types. Do we want to target a quarterback who is a known commodity - a steady, week-in-week-out guy who is priced pretty much fairly, with a high floor and a lowish ceiling? Do we want to try and shoot the moon, taking a QB and WR pair that's very erratic, but has high upside? Do we want to spend up for safety at quarterback, or go cheap at quarterback to spend up elsewhere?

It's truly a dilemma.

Quarterbacks - the most predictable position in fantasy football

Quarterback, thankfully, is one of the more projectable positions in daily fantasy football. That's not to say there's ever a "sure thing" when it comes to projecting athletic performance, but quarterbacks provide at least some sanity in that regard. Since every play starts with the quarterback touching the ball, he has the opportunity to "get to the long run" in terms of sample sizes a lot more quickly than players at higher variance positions.

In fact, if you're thinking about positional variance, you can essentially break it down by how many plays a player rates to be involved in over the course of a given game. A simple example of this can be seen in last year's inaugural game - the Steelers vs. the Patriots.

On the Steelers side of the ball, Roethlisberger threw 38 times, DeAngelo Williams rushed the ball 21 times (and caught a pass), Antonio Brown caught 9 balls, and Heath Miller caught 8 balls. On the Pats side of the ball, Brady threw 32 times, Dion Lewis rushed 15 times, Julian Edelman caught 11 balls, and Rob Gronkowski caught 5.

Unsurprisingly, if we're ranking positions from "most predictable" to "least predictable," quarterbacks rank as the most predictable simply because they are a part of so many more plays. Running backs are concerned to have the next highest level of predictability, then wide receivers, and finally tight ends.

And this isn't anecdotal, by the way. We've run the numbers for the coefficient of variation for each position, and the variability on player performance lines up in exactly the way I described above. I won't bore you with grizzly math details here (though you can find more of them in our free ebook) on how the formula is calculated, but Quarterbacks scored the lowest (and thus safest) number by far, of .36. Running backs were next at .56, and wide receivers were up in the .75 range. So, yeah, quarterbacks were the safest by a mile. Look, a chart!

If quarterbacks are considered to be the safest position in daily fantasy sports, then, the type of quarterback we decide to play on a given day should be directly informed by what sort of contest we are hoping to win. If we're playing a 50/50 contest where we simply need to beat half the field, we'll often want to spend up as much as we can (while still getting good value, of course) at the quarterback position to "lock up" some steady production.

On the other hand, if we're attempting to shoot the moon and win a huge tournament, we often want to spend less at the quarterback position and spend up at more volatile positions, precisely because we're hoping to embrace volatility.

So if we're trying to suss out safety and upside, what other factors should we take into consideration?

Safety vs. Upside - interpreting data from Las Vegas

Let's start by point out that there are a few factors that can contribute to both safety AND upside. The first place to start whenever you're trying to project a given game script is probably the Las Vegas sports books. Since they take massive amounts of action on every single NFL game, they've simply become better than any individual football fan is likely to be at projecting how a given game will play out. They provide publicly available game totals and spreads to help us have a reasonable sense of how the game would play out, on average, if it were played 100 times.

Starting with Vegas totals and lines can help us start to form a narrative around how the game is likely to play. When it comes to game totals, we obviously want quarterbacks in games that rate to be higher scoring. This one is fairly obvious - the more points that will be scored, the safer our players become, and the more upside they stand to have. The next factor - the spread - is a little bit trickier to interpret.

Generally speaking, thin spreads (say, 3 points or less) are going to have a positive influence on the total number of fantasy points scored by fantasy relevant players. Starters tend to play their minutes in close games, for one, but close games tend to have more plays total. Neither team has a great incentive to run down the clock until the very end of the game when one tries to cling to a lead, and teams can be expected to pretty much continue to run their offensive sets. While this helps for upside, it's mostly a factor in considering a quarterback's safety.

Wide spreads, on the other hand, are notoriously more difficult to interpret. If a team is favored by 13 points, for instance, it's often because they have a deadly passing attack (think the Patriots of the last few years). They score all of their points with high flying acrobatics, and many of their fantasy players can have a game's worth of production in 3 quarters. The problem, of course, is that when they do gain that insurmountable lead, most coaches will take to the ground to try and run the clock out. There are exceptions to this of course (the aforementioned Patriots, for one), but by and large this is often the way it winds up playing out. One way or the other, being favored by a large amount often leads to a great deal more volatility for members of the favored team's passing game.

And then there are instances where you'll consider a quarterback on a team that's a big underdog. This is perhaps the most common level 1 argument you'll hear from fantasy football analysts out there: "Well, the Browns will be trailing here - so that means they are going to have to throw the ball!"

And while this is true, the data suggests that this in no way makes the underdog quarterback a sure thing to put up a good fantasy total. DFSR co-founder Doug Norrie did an inquiry into this for last season, and found that pass attempts had just a .36 correlation with total fantasy production. While that's significant, it's nothing like the factors that ranked above it (the highest of which were, you guessed it, total passing yards and passing touchdowns - more on this in a minute).

So while being a big underdog doesn't make a quarterback safe, it arguably does contribute to his upside. We've certainly seen examples of quarterbacks do great things in garbage time with the game essentially decided. Matthew Berry from ESPN cited a fascinating stat about Blake Bortles in his 100 facts column for this year where he pointed out that 29 of Blake Bortles' 35 touchdowns were thrown with his team trailing, for instance. But as a general rule of thumb, counting on garbage time is a lousy way to find safety.

So when it comes to using Vegas data, we'll check the following boxes:

If you're looking for upside AND safety, you want games with high totals.

If you're looking for upside AND safety, you want games with thin spreads.

If you're looking for upside and not worried about safety, target games with wide spreads.

If you want to shoot the moon, take a quarterback in a good passing offense that's a big underdog to a team with a great offense but poor defense.

Moving past Vegas and looking at individual QBs

I nodded to the stats we want to work our hardest to project for quarterbacks in the Vegas section - touch downs, and passing yards. Of the two, passing yards are arguably the easier of the two to project. Drew Brees, for instance, only had 4 games with fewer than 250 passing yards or more than 350 last season.

And while passing yards are more projectable (and should thus correlate more highly with safety), you need to remember that you also get fewer fantasy points per passing yard than for rushing or receiving yards, so the opportunity to simply acquire more yardage doesn’t help a passer’s total quite as much. Additionally, there are many teams who pass a ton primarily because they are playing from behind. If you’re constantly playing from behind, chances are that 1) you stink, meaning you aren’t getting very much out of your pass attempts in the first place and 2) that opposing defenses know you are passing, making your pass attempts even less effective.

So yes, we want to be able to project touchdowns as well as yards. The good news? Touchdowns might be more projectable than you think.

Projecting quarterback touchdowns

A common perception is that touchdown scoring is basically lucky (you could get tackled on the 1 yard line!), and that accumulating yards is not, but denying that we have some ability to predict who will pass for a lot of touchdowns would be silly.

As Raybon astutely points out on 4for4, 66% of all touchdowns are scored in the red zone, and another 25% are scored on the opposing 49-20 yard lines. The biggest predictors of touch down scoring, then, are how often a team will get to the red zone, and what they will do when they get there. More good news for you! We have a chart showing exactly how often teams passed in the red zone last season, and you can feast your eyes upon it now.

Do you know the stat that has the highest correlations with passing touchdowns? You guessed it! Passing attempts in the red zone. We want teams that not only get to the red zone, then, but teams who pass when they do get there. It's no surprise to see pass-first (pass-only?) teams like the Pats atop the list of red zone passing touchdowns, but the .66 overall correlation between passing attempts in the red zone and red zone passing touch downs is a very strong one.

Now, not all red zone passing attempts are created equally. Dallas and Chicago, lacking two of their biggest red zone threats, struggled mightily to convert their attempts to touch downs last season. But with healthier squads to start the season, they might be very strong regression candidates, ready to put up some big numbers. Attempts won't solve everything, of course. Cleveland was awful at converting red zone passes to touch downs last season, and they're quite likely to suck at it again this year, but as a general rule - grabbing passers that are A) capable of getting their team in the red zone and B) entrusted with the ball when they get there will go a long way to helping you find both the safety and the upside you're looking for.

Conspicuously absent in the discussion so far is a quarterback's propensity to throw the long ball. While they're have been edge cases in the past where quarterbacks who thrived on the long ball had huge fantasy upside, by and large we can get just as much upside from taking guys who just move the ball down the fielder. Even in the Tom Brady/Randy Moss days, the flashy bombs were the exception rather than the rule.

Finding safety and upside with our new player lab

Want to know all of the factors we consider when trying to pick a daily fantasy football team on FanDuel and DraftKings? Take a look at the filters for our new

Player Lab comes with default filters that allow you to select for guys against bad defenses, or games with high over-unders, or thin spreads, and also lets you tinker with a number of other settings to seek safety (like low interceptions) or upside (like high team pass TD%). One thing's for sure - finding the right quarterback for the right kind of contest takes WORK. But if you're ready to put it in, we're ready to help you along the way.

At the moment we're offering a free 2 week trial two our NFL Player Lab. Are you in?