DFS Projections Explained: How to Use Them to Build Better Lineups

DFS projections are incredibly useful.

They’re also incredibly easy to misuse.

Pull up a projection model, sort by projected fantasy points, jam the highest numbers into a lineup and congratulations: you have successfully outsourced your entire DFS strategy to a spreadsheet.

Projections should help you make decisions. They shouldn’t make every decision for you.

Here’s how to actually use them.

What Are DFS Projections?

A DFS projection is an estimate of how many fantasy points a player is expected to score.

Projection systems can account for things like:

• Expected playing time
• Player performance
• Matchup
• Team environment
• Injuries
• Usage or opportunity
• Vegas lines and game totals
• Recent information

The exact inputs and methods vary by projection source.

That last part matters.

There isn’t one universally correct projection for a player. Different models make different assumptions, weigh information differently and update at different speeds.

If one site projects a player for 18.7 points and another has him at 20.1, it doesn’t necessarily mean somebody screwed up.

They’re estimates.

And sports remain annoyingly committed to actually playing the games.

Projections Are a Starting Point

This is probably the most important thing in this entire article.

A projection is not a prediction of exactly what will happen.

If a running back is projected for 17 fantasy points, nobody is saying he’s going to score 17.

He could score 5.

He could score 35.

The projection is trying to estimate the center of his range of potential outcomes based on the information available.

That’s enormously valuable.

But DFS contests aren’t generally won by correctly identifying everyone’s most likely outcome.

Especially in tournaments.

Median Projection vs. Ceiling

This is where projections start becoming more useful.

A player’s basic fantasy projection generally represents something around his expected or median performance.

But DFS players should also care about ceiling.

Ceiling is essentially the upper end of what a player could reasonably produce if things go right.

Imagine two players who both project for 15 fantasy points.

Player A consistently lands somewhere around 12-19.

Player B could score 6, but could also score 32.

Those players might look similar in a basic projection column. They are absolutely not the same DFS play.

In a contest where simply finishing above half the field pays, Player A’s stability can be attractive.

In a huge tournament where you’re trying to beat thousands of lineups, Player B’s volatility may be exactly what you’re looking for.

Same projection.

Different job.

Salary Changes Everything

Raw fantasy points only tell part of the story because DFS gives you a salary cap.

If Player A projects for 22 points at $9,000 and Player B projects for 18 at $5,500, you can’t just say Player A is the better play because 22 is bigger than 18.

Thank you for attending today’s advanced mathematics seminar.

The question becomes what those points cost you and what you can do with the salary you save elsewhere.

That’s where value enters the conversation.

One common way to think about value is projected fantasy points relative to salary.

You don’t need to become obsessed with hitting some magical points-per-dollar number. Salary structures, sports and slates are different.

What matters is understanding the tradeoff.

Paying down at one position might cost you three projected points while allowing you to gain eight somewhere else.

That’s lineup construction.

You’re not drafting players independently. You’re buying an entire lineup with a fixed amount of money.

Why the Highest-Projected Lineup Isn’t Automatically the Best Lineup

Most DFS optimizers can generate a lineup that maximizes projected fantasy points while staying under the salary cap.

Sounds perfect.

There’s one problem.

Everybody else can do it too.

If several players have strong projections, good salaries and obvious roles, they’re probably going to appear in a lot of lineups.

That’s especially important in GPPs.

Imagine two tournament lineups:

Lineup A: projected for 151 points and extremely popular.

Lineup B: projected for 148 points but constructed differently from much of the field.

Those three projected points aren’t automatically worth taking the first lineup.

If Lineup B has similar upside while giving you a path to pass a massive chunk of the field when a popular play fails, it may be the more interesting tournament lineup.

This is where projections and ownership begin working together.

Projection tells you how attractive the play looks.

Ownership tells you how attractive everyone else thinks it looks.

The gap between those two can create opportunity.

Using DFS Projections in Cash Games

Cash games generally reward consistency more than uniqueness.

You’re not trying to beat 50,000 people. You’re trying to finish above a payout line.

That makes median projections particularly useful.

You can generally prioritize:

• Strong projected volume
• Reliable playing time
• Favorable salaries
• Stable roles
• High projected value
• Lower uncertainty

You don’t need to fade an excellent play just because everybody else noticed he’s excellent.

In fact, fading an obvious high-value player can sometimes create unnecessary risk.

Cash is usually not the place to get cute just so you can feel smarter than the room.

Using DFS Projections in GPPs

Tournament strategy changes the equation.

Projection still matters. A lot.

But now we’re interested in the relationship between:

Projection + ceiling + ownership + correlation + salary

A slightly lower-projected player may become attractive if he has a strong ceiling and dramatically lower ownership.

That doesn’t mean blindly playing low-owned players.

Low ownership isn’t inherently good.

Sometimes a player is 2% owned because he’s a terrible play. The field is allowed to get things right.

The goal is to find situations where the potential reward is greater than the popularity suggests.

That’s leverage.

And projections give us one of the tools we need to identify it.

Why Different Projection Sites Disagree

If you’ve ever opened projections from multiple sources, you’ve probably seen some surprisingly large differences.

That’s normal.

Models may disagree about:

• Playing time
• Usage
• Efficiency
• Matchup strength
• Injury impact
• Team tendencies
• Game environment
• Expected opportunity

Sometimes the disagreement itself is useful.

If nearly every reputable source sees a player similarly, there’s probably less uncertainty around the basic expectation.

If projections are all over the place, something may deserve a closer look.

Don’t automatically average everything together and assume you’ve created the world’s greatest projection system.

Figure out why the numbers are different when you can.

That information can be more useful than the difference itself.

Projections Can Become Outdated Fast

This matters more in some sports than others, but it’s important everywhere.

A projection is only as useful as the information behind it.

Then somebody gets ruled out.

A starting lineup changes.

Weather turns ugly.

A player’s expected role changes.

Suddenly that beautiful projection you downloaded three hours ago belongs in a museum.

NBA DFS is probably the most obvious example. One late scratch can redistribute minutes and usage across an entire rotation.

NFL inactives can create similar situations.

MLB lineups can change opportunity significantly.

The closer you get to lock, the more important current information becomes.

Always know when your projections were last updated.

Don’t Confuse Precision With Certainty

This is one of the easiest traps in DFS.

A projection says:

18.73 fantasy points

That looks incredibly precise.

It’s still an estimate.

The decimal places don’t make the future more predictable.

There’s usually very little practical difference between a player projected for 18.7 points and somebody projected for 18.4.

Don’t treat projection rankings like standings where the guy in seventh objectively belongs ahead of the guy in eighth.

Think in ranges.

If several players project similarly, that’s your invitation to start considering the other parts of DFS strategy.

Ceiling.

Ownership.

Correlation.

Salary.

Contest type.

Roster construction.

That’s where you actually get to make a decision.

Projections Should Make Your Research Faster

This is where projections are at their best.

A full DFS slate can contain hundreds of players.

You don’t have three hours to conduct a documentary investigation into every one of them.

Projections help shrink the player pool.

They can quickly show you:

• Where the strongest values appear
• Which expensive players stand out
• Where positions are deep or thin
• Which players deserve additional research
• Where your personal opinion differs from the market

Then you investigate the decisions that actually matter.

That’s a much better use of your time than trying to independently analyze every player from scratch.

What If You Disagree With the Projection?

Good.

You are allowed.

The entire point of DFS would disappear if everyone had the same numbers and was required to obey them.

But there should be a reason.

Maybe you think a player’s role has changed.

Maybe you expect more opportunity than the model does.

Maybe the projection isn’t properly accounting for recent news.

Maybe you believe the player’s ceiling outweighs a mediocre median projection in a tournament.

Those are arguments.

“I have a feeling about this guy” is technically also an argument, but we’re going to put it slightly lower on the list.

The key is distinguishing an informed disagreement from simply wanting the projection to be wrong.

The Cover The Slate Approach to DFS Projections

Our approach is pretty simple:

Projection is a starting point, not the decision.

Use projections to understand the slate.

Use salary to understand value.

Use ceiling to understand upside.

Use ownership to understand the field.

Use correlation to understand how your lineup works together.

Use late news to make sure the assumptions underneath all of this still make sense.

Then build the lineup.

The projection is one piece of the puzzle. A very important piece, but still one piece.

Because the goal isn’t to build the lineup projected to score the most points before the games start.

The goal is to build the lineup that gives you the best chance to win the contest you’re actually playing.