Apex Drafting: How to Approach Keepers (Fantasy Football)

In the late 19th century, Italian economist Vilfredo Pareto observed that 80% of Italy’s land was owned by 20% of the population. This led him to the broader concept that a smaller proportion of causes often shapes the larger proportion of effects. Maybe 80% of workplace problems stem from just 20% of the employees; 80% of our cognitive growth may come from 20% of what we read; 80% of a website’s traffic may result from just 20% of its content. Pareto’s work was later expounded upon by Romanian engineer Joseph M. Juran, who dubbed this concept the Pareto Principle, or “the vital few and the useful many.” Many of us know it more commonly as the “80/20 rule.”

In recent weeks, I’ve detailed a similar concept in fantasy football. Most meaningful scoring comes from far fewer players than we care to admit. Generally, about 70 players matter in a fantasy season, give or take, and about half a dozen to a dozen really matter. The rest of the field is flat, and the difference between the best of what’s left and what can be found lying around is almost negligible.

The closer to the top a player sits multiplies the whole effect. We increase our edge over our opponents by hoarding as many of these key players as possible – the closer to the apex at their position the better. I have concluded we should essentially only seek upside, abandoning security and consequentially looking for the commonly correlated hallmarks of unexpected scoring explosions. I refer to this idea as the Apex Drafting Strategy.

Put another way: Why are you spending so much time trying to cling to your bishops and rooks; you play a game where you could have two or three queens. So try to do that!

I have written about this concept in my last two articles, and I’d love for you to go back and check those out if you’ve got the time.

• The Apex Drafting Strategy: Draft Sharks Not Guppies
• Apex Drafting: Finding Upside Round by Round

With all this as the prelude, I am here to tell you that you’ve been doing keepers all wrong for years.

How Most People Select Keepers

Keeper leagues rarely have consistent rules; as such, they are hard to analyze. Most keeper leagues have a tie-in to a past selection enabling a manager to keep a player in place of a certain pick from the current draft. For example, maybe we picked Player X in Round 10 last year, and now we get to keep Player X in exchange for our Round 10 pick this year. I call this forfeited round – however we arrive at it – the Keeper Round Value (KRV).

As a general rule, the goal is simple when most people select keepers. They want the maximum savings on Average Draft Position (ADP). The equation is as simple as KRV-minus-ADP, and the biggest number wins. It’s bargain-hunting or thrift-store shopping. And most good fantasy players are pretty good at it. The problem is, we shouldn’t be doing it at all.

Unlock Exclusive Tools + Bonus Episode

Join the FootClan

The view seems smart based on a flat economy within our league, but as I’ve tirelessly tried to point out in recent weeks, fantasy football scoring does not have a flat economy; it has a very top-heavy one – an 80/20 type thing. If we recognize this and our league-mates don’t, it creates an exploitable advantage for us.

Creating a Stars-and-Scrubs Approach

In high-stakes auction leagues, Stars and Scrubs (sometimes called Studs and Duds) is a common strategy. The idea is to spend massively on two or three premium players, then manage the rest of the draft with limited resources. It works too. And the secret behind the idea is the same: We find our biggest edges at the apexes. By selling out to a few great players, we secure our best shot at this critical advantage and trust that we can fix the rest on the fly.

The Studs and Duds approach in auction leagues has become common in home leagues; most of us know it. I’m a devotee of the strategy, and I would argue that we should apply this same concept to even more fantasy football contexts. When a league is in motion, we should try to trade pebbles for a diamond. If we’re in a redraft league that allows draft-pick trading, we should trade middle-round picks to get more shots at the top. In fact, I would propose that, in any redraft format where we can, we should vacate the middle of the player pool in exchange for the top. Keeper leagues provide an awesome way to do exactly that. It just takes a little reimagining of a certain long-held convention.

Selecting Keepers the Old-Fashioned Way (The Useful Many)

To my knowledge, no database charts the most common keepers in home leagues each season. However, we’re ultimately trying to weigh two viewpoints against each other anyway, so this is not of great concern. My first step was to represent the common KRV-minus-ADP formula.

WARNING: A number-and-acronym spaghetti is imminently incoming…thinking caps must be applied to the crown of the head to pass beyond this point.

I compared 2022 ADP with 2023 in half-PPR scoring formats. Finding the players with the highest 2023 ADP relative to 2022 ADP, I came up with the 10 players who made the biggest jump from one year to the next. Then, I found each player’s half-PPR/G averages in 2023 from Weeks 1-17, and I also found per-game averages for league starters at each position (for home leagues, I used 12 QBs, 24 RBs, 24 WRs, and 12 TEs, considering flexes separate). It is reasonable to assume these were among the 10 most popular keepers from 2023, but even if they weren’t, they test the KRV-minus-ADP approach the best. Here is the dataset (Note: an ADP of 200 was assigned to undrafted players):

The first thing to note is that each player came in below average compared to fantasy starters at their position. This is not overly concerning (honestly, it’s expected), as these players are all tied to Rounds 9-16.

Unlock Exclusive Tools + Bonus Episode

Join the FootClan

In addition, averages were used rather than medians (intentionally, as we are finding apexes), so once again, the players at the top of the positions score a significant amount more than the players below them; as a result, well over half of each position’s fantasy starters will show up below average.

The first thing to establish is what we are gaining with these players relative to what would have been available had we used the KRV to draft a different player. For this, I found all the players from the forfeited round at the same position and found a median. Here are the results (Note: if there were less than three players from a select position in a specific round, I incorporated the last result from the previous round and the first result from the ensuing round):

Our 10 keepers have gained an average of 2.0 points per player over expected in the rounds to which they were tied. However, opportunity cost should be considered; we essentially gain a much earlier pick by associating these players with later rounds, meaning our decision becomes far more profitable than this alone. We could pick any position we wish, but the conversions wouldn’t be compatible across positions since they score different amounts, so, for simplicity, we will only compare apples to apples by using the same position. Once again, I’ll take medians using the round that corresponds with 2023 ADP.

We’ve essentially added 16.74 half-PPR points/G per player in opportunity cost by also spending the pick the player would have exhausted on an entirely new resource. Adding the two figures together, we find that, on average, our top 10 keepers selected by KRV-minus-ADP have netted us a total advantage of 18.74 half-PPR points/G per player over what we would have received in a conventional draft without keepers.

The Other Way Around (The Vital Few)

I wanted to create a relatively objective experiment regarding what might happen if we prioritized excellent players rather than savings on ADP. The best way I could come up with was as follows:

1. Find the top players at a position in 2023 ADP
2. Find those players’ 2022 ADPs to establish KRV
3. Avoid players with a KRV in the first round in which that position was being chosen (as we’ve established already, there is an advantage to having a keeper as opposed to not, so we don’t want to forfeit the top round)

Here are my 10 keepers based on the new parameters (for consistency, I used the same breakdown as above: 4 QBs, 1 RB, 4 WRs, and 1 TE).

Unlock Exclusive Tools + Bonus Episode

Join the FootClan

Our new list is ahead of the fantasy starters at like positions by an average of 1.34 fantasy points. We also have hit the positional apex twice at WR and narrowly missed at QB and TE, where Hurts finished second per game and Jackson and Andrews each finished third, making each a significant contributor scoring well above the trendline: this is without expending a first-rounder at WR or a second at QB and TE.

Let’s put them through the same process we put the KRV-minus-ADP keepers through.

The keepers from the second table scored 2.7 more half-PPR points/G per player than the medians from their KRV, an improvement on the keepers taken using the KRV-minus-ADP formula. Let’s examine the opportunity cost as well.

The added opportunity cost associated with the round we gain is significantly higher at 20.3 half-PPR points/G per player, which makes sense, as these selections are all from Rounds 1-4. When we add the 2.7 half-PPR points/G from the first table, the advantage is 23.0 half-PPR points/G per player, making choosing our keepers this way 4.26 half-PPR points/G more profitable than the old-fashioned method of taking the biggest savings on ADP.

In Conclusion: A Little Perspective and a Lot to Consider

Can I be honest with you? I started this article with no idea how the math would come out. I didn’t fully know how I would frame the argument, which had me tied me up in knots and frustrated. There was a lot of trial and error; I’d start one way, then realize there was a better way – a lot of going back to the drawing board in general. I had a lot of anxiety about the clarity of the argument and kept trying to make it digestible, too (I am still not overly confident about that).

Unlock Exclusive Tools + Bonus Episode

Join the FootClan

Still, I wrote the intro, which I’ve never touched since. I spoke about the conclusions of the entire experiment as confidently as proven facts and never doubted the outcome. I knew it was going to work out, and that’s because I’m confident about emphasizing the accumulation of players with the best pathway to elite outcomes. Time and time again, this has proven more important than possibly anything else we prioritize in this silly game.

Yeah, it’s a small sample size; I get it. The critics are gonna talk; they’re gonna say, “Yeah, but this, and what about that?” And that’s fine. I’m confident because I’ve framed this very same question dozens of ways throughout my life as a fantasy player; it’s given me the same result a thousand different ways, so I now accept it as law. When we can abandon the middle of the player pool so we can gain more assets near the top, we should do that (Stars and Scrubs). When we can take a big upside over security, we should do that. This game is won by the arrangement of a narrow band of players every season – 70 are the useful many and 10 are the vital few – the chess queens.  Make it your effort to exhaust every resource to find those players however you can. Keepers are one such resource.