When Should You Hit the Waiver Wire? (Fantasy Football)

Two weeks of fantasy football is in the books; hopefully, you’re undefeated and everything has gone swimmingly. Most likely, though, you’ve been beset by unexpected obstacles: Saquon Barkley‘s lingering injury, chaos in the 49ers’ backfield or the Cowboys’ insistence on using Ezekiel Elliott and Tony Pollard in something resembling a running back committee.

While ‘draft day’ is the sexiest part of the fantasy season, we’ve seen how the tailwinds of a good draft fade as the season progresses. You need to be active to stay atop your league; specifically, bona fide managers know that you will often earn your championship on the waiver wire. The additions that you make in-season, whether to add depth, replace a player who went down, or take a shot on a high-upside flier – can make or break your campaign. However, of course, the waiver wire isn’t free. You’ll have to compete with the other managers in your league to sign desirable players.

The two main waiver settings are waiver priority and FAAB (free agent acquisition budget). In the former, teams are assigned an order from first to last (can be random, or inverse of draft order). Each team will submit their waiver preferences and if there is a conflict (two managers submit a preference for the same player) the player goes to the manager first in line. That manager then goes to the back of the waiver line; therefore, there is a cost to submitting a preference, since you are sent all the way to the back!

FAAB is a bit simpler (and, in my opinion, a better embodiment of ‘pure’ strategy). Each manager starts with, say, $100 at the beginning of the year (this is usually fake money!) and uses this budget throughout the season. For waiver options, managers will submit bids for players they want and, if multiple managers bid on one player, the highest bid wins. Naturally, there is a built-in cost: once you spend your FAAB, you can’t get it back! You are stuck with what’s leftover for the rest of the season. However, you don’t want to end your season with anything more than zero FAAB, since that means you could have spent more to roster better players (FAAB won’t score you any fantasy points, players will!).

It’s important, then, to determine when to go ‘hard in the paint’ trying to roster players. When should you burn your high waiver priority, or drop a majority of your FAAB? The Fantasy Footballers have discussed how newcomers to the FAAB system are often too hesitant in spending their budget on valuable pieces; on the flip side, though, you don’t want to be left high and dry come fantasy playoff time. Is now the time to stake your claim, or should you play it coy and wait until the season matures?

We will analyze this question – purely based on player performance – in this article. Fantasy data comes from nflfastR, and ADP data from the fantasy football calculator.

Methodology

We’ll consider data, in Half-PPR scoring, since 2015. The crucial phenomenon to define here is a breakout: a player that is on the waiver wire in many leagues, has a high-scoring week, and becomes a target for managers to roster. Of course, it would be best to look at many years’ worth of intra-season roster percentage data and cross-reference that with fantasy scoring, but this is nearly impossible. Instead, a programmatic, inferential approach will have to do.

To start, we will define players that are likely un-rostered (and thus on the waiver wire). The first main pool of these players are those drafted after the 12th round in a fantasy draft; there’s a good chance that, in many leagues, 13th round fliers and beyond are available on waivers. The second pool will be players that underperform: in this case, if a player scores five points or less for four straight weeks, we will assume that they have been dropped and replaced by someone else (in many leagues). Of course, neither of these are perfect measures, but they will do a pretty good job approximating who is broadly rostered and who is not.

We will then define a ‘breakout’ as an un-rostered player (as defined above) scoring 15 points or more in a given week. Once a player breaks out, we assume they are rostered for the rest of the year and thus can’t break out again. This is a fallacious assumption, but having multiple breakouts makes things more complicated and potentially confounds results. Plus, there aren’t that many players that are ‘hot waiver wire adds’ many weeks apart (i.e., Week 1 and then again in Week 12).

Post-Breakout Performances

Let’s turn to the charts! We can consider the average points per game (y-axis) for the rest of a season once a player breaks out. We’ll just consider RBs, WRs, and TEs; QBs are exempt because it’s an oft-streamed position.

Each line here represents a breakout week: for example, the left-most line represents players that broke out in Week 1 and tracks their average performance for the rest of the season. The right-most line represents players that broke out in Week 10 and tracks their performance for the rest of the season. Note that the breakout is included here, which is what gives each line its funny shape (the line starts up quite high before falling back down).

Perhaps a better way to visualize this chart is by giving each week its own box. Further, while I don’t show it here, each position (RB, WR, TE) has a relatively similar ‘line shape’.

The strongest takeaway I have from this chart is regression to the mean. We saw in a previous article how Week 1 wonders tend to disappoint after an exciting start, and it looks like that principle holds for breakouts in every week. Each line starts out quite high (the breakout week) and then rapidly falls off. In fact, players fall to almost unusable levels, scoring well under 10 PPG (especially low because we are taking an average, which is skewed by ‘boom’ performances).

We can visualize this in terms of ‘decay’, or the percent point total of the breakout week that the player manages to replicate going forward. Players that break out in all weeks essentially fall back down to below 50% of their breakout point total the very next week. It’s also important to note that this decay factor is more or less constant: players consistently produce around 35-40% of their breakout week, on average, for the rest of the season, regardless of when they broke out.

Waiver Wire Strategy

While we have seen another harrowing example of regression to the mean, there might still be some useful information here. Specifically, even though breakouts usually fall back to earth, are there some that still outperform players that break out in other weeks? This chart shows the week a player broke out (x-axis) and their PPG the rest of the season (y-axis).

What is clear off the bat is that players that break out in Week 1 have a higher PPG total the rest of season (about 0.75 PPG more) than players that break out in following weeks. What’s more, this difference is statistically significant, whereas the other two big PPG jumps in the data (Week 8 and Week 12) are not statistically significant. This is largely because of sample size: intuitively, more players break out in Week 1 than later in the season. Here is a chart that shows how the rate of breakouts diminishes over time; again, these are totals since 2015.

 

Anyways, this is impactful: even though Week 1 breakouts do regress to the mean, they still score more on average than players that break out later. This is likely an information issue: in Week 1, the fantasy community doesn’t have much information about various rosters and player prospects. An extreme example is Brandon Aiyuk being a healthy scratch to start the season! Therefore, it’s easier for under-the-radar breakouts to emerge in Week 1 because we simply know less. After Week 1, when the fantasy community has a more accurate understanding of the NFL landscape, it’s more difficult for players to break out from obscurity.

Conclusion

The first takeaway, which is worth repeating, is regression to the mean. Players that break out will, on average, fall back very quickly to lower levels of fantasy production. In fact, that level of fantasy production is so low as to possibly be unusable: just 6 – 7 PPG on average.

However, don’t hear what I’m not saying: even with this regression, it is still important to be active and take your shot on the waiver wire. Even if the ‘average’ result is middling, there are potentially league-winning diamonds in the rough (i.e., James Robinson of 2020). What’s more, your team may simply need it: perhaps a starter went down or underperformed, and you need to add valuable depth. It’s just important to set realistic expectations for waiver pickups.

With all of this in mind, we’ve seen in this article a very important result: the decay rate, or how quickly and how much, a player’s production falls off after a breakout is constant relative to when they broke out. This just means that breakout players in Week 4 are no better than breakouts in Week 10, or vice versa. The one exception is Week 1 breakouts, which score about 0.75 PPG more.

This means that you should hit the waiver hire, or spend your FAAB, early in the season. The reasoning behind this is simple: breakouts of all weeks are likely to decay in value at the same rate, so you mine as well roster them now and get as many weeks of production as possible. Waiting for later weeks won’t yield you better breakout options; this is a reasonable theory, since breakouts in the mid-to-late season might have built momentum and the trust of the offense, but one we have shown to be false.

Of course, this is general advice, and there are always team-dependent factors. You might have a stacked roster and have no need to blow your FAAB on an exciting acquisition; in that case, save for a rainy day when your roster is going through a rough stretch. The takeaway here is just a broad principle: breakouts are largely the same across weeks so, all else equal, grab them early.

________

Did I miss something? Message me on Twitter.