When people think of bias, they typically think of something negative and rightly so. Systematic beliefs about something based not on facts and logic, but upon a subconscious (or worse, conscious) force can range from stupid and unfair to dangerous bigotry.
When we consider bias as it relates to fantasy sports and other kinds of decision making, e.g. financial, we can start to see how many common cognitive biases arose and are maintained for good reason. Recency bias helps us repeat actions and decisions that have been beneficial (for our survival as it first evolved) while avoiding those that were detrimental, and novelty bias, today’s topic, helped us keep trying new things to expand our horizons and allowed us to thrive rather than merely survive.
Novelty bias, in simplest terms, is a preference for the new option over familiar options. It exists as deep into the animal kingdom as scientists have been able to look, including the nematode species I study in my behavioral genetics research, Caenorhabditis Elegans. Even these microscopic worms will move toward a novel chemical stimulus over familiar odors they find pleasant. In other species, novelty bias helps males to impregnate multiple different females, increasing the likelihood of them passing their genes on far and wide. Like many biases, release of the neurotransmitter dopamine, which signals reward, is a necessary component of novelty bias. It feels good, so novelty is reinforced in our brains.
Fantasy sports aficionados have their share of cognitive biases, and I’ve written extensively on which of those we want to reign in or overcome to optimize our weekly or daily decision making. Check out my book here. One I haven’t really touched on is novelty. At this time, every year, shiny new rookies await, full of promise, some even with opportunity. One of their most attractive qualities? Aside from gleaming Combine scores and college production, they haven’t yet burned us in fantasy! While we’re clearly drawn to the new talent pool, particularly at low prices in DFS, how should we approach these rookies in 2017? I went position by position, looking at certain trends to help us all make smart decisions about the newest choices on our favorite DFS sites.
Quarterback was most intriguing to me, as it’s arguably the largest and most consistent source of fantasy points each week. One thing we know to pay attention to when selecting a quarterback is the Vegas line and his implied team total. Chris Raybon did the original work on this a year or two ago, and the trends he defined continue to hold up (see graph showing data from all games 2013-2016). The higher the team total, the higher the QB fantasy production.
Rookies are probably more volatile than their more experienced counterparts, with a wider range of outcomes. Their low price reflects this uncertainty, and increases our novelty bias. We don’t really know if he’s good or bad and it doesn’t cost much to find out, might be our DFS lineup building logic. The data indicates that rookie quarterback performance also correlates with implied team total, but the trend is notably weaker (graph shows games in which rookie quarterbacks played 2013-2016).
Taking it one step further, I found that eight of the top 11 fantasy performances by rookies between 2013 and 2016 were by Dak Prescott (2016) and Marcus Mariota (2015). That’s a red flag to me. It says that not any rookie QB has a shot at a great fantasy day. Moreover, there was not steady improvement for rookie QBs as the season went on, meaning that it wasn’t necessarily safer to roster a rookie in Week 11 than it was in Week 3.
Bottom Line: If you’re saving money at QB, it’s probably better to do it with veterans like Jay Cutler, Brian Hoyer or Carson Palmer when they have a high team total. Even less experienced non-rookies like Tom Savage and Carson Wentz represent better value than most rookies will return, and have the advantage of being invisible to those with a strong novelty bias, who will rush to load up on DeShone Kizer in Week 1. Come back to Kizer, and whomever else may have moved into a starting QB position in a few weeks, when their salary will still be low and their ownership even lower. But still follow the team totals for the best chance at a big return on your QB investment.
The trend with running backs and team total isn’t nearly as strong as that with quarterbacks, but it is mildly positive (see graph). It doesn’t hurt to take players from potentially high scoring teams, but as you know, with running backs you want predictable high volume more than anything else.
When it comes to rookies, again the weak trend is even weaker, but there are clearly rookie running backs that have met and exceeded value (see graph below). One notable finding is that the peak rookie running back average fantasy points scored is between weeks 5-12. The adjustment period to the NFL is real, and even studs like Christian McCaffrey, Dalvin Cook and Joe Mixon will likely need to get a few games under their belts before breaking out. Thinking about Kareem Hunt in Week 1? Normally a good idea, but no NFL experience against a Patriots team that allowed seven total touchdowns to running backs all last season has me putting on the breaks. A slow start may take the shine off these rookies for people who can’t wait to slot them into their lineups, which means by the time they’re ready to reach value, ownership could be nice and low.
Wide receiver is one of the most volatile positions in your roster. This sets up a quandary for DFS players with some cognitive biases. On the one hand, you’ve been let down by your receivers. We’ve all rostered players in a great spot, with zero catches, sometimes at a high salary. We’re hungry for novelty! But before we get to the rookies, notice that the correlation between wide receiver production and implied team total isn’t very strong, but it is there. The correlation between quarterback and WR1 is an especially strong one (used in most stacks for DFS), so it’d be hard to imagine that WR1 production wouldn’t correlate with team total. What you’re seeing below is the dilution from WR2 and WR3. Now note that most WR1 are veterans.
Rookie wide receivers are probably the most attractive targets in fantasy, the place where “new meat” is most welcome, but beware. 2014 was the year of the rookie wide receiver and it was an outlier. Eight of the top 10 rookie wide receiver performances over the past four seasons were in 2014 (five were from Odell Beckham Jr.), and 16 of the top 30 were also from that year. Our recency bias will be a little weaker this year than it was in 2015 or 2016, but it still exists.
Rookie wide receivers tend to have lower ceilings and just as much or more volatility as veterans. Most will bust most weeks and you’ll have to be more lucky than good to use them when they go off. One trend I noticed when looking at this data was that like running backs, rookie wide receivers tended to peak mid-season, with an average of 4-5 fantasy points per game more in weeks 7-9 than any other time during the season. Injury to a WR1 or WR2 is the best opportunity to use a cheap wide receiver and it will likely increase his floor if he’s had the chance to play in a few games first.
Tight ends need to score touchdowns to be valuable, and touchdowns to tight ends are difficult if not impossible to predict for most teams. It’s a great spot to save money, but there are plenty of veteran tight ends who are a lock for six targets a game that are just as affordable as rookies. Consider this: I wasn’t surprised that Hunter Henry (recency) and Jordan Reed (stud) each had several games that made the rookie top 15 (fantasy points per game 2013-2016), but I don’t remember Will Tye and Tim Wright each having three of the top 15 rookie tight end games. These are touchdown-dependent players and outside of a few special cases, it is an exercise in frustration to rely on them in daily fantasy.
We all like new things. Trying new things, whether we end up liking them or not, is its own reward. The tingle of excitement we feel in a novel situation is the release of dopamine and norepinephrine, neurotransmitters that signal reward—in our brain and in the pit of our stomachs. As I always say, DFS should be fun, so roster guys you feel good about. If that includes a rookie, so be it. This article and the data within should serve as a guide to when is the best time to do that. Rookie quarterback? Go for one with a high implied team total when you think his ownership will be low. Rookie running back or receiver? Wait until about Week 5 or a significant injury to a starter provides more stable opportunity. Rookie tight end? No thanks.