Quantifying Age Premium III: Age Trends in Dynasty
I have been looking at the difference in dynasty and redraft values in my series on the age premium, which has shown a clear trend of owners failing to prioritize winning championships vs. controlling relevant, but inferior assets. Let’s continue looking at this phenomenon from a macro perspective.
That is to say, what can we learn by looking at age and position?
Age is Not Just a Number
I did not have access to a complete dataset for player ages that I could join to the ADP supplied in the Dynasty App and Best Ball App. I did have access to at least the first 167 names on the list. While this is not a large sample, they are the highest equity names in dynasty so there is still something to be learned. Grouping them by age confirms what you would have suspected for this sample:
The bulk of the age markup occurs for players 25 and younger, levels off between 26-28, and then the heavy discounts start at 29 (actually, it starts at 28 as removing Alfred Morris, who is undrafted in MFL10s but in the top 167 for dynasty, from the 28 bucket brings the discount from 0.02 to 0.12). It’s a good representation of the fact that owners will pay premium prices for rookies, sophomores, and some juniors even if they are largely just rolling the dice on what level of production they may actually achieve.
While many owners obsess over the dreaded “age cliff” at 28, it’s worth remembering that not all players fall into total obscurity. Many of them will decline but can still be highly productive in those later years. When you factor in that you can buy them at a discount, it is only decreasing the odds you’d need for them to stay productive and still be profitable in trading for them.
So while many age-sensitive owners are rushing to the payout window on AJ Green and Dez Bryant before they turn 28 this year, forward-thinking owners are likely recognizing that those two are still among the top seven projected WRs this year by the RotoViz Staff and can probably be acquired for less than they are worth.
When we add the 2016 projected fantasy points to the equation, buying into the age cliff seems even more tempting.
There is some skew to those early ages from the large amount of rookies and sophomores who are being undrafted in MFL10s but still valued in dynasty, but even just looking at dynasty players with a top 100 ADP shows that older players are either, A) more productive, or B) similarly productive but coming at a massive discount.
I cannot stress this point enough: the least difficult season to predict is the next one. Between the higher cost of acquisition and the lower projected output in the most predictive season you can analyze, you can see in a broad sense why buying youth is a negative expected value bet.
I graphed the projected points by age for each position group just to see if there were any other trends that could be exploitable:
It appears the WRs have the steepest line and QBs look to have the flattest. The slopes of the regression lines confirm this (you can read this as, “For every additional year of age, projected points go up by __”).
It should be noted that TEs were not found to be statistically significant at 95 percent confidence, but the other three confirm some common beliefs in the fantasy world:
- QBs are replaceable. While older is still better, the marginal difference between them is significantly smaller than the differences at RB and WR. The model suggests that a five-year age gap in QBs is only slightly more predictive of points than one-year for WRs. Consider Eli Manning and Jared Goff. Eli is 13 years the elder and the projected eighth best QB while Goff is 32nd but their difference in points is 109 (for the same dynasty ADP, by the way). Whereas the difference between say Brandon Marshall and Kenny Bell is almost double at 206 points with only eight years of age difference. The lesson here is obvious: don’t invest too heavily in QBs.
- WRs age better than RBs. There are far more WRs projected for success past 28 than there are RBs. Some of that is due to RBs wearing down from years of hard hits in the trenches and some of that is the advent of more committees (which might sound circumstantial but RBBCs are designed to minimize exposure to those hits to keep RBs fresher; not to minimize the amount of touches older backs get just to mess with fantasy owners). What this does show is that it is OK – if not preferred – to invest in older WRs since they are still scoring enough points (even if it’s at a declining rate) to be drafted highly in redraft.
The ZeroYouth Strategy?
Well that would be a sweet piece of branding, were it not misguided. The goal of constructing a dynasty team is still to balance age and production. So to take from this series, “That guy on the Internet told me to sell all my youth for older players because they’re more productive,” would be a misinterpretation.
Rather what I have attempted to prove is that there really is not that much productive youth to go around, and that it is better to shoot for older, shorter-term solutions than chasing the next big thing under 25. If antifragility is still a buzzword around here, then keep in mind that as high-priced youth flame out their opportunity typically falls back to the incumbents (hello, Kamar Aiken and Steve Smith).
In my last article, I showed a recent startup draft in which I drafted an older team while others bought youth. In the first 11 rounds, I drafted 12 players, all of whom had an MFL10 ADP in the ninth or lower and seven of them in the first four rounds. I have another team where four of the five starting WRs are all “dinosaurs” to the dynasty community: Marshall, Bryant, Larry Fitzgerald, and Doug Baldwin. Other than Bryant, I have a minimal amount invested in them, and I’m still sitting here with more depth and all my trade capital intact for when the right opportunity arises. These may not be the sexiest teams on paper, but they should get me to the playoffs.
A simple guideline to remember is that as age goes up points go up, and as points go up dollars go up.