A Pro Football Focus study found that players labeled as "reaches" underperformed compared to peers taken at the same draft slot, while supposed "steals" did not outperform expectations for their draft position. The explanation rests on asymmetric error likelihood: a single team can err upward and overpay, but for a true bargain to exist most teams would have to mis-evaluate the same player despite access to extensive information. The same logic likely applies to the soccer transfer market, where overpaying by one club is common while collective undervaluation by many clubs is unlikely.
Every year, after the NFL draft, everyone is talking about "steals" and "reaches." The "steals" are the players that the football-watching public thinks went way later than their talent warranted. And the "reaches" are the players we thought went way higher than they should've. It turns out: we're only half right. A 2021 study by Timo Riske of Pro Football Focus looked back at six years of draft data and identified the players who went significantly higher or lower than the consensus of publicly available draft rankings. What he found is that the players who were "reaches" did underperform, on average, compared to the other players drafted at the same pick in other years.
The elegance of this study comes in the rationality of its explanation. For a player to be a true "reach," only one team has to make a player-evaluation mistake. For a player to be a true "steal," almost the entire NFL has to make a player-evaluation mistake, and NFL teams have access to way more information than the general public does.
I bring this up because I think a similar heuristic might apply to the soccer transfer market. It's very easy for one club to lock onto a player and pay way more than any other would have ever considered. It's much harder for every club with the requisite budget to undervalue the same talented player. This is why there's a common refrain among data-based thinkers in the soccer world: Hire me just so I can tell you "no" a couple times a year, and I'll be worth it.
Collection
[
|
...
]