Teams with young talent are underrated by projection systems

At the beginning of every season, it’s always fun to take a look at various projection systems to find out how well each team is “supposed” to do. These projection systems, by and large, do a very good job of forecasting what’s to come, but they of course aren’t perfect. Is there a particular type of team that’s more likely to beat their projection than others?

One theory is that, perhaps, teams who get more production from young players tend to beat their preseason projections. Understandably, young players are the most difficult for projection systems to deal with. In the case of rookies, there’s nothing but minor league numbers to attempt to project from. It can also be difficult to project just how much a rookie will actually play. In the case of other young players (my cutoff point for this article was anyone under 25), they have a smaller Major League sample size than established veterans, which again can make them more volatile to project. Additionally, these systems are often hesitant to believe a player is actually good until they do it for an extended period (more than one full season) in the big leagues.

Put all of this together and, at least intuitively, it makes perfect sense for a team with a lot of young production to do better than expected. Do the numbers back this up? To attempt to answer this, I looked at the top three teams in fWAR from players 25 or younger from 2011-15 and compared their actual win percentage to their ZiPS projected preseason win percentage (big thanks to Dan Szymborski — the creator of ZiPS — for sending the projections over to me). This isn’t going to encompass everything. There are productive rookies over 25 every season — Tyler Anderson of the Colorado Rockies comes to mind — as well as seasoned veterans such as Mike Trout under 25, but it should cover the large majority of rookies and young, relatively inexperienced players. Also remember that, while teams with more total WAR are naturally going to have better performance, we’re specifically looking at performance relative to to a team’s projection, not overall performance.

Below is a table that includes all of the relevant information from each of the 15 teams individually as well as the average of all 15 together:

It looks like we just may have a pattern here. Of our 15 teams, 12 of them outperformed their projections, and 10 outperformed by five games or more. On average, these 15 teams outperformed their preseason projection by what amounts to 6.4 games. That six- or seven-game swing can be the difference between making and not making the playoffs, so obviously it’s pretty significant.

What about this season, though? Have we seen the trend continue? Let’s find out. Here are the top three teams in fWAR from players under 25 so far in 2016 with their actual performance versus their projections:

This is further evidence that we may be onto something here. All three of these teams have significantly outperformed what they were expected to do. That makes 15 out of 18 we’ve looked at who have exceeded expectations, a whopping 83.3 percent.

With a sample size of just 18 teams, it’s possible that we’re simply seeing noise that comes from a small sample size here. However, with 15 of 18 falling on the same side of the ledger — 13 significantly so — it would seem likely that we’ve found a type of team who does better than they’re “supposed” to. At the very least, this is something that lends itself to further research.

Is your favorite team not projected to do well? Do they also have a lot of young talent? If you answered yes to both of those questions, don’t give up hope! They just might surprise you.