Happy Thanksgiving! Regularly Scheduled Articles Will Resume Monday, December 1
February 22, 2013
The Merits of Category Dumping
Every year, there’s usually at least one fantasy player who poses a question along these lines:
What totals should I be targeting in my fantasy league this year? Is there a sweet spot I should be trying to aim for in each category?
Scott Wilderman of the stat service On Roto was kind enough to provide BP with a sampling of league totals from 2012. Below are the average totals for 42 American League, eight National League, and eight mixed 5x5 leagues.
Table 1: American League only 12-team, 5x5 Category Finish Averages
The bolded numbers indicate the highest and second-highest gaps between one category and the category ahead of or behind it. In every instance in Table 1, the biggest gap is between first and second place and 11th and 12th place. I found this surprising; most of the leagues I play in are keeper or carryover leagues, and there is usually a sizeable gap between sixth and seventh place. Half of the teams in my keeper leagues generally go for it, while the other half dump and play for next season. This wasn’t the case here.
Some of the larger gaps can be explained by outliers at the top or bottom of the standings, but not too many of them. There is definitely an inefficiency among owners at the top of the standings, as they fail to trade away their excess in a category. For winning teams, this doesn’t matter, but if you didn’t win and finished with a large excess in a category, a trading opportunity was missed.
At the bottom of the standings, it appears that there is an opportunity to toss a category overboard and play a nine-category game, but a deeper dive into the standings shows that this typically isn’t the case. One drawback to this approach is that success or failure is contingent on the competitiveness of your league. I found a 90-point squad (out of a possible 120 points) that won finishing dead last in ERA/WHIP, but there were not a lot of examples of winning squads with one- or two-point finishes in any category. Steals and saves—as you might expect—saw some low category finishes from winning teams, but no one- or two-point finishes.
Table 2: National League only 13-team, 5x5 Category Finish Averages
The largest gaps in the National League weren’t quite as uniform, but with six of the 10 categories showing the largest and second-largest gaps between first and second and 12th and 13th, the National League wasn’t exactly bucking the trend. The same difficulty with locating a great category-dumping strategy applies. One league winner did a nifty category dump where he finished 11th in batting average and wins and dominated his league, winning by 12 points. I’d love to know if this was by design or merely an accident.
Table 3: Mixed League 15-team, 5x5 Category Finish Averages
The mixed-league data present a more divided picture. Half of the largest and second-largest category differences are between teams at or near the top and at the bottom. However, the other half of significant category differences is between two teams at or near the bottom. The other thing that leaps out from Table 3 is that the category differences are much tighter than in the one-circuit leagues. This makes sense. Even in a “deep” 15-team mixer, there are still more everyday players to be had; it’s less likely that a squad is going to get decimated by injuries and be stuck carrying seven or eight everyday players and more than a few part-time fill-ins.
So are there definitive conclusions that can be drawn?
This is a problematic question. It is easy to go back and examine what happened last year, but harder to predict what will happen in 2013. Many projection systems spew optimistic totals and can’t or don’t predict the significant injuries that will inevitably dash someone’s hopes and dreams. Looking back at past results and attempting to balance this against future projections can and often does create an expectations gap.
Nevertheless, if you’re coming into your auction with limited freezes or a lopsided roster, it helps to know where the categorical soft spots are or might be. I don’t recommend category dumping as a primary option, but if your freeze list is weak, it helps to know where your league’s weak points are. If you’re going to take this approach, I recommend looking at your league data, as well, to ensure that you maximize the use of your rival owners’ trends and data points.