The argument is simple. Steroid testing has been more or less effective for the last several years, power has dropped off, and therefore power hitters should come at a premium in fantasy drafts. It seems like an obvious conclusion as the number of home runs per game has dropped from the 2000 record season of 2.34 HR/game to last season’s 2.00. Less home runs raises the value of each home run – that’s simple math and indisputable.
But does it necessarily follow that the relative value of the typical home run hitter has risen versus the typical speedster? The answer is a fairly resounding no.
Let’s be clear what we’re talking about here. We’re setting out to debunk the idea that guys that contribute in home runs and RBIs are harder to come by than guys that contribute in stolen bases and runs and that you can get those speed guys to fill your middle infield slots easily later on in the draft. For this purpose we will be using data for the top 200 fantasy players from the 2006-2008 seasons with our main comparison being the total of the standard scores for stolen bases and runs on the one hand and home runs and RBIs on the other. For example, a player that was one standard deviation above the mean in steals and two standard deviations above the mean in runs scored would get a total standard score of three for the SB/runs category. Before going forward, I want to be clear that players that score runs don’t necessarily steal bases and vice versa – the average correlation between those two categories of 0.364 is relatively small compared to the 0.823 correlation for home runs and RBIs. This means that our SB/runs guys aren’t necessarily a homogeneous group, but for the purposes of this discussion we can live with that distinction. We’re simply checking to see if guys that can solidly contribute to runs and/or steals are more common than guys that contribute to RBIs and/or home runs. Furthermore, the distribution of steals versus home runs is very similar to the rest of the distributions, as we’ll see in the final chart of this article.
This chart of the 2006 total standard scores gives us a good look at the trends we’ll be discussing here:
The x-axis on the chart is the number of players that perform at the given level and the y-axis is the combined standard score for the two stats in each category (SBs+runs or HRs+RBIs). What we see right away is four trends – the top six or so power hitters are actually better than their speed counterparts. This means that at the very top end power is easier to come by than speed. For the next ten players or so power dips behind in commonality. After that, the large middle ground player either provides slightly more power or about the same until the bottom third of players, who are scarce on power compared to speed and run scoring. The difference is not that drastic at any point though, amounting to less than half a standard deviation’s difference in the top 180 players. Neither speed nor power are more common on a large level – in fact, they’re distributed fairly evenly.
2007 and 2008 provide more of the same:
If anything, the availability of talent has converged more over the last few years. In any case, when we overlay all three years, it’s pretty clear that the overall trend in the data is fairly static:
Lest we think that this chart is somehow colored by its normalized distribution or by our use of composite stats (standard scores of SBs + standard score of runs and standard score of HRs + standard score of RBIs), I’ve included the 2008 charts that show the raw stats – both the composite stats of SBs+runs and HRs+RBIs and the ‘pure’ power and speed stats of SBs and HRs:
No matter how we try to break things down, power is either equally or more available than speed. In fact, the last chart even explains why home runs seem to be more valuable relative to stolen bases at the tail end of each chart. Since there is a fairly decent population of hitters that steal very few bases, the drop-off in the run/SB category is more gradual. Once we’ve lost all meaningful contributions from the stolen base category just after the midpoint in our charts, the HR/RBI categories gradually catch up with the declines in run/SB. This isn’t evidence of a surplus of speed guys; it’s actually proof that they’re in shorter supply than the power guys.
Using the easier to read standard score charts (the first four in the article), I would lean toward saying that a bias toward either power or speed in typical roto leagues is a tactical error. Whatever valuation method you use for players – Standard Gain Points (SGP), Standard Deviations from the Mean, Eyeballing It – you should be placing equal value on contributions in power and speed. If other owners are overvaluing power, then you’ll have a great opportunity to ’steal’ some speed throughout your drafts.
The Loveable Losers (known as Bob in the parlance of those not addicted to fantasy sports) is a computer programmer and numbers junky from Clayton, Ohio.
Questions or comments for Bob? Post them in the Cafe Forums!
Want to write for the Cafe? Check out the Cafe's Pencil & Paper section!