In 2010, I started playing fantasy baseball. A friend was desperate for another owner, and asked me to join. He wanted to see what I knew about baseball, so he asked me if I heard of Hanley Ramirez, the typical first player drafted that year. I had no idea who he was talking about, as I had not followed baseball in many years, so I set out to do research on players. It was a few days before the draft and I only had time to research hitters. When draft day came along I just passed on starting pitchers for the first 15 rounds. I seemed to do very well that year, and repeated this “strategy” the following years.
I had accidentally stumbled on something – don’t draft starting pitchers early. So here I set out to understand why this is, and see if It holds up in other leagues.
Below is a plot of the (average) number of picks spent on hitters in the first 3 rounds versus the team’s standing (2011 data). As is very evident from the plot, those who are among the top of the standings used their early picks to take bats. Roughly a 1/3 of the owners that only took bats in those rounds are in the 1st, 2nd, or 3rd in their league. Similarly, 1/3 of the owners who took a SP are in 10th, 11th, or 12th.
This trend holds very well for Rounds 1 -5—those who concentrate on bats did significantly better. But at some point you need to start taking pitchers. Below is a similar plot to the one above, but rounds 4 through 8. The top two teams still took more pitchers than average, but as whole, it still not a good place to draft too many pitchers. This is the place where draft value of batters and pitchers normalize. At this point owners were drafting Kershaw, Weaver, Hamels, Verlander, etc…, who have not only lived up to their draft position, but have outperformed the pitchers drafted earlier.
This demonstrates that drafting pitchers early on was a bad idea, but it doesn’t answer the questions as to why. Here are the reasons:
- Pitchers are more injury prone. Two-thirds of all time spent on the DL are by pitchers.
- SP is deeper than hitters, and talent drops off slower than with hitters. This is simply because we only need 6 SP per team, but 13 batters per team, while an MLB team has a 5 SP rotation and a 9 batter lineup. Statistically speaking, since there are so many more batters than pitchers, there will be more batters who perform at a higher statistical significance. In other words, suppose you took a group of 200 people and measured their height, and took 100 people and measured their weight. The tallest person in 200 group to likely have a more extreme height than extreme weight of the heaviest person in the group of 100. You can see this effect in the player rater. The best batters are more valuable than the best pitchers.
- The average pitcher is less valuable than an average batter. Simple. Batters produce in 6 categories, while starting pitchers produce in 4. You can see this effect in the player rater, which measures expected points in the standings vs. replacement.
- There is much more luck in pitching. Rarely does a top 5 pitcher from a previous year make it into the top 5 again. But this is also the reason why drafting a pitcher in round 4-8 is such a good idea. There is a great deal of upside that luck can bring.