As a sports fan, you’ve likely seen the trading of draft picks that occurs at the NFL draft and you may have heard about NFL teams using charts indicating the value of those draft picks in order gauge those trades. For fantasy baseball purposes, values haven’t traditionally been applied to draft pick slots though they should be. Not all fantasy leagues allow for pick trading but knowing the value of draft slots can be worthwhile for other reasons as well. In my case, knowing the value of those draft slots can come in handy when doing research on draft trends and draft history. Recently, I started doing research of the past success of sleeper picks here and one issue that came up was comparing the risk of drafting a “sleeper” with the 80th pick versus how much risk there was 200 picks later. It’s obvious that not all draft picks are created equally but there isn’t a good way to make comparisons between them without a chart similar to the one that NFL GMs use. Time for Mr. Cheatsheet to rock the show and come up with that!
My method for this experiment consisted of taking the average draft position (ADP) data for the past five years (hat tip to Fantasy Gameday for his help) and then applying the WERTH roto value of each of those players for that year they were drafted there. Due to our limited data set, it became more worthwhile to evaluate ranges of draft picks as opposed to exact draft slots. In doing this analysis, I was able to come up with “expected” roto value for each range. Using those expected values, I was able to figure out how the value trended from start to finish in the draft and apply values to each draft range. Once I was done with all of that wizardry, here is the chart that I came up with:
The drop in value is pretty steep within the first few rounds, as expected, but starts to become more gradual as we go on in the draft. Granted, these are just arbitrary numbers but they are based on player performance within those draft slots over the past 5 years so it is a good starting point for gauging if you want to trade a draft pick. But, more than anything, these numbers are good for some further analysis on the draft pool in a given year. I’ll be referencing this particular post in some future articles which will utilize these numbers. Stay tuned.
NoisyFlowers12/16/2012 at 9:21 PM
I have a 20 team league where we begin by drafting draft positions. A chart like this allows me to quickly surmise if there is is better value from the top 10 position than in the bottom 10. A pick 1-10 plus a pick in the 30-40 range gives you 186 in total value whereas an 10-20 plus a 20-30 pick only gives you 174 in total value.
Luke12/18/2012 at 3:16 PM
Nice! That's a really interesting way to use the data too. Good idea. That might be worth some further analysis too in a future post.
NoisyFlowers12/19/2012 at 8:13 PM
Small quibble but shouldn't each group begin with a "1" number?
1-10, 11-20, 21-30, etc.
The way you are presenting the data now, it seems as if you're counting the same pick twice which I doubt was the intent.
Luke12/26/2012 at 6:30 PM
Yeah, it comes out funny but the reason I had done it that way was because someone's ADP may be 10.42 so they were in the 10-20 group as opposed to 11-20 group. So, to be fully accurate, it would really be the 10.00-19.99 group but looks a bit uglier in the graphic.
NoisyFlowers12/27/2012 at 4:13 PM
Thanks for the explanation. Happy Holidays!
Steve Slaveck03/09/2013 at 12:13 AM
Very cool indeed … but the standard deviation in those top 40 picks is significantly higher than the remaining 360 picks. Is there any way to finesse the data some more to narrow down the pick range at the top of the draft? Even 5 pick increments for the first 40 picks would be extremely helpful.
Elaine Day01/27/2015 at 6:16 AM
Thank you !!
My Blog: SoftballArena