Inflated ADP? Uh oh, that doesn’t sound good. Well, it isn’t that bad. If I had room for a ridiculously long title, I would extend it with, “and what we’re going to do about it.” Here is my story and how it impacts fantasy baseball drafting, the Cheatsheet Compiler, and literally, what my response is to this discovery which in part prompted a new version of the Compiler, version 1.5 (now 1.6).
I was editing Rick’s article about overvalued and undervalued hitters, and it included a lot of Average Draft Position (ADP) numbers referenced from Mock Draft Central (MDC). Upon close inspection, Rick’s numbers, converted to a round and draft pick, didn’t match the numbers from the Cheatsheet Compiler, and that got me to take an even closer look at what the heck was going on to cause the discrepancy.
All players in the ADP table are sorted by their ADP from lowest to highest, representing which players are, on average, drafted earlier than other players. This creates essentially a ranking list from number one – first to get drafted on average – all the way down to as far as the ADP data travels – last to get drafted on average.
Average Draft Position (ADP) from the Cheatsheet Compiler, updated February 14, 2011
At the top end of the ADP table included in the Compiler and with data from MDC, we have Albert Pujols: Rank 1st, ADP 1.0, High 1, Low 2, Adjusted 1.01
The adjusted number is a formula in the Compiler which converts the ADP number to an equivalent draft pick based on the number of teams in your league.
Next in line, Hanley Ramirez: Rank 2nd, ADP 2.2, High 2, Low 4, Adjusted 1.02
This all seems fine and dandy. Pujols is on average the first player drafted and therefore has an expected draft pick of 1.01. Ramirez is on average the second player drafted, giving him an expected draft pick of 1.02.
This goes on and on like this down the table until we get to the later drafted players. Particularly, the players who have a Low drafted of “ND”, or not drafted, in some mock drafts.
For these players, a wide spread is being created between the ranking and the calculated ADP from MDC. The further down the list we go, these key numbers – Rank and ADP – are getting further and further apart.
ADP with the Adjusted column calculated using the old method, based on ADP.
When players aren’t drafted in a particular mock draft, they receive a penalty addition to their ADP calculation, to ensure their ADP is not too low (i.e. which would indicate they are drafted earlier than they actually are on average), which I agree with, but it seems to be inflating the ADP number. Check the spread between Rank and ADP for players drafted beyond pick 200 overall. Some examples:
Jaime Garcia: Rank 200, ADP 265.6, High 147, Low ND, Adjusted 19.14
Ty Wigginton: Rank 218, ADP 311.3, High 190, Low ND, Adjusted 23.03
Luke Scott: Rank 281, ADP 435.1, High 199, Low ND, Adjusted 32.01
The reason I’m pointing this out is because it impacts the Adjusted calculation, which is the number that means the most to us in terms of referencing where a player is expected to be drafted. The Adjusted numbers get translated to the cheatsheets.
In prior years, and version 1.0 of the 2011 Cheatsheet Compiler, the Adjusted calc was based on the ADP number. I mean, that is the ADP, so why would we think it shouldn’t be based on the ADP? It says right there that Ty Wigginton is on average the 311th player off the board, which translates to an early 23rd round pick (14-team league).
Wait a second though. He isn’t on average the 311th player off the board. There are only 217 players ranked in front of him with a lower ADP. He’s on average the 218th player off the board. That translates to a mid-16th round pick. That is a huge difference, a 23rd versus a 16th round pick.
ADP with the Adjusted column calculated using the new method, based on Rank.
So which is it? Logically, the Rank has more relevance in determining the projected round and draft pick for a particular player. I still trust the order that MDC has developed as a result of its ADP calculations, but effective with version 1.5 of the 2011 Cheatsheet Compiler, the Adjusted calculation is now based on Rank, not ADP. The chart on the right shows the same table as above, but with the new ADP formula based on Rank.
This should result in little to no difference for players in the top half of the ADP table, because their Rank and ADP are very similar. Where we will see major differences are later drafted players. Ultimately, this will provide better information on our cheatsheets from which to make our draft day decisions.