ADP – Draft Position Consistency

Average Draft Position or ADP is very valuable. I weighed NFBC’s ADP into my personal NFBC Draft Champion rankings equally to my position-adjusted composite projections (Baseball HQ combined with RotoChamp composites (Steamer, ZIP’s, etc.)) and the results were great. I used this, our pitcher arsenal scores and these hitter analytics  as my three main draft tools.

I wanted to take evaluating draft position one step further. I want to understand how consistent each player is drafted near their ADP.

Howard Bender, who manages Mock Draft Army provided me with the raw draft results of 19 expert drafts (NFBC format: 2c-1b-2b-3b-ss-ci-mi-5of-9p). It’s not the largest sample, but they are expert drafts, and the drafts that I have taken part in definitely provided great references.

The Approach:

To evaluate the draft position consistency, I worked up the RMSE (Root Mean Squared Errors) for each player’s draft position. The RMSE represents the standard deviation of the differences (errors) between each player’s draft position and the average draft position (ADP). I am using RMSE instead of general standard deviations from a mean draft position because using RMSE assumes that we have a model/estimator or a “line of consistency” versus a simple stray from one single value.

The lower the value, the better — the closer each player was drafted to their ADP.

Draft Consistency Leaderboard:

Below is a matrix ranked by their #MockDraftArmy ADP. The fourth column presents their “Draft Consistency” score. The fifth column depicts how many standard deviations they are from the mean consistency score.

Notes:

• Mike Trout was picked first overall in every draft hence a perfect RMSE of 0.00 (lower the better).
• The last column highlights those in green who are consistently drafted around their average draft position (darker green > .50 SD more consistent than the mean draft consistency). Those in red are drafted all over the place (darker red < .50 SD less consistent than the mean draft consistency).
• You will have to scroll down to Hunter Pence (63), Victor Martinez (77) and Kenley Jansen (101) to see the first few guys with red last columns. No surprise here as all three have injury concerns or diminished PA/IP projections associated.
• If you scroll to the very bottom, you will see #294 (Teixeira) and onward in red. Many of these guys were undrafted at times and were therefore delegated (consistent) arbitrary values. We should therefore ommit them from this evaluation. Mark Teixeira was an arbitrary starting point for me to highlight based on the raw results provided.
• I clustered the players into “rounds” of 15 players (standard NFBC format is 15 teams). Within each round, I highlighted those drafted most consistently around their draft position in yellow: Mike Trout in “round one”; Hanley Ramirez; Justin Upton; etc. Addison Reed’s draft position is remarkably consistent – the only player with a draft position consistency almost a standard deviation more consistent than the mean draft consistency score after “round 11.” Those drafted less consistently than their average draft position are in red: Antony Rendon, Jose Reyes, Nelson Cruz, etc. The Chris Davis mystery is verified by his draft consistency score (again, relatively speaking in that “round”).

Player Notes:

• Mookie Betts is obviously trending upward.
• Some people like Brett Gardner’s 15HR-25SB potential. Some don’t.
• Yasmany Tomas is obviously a mystery. Put me on the Jake Lamb bandwagon.
• Steve Cishek was relatively consistent outside one uber-late draft position so remove him from your “less consistent list.”
• A.J. Pollock is drafted all over the place.
• Kris Bryant is trending upward (times ten).