by Bill Macey

The following is a free sample of a column from Baseball HQ, one of hundreds readers receive during the course of their subscription. Get a competitive advantage in your league by subscribing today!

Average Draft Position (ADP) data has become an increasingly popular tool in draft preparation, as it helps you to develop a general sense of how the market values different players and allows you to identify potential bargains in your upcoming draft. Once you've identified your likely targets, you can formulate a draft plan to build a roster that gets the most benefit from selecting those few key players at a discount.

While useful, over-reliance on ADP data can be dangerous, especially so if the ADP data isn't applicable for your draft. ADP data typically comes from mock drafts held online and ADP results differ, sometimes substantially, from one site to the next.

A primary reason why ADP results can greatly differ between sites is the sites can have significantly different default ranks for those players. This manifests itself in two ways.

First, the mock drafts used to generate the ADP lists might not fully fill, leaving AI drafters to fill the empty spots. Typically these AI teams will select the top player on the draft board, subject to positional need. This phenomenon prevents the top-ranked players from falling too far down the draft board. Also, it is very common for drafters to drop out of the mock early, which typically happens between rounds 8-10. So, subsequent picks will revert to the ADPs and trend to the default values even more.

Second, even if the draft is full of active human participants, it's difficult to not be influenced by the names that are at the top of the draft board, which serve as a constant not-so-subtle reminder that a particular player is still available. Likewise, it's quite easy to forget about a player you like if you have to scroll far down the pre-rank list to find him. Out of sight, out of mind.

To see how strong these effects were, we compared ADP versus default ranking at three major online draft sites: ESPN, CBS, and Mock Draft Central (MDC). Data was collected February 28, 2011.

Consistent with expectations, we found that the ADP values closely mirror each site's particular default rankings.

• At ESPN, among the top 175 ranked players, the average difference between default ranking and ADP was just 2.8 places and 163 of the top 175 ranked players were drafted within 15 places of their ADP.
• At CBS, the average difference between default ranking and ADP for the top 200 ranked players was 4.6 places and 129 of the top 200 players had an ADP within 15 places of their ADP.
• At MDC, the average difference between default ranking and ADP for the top 200 ranked players was the largest at 7.4, but 166 of the 200 had an ADP within 15 places of their ADP.

Clearly, the default rankings greatly influence ADP. Another way to illustrate this is by looking at players with divergent default ranks across the different sites. A good early round example of this is Miguel Cabrera. With the recent news about his arrest for driving under the influence and the subsequent concerns about its implications for his playing time and performance, ESPN moved him down to #13 on their Top 300 list. He remains as the #3 and #4 player on CBS and MDC per their default rankings.

      Rank    ADP

      ====   ====

ESPN    13     11

CBS      3      5

MDC      4      3

We can see that even for a widely known player whose name has been all over the news, the ADP results still closely follow pre-rankings; ESPN drafters have moved him down their draft board, but he still goes within the first 5 picks at CBS and MDC.

The phenomenon continues later in drafts. Consider Adam LaRoche, another player whom the three sites value differently. Again, we see that the mock drafters follow the default ranks closely:

       Rank    ADP

      =====   ====

ESPN    129    147

CBS     236    240

MDC     160    179

This isn't always the case, though. Another player similarly slotted by the sites, Ian Desmond, sees his ADP converge.

       Rank    ADP

       ====   ====

ESPN    133    149

CBS     234    176

MDC     157    179

In both of these players, as well as other players with divergent default rankings, their ADPs trend slightly towards a consensus middle ground. That is, while their ADPs are strongly influenced by their default rank, there is some evidence that they are pulled towards an industry average.

Implications

How can we best use this information on draft day?

• If your league is holding an online draft, the ADPs provided by that service are your most reliable source for identifying likely bargains. Relying on another site's ADPs may influence you to draft a player earlier than necessarily or miss out on him altogether. (Although, if you really want a player and can't afford to not get him, it's best to ignore ADP entirely and take him were you value him.)
• Spend time with the draft software you'll be using on draft day and adjust your pre-rankings to reflect your own preferences such that the players you prefer are closer to the top of the list. You don't want to forget a player you really like simply because time was tight and you didn't scroll far enough down the list of available players.
• If there are any players you are keenly interested in drafting, consider looking at ADPs from other sites. If there are vastly different ADPs for those select players, continue to use the ADPs provided by the site hosting your draft as a rough guide for their value, but place less confidence in that data.
• If your draft is live, it may be worth reviewing the ADPs from multiple sources to see where there is general consensus and where players diverge. Certainly don't take ADPs from a single source as a reliable estimate as to how your draft will proceed.

This entry was posted on Monday, March 28th, 2011 at 11:34 am and is filed under A Word From Our Sponsors, Fantasy. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.