How to Chose Your Keeper

  Guest Submission posted 8/25 by Patrick Prickett, Exclusive to Footballguys.com

I've played in a keeper league for several years and my strategy for picking keepers has always been the well thought out: "select the highest rated RB that you are allowed to keep." This year, due to our league rules, this strategy breaks down for me (unless I want Kevin Jones in the 4th round. Ugh).

So, I started thinking about how one should select their keeper?

Here are the basic strategies that most people seem to employ:

• Keep your Best RB
  • This is the scarcest position, and keeping your best RB allows you to go into the draft knowing you have one stud to build around
• Keep the Player with the Highest VBD
  • Whether it's LaDainian Tomlinson, Peyton Manning, Antonio Gates or Steve Smith. You keep the player with the highest VBD to get your team off to strong start
• Keep the Best Bargain
  • For example, say you selected Reggie Brown in the 13th last year and his ADP is the 6th round this year.
• Keep the Best "Value" Starter Bargain
  • Same as the strategy above, but omits players that are bargains, but who would not be an every week starter.

For the sake of argument, let's assume these are your potential keeper choices (I omitted the first round because many leagues don't allow first rounders to be keepers):

Which strategy should you employ? Who's the best choice in this case? A stud RB like Gore early? A bargain RB like Norwood late? A good starter in a middle round like Brees? A bargain QB late like Leinart or Cutler? The stud TE with a big VBD differential between the next guy? Looking at this list, all of these players seem like they could be potential keepers. But which one should you select?

Step 1: Assign the Selection Number to the Player

For simplicity, let's assume you have the 9th pick in a 12 team league regular serpentine draft. This means you will be selecting #9 in the first round, #16 overall in the second round (i.e. the 4th pick in the second round), #33 in the third round (i.e., the 9th pick in third round), #40 in the fourth, etc. Here's the full list assigning pick numbers to the players:

Step 2: Assign VBD Expected at that Pick Number

If you ignore for a moment the actual player above that could be selected at each pick, you can determine how much VBD you should be getting for each pick. After entering your scoring rules and VBD baseline projection in the Draft Dominator, you can produce the overall list of players - ranked by VBD. Using my league rules as an example, if this was a redraft league and everyone picked off the same projections, at #16 I would select Reggie Bush who has a VBD of 79, at #33 I would select Ronnie Brown with a VBD of 51, at #40 I would select Javon Walker with a VBD of 43, etc. In other words, with your pick #57, assuming everyone drafted perfectly, you should expect to get 35 VBD points (from Deuce McAllister). Thus, this is how much this pick is worth to you.

So assign the VBD for the pick number, to the actual player you have available to choose at that draft spot. Here's the example of what this might look like (using my league scoring system and VBD baseline projections):

Step 3: Assign VBD Expected by that Player

Now you know how much the draft spot is worth, but how much is the actual player you have available in that draft spot worth? Using the Draft Dominator again, I look up Frank Gore and see that under my rules I project him to have a VBD of 125, Antonio Gates is expected to offer 100 in VBD, Plaxico Burress is expected to offer VBD of 29, etc. So I add the VBD of all of my potential selections to the table below:

Step 4: Calculate VBD Gain

Now you know how much VBD you should expect out of the pick number, and you know how much VBD you can expect from that particular player, so you can calculate the VBD you gain (or loss) by keeping that individual player. For example, I expect to get 125 VBD points for Frank Gore but I only expect to be able to draft 79 VBD points from the #16 spot - so if I keep Frank Gore as my #16 pick I gain 46 VBD points (=125-79) beyond what I should have expected with that pick. After repeating this calculation for every player, you can generate the following table:

Step 5: Rank player based on VBD Gained

For convenience, sort the table for VBD Gain and you now have a list of who offers the most VBD in comparison to the spot you could select them in the draft:

So, according to this analysis, the best keeper choice would be Drew Brees, although Jay Cutler isn't far behind. [How many people expected that when they looked at the original list?] By selecting Brees, you get 60 VBD points more than you would expect from choosing someone else in this draft spot. It also tells us, that Plaxico Burress should definitely NOT be your choice because at #40 in the draft, you would expect to get more VBD at that pick than he is expected to provide.

Step 6: Assess your Competition

Now you know your best choice, but this assumes you are the only person selecting a keeper. However, the other 11 teams also select keepers as well. And their selections can affect your choice of keeper.

In an ideal world you would know all of your opponents' selections before you had to make your pick, but most leagues require keepers to be selected on the same day, so you may not know their selections in advance.

Fortunately you can repeat this analysis for your opponents' teams and predict who your opponents should select (and as a by product of this - you can also assess who enters the draft with the best keeper/highest VBD).

For the sake of argument though, let's assume you are in a RB loving league, and everyone selects their best RB, as the table below shows:

Under this scenario, if all 11 other teams keep a RB, you don't pick until #9 overall, and the first eight selections are all RBs - it is possible that the best RB you could draft at #9 would be the projected #19 RB (Thomas Jones). Would this make selecting Frank Gore or Laurence Maroney, both of whom had a VBD of 46 in step 5 above, a better choice than Drew Brees and his VBD of 60?

To determine this, you need to redo steps 1-5 after removing the 11 players listed above from the player pool. [NOTE: These players must be removed from the player pool altogether because their removal changes the VBD for all remaining players.]

I won't redo all five prior steps in detail, but this table summarizes the results of the analysis after removing the 11 opponents' keepers:

As you can see, the value of Drew Brees actually increases from 60 VBD to 61 VBD due to your opponent's keeper choices, but he is no longer the best choice. Because all of your opponents have selected RBs as their keepers, the VBD for the remaining RBs has increased even more, and therefore your keeper should either be Laurence Maroney or Frank Gore.

[NOTE: Many people are tempted to think that Maroney would be the better choice because he is a 5th round pick and Gore is a 2nd rounder. This is incorrect. Both are equally valuable as Gore's nominal VBD of 177 is 92 better than the VBD you should be able to draft at that spot; and Maroney's nominal VBD of 124 is 92 VBD better than what you should be able to draft at that spot.]

One final note, if your opponents don't keep their best players based on VBD gain, it is even better for you as more value remains on the board. However, this does create a problem as the value for the players in step 6 won't be exactly accurate. The good news though is that in most cases this does not matter. In many cases you won't even need to complete step 6, because you only have 1-3 viable choices based on their VBD gain versus their draft position anyway [I cherry-picked the best selections for my entire to develop the potential keeper list for this article]. And as illustrated above, even picking Brees would not be catastrophic.