Introducing Projected Availability

One of the most exciting aspects of drafting is seeing how long you can wait for players you want. Projected availability is a good way to figure that out.

The History of Projected Availability

I think the first person to describe the idea of Projected Availability was Jeff Zimmerman, when he posted a spreadsheet and explanation on RotoGraphs in 2017 ("ADP Availability Workbook"). He uses the NFBC average draft position (ADP), along with their min and max picks, to estimate a normal distribution of picks for each player.

But it seems to be catching on: Rudy Gamble of Razzball has also picked up on this tactic, and he outlined the usefulness of it in a 2020 article, "How to Do a Fantasy Baseball Snake Draft". I also noticed BaseballHQ's Ryan Bloomfield used this idea in his 2022 LABR draft ("Mid-round team building in LABR Mixed").

Implementing Projected Availability

If you read Jeff Zimmerman's article that I mentioned above, he explains that you need two data points to calculate projected availability:

1. Mean (ADP)
2. Standard deviation

Well, we don't know the standard deviation, so we're off to a rocky start. Jeff approximates it as 1/4 of the range between the min and max picks. So now we need three data points:

1. Mean (ADP)
2. Min pick
3. Max pick

Then we hit the second problem: The NFBC is the only site that publishes min and max picks along with their ADP. So Jeff's method works if you have a 15-team league like the NFBC, but is less helpful for other formats.

I figure, since we're already approximating the standard deviation, we might as well approximate the min and max, too. Looking at the NFBC data, I came up with this basic rule of thumb:

The range between min and max picks is approximately equal to a player's ADP.

So a player with an ADP of 6 is usually taken between picks 3 and 9. An ADP of 100 might mean the player goes between picks 50 and 150.

It's actually not that simple, as the range starts out a little tighter, and then gets wider as the draft progresses. There's also usually a wider range for pitchers than hitters. So there's' room to improve this, but it gets us started. And it works for any site that we have ADP for.

So I plug the following into a normal cumulative distribution function:

• X: your next pick number
• Mean: a player's ADP
• Standard Deviation: ADP / 4

That gives you a (very) rough percentage for how likely that player will be taken before that pick. I subtract that number from 1 to flip it around as the probability he is still available.

Using Projected Availability

On DraftKick, I show the projected availability for a player at your next two picks. So if you are about to make pick 2.19, it will show the projected availability of players at picks 3.30 and 4.43.

Knowing that this is just a rough approximation, I don't recommend any strict rules for following projected availability. But here are my guidelines:

• If there's a 60-70% chance or greater chance that a player will be available with the next pick, I'll usually wait to draft him.
• If a player I want or need dips below about a 20-30% availability with the next pick, I'll usually go ahead and take them.

I factor this in with a player's Live SGP, which tells my how important a player would be on my specific team.