This article explains my Valuation for Snake Draft formats.

The method yields a ranked list that improves upon ordinary Value-over-baseline.

There’s lots more detail for people who want it, but in summary TLDR:

• The goal is to explain why my “Snake Draft Value” is an improved single metric for guiding draft decisions. (No need to mentally balance “VORP and PS%”.)

• The new metric helps to address Positional Scarcity, by applying Value Based Drafting (VBD) along with a modified version of “VONA” (Value Over Next Available).

• In my draft sheets, you can see the original Value-over-baseline displayed next to this Snake Value. So you can see if the player is valued relatively more vs. less, in a snake draft scenario. Players are sorted by Snake Value alone because it gives superior guidance for non-auction drafts.

My Motivation for a Valuation Dedicated to Snake Drafts

I created this new snake draft metric because I was aiming to give you a reliable and supremely easy-to-use ranking system.

While I was trying to prepare draft guidance, to make things ultra-clear and usable for you, I realized a few things about Positional Scarcity (PS):

1. The most useful aspect of Positional Scarcity is already incorporated by VBD calculations (because of ranking from positional baselines).

2. “PS%” (as a percent) is not related to points-optimizing your roster.

3. It’s even more important to address Scarcity of limited chances, in snake drafts.

So I devised a new evaluation that provides a modified “Value over baseline” ranking:

• It transforms the usual VORP-like auction rankings into an analogous static ranking list, suited to snake drafts.

• It converts PS% into a more useful form of expected positional “Opportunity Cost”, in fantasy points.

The result is a single metric that’s better for guaranteeing a minimum value in Snake Drafts. In the same way that Value-over-baseline provides a minimax solution to auction drafts.

First Understand Auction Drafts, to Compare Against Snake Drafts

Snake draft valuation is rooted in the same VBD approach used for auction drafts, as explained in this auction draft background.

Auction valuation is simpler. And it’s necessary background to understand how it’s modified when addressing snake drafts.

Snake Draft Prioritization: The Need to Examining Tradeoffs

There’s a special danger in Snake Drafts that doesn’t occur in auctions: There’s a special risk of downside from scarce opportunities to pick.

You all know the feeling: You make your draft pick, and then suddenly things are out of your hands. You’re at the will of your league. For maybe a dozen more picks, you have no influence over who will be left to you.

The limited number of chances means you need to consider tradeoffs between selecting different positions for each pick you decide on. You might rephrase these dilemmas as “tradeoffs in immediate, near-term positional scarcity”. You need to consider these tradeoffs in each round. You’re required to either take a position now, or to sacrifice value by waiting on that position.

These tradeoffs effectively distort the optimal valuation of players. In the Auction draft article, I mentioned that Value-over-baseline metrics provide a “minimax” solution. This means it’s optimized to protect you against getting a below-average roster. But this conclusion doesn’t apply to snake drafts. If you don’t modify the numbers, then your league-mates’ decisions could leave you with a weaker team.

Applying VONA (Value Over Next Available) With Modifications

To address the risk of losing positional value in a snake draft, the most important consideration is Opportunity Cost.

• Opportunity Cost (for a given position) = (Value of taking the best player at the position now) - (Value of the player expected to be remaining one round later)

This calculation is the essence of VONA (Value Over Next Available), which advises you to select whichever position has the biggest drop-off. Here', I’m only referring to VONA applied to sequential drafts; using VONA in auction drafts is different.

VONA is a decades-old idea. For those familiar with it, here’s some relevant context— and a description of how I’m applying it differently:

1. It’s generally acknowledged that you can never exactly know who will be remaining next round, so VONA requires assumptions to approximate it.

   • I assume your league will draft roughly in order of a VBD rankings list…

   • …but with smoothening of future picks, to account for uncertainty in the exact sequence.

2. VONA is usually meant to be calculated live, meaning VONA gets updated after every live draft decision.

   • I instead target an optimized static list, similar to usual Value-over-baseline rankings.

   • My claim is that these valuations will remain accurate unless your league-mates “reach” within a position.

3. The VONA concept can be been applied looking multiple rounds ahead.

   • In my case, you could say I effectively use both current-round VONA and “cumulative future VONA” of all remaining valued players.

   • I believe this is necessary to avoid hastily grabbing a position that actually has enough remaining players carrying decent value. It’s more important to take a position if it has fewer players left with decent value.

Methodology for Snake Draft Value

The basic 1-2 premise involves modifying a VBD draft list until there’s no apparent way to “lose out” against league-mates who make smart picks:

1. Take a set of rankings (Value-over-baseline), and step through each pick in order.

2. Update the list so that local drops in positional scarcity will boost players upwards in the list. Repeat.

So even if your opponent is holding your rankings list, they still don’t have any obvious way to maliciously make your team weaker.

If you want more detail, here’s what I actually do at each step:

1. At each pick, look 1 round ahead at which players might be remaining at the position.

2. Take the expected “value remaining” for players at the specific position, and smoothen them into a weighted average.

   • Early picks are likely to be taken closer to VBD order, so initially smoothen only by ±2 positions.

   • The end of the draft might smoothen by ±12 or so.

3. Use this weighted average of “future expected value remaining” to calculate the player’s opportunity cost (VONA).

4. Add a proportion of this opportunity cost to the player’s original Value-over-baseline.

5. Repeat for all players.

6. Re-order the original list (being careful for any player “cross-overs”).

7. Repeat until there are no more “inefficiencies” in the order.

This Approach Goes Beyond PS% Alone, to Optimize for Points.

I was thinking that some of you might wonder what happened to the percentage metric, labeled “PS%”. Did it disappear?

Yes and No.

In a hand-wavy way, the use of PS% now looks something like this:

Updated Player Value = (Cumulative PS% summed over all players at the position with positive value + PS%[earlier player] - PS%[later player, weighted average]) * (Total sum of all positive values of players at the position)

Unlike “PS%”, it’s necessary to multiply by total positional value here, to make sure you optimize for points across the different positions. Keep in mind: as I stated, PS% is more for a goal of roster balancing and not for points-optimizing. With the above procedure, we’re actually tackling both problems at once.