NSBA Draft Analyticsembargoed · 2026-06-06

A9_cs_strategy.md

A9 — The CS Specialist: A Draft-Strategy Framework for Cornering a Scarce Category

A9 — The CS Specialist: A Draft-Strategy Framework for Cornering a Scarce Category

Agent: A9 (CS draft-strategy, conceptual/strategic) Date: 2026-05-30 Scope: Conceptual framework only. A separate agent quantifies the empirical CS win-share from NSBA2/3 logs. This document tells that agent what to measure and why.

0. The Anomaly in One Paragraph

NSBA swapped the standard Science Bowl "Energy" category for Computer Science. Because CS is non-standard, the cross-over between "elite Science Bowl player" and "competent CS player" is small — CS skill is scarce in the player pool. Every match cycles all 6 categories ~equally, so CS is ~1/6 of points, and a team with zero CS coverage effectively forfeits an entire sixth of the board every match. This converts a normal "fill out your roster" problem into a positional-scarcity + monopoly-value problem that is well-studied in fantasy/real-sports drafting. The strategic question: how much should we reach, and when, to secure a scarce specialist who monopolizes an otherwise-uncovered category?

The short answer is that CS in NSBA behaves like an elite-TE / scarce-QB tier in fantasy: the points-over-baseline math (VORP) is modest, but the scarcity-adjusted, opportunity-cost math (VONA / "snake value") can be enormous — if the empirical drop-off (cliff) is steep and if CS is a meaningful share of decided points. Both are testable.

1. Core Frameworks Borrowed From Drafting Theory

1.1 Value Over Replacement (VORP) and replacement level by category

VORP measures a player's marginal production above a freely-available baseline at the same position — typically the player at the roster/waiver threshold (e.g., the WR65 in a 12-team league). The key insight transplanted to NSBA: value is positional, not absolute. A player is worth his points minus what you'd get for free at that slot anyway. (FantasyPros VORP; Draft Value Analytics)

The scarcity multiplier. Positional scarcity is "the degree to which elite players at a position are concentrated at the top," and "a position becomes scarce when the supply of dependable players starts falling faster than demand." (Yahoo — Positional Scarcity) When supply is thin, the drop-off from the best player to the next is steep, which mechanically raises the VORP of the top names at that position. This is the entire case for CS: if there are only ~6–10 genuinely CS-capable players across 14 teams, the replacement-level CS answerer is near-zero, so the best CS player's VORP is huge even if his raw points look ordinary.

1.2 Three baselines — and why the right one for CS is VONA, not VORP

Drafting theory uses three baselines, each answering a different question (FantasyPros):

Baseline Definition NSBA analogue
VORP (Value Over Replacement Player) Points above the best freely-available player at the position CS points above the worst CS-dabbler / a generalist who occasionally steals a CS tossup
VOLS (Value Over Last Starter) Points above the last startable player at that position CS points above "the CS guy every team is forced to roster"
VONA (Value Over Next Available) Points above the best player you expect to still be available at your next pick The decisive metric: how much better is this CS player than the CS player who survives to my next snake turn?

VONA "continuously shifts throughout the draft based on actual selections" and is the metric that explicitly encodes draft timing and scarcity — it tells you to take a player now if the next-best at that position won't survive the round-trip. For a scarce category in a 14-team snake (a long round-trip of up to 26 picks between your turns), VONA on CS can spike to near the player's full value, because the cliff after the top CS names may leave nothing by your next pick.

1.3 Tier-based drafting and the "cliff"

Tier-based drafting reframes the decision as opportunity cost: "your best pick is often the player who prevents you from falling off a cliff at a position." The operational rule: count the names left in the tier. (DraftSharks — Tiers; Athlon)

"If your tier has six players left and you are drafting soon again, you can usually take a different position now. If your tier has one or two players left, you should decide whether you want to pay to avoid the drop."

For CS this is the whole ballgame. If our empirical analysis finds, say, 3 CS players in Tier 1 and a steep cliff to replacement-level, then once 2 are gone the third is a "reach now or get nothing" decision — exactly the last-player-in-tier scenario. The "scarcity cliff" is "the draft slot after which a position's top-tier talent is exhausted." (Draft Value Analytics — Snake)

1.4 Snake-draft opportunity cost ("Snake Value")

Plain VORP fails in snake drafts because it ignores the immediate trade-off of taking a position now vs. a weaker player later: "It's possible that you can still end up with a below-average team, even when you use your pick on the player with best value-over-baseline." The fix is Snake Value = value-over-baseline + a fraction of the upcoming positional cliff (VONA), looking ahead ~one league round. (Subvertadown — Snake VBD) In a 14-team draft the look-ahead window is ~26 picks — long enough that a scarce category can fully empty between turns, which is precisely why CS deserves a scarcity boost on top of its raw value.

2. Need vs. Best-Player-Available (BPA) — when "fill the empty category" wins

The real-NFL consensus is BPA early, need later: "NFL teams that try to fill needs instead of going best player available most often fail… reaching for a need player most often blows your chances." But the same sources concede the exception: "around the third or fourth round, teams really dial in for depth… or go after players at a devalued position." (NFL.com — Best Available vs Need)

Why NSBA tilts toward need more than the NFL does. The anti-"reach" warning exists because in football a need-reach buys a worse player at a position you already partly cover. NSBA's CS situation is different in two load-bearing ways:

  1. A truly uncovered category is a step-function, not a marginal loss. A team with zero CS coverage doesn't lose "a bit of efficiency" — it concedes ~1/6 of every match's tossups outright. The marginal value of going from zero coverage to some coverage in a category is far larger than going from good to better in a contested category. This is the diminishing-returns logic that makes need-drafting correct precisely when the need is a categorical zero.
  2. The "reach" is cheaper than it looks if rivals undervalue CS (see §4). A reach is only costly if you're bidding against the field. If the field sleeps on CS, the "reach" price collapses toward BPA price.

Synthesis rule: BPA dominates inside contested categories; need dominates at the margin between "category covered" and "category empty." CS is the one category most likely to be empty for the field, so it is the single most need-sensitive pick on the board.

3. Monopoly / Cornering-the-Market Value

Economically, cornering a market means controlling enough of a scarce resource to "control the price" and force rivals out. (CFI; FasterCapital) The relevant transplant:

4. Game Theory: Is CS Mispriced by the Field?

This is the highest-leverage uncertainty, and it is a read on the other 13 GMs, not on the players.

5. The Concrete Framework: How to Value and Time CS Picks

Decision variables (to be filled by the empirical agent and live draft reads):

The CS value rule:

CS_pick_value = CS_VORP  +  scarcity_boost(D, N)  +  monopoly_boost(S)  −  contested_opportunity_cost
where scarcity_boost rises as D rises and N falls,
      monopoly_boost rises as S rises,
      and the reach is discounted by (low R) / penalized by (high R).

Timing (the snake-aware trigger): Reach for the top CS player one pick before the VONA cliff — i.e., the last turn at which we're confident the next-best CS player won't survive the ~26-pick round-trip to our next selection. Use the tier-count rule: act when CS Tier 1 is down to its last 1–2 names and rivals with no CS coverage pick before our next turn.

Conditions under which CS specialists ARE worth reaching for (all should hold): 1. Steep cliff — large D, small N (few viable CS players; sharp drop to replacement). 2. Meaningful share — S near or above 1/6; CS actually decides matches rather than being mop-up points. 3. We're otherwise at zero — we don't yet have any CS coverage (need > BPA at the empty-category margin). 4. Rivals are bidding (high R) OR the survivor pool won't last the round-trip — i.e., genuine risk of getting shut out.

Conditions under which we FADE / don't reach: - Low S — CS is a small or low-leverage share of points. - Generalists cover it — if "early-undergrad CS" is easy enough that strong all-rounders convert CS tossups anyway, replacement level is high, D collapses, and the monopoly is illusory. This is the single biggest threat to the thesis and must be tested first. - We already have adequate coverage — don't draft the #2 CS monopolist; spend on contested upside (avoid over-investment / §3 caveat). - The field is sleeping (low R) and the pool is deep enough to wait — capture the discount; don't pay a reach price for a player no one else wants.

6. What DATA Confirms (or Kills) This Framework

For the empirical agent working the NSBA2/3 game logs:

  1. CS points share & leverage (S). What fraction of tossups are CS? More important: in close matches, how often is the deciding margin attributable to CS tossups? (Win-share, not raw share.) Kills the thesis if CS is a small or non-pivotal share.
  2. Replacement level & cliff (D, N). Rank every player by CS conversion rate. Plot the curve. Is there a steep cliff after a top tier (thesis holds) or a smooth gentle slope (thesis weakens — generalists cover it)? Count the viable-CS tier sizes (N).
  3. The "generalist covers CS" test. What share of CS tossups were answered by players who are not CS specialists? High share ⇒ replacement level is high ⇒ monopoly is weak. This is the decisive falsification test — run it first.
  4. Empty-category penalty. Identify teams that had ~zero CS coverage in NSBA2/3 and measure their actual win rate / point margin vs. the field. Quantifies the step-function cost of conceding the category.
  5. Monopoly realization. For teams that did roster a top CS player — did they actually win the CS category most matches (monopoly held), or did opponents still split CS points (monopoly leaky)?
  6. Market price (R, historical). At what draft slot / ADP did CS players actually go in prior NSBA drafts? Compare to their realized value to estimate whether CS was historically under- or over-priced — directly sets our §4 aggressiveness.
  7. Over-investment check. Did any team roster 2+ strong CS players, and did the second one return value or sit idle? Validates the "secure one, don't hoard" rule.

Sources

NSBA Draft Analytics · embargoed until after the SSB draft · ← hub