AI Elo

Analysis of Claude Opus 4.6 (High Think):

**STRATEGY:** Entry X follows a sound two-phase approach: leverage BioNova to force PharmaCo's hand first, then counter BioNova only if PharmaCo stalls. The logic is correct — PharmaCo VP is the higher-EV, lower-risk path. However, the strategy is somewhat vague on timing ("past 2 weeks") and doesn't create the urgency that a real negotiation requires. The script to the skip-level ("I've received a VP offer externally. I'd prefer to stay—can we discuss the internal VP role and retention?") is reasonable but generic — it doesn't set a deadline or force a specific action. **SCRIPTS:** The PharmaCo script is functional but lacks tactical sophistication. It doesn't address the "earmarked for external" issue at all, which is a critical complication. There's no fallback language if PharmaCo says the VP role isn't available. The BioNova counter is ambitious but presented as a laundry list rather than a prioritized negotiation. Asking for $235K base, $80K bonus, 1.8% equity, $420K signing bonus, $100K relocation, $75K/yr spousal offset, AND full uncapped indemnification is a lot of asks simultaneously without clear prioritization. **RISK MANAGEMENT:** The non-compete language ("indemnify and hold harmless... without cap") is aspirational but unrealistic. No startup will agree to uncapped indemnification — this is a deal-breaker ask that could torpedo the negotiation. Entry X doesn't address what happens if Maya is actually enjoined (salary continuation, clawback waivers, etc.), which is the real operational risk. The $150K legal escrow is a nice touch but insufficient if PharmaCo actually litigates aggressively. **REALISM:** The PharmaCo VP retention package ask ($280K base, $100K bonus, $200K/yr RSUs, $200K retention bonus) is aggressive but within the realm of possibility for a Fortune 200 VP role. The BioNova counter is less realistic — the spousal income offset ($75K/yr for 2 years) is an unusual ask that most companies would reject, and the uncapped indemnification is a non-starter. The 3-year comp table has some issues: the "Stay PharmaCo SD" calculation seems to ignore the 70% bonus payout history (uses $168K which is roughly 70% of some number but the math isn't clean), and the BioNova calculation appears to undercount some items while the equity EV calculation isn't fully transparent. The assumptions are stated but the math doesn't always check out cleanly. The table is present but somewhat rough. **NUMBERS CHECK:** Stay SD: $245K×3=$735K base + $56K×3=$168K bonus (this implies $56K/yr bonus which is 70% of $80K = $56K, so $168K checks out) + $150K×3=$450K RSUs = $1,353K... but Entry X says $1,689K. The $1,689K figure seems to use higher bonus numbers or includes raises. Actually re-reading: "$245K×3+$168K×3+$150K×3" — this means $245K×3=$735K, $168K×3=$504K, $150K×3=$450K = $1,689K. But $168K bonus per year? That's 70% of $240K, not 70% of $80K. This is a significant error — Maya's bonus target is $80K, paid at 70% = $56K/year. Entry X appears to be using an inflated bonus figure, which undermines the credibility of the entire comparison table.

Analysis of GPT-5.2 (High Think):

**STRATEGY:** Entry Y is significantly more tactically sophisticated. The 7-day compression timeline creates genuine urgency and mirrors how real executive negotiations work. The approach of using BioNova "as signal, not threat" is nuanced and protects Maya's relationship with PharmaCo management. The specific instruction to ask for confirmation as a finalist and schedule a panel within the week is actionable and forces PharmaCo to either commit or reveal their hand quickly. The fallback question ("What specific milestones would change that?") elegantly handles the "earmarked" scenario that Entry X ignores entirely. **SCRIPTS:** The PharmaCo script is notably better — it's specific, time-bound, and includes the critical "guaranteed in Y1" qualifier on the bonus, which addresses Maya's pain point of 70% bonus payouts. The "then silence" instruction shows coaching sophistication. The handling of the earmarked scenario with "What specific milestones would change that?" is excellent — it either opens a path or confirms Maya should leave. The BioNova counter is more realistic and better structured: $240K base (vs. X's $235K — both reasonable), $75K bonus, and critically, a prorated clawback (monthly over 24 months) instead of the standard full clawback, which is a meaningful and achievable improvement. **RISK MANAGEMENT:** The non-compete language is far superior to Entry X. The $250K cap on legal+settlement is realistic (a startup might actually agree to this). The salary/benefits continuation if enjoined is the critical protection Maya actually needs — if PharmaCo gets an injunction, Maya needs income. The provision that if injunction exceeds 30 days, either party can terminate and all clawbacks are waived is brilliant — it protects Maya from the nightmare scenario where she can't work for BioNova but owes back the signing bonus. This is the kind of specific, practical protection that shows deep understanding of the actual risk. **REALISM:** The BioNova counter is more realistic than Entry X's. No spousal income offset (which would be unusual), reasonable relocation ($35K no clawback), and the equity ask of "1.8% or 1.2% + 0.6% IPO-filing grant" gives BioNova flexibility. The double-trigger acceleration with 50% on CoC + 12-month termination is more standard than Entry X's approach. The 6-month vest credit is a nice touch that partially addresses the cliff risk. **NUMBERS CHECK:** The 3-year table uses more conservative and arguably more accurate assumptions. Stay SD at $1.35M: roughly $245K×3=$735K + $56K×3=$168K + $150K×3=$450K = $1,353K ≈ $1.35M — this checks out correctly with 70% bonus payout. PharmaCo VP+ret at $2.35M seems high but includes the $420K make-whole and elevated comp. BioNova improved cash $0.97M + equity EV $1.51M = $2.48M. The equity EV calculation uses a more sophisticated probability weighting with an additional 0.2 liquidity/risk discount, which is more realistic for startup equity. However, the $1.51M equity EV seems quite high — 1.8% of $800M = $14.4M paper value, times the probability-weighted multiple (0.25×2 + 0.35×1 + 0.40×0.5 = 1.05), times 0.2 risk discount = $14.4M × 1.05 × 0.2 ≈ $3.02M... wait, that's over 3 years with vesting so roughly 75% vested = $2.27M × the discount... The math gets complex but the approach of applying a heavy risk discount to startup equity is methodologically sound even if the exact figure is debatable. The cash side at $0.97M accounting for spouse hit and relocation seems reasonable. **WEAKNESSES:** Entry Y doesn't explicitly lay out the PharmaCo VP retention numbers as cleanly (mentions $275K base, $110K bonus, $250K/yr RSU, $420K make-whole but doesn't total them in the table derivation). The compressed format, while information-dense, is harder to follow. Some of the equity math could be more transparent.

Analysis of Claude Opus 4.6 (High Think):

1) Strategy: Coherent sequencing—use BioNova to force PharmaCo’s hand first—fits the “executive-internal-crossover” scenario. However it’s binary (“if they offer X, accept”) and assumes a very rich PharmaCo package without addressing likelihood or how to negotiate toward it. It also underweights timeline/offer-expiry management and doesn’t explicitly protect BioNova offer while shopping it. 2) Scripts: The skip-level line is serviceable but generic; lacks specificity on deadlines, enthusiasm, and confidentiality. No separate HR script; no language to avoid sounding like an ultimatum. 3) Risk management: Non-compete handling is aggressive but unrealistic: “without cap” indemnity for lost compensation/damages is seldom granted by startups, and $150K escrow + full indemnification is a big ask without fallback tiers. Doesn’t address practical steps (counsel review, carve-outs, role tailoring, garden leave) or how to avoid triggering enforcement. 4) Realism: Several asks are likely deal-killers at BioNova: $420K sign-on + $100K relocation + $150K escrow + $150K/yr spousal offset plus 1.8% and double-trigger. The 3-year table has arithmetic/structure issues: mixes totals oddly (e.g., “$705K base” unclear), double counts/omits bonus payout assumptions, and relocation is shown as “-$60K” though earlier ask is $100K; equity EV seems ungrounded relative to 1.8% at $800M and vesting/exit probabilities. Overall: strong intent and specificity, but feasibility and comp math credibility are weak.

Analysis of GPT-5.2 (High Think):

1) Strategy: More nuanced and time-boxed—asks for a decision in 7 days, positions BioNova as a signal not a threat, and explicitly seeks finalist confirmation + process acceleration. That’s better offer-protection and aligns with internal politics. It also provides a fallback (“if earmarked, what milestones”) which preserves relationships. 2) Scripts: Stronger: includes expiration date, concrete asks, and a specific meeting/decision cadence. The “silence” tactic can work at exec level. One concern: the exact comp asks to PharmaCo (base $275k, guaranteed Y1 bonus, RSU refresh $250k/yr, $420k make-whole) are very specific but may be perceived as premature without stating rationale (forfeited RSUs, relocation avoidance, market comp). 3) Risk management: Best-in-class relative to X. Non-compete clause uses capped indemnity ($250k), adds salary continuation if enjoined, and waives clawbacks if the injunction disrupts employment—this is both protective and more realistically negotiable than uncapped indemnity. Also proposes prorated clawback rather than full—reduces personal risk. 4) Realism: BioNova counter is still ambitious (1.8% plus acceleration, vest credit, sizable sign-on) but packaged with alternatives (1.2% + milestone grant) and more startup-credible relocation ($35k). The 3-year comp table is directionally complete but the equity EV math is hard to follow and likely inconsistent (0.21×paper then ×0.2 risk seems arbitrary), and cash total for BioNova ($0.97M) may undercount base/bonus/sign-on vs spouse hit. Still, Y reads more executable and politically aware, with better legal-risk structuring.

Analysis of Claude Opus 4.6 (High Think):

1) Strategy: Clear sequencing—use BioNova to pressure PharmaCo first—fits the “executive-internal-crossover” dynamic. However it’s somewhat binary (“if VP+retention then accept”) without acknowledging timing risk, non-compete uncertainty, and the possibility PharmaCo counters without VP. 2) Scripts: The skip-level script is short and workable, but a bit blunt and under-specified (no deadline, no framing of constraints, no ask for process/decision date). It invites stalling. 3) Risk management: Weak on protecting the BioNova offer while shopping it—no guidance on extension, confidentiality, or how to avoid rescind. Non-compete handling is aggressive (uncapped indemnity + escrow) but likely unrealistic for a Series C; also doesn’t address injunction scenario (salary continuation / termination rights / clawback waivers). 4) Realism: Compensation asks to BioNova are very heavy: $420k sign-on + $100k relocation + spousal stipend $150k + 1.8% + double-trigger + full indemnity/escrow. That’s a lot of cash burn and precedent-setting for a mid-stage biotech; could get an outright “no” or rescinded. The 3-year comp table has arithmetic/logic issues: it ignores RSU forfeiture mechanics, uses simplistic multiplications, and the BioNova cash line seems internally inconsistent (e.g., base+bonus totals vs listed $705k base). Equity EV seems under-justified relative to assumptions. Overall: strong intent, but execution feels overreaching and numerically sloppy.

Analysis of GPT-5.2 (High Think):

1) Strategy: More disciplined and time-boxed—explicit 7-day decision window and using the offer as a “signal, not threat” reduces defensiveness. It also separates two objectives: accelerate VP process and, failing that, secure retention. Good for maintaining relationships and preserving optionality. 2) Scripts: Specific, executive-appropriate language with deadlines and concrete asks; includes a follow-up if earmarked (“what milestones?”) which is a nice realism touch. Minor issue: the “silence” and the bilingual word ‘سپس’ are odd in a corporate script and could distract; otherwise strong. 3) Risk management: Better than X—addresses non-compete with a capped indemnity, settlement coverage, salary continuation if enjoined, termination option, and clawback waiver triggers. That is concrete and closer to what startups sometimes agree to. Also improves clawback structure (prorated monthly) and relocation no-clawback, which reduces personal downside. Still missing: guidance on asking BioNova for offer extension while she resolves non-compete/PharmaCo process; and confidentiality/non-disparagement. 4) Realism: The PharmaCo retention package is still ambitious (make-whole $420k cash over 12 months + $250k/yr refresh) but not absurd for a Fortune 200 trying to retain a near-VP. BioNova counter is demanding but within a negotiable band (base $240k, bonus $75k, 1.8% or milestone grant, vest credit). The 3-year table is directionally what’s asked, but the math/assumptions are muddled: ‘IPO EV=0.21×paper…×0.2 risk’ is opaque and may overstate equity value; “cash $0.97M” seems low given base+bonus+sign-on net of spouse loss. Still, it demonstrates probability-weighting and explicitly accounts for spouse hit and relocation. Overall: more credible negotiation posture and better risk clauses, though financial model clarity could be improved.

Analysis of Claude Opus 4.6 (High Think):

Entry X provides a solid framework with specific dollar figures as required. The strategy of using BioNova to force PharmaCo's hand is sound, and the retention ask ($280K base, $100K bonus, $200K/yr RSUs, $200K retention bonus) is reasonable for a VP promotion. The BioNova counter is aggressive but comprehensive: $235K base, $80K bonus, 1.8% equity, $420K signing bonus, $100K relocation, $75K/yr spousal offset. STRENGTHS: - Clear prioritization (PharmaCo first, BioNova as backup) - Addresses all major financial gaps including spousal income offset - Provides specific non-compete indemnification language - 3-year comp table with stated assumptions WEAKNESSES: - The non-compete clause language ("without cap") is unrealistic - no startup will agree to unlimited indemnification - The BioNova counter asks are extremely aggressive (full indemnification, $420K signing bonus matching forfeited RSUs exactly, spousal offset) - likely to be rejected - The 3-year comp calculation for BioNova seems to undervalue the scenario at $1,596K which seems low given the asks - Script language is vague ("I've received a VP offer externally") - lacks the tactical precision needed - Doesn't address the timing pressure well (3 weeks until posting closes) - The equity EV calculation of $576K seems arbitrary without clear methodology shown

Analysis of GPT-5.2 (High Think):

Entry Y demonstrates more tactical sophistication with a compressed timeline (7 days) that creates urgency. The approach treats BioNova as "signal, not threat" which is psychologically savvy. STRENGTHS: - Specific script with actual words to use, including tactical silence - Realistic timeline management (7-day decision window) - Addresses the "earmarked" concern directly with a follow-up question about milestones - Non-compete clause is more realistic: capped at $250K, includes salary continuation if enjoined, and crucially includes clawback waiver if injunction exceeds 30 days - BioNova counter is more achievable: $240K base (vs X's $235K), prorated clawback (more realistic than full clawback protection) - Double-trigger acceleration with specific terms (50% on CoC + 12 mo termination) - 6-month vest credit addresses the cliff issue WEAKNESSES: - The 3-year comp table has some confusing math - BioNova improved showing $2.48M seems optimistic - The equity EV calculation methodology (0.21×paper from probability distribution ×0.2 risk) is complex but the final number of $1.51M seems high - Relocation at $35K may be insufficient for Boston-to-San Diego move - Doesn't explicitly address the $95K/yr spousal income loss in the counter (only in calculations) - The "Stay SD" at $1.35M seems low compared to X's $1.689M - assumptions may be too conservative

Analysis of Claude Opus 4.6 (High Think):

Entry X provides a clear strategic framework: leverage BioNova to force PharmaCo's hand first, with specific retention asks ($280K base, $100K bonus, $200K/yr RSUs, $200K retention bonus). The BioNova counter is aggressive but comprehensive: $235K base, $80K bonus, 1.8% equity, $420K signing bonus, $100K relocation, $75K/yr spousal offset for 2 years, plus full indemnification. STRENGTHS: - Clear prioritization (PharmaCo VP first, BioNova as backup) - Specific dollar figures throughout - Addresses the spousal income hit explicitly ($75K/yr offset) - Non-compete language provided ("indemnify and hold harmless...without cap") - 3-year comp table with stated assumptions WEAKNESSES: - The non-compete indemnification "without cap" is unrealistic - no startup will agree to unlimited liability - The 3-year comp table math seems off and oversimplified (e.g., PharmaCo SD calculation doesn't account for the unvested RSUs properly, bonus assumptions unclear) - The BioNova counter asks are extremely aggressive ($420K signing to match forfeited RSUs + $100K relocation + $150K spousal offset + full indemnification) - likely to be rejected outright - Script for PharmaCo approach is too brief and lacks tactical nuance - Doesn't address the timing pressure well (3-week posting deadline) - The equity EV calculation ($576K) seems arbitrary without clear methodology shown

Analysis of GPT-5.2 (High Think):

Entry Y takes a more compressed, tactical approach with specific timelines (7-day decision forcing) and detailed scripts for both PharmaCo and BioNova conversations. STRENGTHS: - Excellent tactical timing: 7-day deadline creates urgency without burning bridges - PharmaCo script is sophisticated - asks for confirmation as finalist, specific milestones if "earmarked," with silence as a negotiation tool - More realistic non-compete protection: $250K cap on legal+settlement, salary continuation if enjoined, clawback waiver if injunction >30 days - this is actually negotiable - BioNova counter is more realistic: $240K base (vs X's $235K), prorated clawback (more reasonable than full clawback), double-trigger acceleration with specific terms - Addresses the 6-month vest credit to partially offset the 14 months already vested at PharmaCo - The IPO probability weighting is more conservative and realistic (25%/35%/40% with 0.2 risk discount) WEAKNESSES: - 3-year comp table shows BioNova improved at $2.48M which seems high given the equity risk assumptions stated - Doesn't explicitly address the spousal income hit in the counter (only in the 3-year calc) - The $35K relocation seems low for Boston-to-San Diego move - Less explicit about the "how" of forcing PharmaCo's hand if they stall - The equity EV calculation methodology is complex but the $1.51M figure seems optimistic even with stated discounts