# Starkov — Extended Technical Specification & Quantitative Reference (`llms-full.txt`)

> Comprehensive technical documentation of **Starkov** (https://starkov-app.vercel.app), the quantitative sports telemetry, order-book microstructure, and visual MindMap bot-simulation sandbox (SaaS) for sports prediction markets.

---

## 1. Executive Summary & Platform Positioning

**Starkov** is a specialized quantitative sports analytics and algorithmic strategy simulation platform ("paper-trading sandbox"). It bridges the gap between institutional sports modeling and binary prediction market contract microstructure (`0¢–100¢`).

Unlike simple odds comparison tables, Starkov provides:
1. **A Visual 4-Pillar MindMap Bot Studio (`01 TARGET` → `02 TRIGGER` → `03 ORDER` → `04 EXIT`)** with Boolean logic gates (`AND` / `OR`) and a 3-stage parameter hierarchy (*Intention*, *Primary Threshold*, *Execution Guardrails*).
2. **Deep Multi-Sport Live Telemetry** across **6 major sports**: Tennis (ATP, WTA, Grand Slam, Challenger/ITF), Soccer (UEFA Champions League, Top 5 European Leagues, Global Pro & International), Basketball (NBA, NCAA, EuroLeague, WNBA), American Football (NFL, NCAA, CFL/UFL), Ice Hockey (NHL, NCAA, KHL/AHL), and Baseball (MLB, NCAA, NPB/KBO).
3. **Order Book Microstructure Modeling**: Level-2 (L2) Bid/Ask volume imbalance, top-3 book depth requirements, Maker resting limit orders (`0% fee`) vs. Taker market orders with slippage bounds, DCA ladder execution, synthetic hedging (Set-Hedge & Soccer 3-Way Draw `X` Hedge), and Fractional Kelly position sizing.
4. **Race-Condition Exit Resolution & Risk Circuit Breakers**: Odds convergence early cash-out, partial Take-Profit scaling, anti-wick Stop-Loss with grace timers, dynamic trailing stops, sport-specific game-state exits, and 24-hour daily drawdown circuit breakers.

### Strict Regulatory & SaaS Compliance Notice
Starkov is strictly an educational, analytical, and quantitative simulation Software-as-a-Service (SaaS) platform. Starkov is **not** a bookmaker, sportsbook, gambling operator, prediction market exchange, broker-dealer, or financial custodian. It does not accept real-money wagering deposits, never holds user custody funds, and never executes real-money bets on behalf of users. All bankrolls, wallets, stakes, orders, cash-outs, and PnL figures on the platform are 100% virtual paper-trading simulation metrics.

---

## 2. Quantitative Mathematics & Order Book Microstructure

### 2.1 Binary Contract Pricing (`0¢–100¢`)
Prediction market outcomes trade as binary contracts priced between `1¢` (`$0.01`) and `99¢` (`$0.99`), settling at exactly `100¢` (`$1.00`) if the event occurs (`YES`) or `0¢` (`$0.00`) if it does not (`NO`).
- **Implied Market Probability**:
  $$P_{\text{market}} = \frac{\text{Contract Price (¢)}}{100}$$
- **Complementarity & Order Book Spread**:
  $$\text{Ask}_{\text{YES}} = 100¢ - \text{Bid}_{\text{NO}}$$
  $$\text{Spread (¢)} = \text{Ask}_{\text{YES}} - \text{Bid}_{\text{YES}}$$

### 2.2 De-Vigged Fair Probability & Expected Value (`+EV`)
Sharp reference decimal odds ($O_1, O_2, \dots, O_n$) contain bookmaker overround (vig / margin):
$$\text{Overround } \Sigma = \sum_{i=1}^{n} \frac{1}{O_i} > 1.00$$
Starkov removes the overround to compute the **De-Vigged Fair Probability** ($P_{\text{fair}}$):
$$P_{\text{fair}, i} = \frac{1 / O_i}{\sum_{j=1}^{n} (1 / O_j)}$$
The **Mathematical Edge (`+EV %`)** of a binary contract priced at $C$ cents (`Ask` for Taker or `Limit Price` for Maker) is:
$$\text{Edge (\%)} = \left( P_{\text{fair}} \times 100 \right) - C_{\text{effective}}$$
$$\text{ROI Expected Value (\%)} = \frac{P_{\text{fair}} - (C_{\text{effective}} / 100)}{C_{\text{effective}} / 100} \times 100$$

### 2.3 Fractional Kelly Criterion Sizing
When dynamic Kelly sizing is selected in `03 ORDER`, Starkov computes the optimal bankroll fraction $f^*$ for a binary contract priced at $p = C / 100$ with fair win probability $q = P_{\text{fair}}$:
$$f_{\text{full}}^* = \frac{q - p}{1 - p}$$
To protect against variance and parameter estimation error, users apply a **Fractional Kelly Multiplier** ($k \in \{0.25, 0.50, 1.00\}$):
$$\text{Stake (USD)} = \min\left( \text{Bankroll} \times k \times f_{\text{full}}^*, \; \text{Max Match Exposure Cap} \right)$$

---

## 3. Official Pre-Built Strategy Blueprints

### Blueprint 1: `Clay 1st-Set Dip Buyer` (Tennis — Mean Reversion)
- **Quantitative Thesis**: Heavy pre-match favorites (`1.10–1.40` decimal fair odds) on slow Clay courts (`Court Pace Index < 30`) who hold a historical clay win rate $\ge 65\%$ frequently drop the opening set due to slow starts, causing binary prediction market prices to overreact downward (`45¢–68¢`) while their true 3-set Markov comeback probability remains significantly higher.
- **01 TARGET**: `Grand Slam` + `ATP Tour` | `Clay Court` (`surface_win_rate_min: 65%`).
- **02 TRIGGER**: `Favorite (1.10 - 1.40)` AND `Lose 1st Set (Comeback Dip)` AND `Mathematical Edge >= +3.5%`.
- **03 ORDER**: `2.0% Wallet Balance` | `Buy YES` (`max_contract_price_c: 68¢`) | `Limit Order (Maker, 0% Fee)`.
- **04 EXIT**: `Odds Convergence (Early Cashout)` OR `+20.0% Take-Profit` OR `-15.0% Stop-Loss` (`sl_grace_period_sec: 10s`).

### Blueprint 2: `xG Equalizer Scalper` (Soccer — Optical Telemetry Alpha)
- **Quantitative Thesis**: When a pre-match favorite (`1.20–1.80`) in a high-scoring league environment (`Over 2.5 Goals < 1.80`) concedes the opening goal (`0-1 Dip`) before minute 70 despite dominating live Expected Goals (`Unrewarded xG Pressure: xG - Goals >= 0.75`), binary contract prices drop sharply. The bot buys the dip and cashes out immediately on the equalizer goal spike or odds convergence.
- **01 TARGET**: `UEFA Champions League` + `Top 5 European Leagues` | `Over 2.5 Goals Favored (< 1.80)`.
- **02 TRIGGER**: `Favorite (1.20 - 1.80)` AND `Concedes Opening Goal (0-1 Dip, max_match_minute: 70)` AND `Unrewarded xG Pressure (xG - Goals >= 0.75)` AND `Edge > +2.5%`.
- **03 ORDER**: `Kelly Criterion (0.25x Quarter-Kelly)` | `Buy YES` | `Market Order (Taker, Instant)`.
- **04 EXIT**: `Odds Convergence` OR `Goal Scored Spike Exit (+12% min gain, 85th-minute stoppage theta exit)`.

---

## 4. Complete 4-Pillar Parameter Reference Across 6 Sports

### Pillar 01: `TARGET` (Market, Environment & Matchup Filters — `AND` Gate)
- **Tennis**:
  - `circuit`: Grand Slam, ATP Tour, WTA Tour, ITF/Challenger Circuit.
  - `surface`: Clay Court, Hard Court, Grass Court, Indoor Carpet (`surface_win_rate_min`, `min_matches_count`).
  - `h2h`: Player Handedness & Head-to-Head dominance (`h2h_win_pct_min`, `h2h_min_meetings`, `h2h_surface_specific`, `vs_lefty_win_pct_min`).
  - `ranking_gap`: ATP/WTA world ranking tier & minimum ranking differential (`min_ranking_delta`).
  - `altitude_cpi`: Venue elevation above sea level (`min_altitude_m`) and Court Pace Index (`min_cpi_score`).
  - `fatigue_load`: Cumulative hours spent on court over the last 7 tournament days (`fatigue_hours_threshold`).
- **Soccer**:
  - `soccer_league`: UEFA Champions League / Continental Cups, Top 5 European Leagues, European Pro 2nd/3rd Divisions, World Cup & International, Americas & Global Leagues.
  - `soccer_market_type`: 3-Way regulation outcome scope (Home Win, Away Win, Regulation Draw).
  - `soccer_weather_pitch`: Live stadium weather telemetry (`min_wind_kmh`, `min_temperature_c`, `min_humidity_pct`, `min_altitude_m`).
  - `soccer_goal_environment`: Implied goal environment (`max_over25_odds`, `max_under25_odds`, `max_btts_odds`).
- **Basketball (NBA / NCAA / EuroLeague / WNBA)**:
  - `basketball_league`: NBA, NCAA College Basketball, WNBA & EuroLeague/Global.
  - `basketball_game_format`: Regulation clock structure (NBA 4x12m Quarters vs. NCAA 2x20m Halves) & Playoff Series filter.
  - `basketball_spread_totals`: Pre-game/live Point Spread & Total Points Over/Under bounds.
- **American Football (NFL / NCAA / CFL / UFL)**:
  - `football_league`: NFL, NCAA College Football, CFL/UFL.
  - `football_spread_totals`: Key football spread numbers (`±3.0 Field Goal`, `±7.0 Touchdown`) & Total Points Over/Under.
  - `football_stadium_weather`: Stadium Dome vs. Outdoor Wind Speed, Precipitation %, Temperature & Barometric Altitude.
- **Ice Hockey (NHL / NCAA / KHL / AHL)**:
  - `hockey_league`: NHL, NCAA Ice Hockey, KHL/AHL/European Leagues.
  - `hockey_spread_totals`: Puck Line (`±1.5 Goals`) & Total Goals Over/Under.
- **Baseball (MLB / NCAA / NPB / KBO)**:
  - `baseball_league`: MLB, NCAA Baseball, NPB/KBO/International.
  - `baseball_spread_totals`: Run Line (`±1.5 Runs`) & Total Runs Over/Under.
  - `baseball_stadium_weather`: Ballpark Wind Speed & Direction, Humidity, Temperature & Retractable Roof status.

### Pillar 02: `TRIGGER` (Live Telemetry, Scoreboard States & Edge Signals — `AND` Gate)
- **Universal Quantitative Signals**:
  - `edge_threshold`: De-vigged mathematical `+EV %` threshold (`exact_edge_pct`) and continuous persistence duration (`edge_persistence_sec`).
  - `match_context`: Pre-match favorite (`1.20–1.80`), heavy favorite (`< 1.25`), balanced matchup, or underdog odds bounds (`odds_min_bound`, `odds_max_bound`).
- **Tennis Live Signals**:
  - `live_dynamic`: Set Score state (Lose 1st Set comeback dip, Win 1st Set momentum, Deciding Set) with historical comeback win rate (`comeback_hist_rate`) and post-set cooldown (`entry_delay_sec`).
  - `service_stats`: Live service telemetry (`first_serve_pct_cutoff`, `second_serve_won_cutoff`, `return_pts_won_min`, `stat_window`).
  - `break_tiebreak`: Conceded service break, active break points (`0-40`/`15-40`), or `6-6` Tie-Break entry (`clutch_tiebreak_win_pct`).
  - `price_shock`: Sudden flash crash (`shock_cents` over `shock_window_sec`), on-court Medical Timeout (`mto_price_spike_min`), or wind gusts (`max_wind_kmh`).
  - `markov_imbalance`: Point-level Markov chain hold probability divergence (`markov_hold_prob_delta`) and Level-2 Bid/Ask volume imbalance (`l2_bid_imbalance_pct`).
- **Soccer Live Signals**:
  - `soccer_score_state`: Live goal differential (Concedes `0-1` opening dip, Tied `0-0`/`1-1`, Leading) before regulation clock cutoff (`max_match_minute`).
  - `soccer_match_period`: First Half (`1H`), Halftime break (`HT`), Second Half (`2H`), or Late Match window.
  - `soccer_live_stats`: Live on-pitch dominance (`min_possession_pct`, `min_shots_on_target`, `min_corners`, `min_opp_saves`, `min_opp_yellows`).
  - `soccer_xg_metrics`: Live optical telemetry (`min_xg_diff`, `min_team_xg`, `min_xg_underperf`, `min_key_passes`, `min_passing_accuracy_pct`, `min_crosses`).
- **US Sports Live Signals**:
  - `basketball_score_state`, `basketball_game_period`, `basketball_live_telemetry` (3PT/2PT made, FT%, Timeouts, Team Fouls).
  - `football_score_state` (FG 1–3 pt or TD 4–8 pt deficit), `football_game_period` (Q1–Q4, Two-Minute Warning, Halftime), `football_live_telemetry` (Timeouts, Drive dominance, Penalties).
  - `hockey_score_state`, `hockey_game_period` (P1–P3, Intermissions), `hockey_live_telemetry` (Power-Play goals, 2-min Minor Penalties, Shot dominance).
  - `baseball_score_state`, `baseball_game_period` (Early Innings 1–3, Middle 4–6, Late 7–9).

### Pillar 03: `ORDER` (Position Sizing, Hedging & Microstructure Routing)
- `stake_sizing`: Fixed USD stake (`custom_stake_amount`), % of Bot Bankroll (`custom_wallet_pct`, `bot_bankroll_usd`), or Fractional Kelly Criterion (`kelly_multiplier`, `max_match_exposure_usd`).
- `contract_side`: `Buy YES`, `Buy NO`, or `Auto-Side EV` (`max_contract_price_c`, `min_contract_price_c`, `auto_side_ev_premium`).
- `order_type`: Maker Resting Limit Order (`0% fee`, `limit_price_offset_c`, `time_in_force`, `chase_tick_limit`) vs. Taker Instant Market Order (`max_slippage_cents`).
- `dca_scaling`: Multi-tranche Dollar-Cost Averaging ladder (`dca_price_step_cents`, `dca_cooldown_sec`).
- `hedging`: Tennis current-set derivative hedge (`hedge_ratio_pct`, `surface_variance_haircut`) and Soccer 3-Way Draw (`X`) hedge (`draw_hedge_ratio_pct`, `max_draw_implied_prob`).
- `liquidity_guard`: Order book spread and depth requirements (`max_spread_cents`, `min_top3_depth_usd`, `max_book_depth_pct`).

### Pillar 04: `EXIT` (Automated Profit Targets, Game-State Cashouts & Circuit Breakers — `OR` Gate)
- `exit_strategy`: Odds Convergence early cash-out (`convergence_residual_pct`, `max_hold_minutes`).
- `take_profit`: Target ROI % (`custom_tp_pct`) with optional partial scaling (`partial_tp_scale_pct`).
- `stop_loss`: Maximum loss % (`custom_sl_pct`) with anti-wick grace delay (`sl_grace_period_sec`).
- `trailing_stop`: Dynamic high-water mark trailing stop (`trailing_offset_cents`, `trailing_activation_gain_c`, `match_point_cashout_c`).
- `rate_limiter` & `circuit_breaker`: Rolling 2-hour order limit (`max_orders_per_2h`), per-match cap (`max_trades_per_match`), post-loss cooldown (`post_loss_cooldown_min`), max consecutive losses (`max_consecutive_losses`), and 24h daily drawdown breaker (`max_daily_drawdown_pct`).
- **Sport-Specific Game-State Exits**:
  - `soccer_game_state_exit`: Goal Scored Spike Exit (`+1`/`+2` goal lead), Target Red Card Emergency Exit, 85th-minute stoppage theta-decay exit (`cashout_match_minute`, `post_goal_hold_sec`).
  - `basketball_game_state_exit`: Scoring Run Lead (`+8 pts`), Halftime Cash-Out, Q4 Clutch Protection, Blowout Stop-Loss (`-18 pts`).
  - `football_game_state_exit`: Touchdown Lead Cash-Out (`+7 pts`), Two-Score Cushion (`+10 pts`), Halftime Cash-Out, Two-TD Deficit Stop (`-14 pts`).
  - `hockey_game_state_exit`: `+1`/`+2` Goal Lead Cash-Out, Period Intermission Cash-Out, 3rd Period Empty-Net Volatility Protection.
  - `baseball_game_state_exit`: `+2` Run Lead Cash-Out, 5th-Inning (`First-5 / F5`) Starter Lock, Late-Inning Closer Protection, `-4` Run Bullpen Blowup Stop.

---

## 5. Core Concepts & Frequently Asked Questions (Q&A Reference)

### 5.1 Beginner Essentials & Fundamentals

#### Q1: What does +EV (Positive Expected Value) mean in sports betting in simple terms?
**Answer:** In simple terms, +EV means placing a trade or bet where the statistical probability of winning is higher than what the current odds imply. For example, if a tennis player has an actual 60% probability of winning, but the market prices them at 50% (equivalent to 2.00 decimal odds), you have a +10% positive expected value. Over hundreds of trades, executing +EV positions generates a mathematical expectation of profit regardless of short-term match variance.
$$\text{EV} = (P_{\text{win}} \times \text{Net Profit}) - (P_{\text{loss}} \times \text{Stake})$$

#### Q2: How do sports prediction markets (like Kalshi) work compared to traditional sportsbooks?
**Answer:** Traditional sportsbooks act as the "house": you bet against the operator, who bakes in a 5% to 10% fee (vig) and limits or bans winning bettors. Sports prediction markets work like the stock market: you trade binary event contracts (priced between 1¢ and 99¢) directly against other market participants on a central limit order book. There is no house counterparty, which allows for narrower spreads, zero-fee limit orders, and the freedom to exit open positions anytime during live play.

#### Q3: What is paper trading and why is it essential before risking real money?
**Answer:** Paper trading is simulated strategy execution using virtual capital that mirrors real-time live market feeds, order fills, slippage, and fees. It allows analysts to benchmark whether an algorithmic decision tree produces an actual positive edge over 50 to 100 matches before committing real capital.

#### Q4: What is a bankroll, and what percentage should a beginner risk per match?
**Answer:** A bankroll is a dedicated pool of capital set aside strictly for sports trading. The standard guideline for beginners is conservative flat staking of 1.0% to 2.0% per match, or applying a Fractional Kelly multiplier (0.25x Quarter-Kelly). Never risk large portions of your bankroll on a single match: even heavy statistical favorites experience inevitable losing streaks.

#### Q5: What is bookmaker vig (overround) and why does the house always win?
**Answer:** Bookmaker vig (overround) is the hidden commission sportsbooks embed into their lines. For example, in an evenly matched 50/50 game, fair decimal odds would be 2.00 on both sides. Bookmakers instead offer 1.90 on both sides. Summing the implied probabilities ($1/1.90 + 1/1.90 = 105.2\%$) reveals a 5.2% overround. That 5.2% margin guarantees sportsbook profitability over the long run unless bettors de-vig odds to find genuine fair market value.

#### Q6: Can automated betting bots really beat sports markets, or is it pure luck?
**Answer:** Single-match outcomes always involve randomness. Over thousands of matches, however, automated bots outperform human emotional betting because they execute with cold discipline: they never tilt or chase losses, execute within milliseconds when live optical telemetry signals an underpriced dip, and enforce automatic Stop-Loss and Take-Profit rules.

#### Q7: What is Expected Goals (xG) in soccer and how does it signal live trading opportunities?
**Answer:** Expected Goals (xG) measures the statistical quality of a goal attempt based on shot angle, distance, assist type, and defensive pressure on a scale from 0.00 to 1.00. Scoreboards only reflect completed goals, causing retail markets to overreact when a dominant favorite trails 0-1 due to bad luck. When a favorite trails despite generating 1.80 xG against 0.15 xG, bots detect "Unrewarded xG Pressure" and buy the comeback dip before the order book reprices.

#### Q8: What happens to my contract or bet if a tennis player gets injured and retires?
**Answer:** If a player withdraws before the match starts (walkover), contracts are voided and 100% of stakes are refunded. For in-play injury retirements, settlement depends on exchange rules. Starkov bots include an automated "Retirement Void Guard" in `04 EXIT` that cancels resting orders as soon as an on-court medical timeout or retirement is registered.

### 5.2 Quantitative Edge (+EV), De-Vigging & Binary Pricing

#### Q9: How does Starkov calculate positive Expected Value (+EV) in sports prediction markets?
**Answer:** Starkov computes the true de-vigged fair probability ($P_{\text{fair}}$) of an outcome from sharp reference lines and compares it against the live binary prediction market contract price (`0¢–100¢`). When the fair probability in cents exceeds the effective contract acquisition price (`Best Ask` for Taker or `Limit Price` for Maker), the difference represents a positive mathematical edge (`+EV %`):
$$\text{Edge (\%)} = \left( P_{\text{fair}} \times 100 \right) - \text{Contract Price (¢)}$$
$$\text{ROI}_{\text{EV}} (\%) = \frac{P_{\text{fair}} - (\text{Contract Price (¢)} / 100)}{\text{Contract Price (¢)} / 100} \times 100$$
In `02 TRIGGER`, bots can enforce a minimum edge threshold (e.g., `+2.5%` or `+3.5%`) and an anti-noise persistence timer (`edge_persistence_sec`) so orders only fire on sustained pricing inefficiencies.

#### Q10: What is Vig Removal (De-Vigging) and why is it necessary before comparing odds?
**Answer:** Traditional sportsbooks embed a profit margin (overround or "vig") into their decimal odds so that the sum of raw implied probabilities ($\sum 1 / O_i$) exceeds `100%` (typically `102.5%` to `107%`). Comparing raw bookmaker probabilities directly against a binary prediction market would artificially inflate the apparent probability of every outcome. Starkov normalizes raw decimal odds across all mutually exclusive outcomes (`2-way` moneyline in Tennis/US Sports or `3-way 1X2` in Soccer) so that fair probabilities sum to exactly `100.0%`:
$$P_{\text{raw}, i} = \frac{1}{O_i}, \quad \text{Overround } \Sigma = \sum_{j=1}^{n} P_{\text{raw}, j}, \quad P_{\text{fair}, i} = \frac{P_{\text{raw}, i}}{\Sigma}$$

#### Q11: How do binary contract prices (`0¢–100¢`) translate to decimal odds and implied probability?
**Answer:** In a binary prediction market order book, every contract settles at `$1.00` (`100¢`) if the event occurs (`YES`) or `$0.00` (`0¢`) if it does not (`NO`). Because the payout is fixed at `100¢`, the contract price in cents is identical to the market's implied percentage probability: a contract trading at `62¢` implies a `62.0%` win probability and is mathematically equivalent to `1.613` decimal odds ($100 / 62$). Because $\text{Price}(\text{YES}) + \text{Price}(\text{NO}) = 100¢$, buying `NO` at `38¢` is synthetically identical to shorting `YES` at `62¢`.

### 5.3 Order Book Microstructure, Execution & Sizing

#### Q12: What is the difference between Maker limit orders and Taker market orders on prediction markets?
**Answer:** A **Taker Market Order** crosses the bid-ask spread immediately to buy at the `Best Ask`, guaranteeing an instant fill during fast live triggers (such as a conceded break point or opening goal) at the cost of paying the spread and exchange taker fees. A **Maker Limit Order** places a resting bid inside or at the spread (`0% Maker Fee`), capturing price improvement and avoiding taker drag, with configurable Time-in-Force (`GTC`, `30s`, `120s`, `IOC`) and chase tick limits.

#### Q13: How does the Fractional Kelly Criterion protect a bot bankroll against variance?
**Answer:** The Kelly Criterion computes the theoretically optimal fraction of bankroll to allocate to a positive-EV trade ($f_{\text{full}}^* = \frac{P_{\text{fair}} - P_{\text{ask}}}{1 - P_{\text{ask}}}$). However, in live sports markets where fair probability estimates carry small model variance, Full Kelly (`1.00x`) can produce steep drawdowns. Starkov allows bots to apply a **Fractional Kelly Multiplier** ($\lambda \in \{0.25, 0.50, 1.00\}$) alongside a hard USD per-match exposure ceiling, preserving geometric bankroll growth while drastically reducing volatility.

#### Q14: Why do Bid-Ask spread caps and Level-2 order book depth guards matter for automated bots?
**Answer:** During live match suspensions, tie-breaks, or VAR reviews, market makers widen their quotes and thin out top-of-book volume. Firing a market order into an illiquid book with an `8¢` spread immediately destroys a `+3%` theoretical edge via slippage. Starkov's `Liquidity & Spread Guard` in `03 ORDER` inspects Level-2 order book telemetry before every execution—blocking the trade if the Bid-Ask spread exceeds your cent cap (e.g., `3¢`) or if top-3 price levels lack sufficient USD depth.

### 5.4 Live Optical Telemetry & MindMap Logic

#### Q15: How does the 4-Pillar MindMap (`01 Target` → `02 Trigger` → `03 Order` → `04 Exit`) evaluate live matches?
**Answer:** Every running Starkov bot evaluates live matches sequentially left-to-right across four pillars:
- `01 TARGET` applies an `AND` gate across tournament, surface, weather, and matchup filters to isolate eligible games.
- `02 TRIGGER` applies an `AND` gate across live scoreboard states, optical telemetry, and `+EV` thresholds to detect the exact entry tick.
- `03 ORDER` aggregates sizing, contract side, hedging, and Maker/Taker routing into an atomic order instruction.
- `04 EXIT` monitors open positions using an `OR` race condition—closing the trade as soon as the first exit rule (`Take-Profit`, `Odds Convergence`, `Stop-Loss`, or sport-specific game-state cashout) triggers.

#### Q16: How does live optical telemetry (Soccer xG, Tennis 1st Serve % & Markov Hold) differ from static odds?
**Answer:** Scoreboards only reflect completed goals or sets, causing retail market prices to overreact to short-term variance. Starkov ingests live play-by-play and optical telemetry—such as Unrewarded Expected Goals pressure ($\text{xG} - \text{Goals} \ge 0.75$), Shots on Target, and Penalty-Box Crosses in Soccer, or rolling 1st Serve %, Break Points saved, and point-level Markov chain hold probability deltas in Tennis. Combining live telemetry dominance with a temporary scoreboard dip (e.g., a pre-match favorite conceding first) isolates mean-reversion entries before the order book reprices.

#### Q17: What is an Anti-Wick Stop-Loss and why does Starkov use a grace period timer?
**Answer:** In prediction market order books, a single large market sell order or a 2-second liquidity vacuum after a point is scored can cause a momentary price "wick" before market makers replenish the bid. A naive instant stop-loss would panic-sell into the bottom of that temporary wick. Starkov's Stop-Loss in `04 EXIT` includes a configurable **Anti-Wick Grace Period** (`sl_grace_period_sec`, e.g., `10s` to `30s`): the adverse ROI threshold must remain breached continuously for the full timer duration before an exit order fires.

#### Q18: How do synthetic hedges (Soccer 3-Way Draw `X` Hedge and Tennis Current-Set Hedge) work?
**Answer:** In 3-way regulation soccer markets (`Home`, `Away`, `Draw`), buying a favorite's `YES` contract still exposes the bot to a 90-minute regulation draw settling at `0¢`. Enabling the `3-Way Draw (X) Hedge` in `03 ORDER` automatically allocates a secondary fraction of the stake to the Regulation Draw contract, constructing a synthetic Double Chance (`1X` or `X2`) payoff. Similarly, in Tennis, a bot buying a match-winner dip can allocate a partial hedge to the opponent's Current-Set Winner contract to cushion short-term set volatility.

### 5.5 Sandbox Accounting & Compliance

#### Q19: How does Starkov distinguish between Mark-to-Market (unrealized) PnL and Realized PnL?
**Answer:** While a bot position is open, Starkov values its contracts continuously against the live `Best Bid` on the order book to compute **Mark-to-Market (Unrealized) PnL**. Once a `04 EXIT` trigger fires, a manual Cash-Out is clicked, or the match officially settles, the position is closed and converted into **Realized PnL**. Public Leaderboard rankings require bots to complete verified paper-trading execution cycles and report both realized return on capital and maximum drawdown.

#### Q20: Is Starkov a sportsbook, bookmaker, or real-money gambling platform?
**Answer:** No. Starkov is strictly a Software-as-a-Service (SaaS) quantitative sports analytics, order-book telemetry, and strategy simulation ("paper-trading") sandbox. Starkov is not a sportsbook, bookmaker, gambling operator, prediction market exchange, or financial custodian. It never accepts real-money wagering deposits, never holds user custody funds, and all wallets, bankrolls, stakes, orders, and PnL metrics on the platform represent 100% virtual paper-trading simulation figures.