Hungry Horse Simulator

Auto-plays the HHP strategy against the trainer's villain models — no input needed. Hero rebuys to the starting stack whenever a hand starts short.

← Back to trainer

Setup (changing a game's settings resets that game's run)

Scenarios are not picked — each hand deals Hero two fully random cards at the next seat around the table and plays the HHP preflop charts (RFI, blind defends, vs-3-bet/vs-4-bet); pot types emerge from those chart decisions and the villain's range reactions. The buckets below form naturally.

Villains:
Frequency
Passive Caller Fish16%
Passive Folder Fish36%
Aggro Caller Fish12%
Whale4%
Trying Rec12%
Reg20%

How often each enabled type shows up.

Mistakesper 100 chances of each spot — all 0 = the perfect HHP policy

Hover a mistake for what it does. Postflop rates roll per decision of that type; loose opens roll per chart-fold hand dealt, so they fire far more often at the same rate. Each game keeps its own profile — compare bb/100 across games to see what a leak costs at each stake. With any rate on, a perfect-policy baseline runs alongside (roughly doubling run time) and overlays the bankroll curve.

Strategiesuntick a component to measure what it earns — its spots play a simpler default instead

Hover a strategy for what it covers and what replaces it. With anything off, the full-HHP baseline runs alongside and the Strategy impact section shows the EV difference. Toggle one component at a time to attribute the win rate cleanly.

Results

Hands dealt0
…folded preflop (RFI fold / blind fold)0 ()
…opened preflop (RFI chart)0
…folded preflop to a re-raise0
…multiway pots played0
Hands played (voluntarily continued postflop)0 ()
Hands won0 ()
…at showdown / by villain fold / preflop steals0 / 0 / 0
Hands lost (incl. hero folds)0 (0 folds)
Chopped pots0
Showdowns reached / won0 /
All-in pots0
Money bet total (all hero chips in)$0
Money won total (pots collected)$0
Money lost total (chips forfeited)$0
…blinds posted on open-folds$0
…rake paid out of won pots$0
Net$0
…showdown / non-showdown winnings$0 / $0
Win rate (95% CI)0.0 ± 0.0 bb/100
Std dev (per 100 hands)0 bb
Worst downswing$-0
Hourly win rate at hands/hr
…live hours simulated at that pace
Rebuys0 ($0)
Biggest pot won$0

Bankroll curve (cumulative net over dealt hands)

Run some hands to see the curve.

Variance (GBP)

Everything here is in GBP, the bankroll currency. Months are built by resampling per-hand results — 60 hours at 25 hands/hr = 1,500 hands per month. Each resampled hand is dealt to one of the games above per its share and drawn from that game's own simulated hands — so each game's stakes, rake, villain pool and Mistakes setting flow through exactly as simulated; USD results convert at £1 = $1.30. Hands are independent in the model, so this is statistically identical to replaying the sim month by month. The same hand stream is also sliced into 6-hour sessions (150 hands) for the Session view.

Run at least 5,000 hands total — and at least 500 in every game with a share — to build the distribution.

By scenario

ScenarioHandsFoldedNetbb/100

Folded = hands hero gave up facing a bet or raise in that scenario. The open-fold row is the preflop trash mix (its net is the blinds it posted).

By villain

VillainHandsNet

By street the hand ended on

StreetHandsNet

By hero flop hand class

ClassHandsNet$/hand

Bluff catching

Every RIVER spot where hero faced a bet or raise holding a bluff-catcher (SDV or Thin Value) — bluff catching is a river-only concept in HHP. Calls and folds are graded against the villain's actual holding, so you see both bad calls and bluffs hero folded to.

No bluff-catch spots yet — run some hands.

Bluffing

Every hero RIVER bet/raise made with a bluff-class hand — bluffing is a river-only concept in HHP; earlier-street aggression with draws/Air is equity betting, not a true bluff. Fold equity is compared with the breakeven fold rate each sizing needed; called bluffs are graded by how the hand actually ended.

No bluffs yet — run some hands.

Hand history

Nothing here yet — run some hands.

Strategy gaps — placeholders for future work

Spots the HHP docs don't answer. The sim plays a labelled default (value keeps going: bet 50% first-in / call raises with CPFS, Thick Value or Combo Draws; everything else checks/folds). Each is a TODO for the strategy.

None hit yet — run some hands.

Auto-resolved judgment reads

The docs leave these to human judgment; the sim resolves them with the listed heuristics.

None yet.

Emulation notes

• Preflop is played from the HHP charts: RFI by seat, blind defends (depth-split at 150BB effective), vs-3-bet and vs-4-bet continues — pot types emerge from chart decisions plus villain range reactions (3-bet/call/fold coverage from their preflop ranges). 5-bet spots run out as preflop all-ins.

• Cold 4-bet and BB-vs-4×3-bet charts are stored but unused (they need a third preflop actor); the SB defend chart is the 200BB version at all depths.

• Multiway pots play the HHP multiway principles (bet value + strong draws, bluff less) against the multiway villain tables — simplified betting: one raise resolution per street, wagers capped at the shortest live stack (no side pots). Facing a bet, only CPFS / Thick Value / Combo Draw / HE Draw ever continue — a pot that saw a multiway flop never bluff-catches, even once it collapses to one opponent.

• The HHP villain type (trainer-only for now) is excluded here — HHP-vs-HHP simulation isn't wired yet.

• First-in rivers use the §9 two-question framework: villain called a big bet (>60% pot) → sticky range (value jams, bluffs bet medium); only small calls → capped (value milks medium, bluffs small-overbet); thin value always bets tiny.

• Uncovered first-in spots (e.g. river value bets) now resolve from the §6.1 range model: capped = villain hasn't bet/raised, inelastic = fish types, fastplay = board isn't Static Dry. Facing-bet gaps and missing chart streets still play the labelled gap defaults.

• As the preflop caller HHP adopts the raiser's strategy for its position (IP range-bets when checked to; OOP follows the OOP tables with the §2.5 Ax-turn-after-XX PFC amendment). 4-bet pots beyond the flop range bet, and other remaining uncovered spots, play the labelled gap defaults where the range model doesn't answer.

• When the villain was the preflop raiser (hero PFC) their flop continuation-bet comes from the editable per-type C-Bet table (~90%+ frequency), whether first to act or after hero checks.

• Playing hands reach the flop with the scenario's standard pot (preflop always completes).

• Villain hole cards are drawn from that villain type's weighted preflop range for the spot; PFC hands reuse the same villain postflop tables (approximation).

• Villain actions use the trainer's behavior tables (incl. low-SPR shove/rescue transforms).

• When hero is IP the villain may lead the turn per their hand class (first-to-act table) — a probe after a checked-through flop or a lead after check-calling; only villain CPFS ever raises a turn bet (other classes never fold it).

• Hero raise sizings vs bets are approximations (flop XR = 8x the bet in SRP / 6x in 3BP+, turn/river raises 3x, huge XR ≈ 4.5x); villain re-raises collapse to call/fold.

• Blinds: hero forfeits a posted blind on open-folds; when playing a pot the blind is subsumed in the preflop investment. SRP opens steal the blinds when the villain has no continue-range hand; 3-bet/4-bet pots always reach the flop.

• £2/5/10 plays as a 5/10 game (the £10 straddle acts as the big blind, opens to £25). Hero's real £2 small blind is charged as dead money once per orbit on his SB hand, and every pot he contests holds one dead £2 — his own on that hand (recouped when he wins) or a villain's otherwise — so blind steals, all-in pots and postflop pots all carry the extra £2.

• Rake is taken from pots hero collects (no flop, no drop on preflop steals); a seed makes runs fully reproducible.

• The Mistakes model injects human error into the otherwise-perfect HHP policy, per category and per 100 opportunities of that decision type: loose opens (opening RFI-chart folds), loose 3-bet calls, bad calls (paying off policy folds), bad folds (folding policy calls), missed raises (flatting policy raises), missed bets (checking policy bets), and spew (bluffing ~75% pot with SDV/Air the policy checks). Blunders are labelled in the action mix and hand history, and the Mistake impact table tracks chances, blunders made, the net of affected hands, and a realized-counterfactual cost per category (exact for fold counterfactuals, check-down approximation for bad folds/spew, unestimated for missed bets/raises and multiway). With any rate on, a perfect-policy shadow run overlays the bankroll curve.

• The Strategies toggles disable named HHP components (blind defence, preflop re-raising, flop c-bet tables, turn tables, river framework, range model, river bluffing, bluff catching, multiway principles) so each one's EV is measurable: disabled spots play the labelled ABC gap default or an explicit simpler line (fold catchers, give up bluffs, flat instead of re-raise, fold blind defends), and the full-HHP baseline runs alongside for comparison.