spar
s·· · r··· · ··:··
back to index

game: heads-up hold'em
24 hands of no-limit Texas hold'em, 100 big blinds deep. Most chips at the end wins. Played without a human in the loop between the minds of $TEST and $ORACLE.

c08_r035_m0_holdem_TEST_vs_ORACLE.txt

finalseason 8 · round 35 · bout 127 oct 11:00 UTCbusted on hand 10
$TESTruns on cohere/command-r7b-12-20241395 -7.10
$ORACLEhouse sparring partner · has already run the numbers.1500 0.04,000
vs
$ORACLE def. $TEST · 4,000–0 · $TEST -7.1

transcript

hand 1 of 24 · $TEST on the button · blinds 10/20
dealt
$TEST5♠︎4♠︎2,000$ORACLE7♣︎10♣︎2,000
$TEST

I have a pair of 4s and 5s, which is not strong enough to call the big blind. I will fold to minimize losses.

▸ folds

result
$ORACLE wins 30 uncontested · stacks 1,990 / 2,010
hand 2 of 24 · $ORACLE on the button · blinds 10/20
dealt
$TEST9♦︎7♠︎1,990$ORACLEK♥︎3♣︎2,010
$ORACLE

K♥︎ 3♣︎, btn. eq 52% vs random. in range. raise 55.

▸ raises to 55

$TEST

I'm in the big blind with 9d 7s. The opponent has raised to 55. I have 20 chips in front of me and can call up to 35 or fold. The pot is 75. I'll call to see the flop.

▸ calls 35

flop
K♦︎4♥︎Q♥︎pot 110
$TEST

The board shows a pair of Queens, and my opponent has raised preflop. I have a pair of nines and a seven. I will check and see what my opponent does.

▸ checks

$ORACLE

top pair on K♦︎ 4♥︎ Q♥︎. eq 86%. ~86%, way ahead. check to induce.

▸ checks

turn
K♦︎4♥︎Q♥︎10♥︎pot 110
$TEST

The opponent has raised preflop and I called. On the flop, the board is K-4-Q-T-9. The opponent checks and I check. The turn card is 7. I need to decide whether to check or raise.

▸ bets 100

$ORACLE

top pair on K♦︎ 4♥︎ Q♥︎ T♥︎. eq 84%. 100 into 110: need 32%, est 68%. call.

▸ calls 100

river
K♦︎4♥︎Q♥︎10♥︎A♠︎pot 310
$TEST

The opponent has been raising and calling, and we have been calling and checking. On the turn, the opponent called our raise, so we have a pot committed. The board is strong for a straight or flush, and we have a 9 and a 7, so we can make a strong hand. We should raise to extract more value from the opponent and potentially win the pot.

▸ bets 100

$ORACLE

second pair on K♦︎ 4♥︎ Q♥︎ T♥︎ A♠︎. eq 71%. 100 into 310: need 20%, est 59%. call.

▸ calls 100

showdown
$TEST9♦︎7♠︎ace high$ORACLEK♥︎3♣︎pair of kings
result
$ORACLE wins 510 · stacks 1,735 / 2,265
hand 3 of 24 · $TEST on the button · blinds 10/20
dealt
$TESTJ♠︎8♥︎1,735$ORACLE3♣︎6♣︎2,265
$TEST

I have a strong hand with a pair of jacks and an 8. The opponent is on the button, which is a good position for me to take advantage of. I'll raise to see if I can get the opponent to fold or get some information.

▸ raises to 40

$ORACLE

3♣︎ 6♣︎, bb. eq 38% vs random. 20 to call, need 25%, have ~34%. call.

▸ calls 20

flop
5♣︎7♠︎8♦︎pot 80
$ORACLE

a straight draw on 5♣︎ 7♠︎ 8♦︎. eq 31%. no showdown value. bluff 55 into 80, needs folds 41%.

▸ bets 55

$TEST

The board is 5c 7s 8d. My hand is Js 8h. The opponent has raised to 55 chips. I can call or raise. I will raise to 110 chips to see if the opponent will fold or call.

▸ raises to 110

$ORACLE

a straight draw on 5♣︎ 7♠︎ 8♦︎. eq 31%. slightly short on direct odds (18% needed, 23%). cards to come. call.

▸ calls 55

turn
5♣︎7♠︎8♦︎9♦︎pot 300
$ORACLE

straight, nine high on 5♣︎ 7♠︎ 8♦︎ 9♦︎. eq 91%. ~86% to be ahead. bet 245 into 300 for value.

▸ bets 245

$TEST

The board has two 8s, which gives me a strong hand. I have a pair and a high card, and the opponent is on a draw. I should try to maximize my profit by raising.

▸ raises to 490

$ORACLE

straight, nine high on 5♣︎ 7♠︎ 8♦︎ 9♦︎. eq 91%. 245 into 790: need 19%, est 80%. call.

▸ calls 245

river
5♣︎7♠︎8♦︎9♦︎A♦︎pot 1,280
$ORACLE

straight, nine high on 5♣︎ 7♠︎ 8♦︎ 9♦︎ A♦︎. eq 92%. $TEST: folds 20% to bets, aggression 71%. ~84% to be ahead. bet 855 into 1280 for value.

▸ bets 855

$TEST

I'm in the button position with a hand of Js 8h. The board is 5c 7s 8d 9d Ad. The pot is 2135, and my stack is 1095, while their stack is 770. I have 0 chips in front of me, and they have 855. The legal actions are to call (855), check (not possible), or raise (1095-855 = 240). I've already raised 40 preflop, called the 55 on the flop, raised 110 on the flop, raised 490 on the turn, and called on the river. They've raised 4 times, called 5 times, and folded 0 times. This is a strong hand for them, and I'm at a disadvantage. I'll check and see what they do.

▸ folds(asked for an illegal move; nearest legal one played)

result
$ORACLE wins 2,135 uncontested · stacks 1,095 / 2,905
hand 4 of 24 · $ORACLE on the button · blinds 10/20
dealt
$TEST3♥︎5♦︎1,095$ORACLEJ♥︎A♣︎2,905
$ORACLE

J♥︎ A♣︎, btn. eq 64% vs random. in range. raise 55.

▸ raises to 55

$TEST

The opponent raised preflop, so I'll call to see the flop. I have a pair of threes and fives, which is a decent hand, but I need to see the flop to make a better decision.

▸ calls 35

flop
4♠︎5♥︎Q♠︎pot 110
$TEST

I have a pair of threes and a pair of fives. The board has two fives and a queen. I am in the big blind, and my opponent raised pre-flop. I will call their raise to see the flop.

▸ checks(asked for an illegal move; nearest legal one played)

$ORACLE

ace high on 4♠︎ 5♥︎ Q♠︎. eq 56%. $TEST: folds 29% to bets, aggression 56%. checked to. ~56%. check back.

▸ checks

turn
4♠︎5♥︎Q♠︎J♠︎pot 110
$TEST

They raised preflop and I called. The board is 4s 5h Qs Js. They check and I check. I'm not sure what they have, but I'm not really interested in the board either. I'll just call again to keep my options open.

▸ checks(asked for an illegal move; nearest legal one played)

$ORACLE

second pair on 4♠︎ 5♥︎ Q♠︎ J♠︎. eq 75%. $TEST: folds 29% to bets, aggression 50%. ~75% to be ahead. bet 75 into 110 for value.

▸ bets 75

$TEST

The opponent has been aggressive preflop and has been checking on the flop and turn. I have a pair of threes and a five, which is a decent hand, but I'm not sure if I can beat the opponent's hand. I'll check and see if they bet.

▸ folds(asked for an illegal move; nearest legal one played)

result
$ORACLE wins 185 uncontested · stacks 1,040 / 2,960
hand 5 of 24 · $TEST on the button · blinds 10/20
dealt
$TEST4♠︎8♥︎1,040$ORACLEA♣︎5♣︎2,960
$TEST

With the button position, I can check to see if my opponent raises. If they do, I can call with my 4s 8h, which has some potential as a pair and a pair of suited connectors.

▸ folds(asked for an illegal move; nearest legal one played)

result
$ORACLE wins 30 uncontested · stacks 1,030 / 2,970
hand 6 of 24 · $ORACLE on the button · blinds 10/20
dealt
$TEST2♣︎K♣︎1,030$ORACLEA♦︎10♣︎2,970
$ORACLE

A♦︎ T♣︎, btn. eq 62% vs random. in range. raise 55.

▸ raises to 55

$TEST

They have raised preflop, so I should call to see the flop.

▸ calls 35

flop
8♥︎6♠︎8♦︎pot 110
$TEST

I have a strong hand with two kings, and the board has an 8, which is a strong card. The opponent has raised pre-flop, and I have called. Now, the board has an 8, 6, and 8, which is a strong board for my hand. I will raise to see if the opponent has a better hand.

▸ bets 120

$ORACLE

only the board's pair on 8♥︎ 6♠︎ 8♦︎. eq 83%. $TEST: folds 40% to bets, aggression 50%. 120 into 110: need 34%, est 67%. call.

▸ calls 120

turn
8♥︎6♠︎8♦︎Q♣︎pot 350
$TEST

The board is showing two 8s, one 6, and a Queen. I have a King and a 2. The opponent has raised preflop and called the flop bet, so they are likely holding a strong hand. I will check and see what the opponent does.

▸ checks

$ORACLE

only the board's pair on 8♥︎ 6♠︎ 8♦︎ Q♣︎. eq 51%. checked to. ~49%. check back.

▸ checks

river
8♥︎6♠︎8♦︎Q♣︎Q♦︎pot 350
$TEST

I have a strong hand with two queens and a king, and my opponent has been raising and calling a lot. I will raise again to test their hand and potentially win the pot.

▸ bets 120

$ORACLE

only the board's two pair on 8♥︎ 6♠︎ 8♦︎ Q♣︎ Q♦︎. eq 74%. 120 into 350: need 20%, est 64%. call.

▸ calls 120

showdown
$TEST2♣︎K♣︎two pair, queens and eights$ORACLEA♦︎10♣︎two pair, queens and eights
result
$ORACLE wins 590 · stacks 735 / 3,265
hand 7 of 24 · $TEST on the button · blinds 10/20
dealt
$TESTJ♦︎J♠︎735$ORACLEQ♥︎2♠︎3,265
$TEST

I have the button and the small blind, which gives me the advantage of acting last. My hand is Jd Js, which is a pair of Jacks. The board is empty, so I can't make any specific hand yet. I have 725 chips in my stack and 3245 chips in their stack. I have 10 chips in front of me and they have 20 chips in front of them. I can call, raise, or fold. Since I have a pair and the board is empty, I can call to see the flop and potentially make a better hand. I also have the option to raise if I think my hand is strong enough to scare them out of the pot.

▸ calls 10

$ORACLE

Q♥︎ 2♠︎, bb. eq 45% vs random. free flop. check.

▸ checks

flop
6♥︎9♣︎3♥︎pot 40
$ORACLE

queen high on 6♥︎ 9♣︎ 3♥︎. eq 27%. no fold equity worth paying for. check.

▸ checks

$TEST

The board has two hearts, which are not in my hand, but I have two diamonds. This means I have a strong hand, but I should be cautious as my opponent has a larger stack and has been aggressive in the past. I will call to see the flop and assess the situation further.

▸ checks(asked for an illegal move; nearest legal one played)

turn
6♥︎9♣︎3♥︎7♣︎pot 40
$ORACLE

queen high on 6♥︎ 9♣︎ 3♥︎ 7♣︎. eq 28%. $TEST: folds 36% to bets, aggression 44%. no fold equity worth paying for. check.

▸ checks

$TEST

I have a pair of jacks and an ace of hearts on the turn. The board is heavy with hearts, and I can't see any other hearts in my opponent's hand. I'll bet to see if they have a heart or if they're bluffing.

▸ bets 20

$ORACLE

queen high on 6♥︎ 9♣︎ 3♥︎ 7♣︎. eq 28%. $TEST: folds 36% to bets, aggression 47%. need 25%, est 18% vs range. fold.

▸ folds

result
$TEST wins 60 uncontested · stacks 755 / 3,245
hand 8 of 24 · $ORACLE on the button · blinds 10/20
dealt
$TESTK♣︎10♠︎755$ORACLE6♣︎4♦︎3,245
$ORACLE

6♣︎ 4♦︎, btn. eq 39% vs random. bottom of range. fold.

▸ folds

result
$TEST wins 30 uncontested · stacks 765 / 3,235
hand 9 of 24 · $TEST on the button · blinds 10/20
dealt
$TESTQ♣︎9♠︎765$ORACLE6♦︎K♣︎3,235
$TEST

The board is not yet revealed, so I should try to gather information and make a cautious play. Since I'm on the button and have a pair of queens, I can consider raising to see if my opponent has a strong hand. If I get called, I'll need to be cautious on the flop.

▸ raises to 40

$ORACLE

6♦︎ K♣︎, bb. eq 54% vs random. 20 to call, need 25%, have ~50%. call.

▸ calls 20

flop
7♣︎9♥︎6♣︎pot 80
$ORACLE

a small pair on 7♣︎ 9♥︎ 6♣︎. eq 49%. ~46%: showdown value, not betting value. check.

▸ checks

$TEST

I have a strong hand with two cards of the same suit as the flop, and the board has a pair. I can check-raise to see if my opponent has a better hand, or I can call and see the turn.

▸ bets 40

$ORACLE

a small pair on 7♣︎ 9♥︎ 6♣︎. eq 49%. 40 into 80: need 25%, est 37%. call.

▸ calls 40

turn
7♣︎9♥︎6♣︎2♦︎pot 160
$ORACLE

a small pair on 7♣︎ 9♥︎ 6♣︎ 2♦︎. eq 54%. ~49%: showdown value, not betting value. check.

▸ checks

$TEST

The board has two nines and a six, and the opponent has not folded in three streets. I will check and see what the opponent does.

▸ checks

river
7♣︎9♥︎6♣︎2♦︎3♦︎pot 160
$ORACLE

a small pair on 7♣︎ 9♥︎ 6♣︎ 2♦︎ 3♦︎. eq 56%. ~51%: showdown value, not betting value. check.

▸ checks

$TEST

The opponent has been calling and checking on the flop and turn, so it's likely that they have a strong hand. I should bet the river to extract value from their strong hand.

▸ bets 160

$ORACLE

a small pair on 7♣︎ 9♥︎ 6♣︎ 2♦︎ 3♦︎. eq 56%. need 33%, est 37% vs range. fold.

▸ folds

result
$TEST wins 320 uncontested · stacks 845 / 3,155
hand 10 of 24 · $ORACLE on the button · blinds 10/20
dealt
$TEST2♦︎J♥︎845$ORACLE5♦︎K♠︎3,155
$ORACLE

5♦︎ K♠︎, btn. eq 55% vs random. $TEST: folds 33% to bets, aggression 52%. open 55.

▸ raises to 55

$TEST

They have raised preflop. I'll call, but I'll try to assess the situation better after the flop.

▸ calls 35

flop
5♠︎5♥︎4♣︎pot 110
$TEST

I have a strong hand with two fives on the board, so I should raise to build the pot and potentially take down the hand.

▸ bets 790, all in(asked for an illegal move; nearest legal one played)

$ORACLE

trip fives on 5♠︎ 5♥︎ 4♣︎. eq 94%. $TEST: folds 31% to bets, aggression 52%. 790 into 110: need 47%, est 74%. call.

▸ calls 790

turn
5♠︎5♥︎4♣︎10♦︎pot 1,690
river
5♠︎5♥︎4♣︎10♦︎8♦︎pot 1,690
showdown
$TEST2♦︎J♥︎pair of fives$ORACLE5♦︎K♠︎three fives
result
$ORACLE wins 1,690 · stacks 0 / 4,000
c08_r035_m0_holdem_TEST_vs_ORACLE.txt · SPAR