Leggo Poker Every Tool You Need To Win


Analyzing Ungar
Posted in Poker | View Comments (23)

Edit: there was confusion writing this whether Tom Sims meant "Raises to x" or "Raises an additional x". The consensus is that Mr. Sims meant the latter, so I had to edit everything to reflect that. Fortunately it doesn't change too much.


I was always under the impression that poker suffers a little bit when it comes to objectively analyzing the play of past champions. I can look at a replay of a professional sporting event from the 70's and get an idea of how the game was played. Same goes for Chess, where classic games are broken down by every single move.

This transparency seemed lacking when it came to historically reviewing the WSOP. Before Moneymaker, the WSOP was typically shown in the form of a documentary, and the only hands that made it to airwaves were the biggest pots. An overwhelming number of these hands were obvious cooler situations that didn't highlight a player's skill. The only way I could accurately gauge the NLHE talents of someone like Stu Ungar would be the words of his peers.

That was until yesterday, when (thanks to @jesswelman) I was made aware of hand-for-hand coverage of WSOP main event final tables from 1995 until 2000. The website is here.

For this blog entry I'll be focusing solely on 1997, when Stu Ungar won his 3rd main event.

Blinds 5k/10k with 2k ante
Starting stacks

(Players are listed in seating order clockwise, so Ungar has direct position on Stanley)

Stu Ungar - 1.1MM
Peter Bao - 204k
Bob Walker - 610k
John Strzemp - 245k
Mel Judah - 300k
Ron Stanley - 700k

For getting VPIP/PFR I read the hand-by-hand report and used my iPhone app DonkeyTracker.

The first 43 hands:
Stu Ungar - VPIP 33, PFR 23
Peter Bao - VPIP 12, PFR 7
Bob Walker - VPIP 14, PFR 9
John Strzemp - VPIP 33, PFR 16
Mel Judah - VPIP 23, PFR 14
Ron Stanley - VPIP 30, PFR 19

A solid start from Ungar preflop, but I would have liked to see him play even more aggressive. He had the chiplead, direct position on the player 2nd in chips and ubernits to his direct left.Most of Ungar's steals worked and saw little resistance, and the only time we saw a threebet it was when Strzemp 3bet shoved for 50bb effective over a 3.5x open. Ungar balances preflop through the first 43 hands by basically never 3betting. As a random aside, they went from 5k/10k blinds to 10k/20k. Not a Matt Savage-approved structure.

I hear people commentate about this final table and say that the way Ungar played, nobody could have taken the bracelet from him. I would counter that Ungar benefitted from an absurdly favorable table draw. Strzemp could have easily stolen the bracelet with his unorthodox donkbetting style.

Postflop is interesting. NLHE in 1997 sorta reminds me of PLO in 2012. NL in 2012 people are strictly betting for value or betting as a bluff. In 1997 people made a bunch of bets that could not possibly hope to get value, but I guess they vaguely wanted to protect their equity. The continuation bet wasn't really a given yet (I think?), so when a player had Kx and defended versus a LP steal and the flop came KQJ, they figured Kx was crushed by a flop cbet range, and they didn't want to give a free card to Ax/55/Q9o, etc. So they donklead. Weird.

I'm impressed with Ungar's postflop play. He definitely had a winning redline (winnings before showdown).

Ungar hands:

Hand 6: MP (14/9) limps, SB completes and Ungar checks. Flop J33r, Ungar stabs with a potsize bet. Nit folds, SB calls. Turn 5h, check check. River Qs, SB checks, Ungar bets halfpot, SB makes a 2.5x river checkaise and Ungar folds. Not sure if Ungar was vbetting with KJo or bluffing to get a stubborn ace high to fold but seems fine either way.

Hand 22: CO (30/19) opens to 3.5x, Ungar flats OTB with AQss. SB overcalls. Flop AK6r, checks to Ungar and he checks. Turn 7c, checks to CO who fires halfpot. Ungar calls, SB folds. River is 3x. CO checks, Ungar bets 2/3 pot, CO calls and mucks. Another interesting hand which seems fine for value. Because the light 3bet doesn't exist yet, Ungar balanaces (I assume) by flatting all of his playable hands on the button, maybe even including big pairs. This distribution (assumedly like top 30%) doesnt smack AK6r so he checks last to act, though I wouldn't be shocked if he bet halfpot if he had something 87 with a backdoor flushdraw. Turn and river are standard and I like the betsizing.I would have thought most LOL Live Pros would 1/7 pot the river with just one pair, but Ungar balances his postflop sizings pretty well.

Hand 28: Ungar 3.5x's hijack with A5o, Strzemp flats SB. Flop AK4cc and Ungar checks back. Turn 8s putting up two flushdraws, Strezemp bets 70% pot and Ungar calls. River 6s bringing bd flush in and Strzemp bets halfpot and Ungar stations. Strzemp shows KTss for the flush and Ungar shows. Not sure if he's forced to show down or if he voluntarily showed, but hand seems standard and many similar lines have been taken in the last few years on these textures versus flop checkback ranges, although Strzemp did coldcall SB. With as many draws present and Strzemp presumedly getting out of line sometimes, Ungar basically makes a game theory calldown

Hand 36: SB (30/19, villain from AQss hand) limps SB, Ungar checks QTo. Flop A96ss, check check. Turn 8x, SB bets 25k and Ungar raises to 85k. SB calls. River Kx, SB checks and Ungar bets 220k. SB folds, showing a 9, and Ungar shows the bluff. Face! With the lack of the light 3bet and the light continuation bet, postflop in NL sorta feels like PLO in that the ranges being represented are so narrow/polarized when large bets go in. Boards like A968 sorta look like a PLO lockdown board when someone is facing a raise. Here Ungar picked up on a great spot where he didn't have enough value to call the turn, but villain's turn betting range is wide, and lots of pair+gutter hands either fold turn or call turn -> fold river. Good bluff.

Hand 39: Ungar 3.5x's hijack at 10k/20k, Strzemp flats SB. Flop KQJr, Strzemp donks for 75% pot. Ungar folds showing 55, Strzemp shows a king. This is what I was referencing earlier. Strzemp has already shown down KTs in his range from an earlier hand in the exact same spot preflop, and now with Kx he elects to donklead this flop then SHOW THE KING. Bizarre. And another "I fold and show" that would never happen these days, but it worked well for Ungar because he was able to establish a concrete range of Strzemp's SB coldcalls versus up front raises (broadway hands)

Hand 42: BvB, The 30/19 3.5x's SB and Ungar defends with KQo. Flop Q88r, SB checks and Ungar checks. Turn 5d, SB checks, Ungar bets 60k (little over halfpot) and SB calls. River Kd (bringing in backdoor 3flush) and SB leads for 120k (about 60% pot) and Ungar calls. SB shows 77 and Ungar wins.

Here we finally get some consistency with the metagame of Ungar's gameplan. Since light 3betting is not in style, and light cbetting is not in style, Ungar opts to play 30/20ish for sixmax and checks lots of flops. The logic is that people typically fold to his steals, and when people defend versus his steals they will typically have a premium holding that will continue on most flop textures. By checking most flops, Ungar keeps ranges wide while villain will presumedly narrow their range down out of position, allowing Ungar to make better decisions.

Furthermore, Ungar can get away with coldcalling OOP and using his same balance tactics OOP because his villain's aren't auto-cbetting. More hands become playable when you know that flop checkbacks are folding before showdown on 90%+ turn/river runouts.

In this case, this was one of the first times that SB opted to openraise over limping, and ended up showing down 77. Had SB balanced by limping or raising everything, the river bluff could have been more credible (meh), but the preflop action narrowed villains range to good starting holdings, and KQ is more than good enough to call river versus that range given the action, especially when villain checks this particular flop as the PFR and check calls turn. Basically a bad/unbalanced line for SB and a balanced/good line for Ungar

Once the 6th place player (Bao) busts, Ungar goes into ICM pillaging. After 52 hands, stacks look like this

Stu Ungar 1.7MM
Bob Walker 200k
John Strzemp 450k
Mel Judah 150k
Ron Stanley 650k

And their stats look like this:
Ungar: VPIP 34, PFR 25
Walker: VPIP 17, PFR 13
Strzemp: VPIP 32, PFR 15
Judah: VPIP 19, PFR 11
Stanley: VPIP 30, PFR 12

Noteworthy hands:

Hand 46: Ungar 3.5x's it UTG 5handed, Strzemp flats SB. Flop QJ3r and Stremp c/f to Ungar's pot sized bet.

AGAIN Strzemp flats in the SB vs. an Ungar open, and opts to finally not donklead a flop texture that assumedly hits his distribution. Ungar bets pot versus Strzemp's assumedly really weak range and takes it down. Strzemp is now sorely unbalanced in that any time he checks out of position first to act, he is basically giving up the pot. A pretty bad gameplan all around.

If we go back to the hand where Unger checked back A5o and stationed turn/river, hands like this are probably what he had in mind. Villain is checking flop airballs and donking if he has a semblance of a hand, so I'm going to check back my top pair no kicker and induce bluffs. Ungar is smart.

Hand 52: Ungar 3.5x's cutoff, Strzemp defends BB. Flop JJ7cc, Strzemp donks 120k into 150k, Ungar makes it 420k, Strzemp folds. Another spot where Strzemp is playing sorely unbalanced, as every Jx in Strzemp's range would check this flop given the metagame? Also when Strzemp bets 120k he leaves about 450k behind, and there is a 200k stack (Walker) and 150k sack (Judah) at the table. So Strzemp left himself open to get ICM paddled, where his donking distribution doesnt look strong and his range to get in stacks is absurdly narrow. So Ungar puts in the raise with a range that assumedly has lots of semibluffs or bluffs in it.

The ICM pillaging continues for 40 more hands, as Ungar chips up to 2MM. Walker and Judah chip down to 6bb and 4bb, respectively. Strzemp and Stanley chip down to ~450k (around 22BB).

Stacks when Bob Walker busts in 5th (possibly a little off)

Stu Ungar: 2MM
John Strzemp: 450k
Mel Judah: 250k
Ron Stanley: 250k

Stats with 4 left (hand 92)

Ungar: VPIP 38, PFR 26
Strzemp: VPIP 36, PFR 16
Judah: VPIP 22, PFR 13
Stanley: VPIP 27, PFR 22

Noteworthy hands:

At this point stacks are short, so Ungar opens for 3x instead of 3.5x. Obviously would like him to see open even smaller than that, but alas.

Stanley starts kicking it up and raising a lot, while Strzemp calms down a little bit and folds more. The other two shorties spend all their time folding.

Hand 65: Folds to Stanley now playing 27/22, he raises cutoff, Ungar flats BTN and rest fold. Flop A97hh check check. Turn 4x, check check. River Th, check check. Stanley shows ATo, Ungar shows QQ. So this pretty much validates my point from earlier re: Ungar's preflop balance tendencies. Stanley gets ICM-scared with top pair (rivered 2 pair) and misses obvious value even though river brings 3flush.

I noticed that Ungar had a definite tournament weakness in that he chose to limp over shoving in SB v BB spots where the BB was both absurdly nitty and shortstacked.

Hand 49: Folds to Ungar in SB, villain in BB has 10bb and is playing 17/13. Ungar opts to limp, and folds to BB's shove.

Hand 84: Folds to ungar in SB, villain is same player and has 7bb, Ungar limps A7o and calls the BB's shove.

I might could excuse the limp in hand 49 if you think you have such a good handle on the table, but the limp in hand 84 is inexcusable.

It goes from 5 handed to 3 handed rather quickly. Walker calls off 90% of his stack with K9o in the SB versus a Mel Judah 5bb shove in the hijack, and Judah wins with TT. Walker busts next hand. Stanley punts to Strzemp when Strzemp defends and donks over pot on T72cc, and Walker decides to minraise all in with J8o (lol). Strzemp snaps him with AA.

Three handed doesn't last long either. Judah opens BTN, Ungar flats SB with QJo. Flop JT3hh and both check. Turn 2x and Ungar bets pot, Judah raises 3x all in and Ungar calls and beats T9o.

Heads up takes 6 hands, as Ungar opens A4o and Strzemp defends and does another one of his goofy PSB donkleads on A53r. Starting stacks were 800k (40bb) and Ungar decides to just ship it in over the flop lead. I think you could honestly argue for flat flop/fold turn because Strzemp's range for donkbetting does decently well on this flop and it seems like a WA/WB spot where Ungar is either outkicked or Strzemp is trying to protect his equity with 5x which shuts down on most turns. But anyhow, Ungar ships it and Strzemp calls with A8o and Ungar hits the wheel. Had Strzemp held up, stacks would have been even, and we'd get a better idea of how Ungar would have adjusted to Strzemp's donkbetting gameplan.

I'm a little surprised at the outcome of analyzing this final table. If you told me Ungar played 45/10 or 20/15 I would have believed you; it's sorta hard to guess these things. But Ungar played like a solid lagTAG, vpip in the low thirties and PFR in the low twenties. I also like that he kept his raises on the smaller side (3.5x to 3x), as I'm told most live pros would routinely 4x or 5x it and would have inconsistent preflop raise sizing (3x with 55, 6x with AK etc). Ungar stayed consistent and on the smaller side.

When the equities are the least defined (preflop and on the flop), Ungar applied the lowest amount of aggression usually. When ranges became more polarized and people were able to realize their equity is often when Ungar got to making his wagers. Everyone talks about how Ungar was the best gin rummy player in the world and relied upon godly instincts, so it's not surprising that Ungar's overall postflop gameplan revolved around making his decisions when he had the most information available. To get into a "standard cooler" on the flop as a small favorite would be a disservice to Ungar's plan of basically soul reading everyone and chipping up slowly.

If there's more interest in this kinda stuff I'll look into the 2000 final table. I remember reading "Positively Fifth Street" by James McManus, so it will be fun revisiting his final table and objectively criticizing the play of TJ Cloutier, Chris Ferguson, and HASAN HABIB(!!)

Thanks for reading

Digg this Post! Add Post to del.icio.us Bookmark Post in Technorati Spurl this Post! Google Bookmark this Post!
Comments 23 Email Blog Entry  
mantasm is offline mantasm
This was a fun read, I would definitely like more of these please.
grogheadflow is offline grogheadflow
grogheadflow's Avatar
mrlennyman is offline mrlennyman
Great analysis and read.
BobboFitos is offline BobboFitos
This was awesome
aejones is offline aejones
aejones's Avatar
this is cool
RunAndGun28 is offline RunAndGun28
ya i liked this one too
Probability is offline Probability
Probability's Avatar
pretty sweet
ralphcif789 is offline ralphcif789
Was good (am only halfway through cos i have to go to work).

How about doing 1 for an episode of HSP or 1 of the cash episodes of Poker After Dark, involving our friends and heroes, the 'live pro's'?
clayton is offline clayton
clayton's Avatar
hi ralph, i'm focusing on analyzing poker strategy from players between 1995 and 2000, since that's the material that's available yet hasn't been scoured by the general poker public.

the allure in ungar is that a lot of people, including myself, didn't know how ungar played. everyone on HSP or PAD you already know exactly how they play, so me analyzing their play would be kinda pointless.
heAdstroMan is offline heAdstroMan
heAdstroMan's Avatar
would be fun though? you could analyse some of the loose cannons from the big game, the time aejones was on it or something!
Mr.Bragi is offline Mr.Bragi
I think this is fantastic. Great job.
tedjam is offline tedjam
Good job
brandnew342 is offline brandnew342
moar, please
theashman103 is offline theashman103
theashman103's Avatar
Good work Clayton. This is grade A material from you.
theashman103 is offline theashman103
theashman103's Avatar
I think every play is acceptable except for the A4o shove on flop. Clear flop call.
ralphcif789 is offline ralphcif789
Updated 10-25-2012 at 03:04 PM by ralphcif789
yeah, to follow up on the point heastroman made, and to help out aejones if he's invited back for the 3rd of the big game. why doesnt somebody analyse an episode of pokerstars' the big game? (i read what you said clayton, i know you're only studying 95-2000)

i think itd make a good article. the structure of the show, you've got the ante's (which i know seem straightforward to the advanced player but to a beginner, you wouldnt have a clue that these effect the whole nature of the gameflow, and the cards being played from each position etc). you've got the loose cannon whose motivation and skillset is different to the professional players. there's a few other aspects involved in this format.

there's a good thread on 2p2 where jared huggins asks every1 for constructive criticism about how he played on the show. then every1s really nice to him. and jared to satisfy some psychological need, keeps saying i played really well. then by the end of it, every1s fed up of it, and i think it was terp has a go at him lol. it was funny.

i dont blame jared tho. pokers hard

im not saying it would be a revolutionary study or anything but it might be interesting.
bigdog6391 is offline bigdog6391
interesting, thanks for doing this.
smackjack is offline smackjack
Awesome blog post, please do more!
aejones is offline aejones
aejones's Avatar
re: the big game. i'm pretty sure they just kept those stats on the show themselves and even showed them from time to time (even tho sample size was always very small).
cntgetmedown is offline cntgetmedown
cntgetmedown's Avatar
This was a really good read, thanks for sharing.
guitarizt is offline guitarizt
Awesome thanks! For the next one I'd rather see something within the past year than something old again. My vote's for the 2011 ben lamb table before the final table or something from the one drop.
Lefort4848 is offline Lefort4848
Awesome stuff. The one thing I would add is that I think it's pretty unfair to call any individual play bad given no info regarding live tells etc.. Ungar was obviously about 300x smarter than everyone else at that final table so I'd be shocked if he hadn't picked up on a TON of live tell stuff given the fact that he presumably had so much more live experience in high stakes situations AND he'd been playing for days with most of these guys. Also, it was an era where people were far less knowledgable about concealing information etc so hands like the final one with A4o or the limp/call with A7o could easily be validated with some additional tell information.
zrezzo is offline zrezzo
Thank u so much. Brilliant job.
If there's more interest in this kinda stuff I'll look into the 2000 final table.
Of course YES!
Recent Blog Entries by clayton