The situation
5 left in an EPT tourney with a €1k buyin that got a 337 runners, and stacks are
HJ (hero) 6 BBs
CO 7 BBs
BN 10 BBs
SB 12 BBs
BB 30 BBs
Hero is dealt KQo but before we act CO moves all in and Button folds (both out of turn). The EPT rule is if we change the action by raising, CO retains all options (fold, call, raise) but if we fold or limp they must jam. So what are we doing? Hero estimates they have somewhere between 45-50% equity versus CO jamming range.
Analysis
OK, let's consider the options
- Fold: CO now has to jam, and if everyone is calling correctly, he gets it through over 70% of the time, and gets called about 27% of the time (mostly by the big blind with a slightly wider range than he's jamming). So we probably ladder about 12% of the time, the CO doubles 15% of the time (really bad for us and we have to post bb and ante next hand), and the CO picks up 2.5 bbs the rest of the time (also pretty bad for us)
- Limp with the intention of calling the jam if nobody else does: We are committing 6 bbs to a pot that will be 14.5 bbs, so for CEv the equity required is 41.4%. To this we add risk premium which is 6.2%, so we need 47.6%. We are about 2% short of this
- Jam ourselves: this is the play if we didn't know CO has a hand strong enough to jam. Of course if we jam they get their options back, and should only call the hands they're supposed to. The problem is that that's just over 7% of their total range, but about 35% of their jamming range, so we have way less fold equity than normal. KQo doesn't make that much as a jam and knowing we are getting called 5 times more often by a strong CO range almost certainly makes it unprofitable.
However, with future equity knowing we have to post 2 bbs next hand, 2 is almost certainly the best option. I confirmed this by running a node locked HRC sim (so CO has to call his open jamming range) which now has KQo outperforming fold (by a narrow margin) and shove (by a wider margin)
Final question
Do you think you could make this decision in 80 seconds? That's all David Docherty (the player in question) had: 20 seconds, plus two timebanks worth 30 seconds each. I hate shot clocks on final tables for this very reason.
David's view
So I think the first interesting thing is my equity against his range. I considered I was roughly 45-49% in-game which seems to be accurate, but I'd also add that there's no way he's inducing anything or mixing 22 or folding any ace or folding KTo, T9s or any of the suited kings that are mixing. All of that considered, I think my equity will go up a percentage point or 2 from the just over 45% you ran the spot with, which should only make limp calling more favourable and the other two less favourable. I think jamming ourselves is clearly the worst of the three options, not just with this but maybe with any hand? Let's say we have AJo and want to induce vs the misclick but don't want a 3 way all in because we'd have to beat both players. I don't think the range that overcalls both jams will change significantly assuming he still sticks it in, and if he folds to my jam I think I'll get called lighter because my opponents will be able to deduce that my hand wants fold equity.
I think ultimately what these results are telling us is we do best when our opponent is all-in with his widest range and having to run his equity. If we fold with any hand that's in that 45-49% delicate equity range then we give him the opportunity to chip up while we'll be putting a third of our stack in blind next hand. If we jam, we give him the chance to fold and we may bust ourselves while not significantly changing our position within the tournament given we're definitely not getting a walk next hand. If we limp, we get full information on the rest of the table's hands, get to fold our way out of a potential cooler, still have a protected range, and can win the chips of the person at the table whose chips we desire the most.
Results
David won the hand, crippling the other guy. Then next hand was faced with another ridiculously decision (and no timebanks left now).
The next hand
David doubled through the cutoff in the previous hand, so he's now big blind and the stacks are:
UTG 1 BB
CO 10 BBs
BN 12 BBs
SB 30 BBs
BB (hero) 14.5 BBs
UTG puts his last big blind in, and it folds to the chipleader in the small blind who shoves. We have ATs in the big blind? Call or fold?
To make matters worse, there's a shot clock and we burned all our timebanks last hand, so we have 20 seconds to decide.
Analysis
This is another ridiculously close spot. I ran the sims and before I give the results I want to talk about some misconceptions people have about this spot:
- UTG should jam any two: not true. Because the rule is BB first, if he folds this hand, next hand he gets to post his last big blind as a blind (no ante), so it has the same value. Therefore, he has no reason to shove a below average hand. In the ICM sim I ran, he therefore only shoves top 40% (a bit tighter than average because of the chance he might ladder this hand)
- Yes but future game: Some people thinks because of future game UTG should go wider than ICM. The reverse is actually true: in the FGS sim he only shoves top 31%? Why? Because winning next hand is far more valuable than winning this one. If he wins this hand he's still pretty much committed next hand, and one of the big blinds he wins is worth a lot less than a big blind to him because it has to be anted. If he wins next hand he gets another whole orbit to find a good spot, and more chances to ladder
- The small blind should shove any two: Not true. Obviously when he shoves the big blind has to call tight, but the downside of the disaster it is when we run into a top 5-10% hand compared to the upside of taking on URG with lots of dead money doesn't make shoving any two a good thing. Shoving 32o here loses $272 in the long run.
In my experience, when players, even good players, talk about future game, they're only doing it to justify what they want to do anyway. Guys who play for the win will always say "I call because that wins the tournament more often", while guys who love a ladder will say "I fold and lock up a ladder". In actual fact there is a line, which the computer can calculate without factoring in emotions like "I love winning tournaments" or "I hate busting tournaments").
Action clocks
Some organisers enforce the "act out of turn, get an automatic penalty" rule rigidly. EPT apparently does not, as no penalties were imposed after the hand on the two players who had acted out of turn. By contrast, some other rules that I think aren't as important (like the nebulous one that if someone is persistently using too much of their shot clock, floor is called and they're on a 5 second clock from then on: I hate this because it's now on the dealer to define what too frequently and too much is, and does it really matter all that much if someone uses 16 seconds every hand?) are enforced rigidly. There's another rule I hate which is if most of your stack is in, the dealer has to call the floor and you're on a 5 second shot clock. That also is poorly defined (dealers weren't able to tell me what exactly constitutes "most of the stack", it's a matter of judgment). I'm not sure every dealer is qualified to make that judgment. Think of Jon Kyte's spot at WSOP: he genuinely had a tough decision with most of his stack already in (and in fact ended up folding). Imagine if they only gave him 5 seconds to make that decision?
These things falling within the realm of dealer discretion also opens up the possibility of unconscious favouritism. I had a spot ITM of the mystery bounty where I decided the correct play was to just call a raise 7 way keeping slightly less than a BB back, and see if I could flop equity before putting the rest in. I didn't, and now facing a bet I had to work out what equity I needed (roughly 6%) and if my hand had that even after missing. I ended up folding after using all my time. In that case, when the dealer realised I'd put 2 of my 2.95 bbs in pre, she looked towards the floor as if about to call them to put me on a 5 second shot clock, but then looked back at me, and smiled as if to say "I'm gonna let this one slide". This is a dealer I like very much and get on very well with, so nice for me, but not nice if I'm a random the dealer doesn't know or someone they actively dislike, and decide to call the floor. Another dealer I'm friendly with warned me privately about the "using too much of your clock every hand" rule after dealing to me where I was doing that: again I suspect they'd just have called the floor if they didn't like me.
One unintended consequence of these rules is because they are taking up so many dealer brain cycles, and because floor is running around imposing 5 second shot clocks on people, it's distracting from more important tasks. I noticed the amount of basic dealer errors increased dramatically once the shot clocks were introduced (after late reg closed) and they started having to focus a lot on the time rules, and for the first time ever at an EPT I saw floor forget to make important announcements like blind level increases (because they were too busy running around putting people on 5 second shot clocks). I'm in no way blaming the dealers (which I said in my last blog are the best in the business) for these errors: they're inevitable when you now place an additional load on them to operate the shot clock for every single action of every player, and also monitor players over several hands for using a nebulous "too much of their clock", and putting "most of their stack" in.
The worst example of this I saw at EPT Barcelona was in a hyper where every player just had 15 seconds to act (with a couple of 15 second time bank extensions). This was far too little time for many recreational players (and this was the most rec heavy field I played), and even the dealers were getting flustered. Almost every hand had a significant player or dealer error.
The 5 second clock
In my view, this rule is really problematic. It seems to be the view of the floor at EPT that if somebody leaves a small amount of their stack behind then it must be for the purposes of stalling. This is not at all the case: I've written several blogs trying to explain the benefits of the "leave one big blind behind" that aren't just "stall for a pay jump". David told me he witnessed a player bet all but one big blind on the river, be put on a 5 second shot clock, and also be told if they did it again they'd be put on a 5 second shot clock for every decision every hand. This is pretty ludicrous in my view: for example, often the correct play with the nut low on the river is to bet all but one big blind (intending to fold to a raise). Players should be free to use that play without risking being put on a 5 second clock for the rest of the tournament.
The vibes
I understand completely why tournament directors are tempted to introduce action clocks. They tackle the problem of stalling, which is one of the things recreational players hate the most. In fact I'd go so far as to say if you ask recreationals if they are in favour of action clocks, a healthy majority would say yes, to cut down on stalling. They'll go on believing they like action clocks until they suddenly have a huge decision and 15 or 20 seconds to make it, or they get their hand killed before they can make their decision.
It's not lost on me that in the hand I started this blog with, two players folded out of turn. I saw it time and time again: as soon as the action clocks came in, players were paranoid about using too much time, and acting out of turn by mistake if the dealer or another player even glanced at them. It also completely changed the atmosphere at the tables: tables which had been friendly with everyone chatting suddenly became grumpy morgues where players didn't feel they could commit brain cycles to chatter or any other distraction. The last table I played on at the festival was one of the grumpiest ever, with half the table tilted by rulings given against them for acting out of turn, or in one cases taking one second too long to act.
Corners and headers
I love football to the point I used to be a referee. It's a near perfect game, but two of the imperfections is that corners and throw ins are a case where the penalty doesn't fit the infringement. Only 3.3% of corners are scored, so defenders have a strong incentive to concede them if they judge the xG of an attack to be greater than 0.033 (which is pretty much every attack). Throw-ins are even worse: the team awarded the throw in doesn't retain possession from them much more than 50% of the time. This is why some teams are deliberately kicking into touch from kickoffs.
FIFA is obviously aware of these problems, and attempts were made to address them when I was a referee. For corners, they tried to make them tiebreakers when games finished level on goals (so if a game finished 1-1, the team with the most corners won). The problem? When they trialed it at junior football level, they found it completely changed the game, with teams deliberately "scoring" corners to try to win the game on corners (in the same way that when gaelic football changed the rules to make shots that passed over the crossbar the tiebreaker, teams now started shooting for points rather than goals).
The proposed solution to the problem of throw-ins was to allow the team awarded a throw to kick the ball. The problem? This went too much the other way: the allure of effectively getting a free kick from the touchline meant teams adjusted and the game became one of trying to bounce the ball into touch off an opposing player.
Unintended consequences
The above examples show what often happens when the rules are changed to solve a particular issue, but the unintended consequences fundamentally change the game for the worse. I believe this is true of the action clock: once it's introduced, live poker is a different game, rewarding experienced players used to making quick decisions. Bullet chess is fun, but no serious chess player thinks the world championship or any prestigious over the board tournament should be decided in a minute. Action clocks introduce an additional work load on dealers and floor staff, and additional pressure on players, leading to more mistakes, more frustration, more arguments. They remove much of the ability of players to pick up and act on physical reads. How long did it take Mike to pick up on Teddy KGB's Oreo cookie tells in Rounders? A lot longer than 20 seconds.
In my view, action clocks change what used to be a gourmet live poker experience (EPTs) into an unsatisfying fast food one. Talking to one player at the end of the festival, he agreed that action clocks almost completely ruin the live experience, and he went further than me saying
"Given a choice between a festival with action clocks and one without them, I'm picking the latter every time going forward".
As soon as he said it, I thought "I agree".
I agree.
.jpg)
0 comments:
Post a Comment