Matt Matros article in Cardplayer about coinflips

I always held to "waiting for a better spot" but I have seen too many situations where I identified a donk such as this, found myself as a probably favorite (>51% but <60%) and folded. After reading that, I think the math is clear that I am letting a good opportunity slip away.

-Gross

[ QUOTE ]
I always held to "waiting for a better spot" but I have seen too many situations where I identified a donk such as this, found myself as a probably favorite (>51% but <60%) and folded. After reading that, I think the math is clear that I am letting a good opportunity slip away.

-Gross

[/ QUOTE ]

A lot of the theory here assumes that we all have to win coin flips in order to win tournaments? I disagree on this fundatmental premise.

of course... in late rounds when everyone has a limited number of blinds, you have to push every little edge.

[woodguy]

[ QUOTE ]

i call 55 in a heartbeat, then again my tournament entry is pennies to me and life is one long poker game. like doyle i'd bet it all on a coinflip getting 55-45, ive been broke before, its no biggie.

furthermore, he may be the worst player and now you have his sorry ass all to yourself...why give him back to the rest of the table?

and...many top players go broke in the first few orbits because of plays exactly like this...thats why they are top players, theats why you know them and they dont know you.

ask yourself this, would you do it if the buy in was $10 instead of $10,000 of course you would! now go back to the $10 games where you can make good decisions. 10k will be a penny in no time.

[/ QUOTE ]

Nice.

You should post here more often, not just in NL/PL.

This forum would be better for it.

Regards,
Woodguy

[pooh74]

[ QUOTE ]
If you follow Matros' logic, then you call with everything down to about 55, or you fold with everything. Who in their right mind would do that?

[/ QUOTE ]

????

God, I see people make this argument all the time..."dude, would you have called me with 22?!?" "You knew I mightve had AK so why not call with 22?" As though they just figured out that day that 22 and QQ are both flips verus AK.

Anwyay, Matros uses the unlikely example of seeing his opponent's cards (AK) so we know we are a flip..so 55 is fine for his example. This has nothing to do with the theory.

[illegit]

[ QUOTE ]
[ QUOTE ]
If you follow Matros' logic, then you call with everything down to about 55, or you fold with everything. Who in their right mind would do that?

[/ QUOTE ]

????

God, I see people make this argument all the time..."dude, would you have called me with 22?!?" "You knew I mightve had AK so why not call with 22?" As though they just figured out that day that 22 and QQ are both flips verus AK.

Anwyay, Matros uses the unlikely example of seeing his opponent's cards (AK) so we know we are a flip..so 55 is fine for his example. This has nothing to do with the theory.

[/ QUOTE ]
Yeah, but.. dude. Cmon. Queens like.. look real big and stuff. just look at how 5s look. I mean, they're FIVES for pete's sake. And queens, are QUEENS. See what I'm getting at?

[Exitonly]

also, q's are like 3% better.

[Proofrock]

First, I'd like to say that I enjoyed the post very much, and it echoes a question I've often asked: what future edge are we really looking to press?

That being said, I have two comments / questions about the argument.

1) "Let’s reasonably (conservatively, actually) say that when you double up right away, your stack will be worth $22,000 at that hypothetical future point at which you would’ve found your better spot. It’s time to do the math. If taking the “coin flip” gives you a 53.8 percent chance to have a stack of $22,000 later in the tournament, how likely do you have to be to double up later in order to fold your pocket queens? Well, you can answer that by solving this equation: x(20,000) = (.538)(22,000)."

Unless I am mistaken, this assumes that by passing up the coinflip early on your chip stack is static until you do (i.e., that you'll accumulate chips faster if you have a bigger stack). This assumption seems questionable to me, especially during the early stages of the tournament. In my opinion, a more realistic model (to first order) would have you accumulating chips at roughly the same rate in both instances -- in that case, assuming you are in an all-in situation, the equation would become x*24000 = 0.538*22000, meaning x < 0.538 -- i.e., it would be better to wait and take an even SMALLER edge later on.

2) "I just made the argument that very good players should actually take slightly negative EV situations early in a tournament, because if they win the hand, they get to use their skill with their new stack."

Let's say you double up on the first hand. Your table then has you sitting with t20000, and everybody else sitting with t10000. Until somebody else at the table doubles up, those extra chips aren't usable for anything (except the psychological advantage that you can bust anybody else at the table). How does your "skill" with your new stack come into play early in the tournament?

---

Not to say I disagree with the article, but I'm definitely not sold on these two points.

-J.A.

[ QUOTE ]
Matros has done a clear and concise job of explaining it and backing it up with math.

[/ QUOTE ]

Respectfully nath, Matros didn't prove anything with the math in his article.

He says that he has won about 53% of his doubles. Ok fine.

What he has not done is demonstrate any correlation between those doubles and his finishes. Further, he has not done any correlation between those times where he lost the chance to double, but did not bust out and still finished in the money or high up in the money.

I will go out on a limb and speculate that in those tournaments where he has finished very highly that those doubles have come at very key points in the tournament, such as right before the money or at the FT.

I would further speculate that he has probably rarely if ever risked a double very early in a big ticket tournament, and then gone on to a high finish or a win.

If you want to restate the conditions and ask do I take this bet later on in a tournament, then sure, I think everyone and their brother fades the bet. But first hand, I don't think so.

Also, with respect to your point about how to treat tournaments, I made a post recently about generally treating tournaments with the same mindset as a single hand.

In all fairness to you and another poster, I can admit that you have a valid point that to me there is a difference between a $30 tournament and a $10K tournament. Perhaps to both of you there is no difference.

But in all fairness to me, at the small entry MTTs that I play in, I see all kinds of nut-cases and supposed internet rated "top players" busting out or doubling up in the first hand or two. For those that did double, I cannot recall a single instance where the player finished high in the money or won.

You can and have demonstrated the linear increase in expectation by an early double. However, you or anyone else has not established an correlation between an early double and higher finish versus a double at some other point.

The real question is whether or not a first hand double is statistically significant or not. To me, it seems intuitively obvious that the larger the tournament, the less significance the early double would have.

Now if you want to change the parameters and say it is now a 10-player single table $10k freezeout, then I say, heck yeah, I'll fade the bet in a heartbeat. But a 1000-player event, I don't think so.

[adanthar]

[ QUOTE ]
ask yourself this, would you do it if the buy in was $10 instead of $10,000 of course you would!

[/ QUOTE ]

Of course I wouldn't, because in a $10 I can double up 70% of the time by randomly pushing kings UTG and having AJo call.

Matt might be right on second thought, though, because for that matter I wasn't thinking about having a 300% edge at the WSOP when I made that post. How many total donks are there in the main event?

[A_PLUS]

I have some further clarity about his use of 22,000 vs 10,000 when comparing the future edge you would need.

Basically, you have 20,000 after a coin flip, after a set number of hands, you can expect to turn that into 22,000.

When you pass, you have 10K. What % of the time can you turn that 10K into 22K? It isnt meant to be read for only all-in situations.

Your chips are expected to increase (b/c you are a winning player). If you played a million tournaments, you would have a good idea of how your chips are expected to grow given a certain number of hands. Forgetting about the benefits of using a big stack. You have some ground to make up to get your 10K distribution to catch up to your 20K distribution.

Will you reach the 20K distribution levels with a greater frequency than you would win the coin flip? If so, pass on the flip. If not, take the flip.

Basically, people overestimate the % of time they reach the level of chips they would have if they started with 20K.