Finding rhythm in the madness (a theory)

[Irieguy]

I think we all often wonder exactly how beatable the SNGs are. What's the best we can do? Well, for what it's worth... here's what I think. The following pertains to PP single table SNGs (using the $11 limit as an example):

Since poker is a zero-sum game, we can only play well to the extent that our opponents play poorly. In the context of a 50-60 minute SNG, there is a limit to how poorly somebody can possibly play; and therefore, a limit to how much we can possibly win. I think the best place to start is to consider the worst error you can possibly make in a tournament: calling.

When you call, you give yourself no chance to win the pot right then, and you surrender betting impetus to your opponent, making it more difficult to win the pot later. You also surrender chips with the play (a very limited resource). So it's pretty bad to call. The more chips you call off, the worse the mistake is, and the earlier in a SNG you call, the worse you are playing. So specifically, the willingness to call all-in in the first level of a SNG is the single biggest mistake you can make. Fortunately, there is a limitless supply of people willing to call all-in with hands like A-4s, K-10, 66, etc. So, if you are unwilling to call all-in early, you will gain an advantage. Experience has shown me that you should be able to finish in 10th place about 3% of the time, 9th place about 5% of the time, and 8th place about 7% of the time. I think this is as good as you can do, and as bad as you can hope that others will play.

As you move out of the first 3 levels and down to the last 5-7 players, the biggest mistake you can make other than calling too much, is raising too infrequently. If you don't change gears, you will not find big hands often enough to keep up with the blinds, and you will be chronically short-stacked, which degrades your betting leverage and makes it very difficult to make the money without getting lucky. By raising more during these levels, you should finish 5th-7th a much greater percentage of the time than you should 8th-10th, but still less often than your opponents who are not raising enough. Experience leads me to think that the best you can do here with correct strategy is about 8%,9%, and 10% respectively. (More on the 10% number below)

Now the bubble. The difference in net monetary outcome between 4th and 3rd place is bigger than the difference between any other adjacent spots. So, if you are going broke in 4th place too often, you are making a huge mistake. It gets tricky here, because sometimes your opponents are playing too tight, and sometimes they are raising too much. But more importantly, they are almost always too willing to call. So, if you call even less frequently than you think you should, you are probably making money. Even though a 4th place percentage of 12% seems bad since it's worse than the average of 10%, a SNG strategy up to that point that stresses survival more than anything else is likely to land you in 4th place quite a bit. I think that is the right way to play, and I don't think you can do much better than 12% 4ths for those reasons.

So, if you add up all of those theoretically optimal percentages, you get 54%. That means the best In-The-Money (ITM%) percentage you can hope to accomplish at the $11 level is 46%.

Once you get into the money, the blinds are high, your opponents, on average, are better, and you will be short stacked quite frequently due to your survival-oriented/calling-averse (and correct) strategy. I don't think that under these circumstances you can hope to do better than an equal proportion of finishes in 1st-3rd. In fact, I suspect that optimal SNG strategy will result in end-game situations involving shortest-stack play so frequently that the best you can hope for is a preponderance of 2nd place finishes. In any case, the best way to estimate the limits of possibility for Return on Investment (ROI) is probably to assume an equal proportion of finishes 1st-3rd.

So, if the best ITM% you can hope for is 46%, your ROI will be 39% if you assume an equal percentage of finishes 1st-3rd.

That's my story for the $11 SNGs and I'm sticking with it.

At the $22 and $33 levels, you will see slightly fewer opponents willing to make the above mistakes, so the limits of possibility for success will adjust downward a percentage point or two each time you move up.

At the $55 level and up, the skill level continues to improve AND you get more chips. Those considerations require much more complex mid-game adjustments in strategy than you would have to contend with at the lower levels. I think the above theory would have to be modified quite a bit to come up with reasonably accurate estimates for the limits of profitability at those levels. Maybe somebody else would like to tackle that.

Comments and criticisms welcome, please.

Play the right way,
Irieguy

[chill888]

Fair enough.

Main criticism: you say (one of) the biggest mistakes you can make mid game is not raising enough. Strongly disagree.

To many people panic TOO soon, when they still have plenty of chips.

One of the biggest mistakes you can make is committing all your chips too often. A major part of any good S&G strategy is to be able to raise (good aggression) and fold to a reraise (good survival). Too many upcomers read about being aggressive and take it (far) too far.

Need to stay tight although as you say, if you play a hand, play agressively

[Irieguy]

I totally agree with you. I wasn't clear enough. When I say the second most critical error from level 3 onward is raising too infrequently, I mean specifically as the first person in the pot, and with a reasonable raise.

I think the common mistakes at these levels are limping too much, calling too much, and as you say, going all-in too much.

I was trying to point out that the main difference between a skilled SNG player and an unskilled one (besides the baseline error of overcalling) is failing to appropriately adjust to the middle levels by increasing first-in raising frequency. You should still be seeing less than 18% of the flops (including your blinds) at these levels... and most players aren't anywhere near that tight. So, it would be easy for my suggestion to be turned into a giant leak.

I think that understanding that concept is what allows you to keep your percentages for 5th-7th below 10%. It's easy to keep your 7th place percentage way below 10% by just being ultra-tight. But that style will get you 5th and 6th place 20% each if you don't adjust appropriately.

Your points are very sound... thanks.

[chill888]

Yes,

As I once posted, you know you are playing well if - in the same game - someone calls you a rock, and later someone else call you a maniac. Shifting gears is critical.
It seems every 4th tourney or so, I go 30 or 40 hands without winning, and then win a few hands in near succession.

Play early game in first gear, middle game in 2nd and 3rd gear, and late game in 4th or 5th gear.

And as you say calling is (usually) a losing strategy.

But for me my biggest breakthrough was when I learned how to be aggressive, BUT CRITICALLY fold to reraises when it was clear I was beaten. Even if I blow away a decent chunk of my stack - or as some here may put it - even when I guessed that my hand was getting good pot odds.

I believe too many learned only the aggressive part and convince themselves (far too often) that the pot is now too big to fold - when it is clear they have lost. Combine this with tilting away your small stack after a big loss - and I think you have the two major leaks of FAIRLY good players.

Regards

[Crispyk]

I used to be a $30 and $50 player exclusively .. but due to some cutbacks in the bankroll... I am playing the tens hoping to regain some of it. I have to tell you.... i've been pulling my hair out for 4 straight days. Bluffing isn't an option in the $10's no one folds ever.... Raising is hardly a good option becuase it does not thin the field at all. You could have 8 players limp in and you could raise 10 times what they put in and every last one of them will call... This is great when you have AA or KK but anything else is just a gamble to raise with. I've really had to tone down my aggression and wait to play until after i see the flop... Too many times i was raising with AQ and running into bad flops with 5 people seeing it. It just didn't make sense to raise with good starting hands anymore. I do alot more limping .. and i see alot more flops.... i occasionally will run into a 2 pair or tripps but .... thats only normal i suppose. The only thing i can tell you for certain is that if you have the idea that you will go into these tens firing away and taking control of the hands.. you overestimate who your playing against. Patience and less aggressive preflop play is the way to go.

[Irieguy]

While I completely understand your frustration, your reaction to the play at the $10 level is just incorrect.

Lately it seems common for frustrated poker players to make comments about how badly people play online, and then go on to suggest that you can't beat these crazies by playing good poker. "Raising doesn't do any good because nobody will ever fold," etc. If you are raising with better values than the people who are calling you have, then you will win money.

Your plan to limp a lot preflop and wait for the flop is a losing strategy. Mainly because you are then always giving infinite pot-odds to the big blind, and always correct pot-odds and implied odds to the SB. I would love to play at a table where I always got to see my BB for free, and could limp from the SB with just about any 2 and be getting 5 to 1 pot-odds and implied odds of at least the size of my stack.

Despite urban legend suggesting the contrary, horrible players are easy to beat. The variance is a biyatch, for sure, but tight aggressive play is still correct. You seem to be falling victim to the "if you can't beat 'em, join 'em" mentality. Well, if you're joining them...

Play the right way,
Irieguy

[Lori]

So, if the best ITM% you can hope for is 46%, your ROI will be 39% if you assume an equal percentage of finishes 1st-3rd.

I can't agree, no matter how good the reasoning, I've sustained over 50% ITM for over two years on single tables and between 45 and 48% whilst three tabling and I'm confident that my sample is large enough.

Lori

[Crispyk]

Look i understand how it works.... playing tight aggressive.. get the chips in with the better cards.... but by doing this you are also depending on a good flop.... Example .. you have AQ you come in for a 3 or 4 big blind raise.. you get 4 callers.. the flop comes down 7 9 2... You bet out on a bluff maybe 75-100% of the pot... you get called by 1 player the next card comes 2... now where are you... you've drained a quarter of your stack by now.. and you still havn't gotten any closer to winning this pot.... most players will check fold on the turn in this spot.... but the only reason you had to throw the bluff out was becuase the pot was so substantial that it was worth taking a stab at it ..regardless if you missed your flop or not... I'm not saying that im "joining them" in there newbie fest .. i'm strictly talking preflop... its easier to get in a see a cheap flop. I find raising doesn't do anything... Most of the ten dollar games i have played have been substantially weak passive play... Where you can almost always see a flop for 1 bet.... i just find that its much easier to see the flop and play from there instead of trying to muscle them around preflop. I've tried both ways and have found more success in not raising as much preflop.... I can't tell you the number of times .. I have raised AQ just to see a J high flop with someone holding J10o etc... by raising preflop you are depending that much more on a good flop.... your building a pot up but your only going to win it if you hit a better flop then your 4+ opponents that call your raise..... believe me i feel alot better about abandoning a hand with only 1 bet out there then 6 or 7 ..... it just makes more sense to me..

[Irieguy]

That's certainly good to hear. How large, exactly, is your sample?

Assuming your numbers are sustainable, what's your opinion on how you are able to finish ITM 60% more frequently than "average?"

What's the highest limit at which you've played more than 300 SNGs and what is your ITM at the higher levels?

I'm very interested to hear your thoughts on these questions... thank you.

Irieguy

[Lori]

Online the largest I've played is $100, and I don't have a representative sample at that level.

The highest limit where I have a sample that matters is $11 because I live out of my bankroll, and ended up in an expenses = earnings trap.

(Live I have won $900 rebuy events)

My theory on how I've sustained it is that I 'should' be playing at considerably higher limits and that most people who could sustain this level at $10+1 simply don't bother because they can make more with a lower ITM at much higher limits.

I'm hoping to move up rapidly by about December, and I'll let you know how it goes.

Lori

[Prickly Pete]

[ QUOTE ]
The difference in net monetary outcome between 4th and 3rd place is bigger than the difference between any other adjacent spots.

[/ QUOTE ]

As a nitpick, it's typically a 20% bump to get 3rd, 10% to get 2nd and another 20% to get 1st.

[Irieguy]

You use AQ vs. JT as an example of why you would rather limp than raise preflop.

You are a 63% favorite with AQ, so why would you want to bet less? I'm not saying you should shove your stack in anytime you are ahead, but you should usually bet.

There are two fundamental concepts in poker upon which all others are based (these are not my ideas, in case you haven't seen them before):

1. Make the most of it when you have the best of it.
2. Anytime you can get your opponents to play differently than they would if the cards were exposed, you are playing correctly and they are making a mistake.

So if there are 3 limpers in front of you and you limp with AQ and then show your cards, what should 2-2 do behind you? What should A-4 do from the BB? They should call, and they do. So, you are giving your opponents the opportunity to play correctly. If you raise appropriately, they shouldn't call if they know your cards, so they are making a mistake... and they do.

That's just the way it is.

[Lori]

1. Make the most of it when you have the best of it.
2. Anytime you can get your opponents to play differently than they would if the cards were exposed, you are playing correctly and they are making a mistake.


This is flawed in a SNG environment.

It is particuarly flawed against the kind of idiots that play $10 events when you are likely to beat them anyway.

Getting your whole stack in, for instance, with a 55% chance would often be -EV because on the 45% of the time that you lose, you miss out on the times you could have got it all in with an 80% chance of winning later in the tourney.

I haven't checked the numbers on this particular example, however the prinicple is important.

Lori

[Irieguy]

This whole thread applies to $11 PP SNGs. The difference between 4th and 3rd is (-$11 vs. +$20 = $33). The difference between 2nd and 1st is (+$30 vs. +$50 = $20). So, the net monetary outcome between 4th and 3rd is bigger than the difference between any other adjacent spots.

Nitpick back at 'cha.

[Lori]

The difference between 4th and 3rd is (-$11 vs. +$20 = $33).

You mean -$11 and +$9 surely?

Lori

[Irieguy]

I don't remember saying anything about getting your whole stack in, did I?

The concepts are not flawed in a SNG environment, they just need to be applied appropriately. You should often pass on marginal advantages in the interest of survival, but that's not at all contridictory to the fundamental theorum of poker if you understand the context.

In TPFAP, Sklansky illustrates very well the concept of passing on an advantage when you are sure to have an even bigger advantage later. He doesn't feel the need to rescind the fundamental theorum during this discussion.

Lori, you are absolutely correct and your points apply particularly well to why you sometimes shouldn't call with the best of it. I just don't think you need to throw out fundamentally sound poker reasoning in order to make those adjustments. You know what I'm sayn'?

[Irieguy]

Surely I do, thanks. Same to pete... jump from 2nd to 1st offers same net as from 4th to 3rd.

He's a better nitpicker than I.

[LinusKS]

Irie, I haven't thought about it in those terms, but that is an intriguing way of looking at it.

The one thing I'd say about the PP$10 games is that the variance there has gotta be huge. In the first place, the stack:blind ratio is so small there's bound to be an enormous crapshoot aspect to the game. To take one example, if you get to the 50/100 level with 700T, and raise 3bb, you're pretty much pot-committed, and your outcome on that game could very well depend on that one hand.

In the second place, the abundance of loose-aggressive/maniacs at PP increases variance as well. Yes, you'll win more when you're running good, but you'll also lose more when the cards aren't coming the way you'd like.

More importantly, you lose the ability to bluff, which IMO, is the most important survival tool a good player has.

Anyway, I suspect your numbers aren't too far off, and I also think the variance in the PP games is far higher than people think.

[LinusKS]

To take the example one step further, if you got in with the best hand 100% of the time, and your average advantage was 60:40, that would mean you'd have a 40% chance of going out on that one hand alone.

If you had to do that twice to make the money, you'd have only a 36% chance of surviving both.

[Irieguy]

Exactly... but if you got all of your money in with a 60/40 advantage every time, then you would be doing 50% better than your opponents. (Your opponents win 40% of the time, and 50% more often would be an additional 20%, or 60% total.)

If you do 50% better than everybody else, you would finish ITM 36% of the time to their 24%, and that would make you a nice profit over the long haul. Fortunately, there are players who are willing to give you much more than a 60/40 advantage for all of their chips. So, you can do even better than a 36% ITM%.

Yes, variance is huge... but it must be. If you could beat these games any more soundly with any less variance, then the top 2% of the players would have all of the money within a year or two. Then what the hell would we talk about?

Irieguy

[dethgrind]

[ QUOTE ]
you should be able to finish in 10th place about 3% of the time, 9th place about 5% of the time, and 8th place about 7% of the time

[/ QUOTE ]

[ QUOTE ]
Experience leads me to think that the best you can do here with correct strategy is about 8%,9%, and 10% respectively

[/ QUOTE ]

I don't see how you can confidently produce these numbers. Are you just looking at your own results, or the results of a couple successful players?

[ QUOTE ]
I don't think that under these circumstances you can hope to do better than an equal proportion of finishes in 1st-3rd


[/ QUOTE ]
I'm not sure if I agree with this. Some of the more respected SNGers have results that disagree, including Jason Strasser (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&Number=761547&page=&view=&sb =5&o=&vc=1) and Bozeman (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&Number=762160&page=&view=&sb =5&o=&vc=1). I would guess that even if you came into the money with on average a smaller stack than your opponents, highly aggressive play could get you a larger than fair share of 1sts. Jason Strasser gave a pretty good explanation in the above link.

[LinusKS]

I agree with you that variance is the reason why losing losers play (though that doesn't make variance any less of a pain in the butt).

Also, you can also often get in with better than 60/40 -- but sometimes you're going to be the dog, no matter how good you are. Even morons catch big pairs, and they tend to misplay them just like they misplay Axs.

And that's assuming it matters what they do, since the blind structure at Party will often demand an all-in, regardless.

[LinusKS]

That reminds me of something I was thinking about last night.

Suppose you're at PP$10, you have 700T in MP, with 50/100 blinds, it's folded around to you, you're holding AQ, and you decide to make it 300T to go.

Everyone folds to the bb, who happens to have you covered, and happens to be holding K2s -- which makes him right about 40% to win the hand right now.

Question #1: Do you want him to fold, or go all-in?

Question #2: Say he goes all-in. Do you fold, or do you call?

[Marcotte]

You didn't mention your position (ie how many to act after you) or how many total left. But with only T700 left I would either push or fold preflop. If you bet 300 and get call there is 600 (+ sb) in the pot. You have no fold equity on the flop. If you know for a fact that everyone b/w you and BB will fold, then I would push.

[LinusKS]

[ QUOTE ]
I'm not sure if I agree with this. Some of the more respected SNGers have results that disagree, including Jason Strasser (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&Number=761547&page=&view=&sb =5&o=&vc=1) and Bozeman (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&Number=762160&page=&view=&sb =5&o=&vc=1). I would guess that even if you came into the money with on average a smaller stack than your opponents, highly aggressive play could get you a larger than fair share of 1sts. Jason Strasser gave a pretty good explanation in the above link.

[/ QUOTE ]

I agree that a good player can take more than his fair share of firsts, once he gets to the money. However, that affects ROI, not ITM.

In Strasser's post he claims rates of 16-10-12% for each place, which are very good results IMO, but still "only" an ITM of 38%.

[LinusKS]

You haven't really answered the question, Marcotte, so let me ask it again, in a different light:

Are you pushing all-in because you want him to call you all-in, or because you want to pick up the blinds?

[LinusKS]

Btw, you're wrong about the fold equity.

If the BB has 1000T, and calls 200T, he will will think twice about calling another 400T if he misses the flop.

[Marcotte]

[ QUOTE ]
Btw, you're wrong about the fold equity.

If the BB has 1000T, and calls 200T, he will will think twice about calling another 400T if he misses the flop.

[/ QUOTE ]

Well, maybe you don't have zero fold equity, but you are certainly giving him odds to call with a lot of holdings. With a T400 flop bet there there is T1000 + the sb in the pot giving him at least 5:2 odds.

As to your other question, it would depend a lot on the situation. If we were on the bubble, I think I would prefer a fold, but with 5 or more, I think I would want to take my chances to double up and go for first. If I knew he had K2s, I think I would want the call. But then again, I'm not sure how strong a player I am. /images/graemlins/grin.gif

Result
http://twodimes.net/h/?z=347641
pokenum -h ad qh - ks 2s
Holdem Hi: 1712304 enumerated boards
cards win %win lose %lose tie %tie EV
Ad Qh 1043111 60.92 660873 38.60 8320 0.49 0.612
Ks 2s 660873 38.60 1043111 60.92 8320 0.49 0.388

[psymunn]

Couldn't agree more. While I have a 40% ITM, which is respectable, I find that, once I make the last 3, even often as short or middle stack, I still have 60% of my ITM as first. I guess the fallacy the origional poster made is that the best players make the ITM when, in fact, odds are you and, maybe one other good player made it. The reason for this is that, Wild players will double up off each other. The survivor of the two or three erratic players will often have chip lead and then it's just a question of stealing from him when blinds get big. Also, you have to be aware that chip leaders don't want to give it up, especially on the bubble. Bullying a leader before he's in the money is a great way to secure a first when the 4th goes.
At the end of the day a 4th and 1st is better then two thirds, so go for blood.

[dethgrind]

Absolutely.

To repeat though, I believe Irieguy may be wrong when he states that an ideal strategy will yield at best an even split between 1st-3rd finishes.

A better strategy might be to sacrafice some ITM, and play more aggressively on the bubble. This way, you will come into the money with a bigger stack on average, and get a larger share of firsts, where the real money is. Decrease your ITM to increase your ROI.

[kamelion44]

Irie, I'd be very, very curious to see what you'd have to say about $55 and up SnGs. I have had a very poor run at the SnGs, and would like advice at all levels (my ROI is dipping under 30% as we speak). Your post did help me put some stats in perspective, thanks. Now, help me with the higher limit SnGs!

[LinusKS]

kamelion, for what it's worth, I'd argue there's absolutely nothing wrong with a 30% ROI. In fact, I'd argue it's very very good.

I think you can get the wrong idea from browsing through the forums. It's just too easy for people to come on and post numbers, or to post them from non-representative or insignificant samples.

Consider this: if the people who measured such things just took peoples' word for it, the average man would have a 12" dick.

[Irieguy]

I just moved up to the $55's, so I'm in the middle of compiling enough trials to have anything of value to say on the topic.

I'm just over 100 now... I will post my results when I have 300-500 along with any thoughts I have on the major differences between the 33's and 55's.

Irieguy

[kamelion44]

Linus, you mean you DON'T have a 12" dick??? But yeah I know what you're saying...I looked over my results and actually my ITM results are really nice, right on par with what everyone always says and on some of the limits it's actually higher, but my ROI is really low on some of the limits (like $33s it's only like 15%), despite the ITM being on par. Maybe I'm just not closing the deal like I need to. Also, Irie, I'll be waiting anxiously for your results.