Should I steal the blinds? (Greg and Dave spreadsheets)
Should I try and steal the blinds? What hands should I attack with? What if my opponent is tight, medium, or loose?
Greg Mortimer and I have been working on some spreadsheets that try and answer these questions. I have put these spreadsheets on my site in a "secret" location. We are interested in your feedback before I link to these from the public part of the site.
So, please look at this work and comment here.
You can see the work by going to www.CanadianPoker.com and then going to "Who is Dave Scharf?" and then clicking on my kilt.
Or, if you aren't inclined to all that fun, you can click here: http://www.canadianpoker.com/spreadsheets/Spreadsheet.html
Have we made any glaring errors? What should the next generation of spreadsheets include? What do these spreadsheets show you?
Greg Mortimer and I have been working on some spreadsheets that try and answer these questions. I have put these spreadsheets on my site in a "secret" location. We are interested in your feedback before I link to these from the public part of the site.
So, please look at this work and comment here.
You can see the work by going to www.CanadianPoker.com and then going to "Who is Dave Scharf?" and then clicking on my kilt.
Or, if you aren't inclined to all that fun, you can click here: http://www.canadianpoker.com/spreadsheets/Spreadsheet.html
Have we made any glaring errors? What should the next generation of spreadsheets include? What do these spreadsheets show you?
Comments
There's a ton of information there which I'll have to save for one of my study days. It'll likley be a whole day devoted to this topic. On the surface, there's a few things I use in my decision process which are not obvious in your analysis.
First, what does stealing the blinds get me vs. having to fold. When I'm at ten times the BB, it's a very difficult decision, as the risk/reward equation doesn't balance out very well. What it does effectively, is set up your big hands for action. So, a steal here and there gets folks ticked off and ready to explode. Now, the hand range for your opponents is getting quite wide. I'm not sure how it affects a tight player vs. a loose player. However, in my experience the tight player is subject to the worst mistakes. They don't see many hands, and now you're forcing them to see less. If there's alot of stealing going on, the tight player will often make a horrible play with a bad hand like KJs, or A2o. The big question is how future EV factors into the stealing process, as I think this is huge.
The second thing I consider is my image. If I'm relatively quite, then the hand range for my opponents gets very narrow, even with the loosest players. So, if I'm quite, and I need to "buy" a round, I will steal with any opportunity -- usually from the hijack set onwards.Â
One of the things I'm really carefull about these days is stealling from the SB. It's just too obvous a steal. So, the hand range widens for the BB, even for a tight player.Â
Your spreadsheet assumes a 10XBB. I think the matrix needs to focus on avg. stack size. I know there's a correlation of 10X and avg., but can't think of it off hand. I find my steals are most effective on avg. stack sizes. Their hand range is the most narrow. At 1/2 avg. they're looking to gamble and double the stack size, they'll call with a very wide range.
That's my thoughts for now. Great work btw -- you're always getting folks to think. I'll likely pick this up in the new year.
Cheers
Magi
In general, you've got to pick up the blinds at least once per round just to maintain your stack. Given increasing blinds, you've either got to steal a pot more than once per round on average, do a re-steal every couple of rounds or hit a bigger pot every few rounds.
These spreadsheets are a good start at looking at the stealing part of the equation. The re-steal will have to be considered separately.
In terms of big pots, if you are an average stack or smaller, how often do you think you'll be able to take one of these? I would have to guess not more than once every few rounds. Plus, if you factor in failed attempts to win a bigger pot while your stack is dwindling, it looks pretty bleak. Does anyone have ideas of how to model something like this or how to factor it into the game plan with stealing and re-stealing?
I agree that if you have a loose image and try to steal 3 pots every round, chances are you should be looking at the numbers on the 'loose' spreadsheet which aren't nearly as attractive.
I steal whenever there is an opportunity. I rarely let an opportunity go by. But, the factors which create an opportunity really vary with each tournament. So, the frequency of my stealing really changes based on tournament conditions.
I use my image, stack size, and the biggest/shortest stacks I'm facing in my decision making process. I also factor in the type of players I'm facing and their "mental state". Ie. A tight player just fresh off a bad beat is likely going to push hard with marginal hands.
Things to note - I've never won a tournament without successful stealing. It's critical in the transition points in the tournament -- going from early to middle, middle to late, and late to final table. I've found that it's critical because it adds chips to your stack for the double up. So, the steal sets up a later hand. More of a meta-game thing.
Another thing to note, is the re-steal only works on a limited number of players. They must be very strong players, and they must have enough chips left to be able to fold. Otherwise, all their chips will be in the middle before you can say "you called with K2 soooted".
Cheers
Magi
Will you steal with any two cards against tight blinds, or do you only do it with 78s etc.
I final tabled a $50 150 entry on UB two nights ago and tryed the resteal. We where 7 handed and the button raised about 4x, im in the BB with J9o. I know he can lay down a big hand here since we played the final two tables together. We where about even in chips with about 22,000. I reraised him 10,000 more and he went into the tank, he finally pushed and i called for 5000 more. He had AQ and spiked the Ace. My thoughts on this hand are AJ he folds and small PP he folds....because of my image. I wanted my raise to look like I wanted action, and he had enough chips left to fold his hand. My opponent was a pretty solid player but had been raising the blinds a little to much to be holding premium hands. This would be the 4th hand i played against him and I won three of the other showing him AA once and him making good folds(I think) on the others.Now that I think about it....if I would have shown him my winning hands against him he may have folded this hand, but since I didnt he might be thinking that I was pushing him around the whole time......when really this was the first time. This was the first time I tryed the re-steal in this tourney and i didnt work out.
Any thoughts on this play greatly appreciated.
Wader
Was this a good hand to re-steal with?
I usually make me re-steal hands randomized by my cards or after a LONG quiet period. J-9o is not in my re-steal arsenal.
My resteal arsenal is: Any pair, any suited ace, any suited connector (or one or two gapper) down to 7-4s (and excluding suited queens). This is the range of hands that, I think, does the best (30% or better pot equity against a typicall calling hand) when you are called.
Was your play a bad play? I am not suggesting that just because it is not within my resteal range that it was bad. Indeed, if his Sklansky Gap is big enough then ANY two cards may be a nice re-steal opportunity. It looks like he has a big gap since he is "willing to fold big hands." So, probably a good play on your part that got unlucky.
My biggest problem is what to do when you steal and get reraised all in..... you know your done sometimes but you get pot commited on line.....
Now live there's the table image and stuff..... anything I've ever won stealing has been a to a large degree...
I will also (not that I consider it stealing) start pushing with marginal hands in late postition;
Acid Joe's comments are on the mark "If you don't steal you won't win" and Dave S.'s comments are also on the mark "If you don't re-steal" you'll get crushed by top notch players. As my daughter says, "Why?" Don't you just love seven year olds -- they want to understand how the world works. As we get older we get lazy (at least I do) and just want the answer rather than trying to understand "why." Here I think "why" is important.Â
Why you ask? Well, I believe a steal is incredibly complex even though it's roots are simple. It has great ripple effects, and if you're aware of the ripples, you can make more effective use of the tool.Â
1. "What is a steal?" -- for me, that's when I take a pot uncontested pre-flop with a hand that's likely inferior to the hands which folded. As such, I don't consider taking a pot down with Group1/2 hands/pairs a steal.Â
Dave, with your hand-range on the re-steal, I'm thinking you most often have the best hand, so I don't necessarily view that as a re-steal. I'm thinking on the lines of the Harrington squeeze play, as a true re-steal. Why is it important for me to make that distinction? Well, it allows me to think in two streams -- a) where I think I'm best and I can get the most value and b) where I see opportunities to take a pot which doesn't necessarily belong to me. So, any two cards would be in my re-steal arsenal (why did I giggle when I typed that?)-- especially J9o. During this year's WSOP, I pushed in with 44 on what I felt was a steal from Lee Salem -- I wouldn't call that a re-steal. Would you?
2. "Why steal?" -- legitimately, you'll only get around 10-20% of hands which are playable. Obviously, they are not evenly distributed and at key junctures in the game, you cannot wait for a playable hand.Â
I thought winning the blinds didn't amount to much in of itself. But then I developed a spreadsheet which mapped out the progress of a large MTT (say the Sunday stars 4,000 player tournament) in terms of M, and Q. Interestingly, at the start of the second hour 1M will equal 4% of the average stack, but at the third hour 1M will equal 13.5% of Q. I'm not sure what the cross over is, but it's definitely when the antes kick in -- I'll work on that this week.
If you're stealing the blinds with less than 5M, that is truly rewarding. It can be the reward for not playing many hands and developing a tight image. However, to be truly effective at stealing -- where the risk/reward equation is decent, I think you need between 1.5 & 2Q in the middle stages of a tournament. A steal increases your stack by 10-15% of a Q -- I think this is vital for when you pick up an actual hand. At 1.5/2Q there's still enough chips on the table to double up and now your steal is worth as much as 30% of a Q. Very significant.
In the late stages a successfull steal can be the difference between winning and losing the tournament. At this point, the average stack is less than 5M, and there are some massive stacks at the table. So, I think this is where Dave and Greg's spreadsheet comes into play. Not worrying about having to put your chips in, knowing you are +EV. The trouble is how many tournaments are you going to have to play to get to the late stages and make that +EV work out? Trust the math, and instincts -- stealing here is critical.Â
It's getting late in the afternoon and I need to play poker!! I'll catch up with this as we discuss and add in some stuff next week.
Cheers
Magi
I would define it as "a raise to steal the blinds with a hand that is below my usual raising range for the position I am in." Probably not a significant point. When I raise on the button with AK it's not a steal. It is, at that point, an EV play. If I have A-5o I consider it a steal even though it is probably the best hand.
I would call this a re-steal. I have never given the specifics of definition of terms a lot of thought (my philosophy professors would be mad at me). I would call this a re-steal because you do not want him to say, "Call." You want to win it right there.
Aha! We've stumbled upon my definition. If you want action it's not a steal. I note that sometimes the action is OK (e.g. 44 and he has AJ). But if you want to win it now, it's a steal. If you want action, it's not.
I have not responded to the rest of Magi at this time because Harrington's terms are meaningless to me (M and Q). I have only read his first book.
Why steal? Why re-steal?
My simple answer is because you can. Take what positive EV plays that game will give you. Sort out what weapons work against what players and deplot them appropriately. So, steal beacuse it adds chips to your stack. And, re-steal because it adds chips to your stack.
AND AND AND ... do these things because it adds a dimension to your game that is subtle, complex, and hard to play against. This, as much as anything, it important. If you never steal you will be slaughtered by good players. If you never re-steal you will be even more slaughtered by good players. And yes, I know I just wrote, "...even more slaughtered..."
Cheers
Magi
Can you guys ponder on one more question?
Who do you re-steal from? Obviously the guy that has been doing a lot of raising pre-flop, but what about the tight players. Lots of people talk about getting out of a tight players way when he enters a pot, but what about stealing from him. I guess this would not be a re-steal because the tight player has a hand and he is not stealing. Can we get the tight player to fold his hand or do we even want to go there. I seen Scotty Nyugen do this in the a WSOP event. He raised from the cutoff and the bb re-raised Scotty instantly pushed allin and the BB folded AJ or AQ (I can't remember) face up and Scotty showed his 74o or whatever it was.
If your opponent is a tight player when he is faced with a bet for all his chips the only hands he will probabbly play are AA,KK and AK and he will fold AQ and maybe even QQ. So my question is have you ever done the re-steal against tight opponents and if not is it something to consider?
Wader