www.flopturnriver.com

Poker Forum

Over 1,246,000 Posts!

Already Registered? Username: Password: Remember Forgot Password
Tweet

PT Stats, Comparing to Optimal? Just what is Optimal really?

LinkBack

Thread Tools

Display Modes

RiverMonkey

03-04-2005, 06:31 PM Post subject: PT Stats, Comparing to Optimal? Just what is Optimal really?

#1 (permalink)

Flush

Join Date: Nov 2004

Posts: 446

I was reading the thread:

http://www.flopturnriver.com/phpBB2/...highlight=vpip

, and came across Hyper's comments below. This got my rather over-active noggin' thinking ..... Let me know what you think ..... The more comments, questions, answers you have, the better.

Quote:

Originally Posted by hypermegachi

to my knowledge the optimal stats are...

full ring should have a VPIP between 10-20
6max should be between 20-30

full ring PFR should be around 8-10%
6max should be around 15%

aggression postflop over 2 for full ring
over 3 for 6max

i don't have enough hands to give any significant data.

A few contextual points before I get to an idea I want to run by everyone for comments:

(1) Assume we all accept the above stats guidelines from Hyper (or that we could all agree upon some other similar set of guidelines given an assumed game profile (IMHO this is an important point).

(2) If you are in a full ring game, and four people fold in front of the first VPIP'er, then don't you effectively have a 6-max game? This would call for for more aggression and a higher VP$IP, right?

(3) Starting hand requirements go down when you approach shorter handed games. Shorter-handed games also call for more aggression.(IMHO, this is really just another way of stating point (2).)

(4) You should play differently when a tighter player or players open before it's your turn to act pre-flop.

(5) Deriving value from comparing YOUR PT stats to "the optimal" values requires that you are comparing apples-to-apples. To me, this means you have to ask yourself (i) "how do we truely define optimal?" and (ii) "are your stats derived from a large enough sample/do they truely represent the long-run?".

----------------
Let me start by saying that to apply my idea below I think that you would really need a rather large HH-DB. (because the smaller it is, the smaller each of your sub-samples will be, and the more variance will rear its ugly head and cloud the waters).

Anyow, given the above points, would it not be extremely informative to try to take your hand-history DB and segment it up into a number "buckets" of hands (using filters I guess) where each bucket is defined by how many people saw the flop/how many people outside of the blinds either (i) limped, or (ii) raised pre-flop, (and prehaps also use pre-flop aggression indicators of your opponent as another bucket-defining dimension too)?

In other words, try to define better sub-samples of your hand-history DB for comparison to optimal PT stats for each sub-sample. Where by 'better', I mean you can be sure that 'optimal' for each bucket is more reflective of game-conditions. I think 'optimal' would have a more definitive meaning for each bucket than it would if you looked at 'optimal' across all your whole HH-DB where you would have the clouding affect of averaging across those buckets. Without the buckets are you not assuming that all game conditions are equal ... clearly they are not!

It seems to me that it would very useful to know what your std. PT stats (such as your VPIP, aggression numbers) are for each of these buckets compared to an agreed-to optimum for each bucket.

If for example, looking at your entire hand-history, on average you've played in more games where the game-profile was such that should have been more agressive (or put more V-$ in the pot), you'd expect the optimal aggression factors for those games (or VPIP) to be higher and in-kind your AFs (or VPIP) should have been higher too.

Some questions for you are:

Just how large a HH-DB would you need in order to make this kind of analysis practical?

Can you think of better ways to define each bucket?

How would you go about deriving/agreeing on optimal stats for each bucket?

Would doing this be worth the effort and would it be informative? Or would it just be a waste of time that would be better spent earning $$ at the tables?

Building on the previous question, if it is just a waste of time, given that your PT stats averaged across all buckets is more than sufficient to improve your game TODAY, would it be something you'd consider doing the next time you felt your game had plateau'ed?

Play for FREE and practice your game at...

Join the FTR Poker Forum to disable these banners and start posting!

|~|ypermegachi

03-04-2005, 06:51 PM

#2 (permalink)

4-of-a-Kind

Join Date: Aug 2004

Location: emo-kid

Posts: 3,580

the thing about those "optimal stats" is that it only works for the majority. for the most part, most winning players will fall in there, but that doesn't mean we can't have exceptions. there are many successful players that will deviate from those stats.

for example, some people may choose to raise 10% preflop in shorthanded games vs the recommended 15%. what they are effectly doing is lowering their variance for the sacrifice of a little expectation.

the problem with filtering into "buckets" is like you said sample size. i don't think even if someone had 100k hands that it would be a big enough sample for each individual bucket.

at first i used to be obsessed with stats. but i eventually realized that stats don't mean anything really....if you are playing g00t poker your stats will eventually fall into the proper ranges. it's better to fall in there playing good poker rather than purposely incorporating extra call downs or raises just for the sake of "i need to go to showdown or i need to raise more."

RiverMonkey

03-04-2005, 07:35 PM

#3 (permalink)

Flush

Join Date: Nov 2004

Posts: 446

In other words .... what you are saying is: just go play g00t poker and earn $$s you stupid, hairless monkey!

Yeah, when it comes right down to it, the only real "stats" that matters are $$s.

Unfortunately, I'm one of those twisted, sickos that likes the math and the theorectical aspects of the game almost (but not quite) as much as taking other people's chips. Mmmmm ... other people's chips .... *drool*!

Chicago_Kid

03-04-2005, 07:43 PM

#4 (permalink)

Full House

Join Date: Aug 2004

Location: People let me tell you about my best friends...

Posts: 1,132

Well, I definitely think the stats help when you are starting out. I still find myself slipping into passive mode, but regularly reviewing my Agg numbers reminds me to either push my edge or get out.

Once I'm giving correct advice on hands like you Hyper, I might just watch the $$$ and pay less attention to the other details. Plus, like Monkey said, I love poking around in PT to see what trends I might find...there are more leaks in there for me, I assure you!

"Been gone so long, forgot how to poker"

|~|ypermegachi

03-04-2005, 07:54 PM

#5 (permalink)

4-of-a-Kind

Join Date: Aug 2004

Location: emo-kid

Posts: 3,580

about the only "obvious" fixes are VPIP and PFR. if you have 40% VPIP you're clearly leaking chips all over the place. if your PFR is 3% you are clearly missing out on a lot of preflop equity.

now the thing about aggression is most people wrongly "fix" it. they have an aggression of 4. people say they are too aggressive. so now they raise less to "fix" it.

others have aggression of 2.5 (which i think is perfectly fine), and someone will say knock that up to 3. so now they start raising more to drive it up.

the problem that many people fail to realize is that folding will INCREASE your aggression.

aggression is calculated as bet/raises to calls. so, in a sense if your aggression is really high maybe you're just folding too often to aggression and not calling down with second pair.

or if your aggression is too low maybe you should fold more often with mid pair.

but if you play g00t poker and trust in your reads it will all fall into place.

RiverMonkey

03-04-2005, 08:05 PM

#6 (permalink)

Flush

Join Date: Nov 2004

Posts: 446

My PT DB is definately too small (just bought it two weeks ago and loving it) for me to start sharing specific #s and getting much useful feedback on them. I do have a few related questions though .......

I can say (based on about 1.2K hands) that my agression numbers are on the high side on all streets (where I want them really), and my VP$IP is floating around 30% (maybe a little high, but perhaps somewhat justifiably given that I try (and succeed) to blind/flop-steal a fair amount and can often run over weak/tight tables with bluffs and/or marginal hands).

I know full well that 1.2K is not enough hands and short-term variance is probably very prevelant in these results, so I'm not going to make any real conclusions about my numbers or play yet.

I've been making a $$-killing at my $4/8 B&M games for a while now. My online $BBs/100 is on the high side too (since I started really counting with PT) ... but I don't think it's sustainable at that level from what I've seen of others'. I'm reallying looking forward to seeing what level it settles at after say 10K hands or more.

How many HHs would you say I need before I can confidently say, based on my $BB/hr, that I'm 'crushing' my online games?

What would you consider 'crushing'; over 4 BBs/100 after say 15K hands

When using PlayerView ... how many HHs do you feel you need for an opponent before you start to feel that you can use their numbers as ingredients for your at-the-table decisions?

And to all of you THANK YOU! I would not be as far along in my game if it wasn't for FTR, SSH (which I only bought because of FTR recommendations), and of course, many-many hours at the B&M and online tables!!!

Trikflow77

03-04-2005, 08:14 PM

#7 (permalink)

Full House

Join Date: Jan 2005

Location: im so asian

Posts: 1,460

How many HHs would you say I need before I can confidently say, based on my $BB/hr, that I'm 'crushing' my online games?

I think you need at least 50k, closer to 100k hands to get a good solid win rate. You are really not going to be able to sustain a 4/100 for 15k hands without a really good card rush. I have about 27k hands at .5/1 and my win rate is at 3.47bb/100, which is pretty high.

What would you consider 'crushing'; over 4 BBs/100 after say 15K hands

You are crushing a game if you are making over 2.5 big bets, anything over 1.5 is good play. The higher you move up, the less you are going to make bb/100 wise

When using PlayerView ... how many HHs do you feel you need for an opponent before you start to feel that you can use their numbers as ingredients for your at-the-table decisions?

I like at least 30 hands before I really consider the stats.

Your VPIP is too high. If it stays at 30%, you will not win at 2/4 on up in the long run. Look over the tight game chart from SSh and review your play. I think you are playing too many hands out of position.

[/list][/quote]

RiverMonkey

03-04-2005, 09:20 PM

#8 (permalink)

Flush

Join Date: Nov 2004

Posts: 446

re: playing too many hands out of position, I definitely DO NOT do this because .... I know better.

I think my game has finally advanced beyond strictly adhering to starting hand charts. Thanks to FTR and SSH I now understand the conceptual foundations behind those charts.

I did start off that way, and used SSH's charts as a foundation to build on. But now, within a game and game-to-game, I'm always actively tuning/changing up my starting hands based on position, players, reads, etc. Am I always gonna get it just right? Of course not. But, I now know that any little edge I give up pre-flop by playing a few (and I really mean just a few) extra hands, I can make up with 'g00ter' post-flop play. We'll have to see if my PT stats bare that out over the long-run when I've got more HHs. A look at Std.Dev and BB/100 will show if in fact I'm taking on a little xtra variance for higher EV.

Now, that all being said ... if you multi-table you can, and probably should play tighter pre-flop cuz' you lose some of that edge post-flop that you derive from being able to study your opponents more closely. My experience is that you tend to play your cards more when you multi-table and can afford to play the players more when you only play one or two tables.

I think my current VP$IP of 30% after 1.2K hands is attributable to a few factors:

(1) I play some hands that I wouldn't otherwise play because game conditions say I should. e.g. blind stealing, or setting up a pot-steal on the flop (and much more rarely on the turn), I'm getting better at game-selection, and I'm quick to ID players that are weak and/or tight (I'm hoping this will be even easier with PT and PlayerView), so I can take advantage of that. In some games, I can't get away with running these tactics as often (or sometimes at all), so I don't try them anywhere near as frequently. In other games, they let me, so I do it more and run em' over.

As an example, I might open for a raise with A-rag-offsuit (and even KTo) from late-MP if there are only 2 or 3 weak/tighties left to act after me (including the blinds), but I would never even dream of trying to limp in with A-rag-offsuit from EP. Will I run into domination situations every now and then doing this? Of course, that's where and why I'd tend re-raise or bet on the flop (depending on texture) for info. and likely fold (or if I'm ahead, I'll either take er' down there, and/or take another off on the turn and then take er' down).

Similarly, I might raise on the button with A-rag-os with only two loose (but not calling stations) limpers in ahead of me to fold out the blinds. Then I fire at the flop no matter what ... you'd be surprised how many little pots you can take down like this.

BTW, I'm talking about $2/4 PP here. There are enough newish players and weakies that I think you can run these plays for profit every now and then.

(2) Natural variance of getting more than my fair share of good cards in those 1.2K hands.

I do expect my VP$IP to come down with more HHs ... but I see it levelling off at around 20 to 23% because of the factors I listed in (1) above. I just think that's my style, and that's not to say that I won't have to re-visit that thinking in different games/stakes etc. My mind is, and will always remain open.

|~|ypermegachi

03-04-2005, 09:46 PM

#9 (permalink)

4-of-a-Kind

Join Date: Aug 2004

Location: emo-kid

Posts: 3,580

for more detailed information;
http://forumserver.twoplustwo.com/sh...Number=1347418

basically, to obtain a 95% confidence interval of your true winrate you need over a million hands.

Trikflow77

03-04-2005, 09:54 PM

#10 (permalink)

Full House

Join Date: Jan 2005

Location: im so asian

Posts: 1,460

30% is still way to high, even with facy play syndrome

ArcticKnight

03-05-2005, 01:06 AM Post subject: Re: PT Stats, Comparing to Optimal? Just what is Optimal rea

#11 (permalink)

Flush

Join Date: Dec 2004

Location: golf course

Posts: 416

Quote:

Originally Posted by RiverMonkey

(2) If you are in a full ring game, and four people fold in front of the first VPIP'er, then don't you effectively have a 6-max game? This would call for for more aggression and a higher VP$IP, right?
?

Hi River Monkey

I think you still have a 10 person game.

You would have a higher VP$IP if you"always" got to play in the latter positions from which it was possible to see 4 folders in front of you; However, keep in mind that you will be in those first 4 positions 40% of the time yourself in a 10 handed game, and your VP$IP there will be low unless you "lower" your starting requirements. And I don't suspect you are going to do that.

The VP$IP optimum stats are lower in 10H (in large part, I would suggest) because of the frequent folding of early positions. So, it's already being factored in that you are making full orbits and spending a lot of time in poor positions, and putting more money in the pot in later positions.

BTW, great post.. lots of good discussion generated.

Gone golfing ..see ya in the Fall of 2006
PS. What did the snail on the turtle's back say?
Wheeeeeeeee........

Latest Poker News
KoRnholio	05-26-2012, 03:08 PM Australia Legalized Online Poker coming up in next 6 to 12 Months
According to an email sent out by Mark Bryan, a gaming analyst at Merrill Lynch, the Australian government plans to legalize online poker sometime in the next six to 12 months. This move will coincide ... more at Poker Blogs »

« Previous Thread | Next Thread »

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Posting Rules
You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is Off Trackbacks are On Pingbacks are On Refbacks are On Forum Rules

All times are GMT. The time now is 12:42 PM.

Play Texas Holdem Online, Online Texas Holdem Strategy, & Poker Forum

This is not a gambling website.

Poker Forum

Over 1,246,000 Posts!

Poker Forum

All Other (Limit) Poker

PT Stats, Comparing to Optimal? Just what is Optimal really?

Play for FREE and practice your game at...