Profitabilty of party's 1/2 games.

[gvibes]

[ QUOTE ]

... only 35-38% flop rate ...


[/ QUOTE ]

geez

[Greg J]

[ QUOTE ]
Yeah but an R squared value of .043 means that just 4.3% of the varience can be explained by the equation. This is pretty much insignificant.

[/ QUOTE ]
Wrong. This just means it does not explain much variance. This has nothing to do with the signifigance level of the RHS variable's coefficient. That being said, it is certainly possible that the variable is not significant, and while a low R sq can suggest this, you have to look at the significant level to see if there is a relationship.

Not trying to slam you here Brad -- this is actually a VERY common mistake in interpreting OLS results. In the research I do, public opinion, I will often get insanely low R sqs like this but have very significant variables.

[Greg J]

What is the significance level of the RHS variable (time)?

T score? P value?

[Isura]

I have been a profitable player over about 30k hands at 1/2 full ring. But I'm more aggressive and looser than you preflop. I'm around 18/10 in this game over 30k hands, and in my last 10k hands it's more like 20/12. I also play 6 max on the side so that has helped me immensely in the many short-handed pots you play at 1/2.

I open-raise a ton from MP3-Button, hands like A7o and K9o against tight blinds. You have to learn slightly different strategies than .5/1 to be a profitable player. You have to learn to understand the type of mistakes that you must induce in short-handed situations in unraised pots.

Against some players it is often correct to not bet the flop with marginal hands in position (e.g. UI overcards, or bottom pair HU), if you feel that it increases the chance that your opponent will make a bigger mistake by making a bad call on the double sized turn bet. For example, say you open-raise preflop in MP3 and get it heads up against a loose passive. If you bet your UI overcards, and get called on the flop and think that your opponent will fold on the turn with nothing, it then sometimes becomes correct to check the flop if you think it will induce your opponent to call a turn bet when a rag falls on the turn.

Another chip accumulating play against weak passive player is to check the turn in position HU when you spike something like high top pair and then bet the river (no matter what card comes). Say you have A9o and open-raise from late position, get called by the BB and flop is like Q84. You bet the flop and get called. Turn is an A. You should consider checking and betting the river if checked to. You will induce bluffs when you check the turn, and also villian is probably more likely to call you down with dominated or hopeless hands that he would have probably folded the turn to (since you were the aggressor 3 straight betting rounds if you bet, but checking shows weakness).

Another situation that you probably miss out on +EV is open-raising and playing middle to small pocket pairs in LP. Say you open-raise in the CO with 55, get called by both blinds and bet the flop of Q96 with 2 suited cards. SB folds and BB calls. Turn is K (no flush). You should usually bet and check behind the river UI. Betting is easy because you will be getting check-raised by better hands (and many times monsters, since players at 1/2 love to slowplay their sets, 2 pair until the turn against a single opponent so they can check-raise (I really bad play usually)). Hence it's really easy to let go of 55 when check-raised. Also, you'll get called by flush/straight draws and Ace high even. I usually try this against passive players you won't usually bluff the river (which is a concern when you could be against a flush or straight draw that misses).

This is just a flavor of the slight adjustments in strategy one has to make to beat a tighter and passive game. I understand that 1/2 is not the best game for everyone, but I have found it to be a great learning and bankroll building experience. It's also a nice limit to know how to beat when you want to drop down limits to clear a bonus quickly.

Edit: Added paragraphs to the original post. [img]/images/graemlins/tongue.gif[/img]

[CEE]

[ QUOTE ]
[ QUOTE ]
Yeah but an R squared value of .043 means that just 4.3% of the varience can be explained by the equation. This is pretty much insignificant.

[/ QUOTE ]
Wrong. This just means it does not explain much variance. This has nothing to do with the signifigance level of the RHS variable's coefficient. That being said, it is certainly possible that the variable is not significant, and while a low R sq can suggest this, you have to look at the significant level to see if there is a relationship.

Not trying to slam you here Brad -- this is actually a VERY common mistake in interpreting OLS results. In the research I do, public opinion, I will often get insanely low R sqs like this but have very significant variables.

[/ QUOTE ]

Thank you Greg! Actually, my intention was not to illustrate a "significant" trend... just one that I had observed over the last year in my games. Seemed like there were looser games a year ago, and in my experience at least, there were.

I believe that more players are becoming marginally better over time. I have a buddy who plays in my home game and also on Party .5/1... I let him borrow a couple books after he had a bad PP downswing. He was a fish... The books helped him tighten up pre-flop, but he didn't have tha patience to really learn the post-flop concepts well. He doesn't like trying to calculate pot odds, or playing aggressively, and goes with his gut all the time ("I could feel the flush coming"). I think he's typical of many players out there. They've read a little bit, maybe even god forbid(!) this board, but they're out there to have fun damnit.

Total digression, I know.

C

[bottomset]

[ QUOTE ]
[ QUOTE ]
But I did discover a leak, afterall: game selection

[/ QUOTE ]

imho, there's no more important factor than finding worse people to play against.

this has been discussed a zillion times before, but my personal opinion is that I'd much rather play 6 max at Party than full ring. besides being more profitable, it's much more fun, for me anyway.

[/ QUOTE ]

I agree on the 6max part, disagree a little on the finding bad player part .. I think at micros 95% of tables are profitable, and being able to play in a lot of various conditions is always a good thing

I think table/seat selection is best suited for higher stakes where the average player is much better than the average at 1/2 ... granted if you are at a terrible table, or terrible seat you should leave .. but most games are fine

also will people please quit suggestioning ppl to jump straight to 2/4 .. its a tougher game, though more profitable for solid players, if you aren't beating 1/2 likely you'll be in trouble at 2/4

[Greg J]

I have to know -- what is the significance level? Or the T score if you prefer.

[CEE]

[ QUOTE ]
I have to know -- what is the significance level? Or the T score if you prefer.

[/ QUOTE ]

...keeping in mind that this was intended to be a humorous graph, not a statistics lesson... I've done a little more work. It occured to me that one component of the increased tightness I was observing was myself: I've moved from being a 23% VPIP player to a 18% player, that's a pretty big move for someone who is a data point in over half of the hands in my PT db.

I also was aggregating to each whole day, then looking at the trend of the aggregates. Why do that, when you have the full detail at your disposal?

So I pulled all 9 or 10 handed $.5/1 games from my db, removed my action from consideration, and then adjusted up to a 10 person game. For example: If a hand was 9 handed, 4 people saw the flop, and I was one of them, I adjusted the "Saw Flop" players from 4 to 3.75 (calculation was ((4-1)/(9-1))*10).

Of course, my play also affects the play of others. I have also moved from a 5% pfr guy to a 9%+ pfr guy in the course of the past year. This would make my opponents more likely to fold (in theory). I didn't attempt to control for this.

Having made these adjustments, I re-graphed all of my hands (N=12,000), and it displayed a similar trend (although not visually very compelling). A link is below. I also attached a link to an SPSS output file for Greg to dig in to. Have at it man, I'm not an expert.

Players Seeing Flop Over Time Chart

SPSS Output File

C

[kenberman]

I'm assuming the spikes are weekend days?

[Greg J]

Thanks much. The results are pretty interesting! Thanks a lot for making the effort.