Graph [Archive] - Two Plus Two Older Archives

GrunchCan

04-14-2005, 02:14 PM

Interesting.

I think I understand the phenomenon that's at work here.

I would normally say to someone who has a graph like yours that you are in for a rude awakening when your variance kicks in. But with your approach, you seem to snap off variance before it gathers any steam. I can't quite explain why you never have multiple subsequent losing sessions, but for now I'll assume there is a reason I'm just not seeing.

It seems like your habit of picking up from a table where your'e stuck 10BB is actually another form of table selection. The opponents at .5/1 are generally pretty bad, and most tables there are good ones. But it is possible to sit at a table where your relative position is so bad that its nearly unbeatable for you. Usually your'e being outplayed by a small number of opponents, even though the rest of the table is tourists. If the tough opponents are in the right spots, they can beat you.

So you lost 10 BB and pick up, taking yourself to a new table where your position is better.

It's possible that your approach is close to optimal for party .5/1, since the vast majority of tables are good ones. I don't think it will work at higher levels, however. It certianly won't work live.

So, if you plan on remaining at .5/1 forever, you'll probably do fine. If you ever try 5/10, you'll need another approach.

Good luck.

grjr

04-14-2005, 02:21 PM

[ QUOTE ]
Interesting.

I think I understand the phenomenon that's at work here.

I would normally say to someone who has a graph like yours that you are in for a rude awakening when your variance kicks in. But with your approach, you seem to snap off variance before it gathers any steam. I can't quite explain why you never have multiple subsequent losing sessions, but for now I'll assume there is a reason I'm just not seeing.

It seems like your habit of picking up from a table where your'e stuck 10BB is actually another form of table selection. The opponents at .5/1 are generally pretty bad, and most tables there are good ones. But it is possible to sit at a table where your relative position is so bad that its nearly unbeatable for you. Usually your'e being outplayed by a small number of opponents, even though the rest of the table is tourists. If the tough opponents are in the right spots, they can beat you.

So you lost 10 BB and pick up, taking yourself to a new table where your position is better.

It's possible that your approach is close to optimal for party .5/1, since the vast majority of tables are good ones. I don't think it will work at higher levels, however. It certianly won't work live.

So, if you plan on remaining at .5/1 forever, you'll probably do fine. If you ever try 5/10, you'll need another approach.

Good luck.

[/ QUOTE ]

My God, Grunch, I never thought I'd hear anything like this from someone on this board. I actually got goosebumps while I was reading. /images/graemlins/laugh.gif

Thanks for your comments and I agree that this would not work as well at higher limits where I imagine the skill/luck factor is much higher. When I was talking before about poker being a gambling game I was mainly talking about the micro limits where many compare it to "bingo".

NAU_Player

04-14-2005, 02:52 PM

This is the graph of the last 1.5 months of my life /images/graemlins/wink.gif

http://img202.echo.cx/img202/5742/statgraph6pb.png

that last downswing really sucked, so i'm taking a break. (I also have papers to do and finals to study for)

topspin

04-14-2005, 03:01 PM

[ QUOTE ]
I would normally say to someone who has a graph like yours that you are in for a rude awakening when your variance kicks in. But with your approach, you seem to snap off variance before it gathers any steam.

[/ QUOTE ]

You can't "snap off variance" by switching tables, and I'm surprised to see you make the claim. If he's due to hit a huge bunch of suckouts, it's going to happen regardless of if he's table-hopping or not.

[ QUOTE ]
It seems like your habit of picking up from a table where your'e stuck 10BB is actually another form of table selection.

[/ QUOTE ]

I thought the same when I read about his strategy. As long as he's finding equally loose tables to move to, table-hopping can't hurt his position, and it might help him if the previous table had many people (e.g. manaics etc) that he plays poorly against.

If this is true, his win rate should be better than it otherwise would be. It still doesn't explain the total lack of cumulative losing sessions in his graph. My thought here is that it's the monkeys-typing Shakespeare effect: with so many people coming thru here these days, someone is bound to have a nice smooth-looking win rate over 20k hands. I find this more likely than the alternate explanation, which is that this strategy can sustain the 6BB/100 win rate his graph shows with no downswings in the long term.

grjr

04-14-2005, 03:11 PM

I'll be sticking around here for a while so we'll see how the monkeys are doing in another 20K hands or so. /images/graemlins/grin.gifNot sure how long that will take since I only play on Party and skins during the reloads. I make more money on another site the rest of the time.

GrunchCan

04-14-2005, 04:55 PM

[ QUOTE ]
You can't "snap off variance" by switching tables, and I'm surprised to see you make the claim.

[/ QUOTE ]

I know, but think about it. He guarantees that no single session can be a loser for more than 10BB. The only think I dont understand is how he hasnt had multiple cumulative losing sessions.

I wonder what Sklansky would say about grjr...

David? Are you listening?

grjr

04-14-2005, 05:36 PM

[ QUOTE ]
[ QUOTE ]
You can't "snap off variance" by switching tables, and I'm surprised to see you make the claim.

[/ QUOTE ]

I know, but think about it. He guarantees that no single session can be a loser for more than 10BB. The only think I dont understand is how he hasnt had multiple cumulative losing sessions.

I wonder what Sklansky would say about grjr...

David? Are you listening?

[/ QUOTE ]

I always seem to find a table or two where I make real good money in a short time.

Here are all my sessions from 4/8/05 starting at 1 am

2.63....37min
(2.50)..36min
(14.25).16min (had 2 real good draws that didn't hit)
(1.75)..6 min
I quit here because I didn't feel I was playing well
Started again at 11:53
(9.50)..9min
(5.75)..33min
(8.25)..54min
7.00....60min
(15.00).14min (got sucked out on 3 straight times)
(11.00).60min
It's now about 12:30 and I'm close to giving up
I'm not used to losing right off the bat like this
11.50...118min
24.75...97min
(3.00)..13min
(7.59)..4min
(5.00)..5min
9.64....48min
(4.50)..1min (i was getting a short fuse with the losers
(9.50)..4min (although I was winning on the other 2 tables)
10.25...23min
Stopped here and restarted at 20:30
38.00...43min
(10.50).23min
(10.12).14min
(6.00)..88min
23.75...183min
(9.50)..8min
(4.50)..3min
47.25...141min
(7.50)..1min (lol)
45.00...59min
13.25...40min
32.50...88min
4.01....63min
6.75....12min
Stopped for the day
4/8 is the next to last day on the chart. I played 6 more
tables on the 9th to finish out my bonus.

I feel I could have had a pretty big downswing when I started out this day if I had stayed at all those losing tables. Maybe things would have turned around if I had stayed there but I really think not.

At the tables where I ended up winning some good money I was usually ahead 15 to 20BB within 15-20 minutes. These are the ones I was talking about in my other posts.

I know this helter-skelter stuff isn't for most people but it seems to keep me from tilting so that's a plus right there.

Aaron W.

04-14-2005, 06:22 PM

[ QUOTE ]
[ QUOTE ]
You can't "snap off variance" by switching tables, and I'm surprised to see you make the claim.

[/ QUOTE ]

I know, but think about it. He guarantees that no single session can be a loser for more than 10BB. The only think I dont understand is how he hasnt had multiple cumulative losing sessions.

I wonder what Sklansky would say about grjr...

David? Are you listening?

[/ QUOTE ]

It's impossible to "predict" a downswing, unless you set yourself up for one by playing at a table where you have too many factors working against you.

I suspect that there are a few effects going on. The first is that grjr doesn't suck (said positively: plays well). A sustained win streak of that length and magnitude generally does not come on luck alone.

The nonexistence of a major losing streak is simply a matter of variance. If you believe that you can have a run of 10000 hands where you play well, but come out roughly even, then a run of 20000 hands without a major downswing is also not unrealistic. It's even more likely if grjr plays a low-variance slightly passive strategy.

You'll notice the break-even stretch from 4000-9000, which is about 25% of the chart. Turn that into a downswing and nobody says anything unusual about the overall chart.

The table selection thing may be in play, or maybe not. It's not hard to imagine that some of us have downswings because we refuse to leave a table where we don't have the best of it. grjr's method will get him off such a table (as well as a number of tables where he should have the best of it). His 'anti-tilt' reasoning is definitely worth consideration.

[Another possible explanation is that the graph is rigged by removing a few select losing sessions, but I don't get that impression from the graph, and I don't want to accuse someone of cheating without evidence.]

topspin

04-14-2005, 06:33 PM

grjr

04-14-2005, 06:40 PM

I would have no reason to falsify my graph here. I'm not trying to prove anything to anyone. I like to try and be helpful if I can.

That dip at the 7.6K level was the result of my trying to stick with some "good" tables after reading that's what I should do here. It didn't work out and I haven't tried it again.

Granted that was only one day but I feel if my bailing method works for me then I don't have a need to stick with something else I'm not comfortable with.

I REALLY hate to lose money. Once I've won it it becomes mine and I don't want to give it back. This , of course, can lead to a big tilting problem which you guys seem to agree is somewhat solved by leaving the tilting situation.

I still believe in "hot" and "cold" tables also although I don't expect to garner any support for that notion here. /images/graemlins/wink.gif

grjr

04-14-2005, 06:46 PM

[ QUOTE ]
[ QUOTE ]
I'll be sticking around here for a while so we'll see how the monkeys are doing in another 20K hands or so. /images/graemlins/grin.gifNot sure how long that will take since I only play on Party and skins during the reloads. I make more money on another site the rest of the time.

[/ QUOTE ]

Hey grjr,

Just thought I'd add in case it wasn't clear from my post that I'm not knocking your positive results - it seems clear that you're a winning player, and your strategy has the merits that Aaron mentions above which I think could be doing good things for your win rate too. My only contention is with the claim that using the strategy will result in the lack of downswings you're showing in the graph.

Out of curiousity, what's your SD according to PT?

[/ QUOTE ]

WTSD%..34.48
W%SD...53.52

I think these are in the accepted range, yes?

waynethetrain

04-14-2005, 06:47 PM

I would show you my graph but it would save time if you just looked at everyone else's graph upside down.

Greg J

04-14-2005, 06:48 PM

He meant yr standard deviation.

grjr

04-14-2005, 06:51 PM

[ QUOTE ]
He meant yr standard deviation.

[/ QUOTE ]

Where do you find that?

Edit: nevermind, I found it

SD/Hour $17.7154
SD/100 hands $22.2798

Can you tell me what that means?

Greg J

04-14-2005, 07:04 PM

A standard deviation is the root of the sum of the squared variation from the mean divided by the number of observations (I think I said that right).

In English: it's a way to tell how spread out yr results at the table are. The higher yr SD the more volitility there is in yr results.

I assume this is for .5/1?

grjr

04-14-2005, 07:23 PM

[ QUOTE ]
A standard deviation is the root of the sum of the squared variation from the mean divided by the number of observations (I think I said that right).

In English: it's a way to tell how spread out yr results at the table are. The higher yr SD the more volitility there is in yr results.

I assume this is for .5/1?

[/ QUOTE ]

Yes, .50/1

So how do you interpret those numbers then?

shadow29

04-14-2005, 07:38 PM

You guys will love this:

http://img.photobucket.com/albums/v332/shadow29/graph.jpg

Greg J

04-14-2005, 07:47 PM

[ QUOTE ]
SD/Hour $17.7154
SD/100 hands $22.2798

[/ QUOTE ]
So yr bb/100 is 22.3. How many hands are we talking here?

Mine is 13.3 bb/100 over 18k hands at 1/2 (I recently lost my PT database and hasd to start over). This means that over the course of a hundred hands it's not unusual for me to up or down 13 bbs, plus the mean (my winrate). About 95% of the time I will be withing 2 standard deviations of my winrate, for better or worse. Poker is a volitile game, but my numbers are lower than most.

Yr SD indicates you have a highly volatile game. Eyeballing yr graph and comparing this to yr graph, this frankly confuses the living shi[/i]t out of me. But it goes along with my impression of grjr: you, my friend, are a complete anomoly. (You are a good poster though, despite what I think are yr silly superstitions.)

Greg J

04-14-2005, 07:50 PM

You bitch! How dare you hack into my computer and steal my PT stats!

Duerig

04-14-2005, 08:02 PM

I remember reading grgrgsgh's (or however you spell it) response and the subsequent argument. Here are my thoughts.

- I think it's fairly likely that a decent amount of collusion goes on at online poker sites. Even with this collusion, it is possible to beat the game (sometimes it makes it easier). However, it's possible that the very soft table you are playing at has a couple of colluders, reducing your EV for every hand that you play.

- Tilt. If it helps you play better, then go for it. Switching tables is probably better than trying to win back a huge pot that some chump sucked out on you for, and playing like a moron because you're trying to exact revenge.

- Table image. While many people discount this part of the equation, I think it is important. If you first sit down at a table and get dealt AA. You can bet to the river and someone will call you down, trying to catch you on a bluff. If however, you've been playing at the table for a while, seeing very few hands, your opponents are going to be (slightly) more prone to fold and the pots are going to be smaller.

- RELATIVE position to the poor players. Grunch already mentioned this, but I thought I'd mention it again because it's important.

- You're more tired / drunk / hungover than the fish at the table and you don't know it, or the players with a VPIP of 30 are fantastic postflop. The point is that there may be some other reason that you aren't beating the table THAT YOU HAVEN'T EVEN CONSIDERED.

Sticking to this -10BB rule is be a bad idea at times, if you get sucked out on and leave a good table, but it acts as a sort of very crude first order filter against possible negative aspects of any table you might be playing at.

I think the reason some posters took issue with you previously was because you didn't give any logical reasoning. Many people (including myself) consider a downswing a CARD BASED phenomenon, that is, the reason you lose money is due to cold cards / suckouts. Note that changing tables has absolutely no effect on cards whatsoever. I think your experience has demonstrated (to some extent) the value of other factors in the game which may result in downswings.

My point is, I agree that switching tables can be a good idea if you hit a downswing. I do not agree that your "bad run of cards" will ever be changed by this strategy. The reasoning behind your decisions is just as important, while learning, as the decision itself. It would be like someone saying that they liked to raise preflop with AA because AA are their initials.

This is by far the longest post I have ever written. I hope it made some sense.

shadow29

04-14-2005, 08:05 PM

[ QUOTE ]
- RELATIVE position to the poor players.

[/ QUOTE ]

I have no clue what y'all are talking about, but this and table selection make up about 1.5bb/100 of my game.

No joke.

I would be hugely in the red at 2/4 if I didn't table select and seat select.

The other stuff is just fluff.

Saint_D

04-14-2005, 08:12 PM

[ QUOTE ]
[ QUOTE ]
I would normally say to someone who has a graph like yours that you are in for a rude awakening when your variance kicks in. But with your approach, you seem to snap off variance before it gathers any steam.

[/ QUOTE ]

You can't "snap off variance" by switching tables, and I'm surprised to see you make the claim. If he's due to hit a huge bunch of suckouts, it's going to happen regardless of if he's table-hopping or not.

[ QUOTE ]
It seems like your habit of picking up from a table where your'e stuck 10BB is actually another form of table selection.

[/ QUOTE ]

I thought the same when I read about his strategy. As long as he's finding equally loose tables to move to, table-hopping can't hurt his position, and it might help him if the previous table had many people (e.g. manaics etc) that he plays poorly against.

If this is true, his win rate should be better than it otherwise would be. It still doesn't explain the total lack of cumulative losing sessions in his graph. My thought here is that it's the monkeys-typing Shakespeare effect: with so many people coming thru here these days, someone is bound to have a nice smooth-looking win rate over 20k hands. I find this more likely than the alternate explanation, which is that this strategy can sustain the 6BB/100 win rate his graph shows with no downswings in the long term.

[/ QUOTE ]

I remember reading about a "Ratchet" effect in gaming. Science News Week of Jan. 15, 2000; Vol. 157, No. 3 , p. 47.

[ QUOTE ]
Researchers have demonstrated that two games of chance, each guaranteed to give a player a predominance of losses in the long term, can add up to a winning outcome if the player alternates randomly between the two games.

This striking new result in game theory is now called Parrondo's paradox, after its discoverer, Juan M.R. Parrondo, a physicist at the Universidad Complutense de Madrid in Spain.

[/ QUOTE ]

It goes on to say that it can't be any old two games. The games have to fit some mathmatical model that I am too dumb to understand. If they have found 2 casino games that fit the bill, they ain't tellin'.

If you subscribe you can see the article here. (http://www.sciencenews.org/articles/20000115/note16.asp)

Perhaps his method has used this or a similar effect by accident. Perhaps it's just an anomoly.

-D

grjr

04-14-2005, 08:17 PM

[ QUOTE ]
[ QUOTE ]
SD/Hour $17.7154
SD/100 hands $22.2798

[/ QUOTE ]
So yr bb/100 is 22.3. How many hands are we talking here?

Mine is 13.3 bb/100 over 18k hands at 1/2 (I recently lost my PT database and hasd to start over). This means that over the course of a hundred hands it's not unusual for me to up or down 13 bbs, plus the mean (my winrate). About 95% of the time I will be withing 2 standard deviations of my winrate, for better or worse. Poker is a volitile game, but my numbers are lower than most.

Yr SD indicates you have a highly volatile game. Eyeballing yr graph and comparing this to yr graph, this frankly confuses the living shi[/i]t out of me. But it goes along with my impression of grjr: you, my friend, are a complete anomoly. (You are a good poster though, despite what I think are yr silly superstitions.)

[/ QUOTE ]

That's for 17,000 hands. I imagine mine is high because I play a lot of hands. Remember my VP$IP is 27.54%. I think I straighten the graph because I'm constantly finding winning tables to more than balance out the losing ones. My best winning table was $103 net in 108 minutes. Was sure sad to see that one break up. You can't win that much in a short period without playing a lot of marginal hands.

I've done the same thing on a couple of the 1/2 tables at Crypto. $155, $140, and $74. Oddly enough, those were all at the same table (Mobile) over 3 different days during the same week. That's kinda weird in of itself.

Can't see the word "anomoly" without thinking of Bruce Willis. heh

droolie

04-14-2005, 08:19 PM

As long as we're busting out our "growth charts" for all to see...

http://i4.photobucket.com/albums/y114/Thedroolies/graph4.jpg

Greg J

04-14-2005, 08:20 PM

Okay, if this includes 1/2 and .5/1 tables then you need to give yr SD in bb/100, not $/100.

Oh, and 27 vpip is insane.

Aaron W.

04-14-2005, 08:24 PM

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I would normally say to someone who has a graph like yours that you are in for a rude awakening when your variance kicks in. But with your approach, you seem to snap off variance before it gathers any steam.

[/ QUOTE ]

You can't "snap off variance" by switching tables, and I'm surprised to see you make the claim. If he's due to hit a huge bunch of suckouts, it's going to happen regardless of if he's table-hopping or not.

[ QUOTE ]
It seems like your habit of picking up from a table where your'e stuck 10BB is actually another form of table selection.

[/ QUOTE ]

I thought the same when I read about his strategy. As long as he's finding equally loose tables to move to, table-hopping can't hurt his position, and it might help him if the previous table had many people (e.g. manaics etc) that he plays poorly against.

If this is true, his win rate should be better than it otherwise would be. It still doesn't explain the total lack of cumulative losing sessions in his graph. My thought here is that it's the monkeys-typing Shakespeare effect: with so many people coming thru here these days, someone is bound to have a nice smooth-looking win rate over 20k hands. I find this more likely than the alternate explanation, which is that this strategy can sustain the 6BB/100 win rate his graph shows with no downswings in the long term.

[/ QUOTE ]

I remember reading about a "Ratchet" effect in gaming. Science News Week of Jan. 15, 2000; Vol. 157, No. 3 , p. 47.

[ QUOTE ]
Researchers have demonstrated that two games of chance, each guaranteed to give a player a predominance of losses in the long term, can add up to a winning outcome if the player alternates randomly between the two games.

This striking new result in game theory is now called Parrondo's paradox, after its discoverer, Juan M.R. Parrondo, a physicist at the Universidad Complutense de Madrid in Spain.

[/ QUOTE ]

It goes on to say that it can't be any old two games. The games have to fit some mathmatical model that I am too dumb to understand. If they have found 2 casino games that fit the bill, they ain't tellin'.

If you subscribe you can see the article here. (http://www.sciencenews.org/articles/20000115/note16.asp)

Perhaps his method has used this or a similar effect by accident. Perhaps it's just an anomoly.

-D

[/ QUOTE ]

Here is a link (http://seneca.fis.ucm.es/parr/GAMES/inbrief.html) to a pretty good description of the game where this happens. After looking over it, I don't see how it applies to this case. In that case, the game changes according to the player's capital (in Game B). It's entirely possible that meta-game considerations could lead to such a game situation, but I have my doubts.

The way opponents play against a new player is (theoretically) different from how opponents play against a someone who has been at the table for 100 hands. For example, I am more willing to call off a strange river bet against an unknown than I am to call it off against a player who is known not to bluff. It could be that the effect of being new to the table will change the overall strategy against Hero.

As I said, there's potential there, but I have my doubts. I am more inclined to chalk this up to variance rather than ratcheting effect that has yet to be demonstrated in a complex system.

Aaron W.

04-14-2005, 08:27 PM

[ QUOTE ]
[ QUOTE ]
SD/Hour $17.7154
SD/100 hands $22.2798

[/ QUOTE ]
So yr bb/100 is 22.3. How many hands are we talking here?

Mine is 13.3 bb/100 over 18k hands at 1/2 (I recently lost my PT database and hasd to start over). This means that over the course of a hundred hands it's not unusual for me to up or down 13 bbs, plus the mean (my winrate). About 95% of the time I will be withing 2 standard deviations of my winrate, for better or worse. Poker is a volitile game, but my numbers are lower than most.

Yr SD indicates you have a highly volatile game. Eyeballing yr graph and comparing this to yr graph, this frankly confuses the living shi[/i]t out of me. But it goes along with my impression of grjr: you, my friend, are a complete anomoly. (You are a good poster though, despite what I think are yr silly superstitions.)

[/ QUOTE ]

A larger VPIP would account for some increase in SD, as now more of the hands have a non-zero return (a 20 VPIP would have 80 zeros whereas a 30 VPIP only has 70). I'm not sure if this is enough to account for the entire difference. There are too many variables.

grjr

04-14-2005, 08:37 PM

[ QUOTE ]
I remember reading grgrgsgh's (or however you spell it) response and the subsequent argument. Here are my thoughts.

- I think it's fairly likely that a decent amount of collusion goes on at online poker sites. Even with this collusion, it is possible to beat the game (sometimes it makes it easier). However, it's possible that the very soft table you are playing at has a couple of colluders, reducing your EV for every hand that you play.

- Tilt. If it helps you play better, then go for it. Switching tables is probably better than trying to win back a huge pot that some chump sucked out on you for, and playing like a moron because you're trying to exact revenge.

- Table image. While many people discount this part of the equation, I think it is important. If you first sit down at a table and get dealt AA. You can bet to the river and someone will call you down, trying to catch you on a bluff. If however, you've been playing at the table for a while, seeing very few hands, your opponents are going to be (slightly) more prone to fold and the pots are going to be smaller.

- RELATIVE position to the poor players. Grunch already mentioned this, but I thought I'd mention it again because it's important.

- You're more tired / drunk / hungover than the fish at the table and you don't know it, or the players with a VPIP of 30 are fantastic postflop. The point is that there may be some other reason that you aren't beating the table THAT YOU HAVEN'T EVEN CONSIDERED.

Sticking to this -10BB rule is be a bad idea at times, if you get sucked out on and leave a good table, but it acts as a sort of very crude first order filter against possible negative aspects of any table you might be playing at.

I think the reason some posters took issue with you previously was because you didn't give any logical reasoning. Many people (including myself) consider a downswing a CARD BASED phenomenon, that is, the reason you lose money is due to cold cards / suckouts. Note that changing tables has absolutely no effect on cards whatsoever. I think your experience has demonstrated (to some extent) the value of other factors in the game which may result in downswings.

My point is, I agree that switching tables can be a good idea if you hit a downswing. I do not agree that your "bad run of cards" will ever be changed by this strategy. The reasoning behind your decisions is just as important, while learning, as the decision itself. It would be like someone saying that they liked to raise preflop with AA because AA are their initials.

This is by far the longest post I have ever written. I hope it made some sense.

[/ QUOTE ]

Yes, it made a lot of sense, thanks! I hadn't considered the table image angle before. That made a lot of sense too(although you won't see me sitting at a table playing very few hands. Heh).

The thing I don't understand is everyone acknowledges that there are "bad runs of cards" but why should that bad run follow me around to different tables? Every time I sit down at a new table it's like starting all over again. Sooner or later I find a table where I win 2 or 3 big pots in the first orbit. I guess the bad run didn't see me there. /images/graemlins/wink.gif

Gamblers are a superstitious lot by nature (and there aren't many gamblers that post here). Just because something might be considered a "superstition" doesn't mean there isn't some underlying foundation for it.

One other thing I wanted to mention again. I've seen people say they have had a bad run of cards for 20K or 30K hands. I think they're fooling themselves if they believe that.

grjr

04-14-2005, 08:42 PM

[ QUOTE ]
Okay, if this includes 1/2 and .5/1 tables then you need to give yr SD in bb/100, not $/100.

Oh, and 27 vpip is insane.

[/ QUOTE ]

No, all the figures I give in here are only for the .50/1 limit at Party.

I've wondered what the optimal VP$IP is for Party .50/1 if you assumed you could have perfect play after the flop. I really think it's somewhere between 25-26% but I have no way of proving that.

grjr

04-14-2005, 08:52 PM

Very nice, Droolie!

Greg J

04-14-2005, 08:53 PM

Okay here we go again:

I think I might make some headway this time -- or I hope so.

[ QUOTE ]
The thing I don't understand is everyone acknowledges that there are "bad runs of cards" but why should that bad run follow me around to different tables? Every time I sit down at a new table it's like starting all over again.

[/ QUOTE ]
This is flat wrong. You chances of being dealt AA are just as good from one table to the other. It doesnt matter if you were just dealt AA at 3 other tables. It doesnt matter if someone hit a runner runner suckout where they were 800:1 underdogs. The chances of it happening again the next hand are the same. The cards don't change depending on the table. Other factors might change, but the distribution of cards will be the same.

[ QUOTE ]
Gamblers are a superstitious lot by nature (and there aren't many gamblers that post here).

[/ QUOTE ]
Dude, everyone here is a gambler. I am. I play poker. Poker is gambling. Bad gamblers tend to be overly superstitious. Good ones might be also, but they know cognitively that yr distribution of winning does not depend on what yr cards were the previous hand, or if you are running "hot", at least from a strictly statistical sense. Good gamblers know there is math involved, and gambling profiably means taking advantage of probability. You seriously need to read some of Malmuth's stuff on probability and gambling theory.

[ QUOTE ]
Just because something might be considered a "superstition" doesn't mean there isn't some underlying foundation for it.

[/ QUOTE ]
No, but it means that the "underlying foundation" is worthless. Sorry to be blunt but it is. My theory of sacrificing chickens in the graveyard is just as logical as yr theory of changing tables is.

grjr

04-14-2005, 09:21 PM

[ QUOTE ]
Okay here we go again:

I think I might make some headway this time -- or I hope so.

[ QUOTE ]
The thing I don't understand is everyone acknowledges that there are "bad runs of cards" but why should that bad run follow me around to different tables? Every time I sit down at a new table it's like starting all over again.

[/ QUOTE ]
This is flat wrong. You chances of being dealt AA are just as good from one table to the other. It doesnt matter if you were just dealt AA at 3 other tables. It doesnt matter if someone hit a runner runner suckout where they were 800:1 underdogs. The chances of it happening again the next hand are the same. The cards don't change depending on the table. Other factors might change, but the distribution of cards will be the same.

[ QUOTE ]
Gamblers are a superstitious lot by nature (and there aren't many gamblers that post here).

[/ QUOTE ]
Dude, everyone here is a gambler. I am. I play poker. Poker is gambling. Bad gamblers tend to be overly superstitious. Good ones might be also, but they know cognitively that yr distribution of winning does not depend on what yr cards were the previous hand, or if you are running "hot", at least from a strictly statistical sense. Good gamblers know there is math involved, and gambling profiably means taking advantage of probability. You seriously need to read some of Malmuth's stuff on probability and gambling theory.

[ QUOTE ]
Just because something might be considered a "superstition" doesn't mean there isn't some underlying foundation for it.

[/ QUOTE ]
No, but it means that the "underlying foundation" is worthless. Sorry to be blunt but it is. My theory of sacrificing chickens in the graveyard is just as logical as yr theory of changing tables is.

[/ QUOTE ]

Greg, I understand what you're saying. I've understood it every time I've read it in all the poker books too.

It's just that the "variance" term in poker mathematics doesn't sufficiently explain the "hot" and "cold" decks for me. I'm sure it does for you and that's great but I'll always be a little skeptical in accepting that at face value.

Thanks for all your time put into our discussions. (where's the thumbs-up smiley when you need it?)

Duerig

04-14-2005, 09:31 PM

[ QUOTE ]

Yes, it made a lot of sense, thanks! I hadn't considered the table image angle before. That made a lot of sense too(although you won't see me sitting at a table playing very few hands. Heh).

The thing I don't understand is everyone acknowledges that there are "bad runs of cards" but why should that bad run follow me around to different tables? Every time I sit down at a new table it's like starting all over again. Sooner or later I find a table where I win 2 or 3 big pots in the first orbit. I guess the bad run didn't see me there. /images/graemlins/wink.gif

Gamblers are a superstitious lot by nature (and there aren't many gamblers that post here). Just because something might be considered a "superstition" doesn't mean there isn't some underlying foundation for it.

One other thing I wanted to mention again. I've seen people say they have had a bad run of cards for 20K or 30K hands. I think they're fooling themselves if they believe that.

[/ QUOTE ]

I didn't want to get sucked into this but I just wanted to say that this is exactly the problem I was talking about. Switching tables if you lose 10 BB isn't in itself a bad strategy, there are a few reasons why it might even be a good one. But rationalizing it by saying that it cuts short your run of cold cards is just wrong. I hate to be blunt, but this argument is just absurd.

Imagine you are playing at a B&M casino and are getting poor cards. What if you switched tables but the dealer passed the deck of cards you were just playing with to the table you have just moved to? Will you stop getting bad cards now because you have moved 15 feet? Will you continue to get bad cards because the deck 'remembers' you? What if you move to a new table and get a whole new deck, except for the aces from the old deck? Are those enough so that the deck 'remembers' you? What if you took all the hearts from the old deck, would that make a difference? Maybe you could extend this theory to roulette: let the wheel spin 10 times, if it lands on red 6 times, it must be hot! Put some money down on red! Clearly (I hope) this sounds ridiculous. Cards (and roulette wheels) are inanimate objects, they have no memory, every 'shuffle up and deal' to them is the same, from the last hand in a huge tournament to a home game you play for nickels.

On the other hand, what if you switched tables, but all the players from your old table moved with you. You probably wouldn't expect the game to change very much, would you? If only half the players from your old game came into your new game, you might expect it to change to some degree. If only one player came with you to your new table, you would probably expect a game that was totally new. And you would be right.

My point is that by switching tables you are getting a new set of players AND a new set of cards. Clearly you have recognized that switching tables seems to be benificial to you, and this is great, but you are a bit off as far as the reason why. A new set of cards means nothing! A new set of players means everything. Don't confuse the two.

I hope this helped.

grjr

04-14-2005, 09:34 PM

[ QUOTE ]
Here is a link to a pretty good description of the game where this happens. After looking over it, I don't see how it applies to this case. In that case, the game changes according to the player's capital (in Game B). It's entirely possible that meta-game considerations could lead to such a game situation, but I have my doubts.

[/ QUOTE ]

That was very interesting. Thank you, Aaron. And Saint D also.

Greg J

04-14-2005, 09:40 PM

[ QUOTE ]
Greg, I understand what you're saying. I've understood it every time I've read it in all the poker books too.

[/ QUOTE ]

Sweet baby Jesus!

Okay that's it. I officially give up.

If anyone wants to order my official guide to ending downswings through animal sacrifice, pm me. Videos are only $200 plus shipping and handling.

handsome

04-14-2005, 09:45 PM

[ QUOTE ]

It's just that the "variance" term in poker mathematics doesn't sufficiently explain the "hot" and "cold" decks for me.

[/ QUOTE ]

Whoa... do you realize what you are saying? Variance explains "hot" cards, "cold" cards and everything in between. It exists whether you like it or not. It watches you while you sleep, eat and jerk off. If you've been getting "lucky," it will catch up to you and bite your penis.

grjr

04-14-2005, 09:47 PM

[ QUOTE ]
[ QUOTE ]
Greg, I understand what you're saying. I've understood it every time I've read it in all the poker books too.

[/ QUOTE ]

Sweet baby Jesus!

Okay that's it. I officially give up.

If anyone wants to order my official guide to ending downswings through animal sacrifice, pm me. Videos are only $200 plus shipping and handling.

[/ QUOTE ]

Now you're being silly. I've tried animal sacrifice and it didn't work. All it did was piss off the wife (which was very -EV)

grjr

04-14-2005, 09:53 PM

[ QUOTE ]
Whoa... do you realize what you are saying? Variance explains "hot" cards, "cold" cards and everything in between.

[/ QUOTE ]

FOR ME, variance is just a term that means "We can't really explain why the hell this happens." Don't roll your eyes Greg, they might get stuck like that. /images/graemlins/laugh.gif

Saint_D

04-14-2005, 09:55 PM

[ QUOTE ]
That was very interesting. Thank you, Aaron. And Saint D also.

[/ QUOTE ]

Your welcome. I want to point out I am not saying this is a situation like the game I link to. It's just interesting to think about.

Greg J

04-14-2005, 10:00 PM

Don't even try. Seriously, i like this guy, but it's like banging yr head into a wall.

topspin

04-14-2005, 10:25 PM

[ QUOTE ]
[ QUOTE ]
Just because something might be considered a "superstition" doesn't mean there isn't some underlying foundation for it.

[/ QUOTE ]
No, but it means that the "underlying foundation" is worthless. Sorry to be blunt but it is. My theory of sacrificing chickens in the graveyard is just as logical as yr theory of changing tables is.

[/ QUOTE ]

I also think the "getting better cards by changing tables" argument is specious. At least online, since all the cards dealt are truly random, changing tables isn't going to help you. You could make the argument that at B&M cards might "clump" (since usually dealers don't shuffle sufficiently to truly randomize) and possibly deal out a bad run. I don't know enough about this to speculate, although I have seen some BJ books make the claim.

But, and I think a lot of the replies (aside from Grunch's) have been overlooking this, I think there are other reasons grjr's strategy could have a positive effect - among other things, it forces him to leave a table where he may be getting schooled and not realize it, or it may help him avoid tilt and play better psychologically to leave a losing table (the placebo effect). Both of these would have positive effects on his win rate.

He may also leave a profitable table due to a bad run of cards, but as long as he replaces it with an equally loose table (an easy task at 0.5/1) he is no worse off. Therefore, assuming he can always hop to an equally loose table, his strategy can't be worse than staying at the same table, and may be better.

topspin

04-14-2005, 10:40 PM

[ QUOTE ]
SD/Hour $17.7154
SD/100 hands $22.2798

Can you tell me what that means?

[/ QUOTE ]

Is this mostly 0.5/1? If so, I'll assume your true win rate was the 6BB/100 you show in the graph (a superb rate, but which is still within what some posters have claimed is at the upper reaches of what is sustainable at 0.5/1) with an SD of 22.3BB/100.

This means that over 1k hands, your expected win rate is 10*6 = 60BB, with a standard deviation of sqrt(10)*22.33 = 70.5BB. Over any 1k block of hands, your probability of being breakeven is Q(60/70.5) = 20%. Your probability of experiencing a 25BB downswing is Q(85/70.5) = 13%. Go for a 50BB downswing, and it's Q(110/70.5) = 6%.

See why many people on the boards have been a bit skeptical? Over the huge sample you've shown, there should be quite a few small dips, and one or two larger dips, even assuming a superb true win rate. I don't think I can even pick out a single 1k block of hands in your graph that was breakeven, much less a 25BB dip. That's what led to my "statistical anomoly" comment. It's not a comment on your skill, just the fact that ultimately your cards have a lot to do with how you win, and the odds have to catch up sooner or later.

grjr

04-15-2005, 12:04 AM

[ QUOTE ]
[ QUOTE ]
SD/Hour $17.7154
SD/100 hands $22.2798

Can you tell me what that means?

[/ QUOTE ]

Is this mostly 0.5/1? If so, I'll assume your true win rate was the 6BB/100 you show in the graph (a superb rate, but which is still within what some posters have claimed is at the upper reaches of what is sustainable at 0.5/1) with an SD of 22.3BB/100.

This means that over 1k hands, your expected win rate is 10*6 = 60BB, with a standard deviation of sqrt(10)*22.33 = 70.5BB. Over any 1k block of hands, your probability of being breakeven is Q(60/70.5) = 20%. Your probability of experiencing a 25BB downswing is Q(85/70.5) = 13%. Go for a 50BB downswing, and it's Q(110/70.5) = 6%.

See why many people on the boards have been a bit skeptical? Over the huge sample you've shown, there should be quite a few small dips, and one or two larger dips, even assuming a superb true win rate. I don't think I can even pick out a single 1k block of hands in your graph that was breakeven, much less a 25BB dip. That's what led to my "statistical anomoly" comment. It's not a comment on your skill, just the fact that ultimately your cards have a lot to do with how you win, and the odds have to catch up sooner or later.

[/ QUOTE ]

This is all .50/1 and I just realized that the graph program would include the 1500 hands of .50/1 that I had at Paradise so they weren't quite all at Party.

The lack of dips will be even more puzzling if you take out the first dip at 1K where I hadn't started bailing at -10BB yet so I lost $30 and $25 on 2 tables. Plus the dip at 8K corresponded to a session where I decided to try and stay at a lively table and lost $28 and $18 on 2 tables.

So I don't think it's a coincidence that the 2 worst dips I had corresponded with not following the -10BB strategy. I checked back on that first losing table at 1K and after losing the first hand there my play was definately tilting. Probably $20 of the $30 loss was tilted off.

The dip at 8K was mostly getting sucked out on and missing all my draws. I'm sure I tilted off some after the suckouts though. So I guess what I'm saying is that probably the main benefit to my strategy is lowering the tilt factor.

I'm confused at the surprise of an around 6BB/100 win rate at Party .50/1. I thought many people maintained a higher rate than that over many more hands. Am I mistaken there? I don't keep records at the other site where I play most of my hands but I'm almost certain I win at a higher rate there on the .25/50 and .50/1 tables. Maybe I'm just imagining that though since I play many more hours there per week.

Aaron W.

04-15-2005, 02:24 AM

[ QUOTE ]
[ QUOTE ]
Is this mostly 0.5/1? If so, I'll assume your true win rate was the 6BB/100 you show in the graph (a superb rate, but which is still within what some posters have claimed is at the upper reaches of what is sustainable at 0.5/1) with an SD of 22.3BB/100.

This means that over 1k hands, your expected win rate is 10*6 = 60BB, with a standard deviation of sqrt(10)*22.33 = 70.5BB. Over any 1k block of hands, your probability of being breakeven is Q(60/70.5) = 20%. Your probability of experiencing a 25BB downswing is Q(85/70.5) = 13%. Go for a 50BB downswing, and it's Q(110/70.5) = 6%.

See why many people on the boards have been a bit skeptical? Over the huge sample you've shown, there should be quite a few small dips, and one or two larger dips, even assuming a superb true win rate. I don't think I can even pick out a single 1k block of hands in your graph that was breakeven, much less a 25BB dip. That's what led to my "statistical anomoly" comment. It's not a comment on your skill, just the fact that ultimately your cards have a lot to do with how you win, and the odds have to catch up sooner or later.

[/ QUOTE ]

This is all .50/1 and I just realized that the graph program would include the 1500 hands of .50/1 that I had at Paradise so they weren't quite all at Party.

The lack of dips will be even more puzzling if you take out the first dip at 1K where I hadn't started bailing at -10BB yet so I lost $30 and $25 on 2 tables. Plus the dip at 8K corresponded to a session where I decided to try and stay at a lively table and lost $28 and $18 on 2 tables.

So I don't think it's a coincidence that the 2 worst dips I had corresponded with not following the -10BB strategy. I checked back on that first losing table at 1K and after losing the first hand there my play was definately tilting. Probably $20 of the $30 loss was tilted off.

The dip at 8K was mostly getting sucked out on and missing all my draws. I'm sure I tilted off some after the suckouts though. So I guess what I'm saying is that probably the main benefit to my strategy is lowering the tilt factor.

I'm confused at the surprise of an around 6BB/100 win rate at Party .50/1. I thought many people maintained a higher rate than that over many more hands. Am I mistaken there? I don't keep records at the other site where I play most of my hands but I'm almost certain I win at a higher rate there on the .25/50 and .50/1 tables. Maybe I'm just imagining that though since I play many more hours there per week.

[/ QUOTE ]

First, the 6 BB/100 win rate at $.50/$1 takes into account the rake. This is why players at the nanos (where there is no rake) can have substantially higher BB/100. I don't know where you play, nor do I know the rake structure. But these are things to keep in mind whenever you try to compare statistics.

Second, the longer I think about your stats, the more it seems to be that nature is simply taking its course. In any 1500 hand block (using the hash marks provided across the bottom), there seems to be nothing out of the ordinary. In particular, my attention keeps getting drawn back to the flattening out period from 4.6K hands to 9.1K hands, where you have what looks to be a modest 75 BB gain. You may have simply been lucky enough to avoid the dreaded slide for 18000 hands. It's not impossibly unlikely. There's a 92% chance of NOT having a slide of 25 BB over 18000 hands and there's a 33% chance that you will NOT have had a slide of 50 BB by this point. Odds are against it, but not prohibitively so.

Third, it's clear that you play/think about poker in a very "intuitive" manner (read: non-mathematical, less formal), so no amount of reasoning along those lines will come close to changing your mind on most of your claims (especially your thoughts on denying certain drops because you were "tilting off" money). All that can be said is that truth will bear itself out in the long run. In poker terms, the "long run" is around 100K hands. Maybe you've found something that everyone else has missed, or maybe the mathematically minded are correct and you are looking at nothing more than an anomaly.

grjr

05-15-2005, 07:39 AM

[ QUOTE ]
I'll be sticking around here for a while so we'll see how the monkeys are doing in another 20K hands or so. /images/graemlins/grin.gifNot sure how long that will take since I only play on Party and skins during the reloads. I make more money on another site the rest of the time.

[/ QUOTE ]

Thought I'd update the graph at 30K and see if anyone has any new comments. I got quite upset about the big loss I had at the 20K mark. I spent the early morning hours on a +170BB in 400 hands streak and got tired so I got some sleep. Played again that evening and promptly lost it all back. Usually when I'm losing I'll quit and go watch the big-screen but my daughters were watching it so I kept coming back to the poker and kept losing. Never did win that evening. I blame them. /images/graemlins/wink.gif

I'm happy to report that my tilt problem seems to be getting better. I haven't broken anything since the 20K mark. I've also found a table selection pattern that seems to fit my game well. Plus, I'm on a really good card run right now that will end immediately since I posted this. /images/graemlins/wink.gif

Oh, this is all at Party at the .50/1 limit.

http://home.satx.rr.com/alphadeals/graph051505.jpg

ArturiusX

05-15-2005, 07:51 AM

If you tilt at .50/1, wait till you move up and get sucked out of $300 pots /images/graemlins/shocked.gif

grjr

05-15-2005, 07:57 AM

[ QUOTE ]
If you tilt at .50/1, wait till you move up and get sucked out of $300 pots /images/graemlins/shocked.gif

[/ QUOTE ]

Which is exactly why I'm still at .50/1. /images/graemlins/laugh.gif

handsome

05-15-2005, 07:59 AM

Your bankroll > $1,872? If so, move up and play $1/2. You'll make more.

grjr

05-15-2005, 08:14 AM

[ QUOTE ]
Your bankroll > $1,872? If so, move up and play $1/2. You'll make more.

[/ QUOTE ]

I've played around 600 hands at Party 1/2 and have done really well but I've gotten a lot of very good cards and it feels like I've been lucky. I have a certain comfort level at .50/1 that is hard to leave. Especially when clearing the ever present bonuses.

Actually my bankroll is around $5000 now thanks to all the bonuses and winnings from TigerGaming so that's not a problem. It's just hard for me to leave something that seems to be guaranteed winnings.

I guess once I'm fairly sure I've calmed down the tilt problem I'll go ahead and move up.

BTW, has anyone checked out the 30/60 games at Party lately? (not that I'll ever play there). I've been watching some over the weekend and frankly they don't look that tough. There seems to be as many fish there as the 1/2 level. If I was independently wealthy that would be a fun level to play.

QTip

05-15-2005, 08:33 AM

I don't agree with this. I define a session as playing any given period of time without a substantial break, who cares if you switch tables or not.

2 more things to add:

1. I went a very long time without having a downswing larger than 40BB. Now, 100BB drops are quite common for me.

2. His skipping tables indicates a larger problem in my opinion. You can drop 10BB before an orbit is over without trying. You go around hopping from table to table like that, you're not getting reads on anyone and playing like everyone is an unknown. You're not helping your game at all. You know nothing about your opponents, and your not adjusting your game accordingly. This sucks.

oh..the other thing that sucks, is me going around bumping threads that are a month old - my bad.

aron

05-15-2005, 09:52 AM

I've read most of the replies here and don't reall know who to address.
But I'm thinking of Grjr lack of variance.
I'f a table is bad for you? What results can we expect?
Well, variance tells us that we can pretty much expect anything too happen.
Wins or Losses.
But in the looooong run we should go broke, right?
So the most likely think to happen is that we lose money, say 60% chance?
Thus leaving a table were you drop 10bb is in a way more likely to be a bad table than a table where you make 10 bb. But vice versa will occur a lot too.

Still this don't explain lack of variance since you ought to have big downswings even if you are playing the juciest table there ever was.

I think I agree to the poster who said this is explained by variance too. If we pick graphs from 10.000 solid players most will look normal but a few ones will look odd too.

-aron