Plus obviously subtract 20 bb/hr off your SD/hr if your shirt is tucked in while playing.

Gwait,G'sSD/hrisnittierthanmine?lolG

So the biggest factor (IMO) in how high variance you are is VPIP.. pretty intuitive too, someone who sees 4x the # of flops is going to have a broader range of results.

Loose is bigger factor than aggression AINEC. Also, getting folds is low variance, so all aggression is not created equally. Thin value would be high variance... break even bluffs would be high variance.

Hopeless calls and fat value are not high variance.

*one other thing that is highest variance of all .. the deck.

That makes sense, as I am very tight (by live standards) PF, but sometimes too aggro post.

Still very curious what my +/- to my perceived WR (RR) is likely to be.

^ indeed looks like the solution... thanks for the link. (but I don't understand it on initial reading. The images are lost and perhaps I need pictures to learn.)

bip, off topic but is the theory behind break-even bluffs that it gets you more calls on your value hands? while i am fairly sure my river bluffs are eliciting folds >50% of the time, the thought of them being a zero-sum proposition is disheartening. i should probably start tracking my river bluffs

This is probably a weird way to look at it, but if my SD is slightly over half of the example of the chart, can I multiply my number of hours by 1.852 to get an "equivalent number of hours" if my SD were 100BB/hr? By that "logic", I get a 95% confidence that my "true WR" is +/- 6BB/hr of my recorded results. Does that work at all, or am I over extrapolating?

Find your hours, find the error at that hour, divide that error by 1.852

(wrong axis Garick ... scale the Y axis to suit your needs)

NVMD... figured it out

Interesting. So my 95% confidence rate is within about 4BB/hr +/- recorded results. That's a tighter range than I would have guessed.

Ok, I'll try again, with a better worded question I think lol.

I found the standard deviation taking into account varying session lengths.
Now it seems dividing my found SD by SQRT(Average session length is appropriate) to get SD/hr?

Then graph (2*SD/hr)/sqrt(total hours) to produce the graph BIP created for 95% confidence in hourly win rate bounds.

That was fun Solid tuesday afternoon lol

Working backwards, does that mean you've only got about ~1000 hours?

If I've done my math right (lol?) I'm 95% confident my winrate falls between 6.1 and 10.7 bb/hr.

GandI'mlike20%confidentinthat,sofiddyfiddy?G

Your stdev * 2 / sqrt(hrs) is that plot

So - any bluff that breaks even should hopefully move the opponents tendency to call you more. At some point though you run out of value hands to sustain this. (NL is great because you can be bluffing nearly 50% of the time if you bet huge versus the pot... PLO you can't).

Counterintuitive poker... If your bluffs are always getting folds, you are not a good bluffer. It means you are bluffing way too infrequently and losing value on your value hands.

I would bluff in every situation I feel is better than break even. Which means at the extreme, it should feel break even when you are making a bluff.

For those trying to do stdev on your results - don't mix dissimilar games. PLO does not mix with NL, limit doesn't mix, etc, etc.

Even games with significantly different buy in formats, radically different player pools, slot limit vs NL.. etc.. those don't mix

To use that table I'm assuming I look up my ~2400 hours, get the ~4, and then divide that by 100 / (my standard deviation per hour) = 100 / ~57.7 = 1.73, therefore +/- 2.3 bb/hr?

Gcourse,thepastisthepast,somightbemeaninglessmovin gforwardG

Yep, exactly.

You can directly calculate too,
57.7 * 2 / sqrt(2400) = +/- 2.36

The plot is more to show people the trend over hours.. Relative meaning of your results over sample size

Ah, ok, makes sense.

Hurray, I wasn't a complete fish relative to the field.

Gnext2400hoursmightbeacompletelydifferentstoryG

I just figured slot limit StDev was *infinity sign*

Or you are just a wizard that always folds out better when bluffing and gets called by worse with da nutz

But in all seriousness, don't get too worked up about being picked off if your general success rate is >50%?

I've been working a lot on my river bluffs when I think we are both weak but I am weaker - usually only getting picked off by strong nutted hands that played it so incredibly passively that it annoys the hell out of me that I was doing their betting for them

So the biggest part of it is keep it player dependent. I would only worry about balance against good players who can range and hand read and who don't have face up ranges themselves.

Also, a break even bluff does not mean 50% get called. If you bluff half pot, you can get called 67% of the time and still break even.

There are players who call frequently enough even if you never bluff... don't bluff them ever. There are players who always fold... bluff them always. Use balance against aware players.

50% be would be correct (Beeak even) for pot sized river bluffs.

It seems obvious but yes the math also shows that when a player folds more than what is correct mathematically you can bluff more of your range approaching 100% and show a profit. Conversely if a player calls more than is correct it becomes profitable to reduce the number of bluffs approaching 0%. Hence the general adage in LLSNL that since there are more callers than folders you shouldn't bluff.

But as Bip points out it is player dependent and I really like his point that it should feel break even. I agree it's one of those things where the negative reinforcement of failed attempts works against the actual profitability results. Somewhat like thin value betting and overbetting IMO.

cAmweaybehindtheracetothebottomoftheSTDEVhillwitht heGbrothersAm

Great job itt bip!

A few people have hinted in this thread recently that crushers often (always?) have higher stddev. I was guessing thin value was part of the answer, but high vpip sure makes sense.

My first 1500 hours was mostly snug ABC stack-a-donk, and my stddev never strayed far from 60bb / hr.

IMO, a better way of looking at bluffs is what bip has suggested, manipulation of betsizing.

If V is prone to call a lot, rather than reducing number of bluffs approaching 0%, one may consider reducing the size of bluff so the impact is much smaller.

Such strategy can benefit in two ways:

1. Reaffirming V's idea that he's making the correct decision, and we can set him up for bigger value bets in future.

2. Potentially setting up our image against other players and increase EV of our value bets.

Lastly, most players fail to notice that even calling stations have limitations and they do respond to different bet sizing with strength of their hands. Range matters and good players understand how to manipulate betsizing to increase EV.

Anyone that can comment on T/T? Multiple tables run once a week and occasionally a bigger T/25 goes on simultaneously. No $5 chips allowed so it plays bigger. I feel like 100 BB's may not be deep enough, or if so you really gotta nit it up. I recognize a couple of players from 2/5 (that I consider to be average players) so the waters don't appear to be shark infested observing from a distance.

Don't feel comfortable bringing 3k with me just yet. Working my way up to bringing 3 bullets $1000 each at 2/5 first, but the game appears juicy.

I know it could be good but the few times I played it was always reg infested. Once it was 5 handed with all good regs think that was the hardest line up I've ever played anywhere