The Life of a Shortstack Reg | Jake's 2014 Cash Thread

[Matt Vaughan]

bb/100 and evbb/100 are both fundamentally flawed measures. fwiw.

 

[ScottishMatt]

bb/100 and evbb/100 are both fundamentally flawed measures. fwiw.

This. I can be 10BI from being two-outered after getting 90% of the money in before the river and it wouldn't show up on the graph.

 

[Logan2]

Agree that both have flaws, but still evbb/100 is a better measure than bb/100, and if bb/100 was posted don´t see why could not ask for evbb.

If he don´t want to post it is fine thought but i don´t see the big deal.





​

 

[duggs]

bb/100 and evbb/100 are both fundamentally flawed measures. fwiw.

Perfect is the enemy of good?

 

[Matt Vaughan]

k, sorry I said anything... I was just pointing it out, cause it's debatable that evbb is even significantly better. you need a similar sample size from both (orders of magnitude-wise) before they become meaningful.

I'm not saying never look at a graph, just think people get too into evbb.

 

[duggs]

k, sorry I said anything... I was just pointing it out, cause it's debatable that evbb is even significantly better. you need a similar sample size from both (orders of magnitude-wise) before they become meaningful.

I'm not saying never look at a graph, just think people get too into evbb.

surely its at least a little bit better? it accounts for one form of variance, it would be nice to have a measure that incorporated when most of the money went in tho.

 

[Matt Vaughan]

I agree it's a bit better, but in terms of sample size before it's significant? I doubt it's even 2x better. You need at LEAST 100k hands before you consider winrate at all, and I think frankly you need more like 250kish hands. If evbb lets you look only 20 or 30% sooner then I still don't really care about it in the short run.

Granted I don't have any math to pull out, but what is the average standard deviation of evbb/100 as compared to bb/100? Cause I guess that would be useful to know. Dunno if PT4 gives that stat though. I'm also not sure if that'd be the best comparison, but would be interesting to know.

Sorry for derail, btw

 

[duggs]

i don't really think of it like that tho, like for any given sample size EVbb should be a better unbiased estimator than bb

 

[Matt Vaughan]

Agreed it's strictly better.

 

[duggs]

like yea its super flawed, and yea everything is only an estimate, but i still think we can draw conclusion quicker than we should be able to purely based on numbers especially when we have other means of confirming correct play (sweats/HA/etc). Biggest barrier to really examining win rate is variability in our own play, variability in field toughness, and variability in the situations come up. like the fact we improve over time means win rates aren't static.

lol ending rant now,

 

[Matt Vaughan]

like yea its super flawed, and yea everything is only an estimate, but i still think we can draw conclusion quicker than we should be able to purely based on numbers especially when we have other means of confirming correct play (sweats/HA/etc). Biggest barrier to really examining win rate is variability in our own play, variability in field toughness, and variability in the situations come up. like the fact we improve over time means win rates aren't static.

lol ending rant now,

That was my whole point to begin with Sigh when we somehow argue over something we agree on.

 

[duggs]

but this is fun, my alternative is reading textbooks.

 

[Matt Vaughan]

Ha, touche.