Calling Vernon N, Engineers and Mathematicians.


 
Status
Not open for further replies.
Rob,
It's done when it's done. But to address your comment about a regression model, I'd say that work has already been done. It may not have been scientific, but I think if you did a regression on all of our experience, you'd come back to the 1.5 hours per pound that is almost always used when talking about bbq. As you know, after you do a regression, you still only have an expected value, not a certain predictor of the outcome. There's always going to be outliers.
 
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> ... For anyone who's still interested I think this problem might be handled if we approached it as a statistics problem. There's a technique called regression analysis that allows you to create mathematical models that predict an end result provided you have sufficient historical data.

If anyone's still interested in the discussion there are also ways of handling things like the varying thickneses and fat contents of meat by using something called a normalized distribution curve.

But all of these techniques would require sufficient data and I admit I'm a long way from figuring out what data to get and how to get it... <HR></BLOCKQUOTE> I was thinking myself using statistics and probability on a sizeable database of results might provide good estimates. sounds like the regression analysis attempts to make the models from the statistical results. Way beyond my skills.

Anyhow, if you need data and can provide a template and instructions Rob I will be more than happy to strap on a crash helmet and log my cooks to assist your project.


Regards,

Shawn
 
Jim, I am not a competitive BBQ'er but I can appreciate that different pieces of meat cook differently based composition and shape.

Differrent amounts or percentages of water, fat, connective tissue, and muscle would affect the length of the cook. Certaily their are other factors, wind, temp, humidity....

I would be willing to bet that you could more reliably predict cook times by knowing the density of the meat. If the density of meat varies much then the the mix of fat, muscle, connective tissue, and water are varying....

To test, get two pieces of different density but similar size and shape and cook side by side on the same level in the wsm...Jim as an expert, let me know if I'm all wet or worse.
 
Jerry, Shawn, Steve,

Good input from all of you. I'm going to have to think a little bit about what problem it is exactly I'm trying to solve.

What I have in mind would be something along the lines of a tool that could answer the question (hypothetical example):

My 15 lb brisket has been cooking at 225* on a WSM for 9 hours and has only reached 159*. How much longer (based on the observed populations of briskets) should it take to reach 195?

If we could develop an adequate model it could give an answer: Based upon the observed population of briskets it should take you 12 hours +/- 2 with a 95% confidence interval. With 99.999 % it will take no longer than .........

Guys like Jim can do this based on experience. No tool is ever going to replace that. Think about all the technology in NASCAR these days. In the end it still comes down to the driver. Still what do you do when you can't find a guy like Jim?

I'm going to run this idea by one of the guys I know who does quant work at one of the financial houses up here. See if I can get this thing defined a little better. If I can I can probably scrounge some computer time w/one of the statistical modeling packages.

If I make any progress I'll update the post. In the meantime if anyone comes up with a brilliant idea here let me know.
 
I agree with Larry, you learn by doing and talking to others who "do". I think way too often people try to over complicate things when really it comes down to lighting fire, putting some meat on, watching the fire, checking the temperature of the meat, taking the meat off when it's done, and eating.

You could add; having a beverage or two in there. Anything above and beyond that tends to cloud the issue. But have at it.

IMHO
 
Status
Not open for further replies.

 

Back
Top