Excellent FP on modeling

[Paul Arden]

AN excellent FP today with some nice photos from James - https://www.sexyloops.com/index.php/ps/ ... erfections

Cheers, Paul

[James9118]

I probably should explain that the photos are of a single roll cast with a MPR. I perhaps should have put some parallel lines along the fence, however I think it's pretty clear that the leader ends up higher than the fly line. Now, to me, how it ends up there is blatantly obvious - and it's got nothing to do with aerodynamic lift or pixies. The great thing about this is that everyone here can replicate this if they choose to.

James.

[grunde]

Just want to thank James for today’s excellent FP on the limits, challenges and usability of models
https://www.sexyloops.com/index.php/ps/ ... erfections


I just want to add a small comment about the different mindset of scientists and engineers, which James touches upon.

Caricature description of the different mindsets:
The scientist wants to describe and understand the main effects (and limits) in the problem, and are quite happy to not include 2nd and 3rd order effects.
The engineer on the other hand wants a “perfectly fitted model” which can model the full problem, often by using non physical (free)fit parameters and existing “domain specific” models and equations. Another variation of this mindsets could be found in domains and for problems where verification is very difficult, leading to overly complex models which are trusted more than “reality”

In my experience there can be problems when these mindsets “collides”...

Cheers,
Grunde

[Paul Arden]

We had a very similar thing occur with the 5 Essentials Model. Namely the Straight Line Path. Some instructors have taken that completely out of context, indeed they will tell you that a Casting Stroke must have a straight tip path, and it doesn't then it's not a Casting Stroke! We have a similar thing of course with the Casting Analyzer.

I see a model as a matchstick man filter that you can apply by placing over the real cast, a bit like tracking paper. What you can't do is substitute it for the real cast because then you find it has many limitations. I love the explanation that it works because it's been fudged to work.

Thanks,
Paul

[Paul Arden]

Can you be sure that it's not pixies, James? I mean you are in a garden.

[James9118]

To be fair, pixies is the more plausible of the two
I was waiting for a comment about how well kept my lawn looks

[Paul Arden]

Yeah it’s lovely. I’m betting the pixies drive the lawnmower.

[gordonjudd]

caricature description of the different mindsets:
The scientist wants to describe and understand the main effects (and limits) in the problem, and are quite happy to not include 2nd and 3rd order effects.
The engineer on the other hand wants a “perfectly fitted model” which can model the full problem, often by using non physical (free)fit parameters and existing “domain specific” models and equations. Another variation of this mindsets could be found in domains and for problems

Grunde,
Interesting that a physicists would see the caricature from that perspective.

I think most engineers see it the other way around.

Is "close enough for government work" also an expression that is heard in physicist's circles?

Gordy

[Graeme H]

I agree. An excellent article on the relevance of modelling.

I get to see it every day in my real job. I have made geological models for the last 15 years. They are always released with the caveat of being fit for purpose - and ONLY the purpose for which they were built. The mine engineers always wanted to build their infrastructure based on a geological model, and if the only one available was built for exploration, they would crack the shits when the model didn’t reveal faults and fractures that we didn’t know were there.

Models are fit for a purpose. To try applying them outside that purpose is fraught with danger.

Cheers,
Graeme

[Walter]

Excellent points Grunde!

Adding to the mindset, any useful models I’ve ever used stated the assumptions and boundary conditions. Of course, as Graeme pointed out, these often get ignored until something goes wrong.

One of the things I’ve advocated for in the past is some sort of statement of assumptions when models are being discussed here. It gets somewhat annoying when a thread waffles back and forth between tethered casts and shooting casts for example.

Maybe in addition to the recent efforts to keep threads on topic we could look at implementing a little bit more scientific methodology in our scientific methodology in this forum? I think Torsten’s roadmap for validating Gatti-Bono’s paper is an excellent example of that. I also think Graeme’s attitude of asking Gordy to help him design an experiment to prove or disprove if lift is significant in fly casting is another excellent example.

Keep up the good work!

PS: James - the well kept garden was the first thing I noticed in your pictures.