The hedgehog, the fox, and the chaos they've created.

Sarlacc83

Active member
Joined
May 1, 2009
Messages
17,110
Reaction score
1
Location
Portland, OR
BlueTalon":1bqz4eyz said:
Throwdown":1bqz4eyz said:
A butterfly in Brazil can cause a tornado in Texas?
No. There are millions more butterflies in Brazil than tornadoes in Texas. If a single wing flap of a butterfly in Brazil could cause a tornado in Texas, there would be nothing left of the entire southern half of the country.

And why would a butterfly wing flap in Brazil be more likely to cause a tornado in Texas than a butterfly wing flap in Texas?

People take the butterfly statement waaaaaay too literally. It's a metaphor for how a small deviation in a chaotic system causes huge variance at the other end of the spectrum when working with the math. As usual, people didn't bother to check the meaning because the analogy sounds cool.
 

BlueTalon

Well-known member
Joined
Mar 3, 2007
Messages
8,944
Reaction score
1,591
Location
Eastern Washington
Buzzkill. I was having fun with it. I know it's a metaphor. And I know people take it way too seriously. People are idiots.

But just imagine the havoc we could create in other countries if we could harness the climate-altering power of butterflies!
 

ZeroTrunks

New member
Joined
May 13, 2013
Messages
4
Reaction score
0
Great post. Ultimately we are talking about the big show. Winning the Super Bowl. And sometimes the best team doesn't always win. With the age of our team that does decrease the likelihood of injury, but really you have to get lucky. Those couple calls that could have gone your way, that holding call that got missed, that questionable pass interference. The good news is- that after this year, with the popularity of our team, those highly questionable ref calls (wether for or against) will get more media coverage, and that is a good thing. The reason is because we are a great team, and if officiated correctly (05 SB) we will win!

GO HAWKS!
 

olyfan63

Well-known member
Joined
Apr 17, 2012
Messages
5,662
Reaction score
1,696
bestfightstory":12p9en0w said:
olyfan63

I would say that the drop off when we lost Clemons was extremely significant.

Excellent point. In its own way, the loss of Clemons was pretty similar (to the loss of Marquand Manuel in SB XL) in its devastating impact on the Seahawks overall game. With Clemons in the game, we beat Atlanta, no question. We get either a crucial sack, and/or a crucial, pressure-related pick, or both. Our run defense is stronger, and Atlanta doesn't go out of their way to run at Clemons, and our run D doesn't allow a washed-up Michael Turner to look quite so much like Emmett Smith in his prime. Losing Clemons = similar impact, but very different reasons.

Clemons was the guy whose play partially covered up for the lack of interior pass rush from our DT's. Losing Clemons highlighted the Seahawks lack of pass rush from the DT spots. It highlighted Bruce Irvin's overall inexperience, and tendency to get locked up with, and mowed over by, big O-linemen. Losing Clemons exposed that too much of the Seattle D was held together by chewing gum and baling twine. Read: Jason Jones(out), Alan Branch(very little pass rush), Red Bryant (with a bad wheel), and toss in Bruce Irvin and Gregg Scruggs (rookies still getting schooled), then toss in Leroy Hill (old and slowing)

The difference is that the PC/JS front office isn't taking it lying down. They have aggressively addressed each of those issues. Clemons out? No problem, we got Cliff Avril, a near Clemons clone. We got Michael Bennett, who also can pick up a lot of that slack as a pass rushing DE or even rush from a DT spot in some packages. PC/JS addressed the DT issues by drafting Jordan Hill and Jesse Williams, and we can hope Greg Scruggs and/or Jaye Howard show us something this year. Marcus Trufant slowing/getting beaten? OK, bring in Antoine Winfield, an upgrade at slot corner.
PC/JS seem to have a similar incremental improvement program going in the secondary, and, with TC, in the O-line area as well.

Another huge, huge risk area is Marshawn Lynch, and the overall beating he takes. I love, love, love the Christine Michael pick. Lynch goes down, and other teams still need to fear and adjust to our featured running back. They even addressed the MRob aging dropoff/injury risk with the Spencer Ware pick.

Right now, the biggest risks to this team are: 1) Losing Russell Wilson; 2) Losing Earl Thomas. In both cases, the dropoff would be huge.

I'm not seeing the chaos theory as much as shrewd, probability-theory-based gamblers, looking to improve the percentage of 1-on-1 battles their side wins, with whatever combination of players are on the field, knowing that on any one given play during a game, one player winning his 1-on-1 battle could make the difference between winning or losing the game. (And yes, more like combination of scheme and player executing within that scheme)

I'm not really trying to put the chaos theory explanation down... in truth, I can't say I really understand chaos theory or how and when it can be useful, and especially its usefulness in predictions and in guiding decisions. I'm seeing it more as an after-the-fact, gee-whiz-that's-fascinating explanation of what happened. I acknowledge my ignorance in that area, and if someone can shed more light on how the chaos theory model is predictively useful, I'm all ears. For now, applying Occam's Razor, the simpler probability-based explanation is one I can make sense of.
 

nbk35zw

New member
Joined
Feb 28, 2012
Messages
222
Reaction score
0
This is perhaps the best piece of sports psychology/analytics I have ever read.

Thank you.
 

Sarlacc83

Active member
Joined
May 1, 2009
Messages
17,110
Reaction score
1
Location
Portland, OR
BlueTalon":2nemnu7n said:
Buzzkill. I was having fun with it. I know it's a metaphor. And I know people take it way too seriously. People are idiots.

But just imagine the havoc we could create in other countries if we could harness the climate-altering power of butterflies!

I was adding onto your statement, but I'm glad to see I harshed your vibe. :thirishdrinkers:
 

HawkHack

New member
Joined
May 2, 2010
Messages
269
Reaction score
0
A public service announcement for anyone who might possibly care. [I am a mathematician by education only.]

A linear equation always "looks like" a line if it's graphed. If you graph the thing and it doesn't look like a line, it's not linear. Guaranteed.

The confusion in the general public arises because mathematicians and scientists often use the term "non-linear" to mean "not simple". If you can describe something with a linear equation, it's dead simple -- and boring -- so "non-linear" has come to mean "more interesting to math and science geeks".

Exponential functions are certainly non-linear, but are actually pretty simple. As an example, consider the classic story about the smart Chinese guy who asked the Emperor for one grain of rice on the first square on the chessboard, two grains of rice on the second, four on the third, eight on the fourth, and so on. By the time you get to the 64th square, you're at something close to 2 quintillion grains of rice. Exponential functions grow (or decay) very quickly -- but once you understand how they work, they're entirely predictable.

A chaotic process is one where (as someone else pointed out) small changes in the initial state can result in a state that is difficult to predict down the road. This is the "butterfly effect" also mentioned, where killing one butterfly might result in tornadoes in Texas. However, chaotic processes aren't non-deterministic: in theory, we can calculate the weather with 100% accuracy. It's just difficult to do. Chaotic processes typically cannot be represented in what we would call "closed form", which means they evade analysis and have to be simulated using software. Climate change models are chaotic in the sense that they're very sensitive to small changes in initial conditions and assumptions, which is one reason why they still generate controversy.
 

olyfan63

Well-known member
Joined
Apr 17, 2012
Messages
5,662
Reaction score
1,696
In other words, there are so many variables, that by the time you get done identifying and analyzing and measuring them, and putting them into a predictive model, whatever event they were to be used to predict is long over. And not only that, but in the meantime, it's become clear that only half the actual variables have been identified.
So I think you are saying chaos theory is not practical or useful as a reliable predictive tool. (Though I think you are allowing for models that simulate many variables to be *useful* for analyzing *potential* outcomes of different choices)

It seems "injuries" are a big part of the chaos of NFL football that make things so hard to predict.

Speaking of additional variables, in my writeup above, I completely neglected to mention the free agent signing of DT Tony McDaniel. Maybe he will turn out to be the X-factor that turns a few games. Impossible to predict at this time... but he gives us a few more chances.
 

DaveyP

New member
Joined
Apr 17, 2012
Messages
48
Reaction score
0
HawkHack":2lc8uelf said:
A public service announcement for anyone who might possibly care. [I am a mathematician by education only.]

A linear equation always "looks like" a line if it's graphed. If you graph the thing and it doesn't look like a line, it's not linear. Guaranteed.

The confusion in the general public arises because mathematicians and scientists often use the term "non-linear" to mean "not simple". If you can describe something with a linear equation, it's dead simple -- and boring -- so "non-linear" has come to mean "more interesting to math and science geeks".

Exponential functions are certainly non-linear, but are actually pretty simple. As an example, consider the classic story about the smart Chinese guy who asked the Emperor for one grain of rice on the first square on the chessboard, two grains of rice on the second, four on the third, eight on the fourth, and so on. By the time you get to the 64th square, you're at something close to 2 quintillion grains of rice. Exponential functions grow (or decay) very quickly -- but once you understand how they work, they're entirely predictable.

A chaotic process is one where (as someone else pointed out) small changes in the initial state can result in a state that is difficult to predict down the road. This is the "butterfly effect" also mentioned, where killing one butterfly might result in tornadoes in Texas. However, chaotic processes aren't non-deterministic: in theory, we can calculate the weather with 100% accuracy. It's just difficult to do. Chaotic processes typically cannot be represented in what we would call "closed form", which means they evade analysis and have to be simulated using software. Climate change models are chaotic in the sense that they're very sensitive to small changes in initial conditions and assumptions, which is one reason why they still generate controversy.


Thanks!! I figured a mathematician that actually knew what they were talking about would be lurking on this board someplace.
 

Lords of Scythia

Well-known member
Joined
Sep 19, 2011
Messages
2,607
Reaction score
167
Damn you. I thought this thread was about the superb French book "The Elegance of the Hedgehog".
 
Top