Thread: The Fallacy of Training Machines
12/05/2005 3:46pm, #371
12/05/2005 3:51pm, #372
Originally Posted by Steve Richards
- Join Date
- Feb 2004
(which, incidentally, reminds me of Ronin's schlong...)
TomasCurrent stage of death: denial
12/05/2005 3:55pm, #373
12/05/2005 4:11pm, #374Originally Posted by Steve Richards
12/05/2005 6:50pm, #375
Originally Posted by Ronin
- Join Date
- Aug 2005
- Baltimore, Maryland Area
- H'ung Ga & SPM
Good enough "method" for fighter pilots ... probably good enough for unarmed combat. :new_astha
12/05/2005 8:38pm, #376
OK, so are there any basic exercises you'd recommmend for entrainment? Would you suggest mentally visualizing a basic but anatomically accurate stick figure shape, maybe similar to one of those 3D motion capture systems (like the lightbulbs-on-silhouette experiment) you mentioned with larger rotational nodes at key joints? Some more nodal complexity at usual target areas like the solar plexus, neck, nose, throat, etc?
Although the lines are simple and entirely complimentary to the movements of your opponent, it seems like it would be a hard thing to do although, if I am correct in following what you've said, it should eventually help with the `sensory overload' that often accompanies a determined attack and allow you to see the critical movements as part of the whole which would, presumably, improve your ability to accurately predict movements earlier on.
Am I way off base on this, or have I understood things correctly? Are there any papers you could point us to that focus on the practise as well as the theory? Anything to look out for?
12/06/2005 5:05am, #377Originally Posted by BenwaMandelbrot
Yes, you've summed that up very well.
For a visual read in 2-D of a basic outline, presented for a specific system, you could have a look here:
You'll see the major jouint articultaions modeled and the torso-clock, transverse and vertical control planes etc.
Style/system is just a delivery platform, the 'doing' has to be subjective, and earned. What can be done is to find out how well various 'platforms' stand up to application of this approach. Stand-off systems are less likely to work well with it, because there is a basic contradiction between them and the HUD. Its not what a TCMA aproach would regard as a 'bridge' methodology, it about passing the bridge, and working directly on the target as a whole.
One common problem early on, is when folks try to force specific techniques or 'style rules' into it. My answer to them is to let the HUD 'tell' you what to do rather than try to make the HUD fit a style. The map has its own logic, so it 'suggests' what are workable motions. Refinement in techinque comes from this, as The HUD is about natural human form modeling, not about progressively refined abstractions of it.
The HUD as an overlay should only be used in the initial entrainment phase, it is a MODEL and it has the limitations of all physically modeled maps. However, once internalized and made tacit, it runs itself - and thats a major goal of its use, a devolved, non-conscious, fast, system of information processing.
You should find very quickly that the 'complementary' aspect - the symetry between the two basic sides of the body mean that you can model the whole human form from as few as two, or even one map point. You can do this because proportion and relative distance are fairly constant, even with differences in height and size in an opponent.
Because this is a probablility model, it does not rely on sight, at this stage, but of an internal spatial and temporal model, which is imposed thru action against the target. There is for example, a distance between the shoulders, along a plane that includes the head and neck. If one shoulder is facing you (laterally), then there is potential for the other shoulder to come into play thru a fairly predictable range of angles. The variations in that angle can be accomodated by action against the opposite shoulder ahead of its potential for motion. Because this is along the shoulder plane, and also takes the head/neck into consideration, you can 'follow the line' and put in a head shot/crank/choke-whatever, along the way - according to the specifics of the situation. This is a very reduced example, its not 'whole-context' but its an example to build up from.
What should be done to build up the map, is to run scenario's like the one above, pressure test them, find as many failure-points as possible, work with these, so that you can begin to take the loading of failure, and also look at new 'firing-off' or 'refiring' points either in that plane, or somewhere else on the opponent, according to your position/structure/motion relative to his. Its a dynamic model, and its one that best works thru the experience of failure rather than success, because being closed-down is where the most mental pressure is felt, and its there that you need spare processing capacity to switch and reverse the situation.
This works better if you have a mental model that is mapped onto the basic human form, and its potential for action in any plane elative to you.
12/06/2005 5:13am, #378Originally Posted by WhiteShark
Not everyone is gifted with natural speed. It can be trained for, but so can processing capacity.
As an example, it is possible to be balistically slower than your opponent - and appear to an observer to be 'slow' relative to him, but because modeling is better entrained, you can be functionally faster.
With the getting there first scenario, they are effectively qunatum-jumping ahead of what 'presents' to a future potential position. This must be a probablility computation as it is not an immediate, sensory-data calculation. We all 'hypothesize' about future positions and relative motions, but if we have a conscious methodology to train it, then more of us can improve the more, rather than having the few outstanding naturally gifted folks always getting there first.
12/06/2005 7:28am, #379
12/06/2005 8:58am, #380Originally Posted by Ronin