Baseball Blues

CSGnet Archive CSG2003
1

[From Rick Marken (2003.01.27.1400)]

The relative ease with which I got my "Control Theory Glasses" published had
lulled me into thinking that conventional psychology was getting more receptive to
the idea that behavior is the control _of_ perception. But I have been reawakened
"Fielder's Choice" paper published. The "Fielder's Choice" paper (it should have
been enthusiastically accepted on the basis of the title alone;-) is a critique of
current models of how baseball outfielders catch fly balls. It has now been
rejected by three journals and I plan to give up any further efforts to publish
it. I'll just publish it on the web. A PDF version is posted at my web site at
http://www.mindreadings.com/baseball.htm. I would appreciate hearing comments on
the paper from fellow PCTers. This might make a nice topic for discussion on
CSGNet.

The negative reviews of the paper were priceless and I plan to scan those reviews
and post them, along with my replies, in the near future. For now, I have just
added a new baseball demo, called "Baseball model comparison", along with the
"Fielder's Choice" paper at http://www.mindreadings.com/baseball.htm. I built the
demo (which still needs some work; suggestions for improvements will be gratefully
accepted) to show that the most popular current model of baseball catching, the
Optical Acceleration Cancellation (OAC) model, which was constantly thrown back at
me as the best existing model of baseball catching, doesn't work nearly as well as
the Optical Velocity Control (OVC) model that I have been advocating. The demo is
also designed to show that optical acceleration and optical velocity are two
independently controllable optical variables. And that it makes a difference (in
terms of matching the behavior of model to the behavior of the fielder) which
variable one takes to be the _controlled variable_ -- the variable controlled by
the fielder. By the way, the closest any of the reviewers could come to saying
"controlled variable" was "control variable". It was clear that even the reviewers
who claimed to be experts in control theory assume that perceptual variables
_control_. They had no idea that I was talking about the fact that perceptions are
_controlled_. They couldn't even _say_ it. And that was what my paper was about,
to some extent. Very interesting!

Anyway, please visit the site and let me know what you think of the paper and the
demo.

Best regards

Rick

···

from my dogmatic slumbers by the incredible difficulty I've had getting my
--
Richard S. Marken, Ph.D.
Senior Behavioral Scientist
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

2

[From Dick Robertson,2003.01.27.1705 CST]

Richard Marken wrote:

[From Rick Marken (2003.01.27.1400)]

The relative ease with which I got my "Control Theory Glasses" published had
lulled me into thinking that conventional psychology was getting more receptive to
the idea that behavior is the control _of_ perception. But I have been reawakened
from my dogmatic slumbers by the incredible difficulty I've had getting my
"Fielder's Choice" paper published. The "Fielder's Choice" paper (it should have
been enthusiastically accepted on the basis of the title alone;-) is a critique of
current models of how baseball outfielders catch fly balls. It has now been
rejected by three journals and I plan to give up any further efforts to publish
it. I'll just publish it on the web. A PDF version is posted at my web site at
Baseball Catching Research. I would appreciate hearing comments on
the paper from fellow PCTers. This might make a nice topic for discussion on
CSGNet.

The demo still looks good to me. About the rejects: sort of reminds me of all the
articles about prosecuting attorneys who refuse to admit they could have been wrong,
even though someone else says he did it, and his DNA is all over the victim and the
witnesses say they lied under duress.

Controlling a perception sure fights disturbance in a lot of people doesn't it?

But, look at mine and Dave's paper, that got published in an obscure journal only ten
years after the fact. (Thanks to Martin BTW). All is never lost.

Best,
Dick R

···

The negative reviews of the paper were priceless and I plan to scan those reviews
and post them, along with my replies, in the near future. For now, I have just
added a new baseball demo, called "Baseball model comparison", along with the
"Fielder's Choice" paper at Baseball Catching Research. I built the
demo (which still needs some work; suggestions for improvements will be gratefully
accepted) to show that the most popular current model of baseball catching, the
Optical Acceleration Cancellation (OAC) model, which was constantly thrown back at
me as the best existing model of baseball catching, doesn't work nearly as well as
the Optical Velocity Control (OVC) model that I have been advocating. The demo is
also designed to show that optical acceleration and optical velocity are two
independently controllable optical variables. And that it makes a difference (in
terms of matching the behavior of model to the behavior of the fielder) which
variable one takes to be the _controlled variable_ -- the variable controlled by
the fielder. By the way, the closest any of the reviewers could come to saying
"controlled variable" was "control variable". It was clear that even the reviewers
who claimed to be experts in control theory assume that perceptual variables
_control_. They had no idea that I was talking about the fact that perceptions are
_controlled_. They couldn't even _say_ it. And that was what my paper was about,
to some extent. Very interesting!

Anyway, please visit the site and let me know what you think of the paper and the
demo.

Best regards

Rick
--
Richard S. Marken, Ph.D.
Senior Behavioral Scientist
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

3

Hi, Rick –

I’ll check out the paper, but I have a suggestion. When you read the
reviews, does it seem that any (or all) of them do not want your paper
published simply because they disagree with your model? If that’s the
case, you would be justified in putting up a very strong objection to
“gatekeeping” – using one’s position as a referee to keep
competing ideas from being heard. There are legitimate reasons for
recommending rejection of a paper, but preferring your own ideas over
those of the author being reviewed is not one of them.

Bill

4

[From Rick Marken (2003.01.28.0840)]

Bill Powers wrote:

Hi, Rick –
I’ll check out the paper, but I have a suggestion. When
you read the reviews, does it seem that any (or all) of them do not want
your paper published simply because they disagree with your model? If that’s
the case, you would be justified in putting up a very strong objection
to “gatekeeping” – using one’s position as a referee to keep competing
ideas from being heard. There are legitimate reasons for recommending rejection
of a paper, but preferring your own ideas over those of the author being
reviewed is not one of them.

That would be great, Bill. Thanks.
When you read the paper, keep in mind that it was originally submitted
to the Journal of Experimental Psychology: Human Perception and Performance
(JEP:HPP) as a commentary on the following article that had just
been published in JEP:HPP:

McLoed, P., Reed, N. and Dienes, Z. (2001) Towards a unified fielder
theory: What we do not know about how people run to catch a ball. JEP:HPP,
120, 135-987

So my Fielder’s Choice paper was intended as a commentary on McLoed,
et al (2001). JEP:HPP does publish such commentaries as “Observations”.
I crafted my paper into the form of an Observation. It’s a bit shorter
than the usual journal paper and it introduces no new data. It’s
really just a theoretical observation. Therefore, I think my paper is most
useful if you read the McLoed et al paper first. I don’t have an
electronic copy of that paper, I’m afraid, but they probably have it in
the library.

The model behavior shown in my paper (in Figure 2) relates to data presented
by McLoed et al. For example, My figure 2a shows how a two control system
PCT model can produce running patterns like those shown in Figure 2 of
McLoed et al. And my Figure 2b shows that, will producing these running
patterns, the PCT model can produce “curved” optical trajectories like
those shown in Figure 6 of McLoed et al. In other words, my paper shows
that the PCT model can account for the catching behavior observed by McLoed,
et al – behavior that, according to McLoed et al, cannot be handled by
any existing model of catching (that’s why they say there is no “Unified
Fielder Theory”).

I submitted my paper to JEP:HPP in early 2002 and had it rejected in
April, 2002. I have electronic copies of the action letters. I will send
you (and anyone who wants them) the address of those letters once you post
your own criticisms of the paper. There were three reviews. Two recommended
rejection and one recommended acceptance. The editor (David Rosenbaum,
who has commented favorably on my work in the past) went with the majority.
I wrote a protest letter because I thought the reviews reflected “gatekeeping”,
but to no avail.

I’m sure the paper is not perfect and I welcome suggestions for improvement.
The third reviewer made some helpful suggestions, for example. But I’ll
let you decide , when you see he reviews, whether the paper was rejected
based on scientific merit or theoretical point of view.

After it was rejected by JEP:HPP I sent it to two other journals simply
because I had put what, for me, seemed like a lot of work into it and I
thought it was a pretty good paper. The reviews from these other
journals seem to me to reflect gatekeeping just as strongly as the
JEP:HPP reviews. But I don’t have electronic versions of those reviews
handy. And the paper really was written specifically for JEP so I don’t
think the rejections from the other journals matter much.

Fred Nickols (2003.28.0845 EST)
Where would you like to see it published?
As I said, I had wanted to publish it in JEP:HPP. Now I’m just going to
publish it in my next book (which will come out as soon as someone is nice
enough to buy up all remaining copies of Mind Reaings and More Mind
Readings; that’s what I would do with my dividend tax savings, if I were
rich;-) It would help the economy and the development of the life sciences).
However, I plan to write a more detailed theoretical paper on catching
(really, intercepting) baseballs because I continue to learn more about
how these baseball catching control works as I continue to play with my
programs. I might try to publish that paper in a theory oriented
psychology journal such as Psychological Review.

Best regards

Rick

···

–

Richard S. Marken, Ph.D.

Senior Behavioral Scientist

The RAND Corporation

PO Box 2138

1700 Main Street

Santa Monica, CA 90407-2138

Tel: 310-393-0411 x7971

Fax: 310-451-7018

E-mail: rmarken@rand.org

5

[From Fred Nickols (2003.28.0845 EST)] --

[From Rick Marken (2003.01.27.1400)]

The relative ease with which I got my "Control Theory Glasses" published had
lulled me into thinking that conventional psychology was getting more
receptive to
the idea that behavior is the control _of_ perception. But I have been
reawakened
from my dogmatic slumbers by the incredible difficulty I've had getting my
"Fielder's Choice" paper published.

To Rick:

Where would you like to see it published?

To All:

Where is it best published?

Fred Nickols
nickols@safe-t.net

6

[From Bill Powers (2003.01.28.1022 MST)]

Rick Marken (2003.01.28.0840)

Part of the problem may be that the other theorists believe that
prediction and trajectory planning are necessary to explain baseball
catching. That’s what most people seem to believe these days. If so, you
need to makew it plain that you are proposing an alternative to that
idea, a negative feedback control model that does not make use of
explicit predictions or planning.

Has anyone heard from Bill Williams? I’ve been corresponding with him
about an economic model, but haven’t heard anything from him for a couple
of weeks. I hope nothing’s wrong.

Best,

Bill P.

7

[From Rick Marken (2003.01.28.1350)]

Bill Powers (2003.01.28.1022 MST)
Part of the problem may be that the other theorists believe
that prediction and trajectory planning are necessary to explain baseball
catching. That’s what most people seem to believe these days. If so, you
need to makew it plain that you are proposing an alternative to that idea,
a negative feedback control model that does not make use of explicit predictions
or planning.

I think prediction and trajectory planning are examples of
what I called the “control of action” approach to explanation of baseball
catching. Perhaps making this clearer would have helped (it might have
helped Reviewer B in particular). But you (and anyone else who is interested)
can judge for yourself. The reviews are posted at:
http://www.mindreadings.com/action_letter_02-028.pdf

Best regards

Rick

···

–

Richard S. Marken, Ph.D.

Senior Behavioral Scientist

The RAND Corporation

PO Box 2138

1700 Main Street

Santa Monica, CA 90407-2138

Tel: 310-393-0411 x7971

Fax: 310-451-7018

E-mail: rmarken@rand.org

8

[From Rick Marken (2003.01.30.1200)]

Here is the abstract of a recent "Observation" published in JEP:HPP that was
written by the very people about whose article I had written my "Fielder's Choice"
observation. I think a PCTer would have no problem seeing what's wrong with this
observation, even without knowing much about the details of catching balls. I
don't have access to the article itself yet but I plan to respond to this
observation with another observation; what the heck. The funny thing is that both
OAC and LOT theory are wrong but no one seems to want to learn _why_ that's true.

···

---

The Optic Trajectory Is Not a Lot of Use if You Want to Catch the Ball

             Peter McLeod and Nick Reed
             Oxford University

             Zoltan Dienes
             University of Sussex

             According to linear optic trajectory (LOT) theory, fielders use the
direction of curvature of the
             optic trajectory to control the way they run to intercept the ball.
Data presented by D. M. Shaffer
             and M. K. McBeath (2002) as support for LOT theory show that the
optic trajectory of balls that
             will fall behind the fielder provide the cue that LOT theory predicts
would send the fielder
             running forward, not backward. In this article, the authors show that
watching these balls would
             provide the fielder with the cue that the optic acceleration
cancellation (OAC) theory of
             interception predicts would send the fielder running backward. It
appears that the fielders studied
             by Shaffer and McBeath were following the cue predicted by OAC
theory, not that predicted by
             LOT theory.

---

How do you explain to people who are using control theory and believe that they
understand control theory better than god that they don't understand control
theory? At least, they don't understand that control systems _control_
perceptions; they don't _use_ these perceptions to determine ("control") their
actions.

Best

Rick
--
Richard S. Marken, Ph.D.
Senior Behavioral Scientist
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

9

[From Bruce Gregory (2003.0130.2024)]

Rick Marken (2003.01.30.1200)

How do you explain to people who are using control theory and believe that they
understand control theory better than god that they don't understand control
theory? At least, they don't understand that control systems _control_
perceptions; they don't _use_ these perceptions to determine ("control") their
actions.

In his book "The g Factor: The Science of Mental Ability," Arthur Jensen
concludes that, "there are nearly as many definitions of "intelligence"
as there are experts." The situation with regard to "control" is not
quite that bad, but is clearly a term that invites confusion. I admit
that "controls perceptions" is a lot easier to say than "maintains
perceptions at reference levels." One advantage of the longer expression
is that it allows one to agree that organisms "control their actions"
and add that they seem to do so by maintaining certain perceptions at
reference levels.

10

[From Rick Marken (2003.01.31.0925)]

Bruce Gregory (2003.0130.2024)

I admit that "controls perceptions" is a lot easier to say than "maintains
perceptions at reference levels." One advantage of the longer expression
is that it allows one to agree that organisms "control their actions"
and add that they seem to do so by maintaining certain perceptions at
reference levels.

I think the confusion about control in the baseball catching literature may be more
than a matter of word choice. As I said in my "Fielder's Choice" paper, I think the
way McLeod et all talk about control reflects the way they conceive of the process
of control. In the Abstract I posted, McLeod et all talk about "using" optical
trajectory as a basis for selecting a path to the ball. Perhaps this is just an
unfortunate turn of phrase -- if anything, optical trajectory is _controlled_ by the
path taken to the ball -- but I won't be able to tell until I see their article. If
the article presents data showing, for example, that characteristics of the optic
trajectory (such as optical acceleration) are correlated with or can serve as cues
for path selection (as is suggested in the Abstract) then McLeod et al are clearly
thinking of control in S-R or information processing terms: the optical trajectory
provides the information that is the basis for movement. And, of course, that
would be wrong.

Best

Rick

···

--
Richard S. Marken, Ph.D.
Senior Behavioral Scientist
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

11

[From Rick Marken (2003.01.31.0950)]

I.Kurtzer (2003.01.31.0930)

After having read your comment I am not suprised it was rejected.
I liked it, but I don't think that explaining that they don't understand
control theory is the best approach or citing their ambiguities in
percentage.

You're probably right. Though none of the reviewers cited this as a reason for
rejection. Read the JEP reviews and see for yourself. The main reason for rejection
was that my model had already been published (in AJP). Apparently, advocates of OAC and

LOT can publish all they like but advocates of PCT can only publish once and then must
disappear.

I feel that explaining how your model
1. makes specific predictions that either cannot be accomodated by the
previous models, or
2 without additional assumptions that make it equivalent,
or how
3. it does the same job in a simpler way

Actually, I did that. I showed how my model accounts for the running patterns and
optical trajectories that McLeod et al said could not be handled by existing theories.

And particularly if you avoid harping on how they're
wrong or misguided or irrelevant prima facia.

McLeod et al said that current theories could not account for the data they obtained in

a catching experiment. I "observed" that they were wrong; a control model controlling
optical velocity and lateral angle can account for the data (at least qualitatively; to

account for it quantitatively I'd have to have the ball trajectories associated with
the optical trajectories and running patterns and I they didn't publish that data.

If they use control of perceptions language 60% of time then just use it
100% of time _without explanation_.
Basically, you just talking nuts and bolts and work from there.

Perhaps. But I think that there is a problem with the way control theorists conceive of

control (we saw this in the Jagacinski/Flach book on control theory). I think it's
important to try to explain to non-PCT control theorists and to their audience what
they are getting wrong. What they are getting wrong (at least a large proportion of the

time) is their concept of control. They see control as an input-output rather than a
control of input process. This influences the way they do research and inhibits their
ability to make significant contributions to our understanding of human behavior using
control theory.

I think that a more restricted version should be publishable in oodles of
journals:
Journal of motor behavior
motor control
perception
perception and psychophysics
experimental brain research
....

It was already rejected by JMB and Perception. It was rejected for the same reasons as
those given in the JEP reviews. I don't mind, though, because it was really only
appropriate as an Observation in JEP. And the entertainment value of some of the
reviews is sufficient pay back for the effort that went into writing the paper (and
doing the modeling).

and hope you give it another shot.

With a different paper, perhaps.

Best

Rick

···

--
Richard S. Marken, Ph.D.
Senior Behavioral Scientist
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

12

I.Kurtzer (2003.01.31.0930)

After having read your comment I am not suprised it was rejected.
I liked it, but I don't think that explaining that they don't understand
control theory is the best approach or citing their ambiguities in
percentage.
I feel that explaining how your model
1. makes specific predictions that either cannot be accomodated by the
previous models, or
2 without additional assumptions that make it equivalent,
or how
3. it does the same job in a simpler way

would be much better.
And particularly if you avoid harping on how they're
wrong or misguided or irrelevant prima facia.
If they use control of perceptions language 60% of time then just use it
100% of time _without explanation_.
Basically, you just talking nuts and bolts and work from there.
I think that a more restricted version should be publishable in oodles of
journals:
Journal of motor behavior
motor control
perception
perception and psychophysics
experimental brain research
....
and hope you give it another shot.

Isaac

, > [From Rick Marken (2003.01.30.1200)] >

···

Here is the abstract of a recent "Observation" published in JEP:HPP that was
written by the very people about whose article I had written my "Fielder's Choice"
observation. I think a PCTer would have no problem seeing what's wrong with this
observation, even without knowing much about the details of catching balls. I
don't have access to the article itself yet but I plan to respond to this
observation with another observation; what the heck. The funny thing is that both
OAC and LOT theory are wrong but no one seems to want to learn _why_ that's true.

---

The Optic Trajectory Is Not a Lot of Use if You Want to Catch the Ball

             Peter McLeod and Nick Reed
             Oxford University

             Zoltan Dienes
             University of Sussex

             According to linear optic trajectory (LOT) theory, fielders use the
direction of curvature of the
             optic trajectory to control the way they run to intercept the ball.
Data presented by D. M. Shaffer
             and M. K. McBeath (2002) as support for LOT theory show that the
optic trajectory of balls that
             will fall behind the fielder provide the cue that LOT theory predicts
would send the fielder
             running forward, not backward. In this article, the authors show that
watching these balls would
             provide the fielder with the cue that the optic acceleration
cancellation (OAC) theory of
             interception predicts would send the fielder running backward. It
appears that the fielders studied
             by Shaffer and McBeath were following the cue predicted by OAC
theory, not that predicted by
             LOT theory.

---

How do you explain to people who are using control theory and believe that they
understand control theory better than god that they don't understand control
theory? At least, they don't understand that control systems _control_
perceptions; they don't _use_ these perceptions to determine ("control") their
actions.

Best

Rick
--
Richard S. Marken, Ph.D.
Senior Behavioral Scientist
The RAND Corporation
PO Box 2138
1700 Main Street
Santa Monica, CA 90407-2138
Tel: 310-393-0411 x7971
Fax: 310-451-7018
E-mail: rmarken@rand.org

13

[From Bruce Nevin (2003.0131.1153 EDT)]

Bruce Gregory (2003.0130.2024)--

... "controls perceptions" is a lot easier to say than "maintains
perceptions at reference levels." One advantage of the longer expression
is that it allows one to agree that organisms "control their actions"
and add that they seem to do so by maintaining certain perceptions at
reference levels.

I.Kurtzer (2003.01.31.0930)--

···

At 08:25 PM 1/30/2003, Bruce Gregory wrote:
At 10:04 AM 1/31/2003, Marc Kurtzer wrote:

If they use control of perceptions language 60% of time then just use it
100% of time _without explanation_.
Basically, you just talking nuts and bolts and work from there.

There is considerable wisdom in this.

         /Bruce Nevin