The difficulty of evaluating "EA"
Nice to meet you, this is Doctor Neko.
Recently, developers have been
evaluating (recommending) other developers' new EAs,
which is something I feel happens quite often. (^^;)
(It seems like this didn’t happen much in the past,
has the era changed significantly recently...)
By the way, on my end,
I don’t usually evaluate other developers’ EAs,
and I feel a bit sorry about that,
but there is a substantial reason for this,
[Judging the future potential of an EA is actually very difficult for developers as well]
that is the reason.
What this means is,
even developers
cannot easily judge the future potential of “other people’s EAs (EA you didn’t develop yourself),”
to be honest, it’s very hard to read.
If it’s an EA that I created,
I understand the internal logic,
the development process, the degree of curve fitting, etc.,
in detail,
so to some extent I can predict the characteristics, strengths, and future prospects of each EA,
but when it’s “another developer’s EA,”
it becomes, in a sense,
the same level as the EA buyer’s information (i.e., what’s on the product page),
so
reading that an EA has future potential is quite difficult.
By the way, I think it’s possible to make some negative readings like,
[this EA seems a bit risky...],
to a certain extent.On this point,
I think many EA buyers also feel, “this EA is a bit...,”
・“An EA that will win in the future is hard to read,”
・“An EA that doesn’t seem to win in the future is relatively easier to read.”
That is probably the reality.
In this sense,
if I were to evaluate others’ EAs,
my assessment would probably be, “This EA probably won’t lose badly...?”
I think.
Also, if an EA has maintained results over a very long forward period,
then, using the forward results as a yardstick,
it would be possible to evaluate it as a good EA, of course.
Creating an EA that can win is quite difficult, but
equally, or even more so, predicting future success of an EA is an extremely challenging task for anyone.
I think this is something users should keep in mind as well,
and it might be better to keep that in mind a little:
especially when the forward period is short and
for some reason the surrounding people’s evaluations are unusually high,
or it’s heavily recommended,
in many cases, that means the recommendation is based on something other than the EA’s true ability,
so,
don’t be swayed too much by such voices, and
consider the developer’s past EA track record and credibility,
and,
evaluate the EA with your own eyes,
and develop a habit of doing so,to achieve better results in the future.
Doctor Neko
