From: Deeppe on 7 Jul 2010 22:07
On Jul 7, 6:42 pm, HASM <netn...(a)invalid.com> wrote:
> Evan Kirshenbaum <kirshenb...(a)hpl.hp.com> writes:
> >> it's starts to be a bit doubtful that they would help awarding a
> >> disallowed goal as the defenders could claim they saw the flag go up and
> >> stopped playing
> > ODF. You play to my whistle. I might wave it off in any case.
> I know that, but that won't stop the players and people on rss from
> complaining, that was my point.
> > I've been assuming that it's someone going frame by frame, identifying
> > when the ball is touched,
> Yes, but even viewing frame by frame sometimes it's hard to tell when the
> ball was touched. It is easy when it changes directions sharply, not so
> easy otherwise.
And what's the point of this level of exactitude again? People need to
re-learn to deal with certain levels of uncertainty. Today's tech
gives too many the illusion they have control where they do not. It's
a good reminder for the controlling type amongst us.
US baseball fans have it all over football fans in this regard. One
hears the occasional cry for a hi-tech strike zone, but nothing that's
taken at all seriously by any other than a fringe group. Baseball fans
know all about uncertainty and humanity. Maybe football fans can learn
as well? A dose of humility helps.
A nice baseball forum for fellow fans:
Goal in/out? Eh, okay. A big rasberry on anything more.
From: JCQ on 7 Jul 2010 22:15
On Jul 7, 9:53 pm, HASM <netn...(a)invalid.com> wrote:
> JCQ <zelig9...(a)gmail.com> writes:
> > I'm not talking about people watching a replay on TV. I'm talking about
> > computers making the calls instantly.
> I think we're a few years away for such technology to work reliably. And
> then you still need to get enough viewing angles to account for all
> player's body parts that count, that can be obscured by body parts that
> don't count, the ball, other players, etc.
> -- HASM
The computers can figure out different angles by using just one
camera. The program does the rest and this has been shown on some
replays used by ESPN and others. With multiple camera inputs I'm sure
they could be over 90% accurate on close calls which is a much higher
percentage than with the human eye. I agree that it should be tested.
FIFA could easily do this if they wanted and at very little cost. The
problem is that they are dinosaurs at FIFA. Afraid of any innovation.
I'm suspecting that it's because they have been able to control
results for years this way. The refs and linesman are afraid of
getting calls wrong against the big teams and this influences the
results of the matches. FIFA loves when the big teams move on through
the tournament. In this world cup it hasn't been a huge issue but it
has been in the past and will be in the future unless technology takes
From: HASM on 7 Jul 2010 23:41
JCQ <zelig9999(a)gmail.com> writes:
> The computers can figure out different angles by using just one camera.
I've worked with computers and computer vision for over 30 years. Figuring
out different angles with just one camera is going to be mighty difficult.
> With multiple camera inputs I'm sure they could be over 90% accurate on
> close calls which is a much higher percentage than with the human eye.
Well, 90% accurate is not enough, and hardly worth it.
From: HASM on 7 Jul 2010 23:44
Deeppe <tutall(a)hotmail.com> writes:
> And what's the point of this level of exactitude again?
No point at all. This was just an aside from a previous comment.
I have no trouble with the +/- a yard of half yard or so referees can call
these things with, more than 90% of the time. Within a yard, referees will
get 50% right, 50% wrong, and let's move on.
From: Mart van de Wege on 8 Jul 2010 04:08
Evan Kirshenbaum <kirshenbaum(a)hpl.hp.com> writes:
> For those who haven't done it, keep in mind that it's absolutely
> critical that the AR (1) be lined up *precisely* with the (moving)
> offside line, (2) be looking down a line precisely parallel to the
> goal line, and (3) be looking directly at the ball to tell when it was
> last played by a teammate.
> (2) and (3) are, of course, impossible to do at the same time, and (1)
> and (3) are impossible to do at the same time if the defender is
> moving (even if it's just his upper body that's moving). And if
> you're willing to trust your ears as to when (3) happened, you have to
> take into account that the sound takes a tenth of a second to go 37
> yards, while what you're looking at is instantaneous. So you have to
> listen and decide "Was he offside a tenth of a second ago?"
I've been wondering if the following wouldn't work better:
1. The AR stays on the offside line, and concentrates on the players
2. The AR always raises the flag when he sees an offside, no matter what
the ball is doing nor the position of the offside player.
3. The referee is already watching the ball. As soon as it is passed
forward, he glances at the AR to see if there's a flag up. If yes,
then he calls for offside if he decides there's a player profiting
from that position.
"We will need a longer wall when the revolution comes."
--- AJS, quoting an uncertain source.