Whatdya think about computers playing football?

#1
hmanvolfan

hmanvolfan

Volmeister extraordinaire
Joined
Oct 23, 2004
Messages
103,068
Likes
59,083
#1
Here is an ESPN site that has supposedly played every game on every teams schedule and the computers results. Click on the Bulldog where it says 2006 Season Preview. Should be food for thought. By the way check out Memphis's prediction.

http://www.collegefootballpoll.com/
 
#2
#2
(hmanvolfan @ Aug 10 said:
Here is an ESPN site that has supposedly played every game on every teams schedule and the computers results. Click on the Bulldog where it says 2006 Season Preview. Should be food for thought. By the way check out Memphis's prediction.

http://www.collegefootballpoll.com/
Click to expand...

At some point in time it might be relevant. But I dont think there is enough accurate input to develop a computer model to predict the outcome of a game. Computer programs are only as good as the information you put into them.
 
#4
#4
(oklavol @ Aug 10 said:
At some point in time it might be relevant. But I dont think there is enough accurate input to develop a computer model to predict the outcome of a game. Computer programs are only as good as the information you put into them.
Click to expand...
Of course I agree, but like I said it's food for thought. Why would anybody spend a bunch of time loading all this data into a computer and let it play all these games before the season even starts? Or is this what we've been doing all summer? Anyway, ESPN sucks!!! :moon2:
 
#5
#5
This is retarded. They have at least 4, maybe 5 teams going undefeated. That may have NEVER happened. They've only guessed the Vols record correctly once.
 
#10
#10
Well, they've never been good enough to take me at Atari, Tecmo, Blitz, EA College, or Madden football. Why should I bet on Vandy over South Carolina?
 
#11
#11
It is correct to say that the program used to estimate the game outcomes is only as good as the input. With no games played this season yet, there is no way they can input data on the current team. This 5-7 record is based upon statistical data gathered from last season.... basically, they are giving statistical results based upon the 2005 Vols and all the other 2005 teams playing a 2006 schedule.

Based upon the previous prognostications, whenever the computer predicts a 10+ win season for the Vols, the Vols fell short 5-out-of-6 years. However, when the computer selects less than 10 wins, the Vols exceeded expectations in 5-of-6 seasons, winning 10+ games in 4 of those seasons.

The better way to interpret their selections is to compare the actual results from the previous season to the preseason projections, ie the computer projected 11-0 for 1999 and the Vols were actually 11-0 in 1998.

YEAR Proj. ---- Previous(actual)
1994 10-1 ---- 10-1
1995 9-2 ---- 7-4
1996 10-1 ---- 10-1
1997 10-1 ---- 9-2
1998 9-2 ---- 10-1
1999 11-0 ---- 11-0
2000 9-2 ---- 9-2
2001 9-2 ---- 8-3
2002 10-2 ---- 10-1
2003 8-4 ---- 8-4
2004 8-3 ---- 10-2
2005 10-1 ---- 9-2
2006 5-7 ---- 5-6

Basically, the computer program is good at estimating a previous year's team against the following year's schedule.
 
#12
#12
(allvol @ Aug 11 said:
YEAR Proj. ---- Previous(actual)
1994 10-1 ---- 10-1
1995 9-2 ---- 7-4
1996 10-1 ---- 10-1
1997 10-1 ---- 9-2
1998 9-2 ---- 10-1
1999 11-0 ---- 11-0
2000 9-2 ---- 9-2
2001 9-2 ---- 8-3
2002 10-2 ---- 10-1
2003 8-4 ---- 8-4
2004 8-3 ---- 10-2
2005 10-1 ---- 9-2
2006 5-7 ---- 5-6

Basically, the computer program is good at estimating a previous year's team against the following year's schedule.
Click to expand...

That doesn't exactly appear to be "rocket science," given that our schedule usually varies by AT MOST 1-2 "quality" games per year.
 
#14
#14
They have Memphis going 11-1. :lol:

Seriously though... this looks good for the Vols if you look at what allvol has stated... they predict us to win less than 10 games, 4/5 chances are we'll win MORE than 10. :biggrin2:
 
#15
#15
As Kenny Mayne used to say, "That's why the games aren't played on paper (or comps)."

Memphis 11-1 is a little off after losing DW.
 
#17
#17
They are only as good as the people entering the variables and the data. In this case both suck at their jobs.
 
New posts
You must log in or register to reply here.

VN Store



Back
Top