Accuscore: Miami-GT

cajunjacket

cajunjacket

Dodd-Like
Joined
Dec 5, 2007
Messages
11,704
Accuscore has the final score being 23 to 23.5, with Miami winning. I know that a team can't score 23.5, but it's a computer simulation that runs averages. With the game simulated over 10,000 times, it has Miami winning 50.8% of the time and GT winning 49.1% of the time.

Just thought I would throw this out there. There are a lot more stats on there: rushing yards, passing yards, etc. You can get a free membership (I just signed up, takes 30 seconds) and can see the analysis. It's: accuscore.com

What does this mean? Statistically, it should be a close one.
 
And that Roddy Jones had a hard cast last week, and will have that off for this week? Most prediction things I've seen don't account for players that were hurt returning, or players getting hurt.
 
Something positive:

In the simulation, they estimated that we would pretty much be shut down on the ground:

Top 3 Rushers
Dwyer - 81.5 yds
Jones - 43.2 yds
Nesbitt - 33.7 yds

Then for passing they have a good estimate:
Nesbitt - 89.6 yds

With this low rushing simulation, they still have us only losing by 0.5 points. Computers are stupid. I just posted this for the fun of it.
 
Very interesting. Thanks for the info.

I do find it odd that it is averaging only 33.7 rushing yards for Josh when he's had 93 yards and 91 yards on the ground against Jacksonville State and Clemson respectively.
 
I don't understand why they run 10,000 simulations. They should only need to run one.
 
GTKyle said:
I don't understand why they run 10,000 simulations. They should only need to run one.
Click to expand...
...You just made me question the validity of the education I received at Tech. Thanks Kyle.

Does this only take this season's results into account?
 
cajunjacket said:
Something positive:

In the simulation, they estimated that we would pretty much be shut down on the ground:

Top 3 Rushers
Dwyer - 81.5 yds
Jones - 43.2 yds
Nesbitt - 33.7 yds

Then for passing they have a good estimate:
Nesbitt - 89.6 yds

With this low rushing simulation, they still have us only losing by 0.5 points. Computers are stupid. I just posted this for the fun of it.
Click to expand...

If they have us winning almost half the time with these numbers, I feel pretty good about Thursday night.
 
Allen Koholic said:
Does this only take this season's results into account?
Click to expand...

Great question. I don't really know. But I would have to assume it takes last season's numbers into account otherwise, Anthony Allen should have been one of the top rushers in the game based on averages.
 
Allen Koholic said:
...You just made me question the validity of the education I received at Tech. Thanks Kyle.

Does this only take this season's results into account?
Click to expand...

If it works anything like Kyle's coach ranking model, it must be super accurate :rolleyes:
 
cajunjacket said:
Something positive:

In the simulation, they estimated that we would pretty much be shut down on the ground:

Top 3 Rushers
Dwyer - 81.5 yds
Jones - 43.2 yds
Nesbitt - 33.7 yds

Then for passing they have a good estimate:
Nesbitt - 89.6 yds

With this low rushing simulation, they still have us only losing by 0.5 points. Computers are stupid. I just posted this for the fun of it.
Click to expand...

I like this poll better.

http://www.solecismic.com/frontier/cfa.php
 
GTKyle said:
I don't understand why they run 10,000 simulations. They should only need to run one.
Click to expand...

Because real life is random and has a lot of risk. While I'm sure you think you are absolutely correct on every last thing under the sun, there's a inherent randomness in everything.

I'm sure they used a Monte Carlo simulation on a number of distributions of variables. Ideally you would combine all the distributions into a single mathematical distribution of the forecasted variable (i.e. points scored). However, that's very difficult/mathematically impossible a lot of the time and the Monte Carlo simulation does a bunch of different random scenarios instead, seeing the resulting distribution.
 
You must log in or register to reply here.
Back
Top