Re-running single player seasons

[Stevebsfan]

I see a lot of people here who re-run single player season to build up data. How is this done?

What I'd like to do is create a roster based loosely on my MP roster, and be able to take that and re-run the same season over and over so I can learn the ins and outs of gameplanning. What is the best way to do this?

And it really doesn't have to be based on my MP roster. My goal is to learn the gameplay mechanics because it's becoming quite clear that I suck in that dept.

Thanks for any help!

[QuikSand]

You might find this helpful:
Excel-Based QuikTest Template - Front Office Football Central (http://www.operationsports.com/fofc/showthread.php?t=60105)

[QuikSand]

Interesting observation from that old thread:

A generic comment -- not really related to the gameplan being tested here, but to the nature of this sort of testing.


Even when I have run TEN SEASONS of the exact same gameplan, with the exact same players, against the exact same opposition -- there still exists wild variation. If you are not yet swayed by people like me arguing about things like "sample size" -- maybe you will start to see what the hell I'm talking about with some of the stats here.

--The Rex team had a record that varied from 9-7 to 14-2, running precisely the same season and the exact same setup, with no injuries to anyone. PLEASE think about this.

--On the Rex team, the exact same QB with the exact same opposition and surrounding cast -- he posted a season-long passer rating that varied from 79 up to 111. Read that again. SEVENTY NINE. ONE HUNDRED ELEVEN.

--Even a global, teamwide stat can very wildly. Look at that Rex team's Yards per carry allowed per rush. It varies from a very impressive 3.54 to a very unimpressive 4.44 -- over an entire season. Again, same exact players, same exact gameplans.

And so forth and so on... the example just from this tiny data set are staggeringly wide.



Next time you try to talk about how much you are "learning" from your players because of the stats they put up in the last few games... or from the stats that they put up on the preseason... or that sort of thing. Please keep this sort of things in mind. This is a complex system. There are a lot of variables involved. Even holding LOTS of them constant, and running WAY more trials... I'm still seeing MASSIVE variation in results.

When your backup running back posts 21 carries for 148 yards and 2 TD in his limited share of four preseason games, I still believe that the proper and reasonable conclusion to be drawn from these stats is: absolutely nothing. If you want to believe that you're really learning something there, then go on doing so. Butt if there were a technical glitch, and the preseason had to be re-run, be prepared to see the exact same guy, in the exact same trivial sample size, post something like 26 carries for 38 yards and a lot of resulting punts. Or any number of results in between.



Indeed... the biggest question this raises is whether we can even learn anything from a sample of 160 games? If I were really serious about this, maybe I'd expand the testing samples to more like 30 seasons (easy enough to do with this Excel tool, incidentally).

Just something to chew on. I realize I'm most likely tilting at windmills here, but it's a pretty important part of dealing with numbers.

[Stevebsfan]

That is pretty interesting, and makes me wonder how much gameplanning actually matters versus just putting out a typical close-to-rex gameplanning and hoping your players ratings allow for the better 'coin flip' on plays.

Although, QB rating and wins/losses are not really good indicators of anything. Yes it varies, but that's coin flip stuff. I do believe there are some things you can pull out of running the same season over and over. It's worth taking a look because at this point I'm just stumped in this game.

[QuikSand]

Gameplanning matters a ton.

Looking at a few games, one full season, or even ten, is just a suspect way to try to prove it.

[Dutch]

If you run a GP 10 times and it ranges from 8-8 to 16-0 and you run another GP 10 times at it ranges from 0-8 to 8-8 then at least that gives you a clue.

[Joker]

Maybe its not the game plan
Maybe its your kicker.

My kicker tried to hang himself.
He went to kick the chair out.
He missed wide left!

[aston217]

--On the Rex team, the exact same QB with the exact same opposition and surrounding cast -- he posted a season-long passer rating that varied from 79 up to 111. Read that again. SEVENTY NINE. ONE HUNDRED ELEVEN.


This makes me feel a lot better.

My star QB on my 100 roster rating team in the OSFL went from 117.5 QBR to 86.5. A differential of 31, which is almost as much as the 32 shown here.