Too much of one thing?

[QuikSand]

I think there has been some exploration on this matter before -- I'm just fishing for good input.

I recently launched an experimental FOF 2004 career, where I thought it might be interesting to really focus the team on things that I rarely do in my regular FOF careers. I decided to try an offense of all screen passing, and a defense with all DB blitzing -- mostly just for kicks. I built a team with the requisite skills (I think) and got my first season up and running.

I expected my offense to be mediocre, since every single play was designed to be a screen pass -- but thought I'd probably see my QB end up with a pretty high competion %, though a pretty low yards per attempt figure. Instead, I was surprised to see the competion percentage abysmally low, in the 30% range.

I can't reach any conclusion other than the AI defenses "figured out" this team, since we basically did teh same thing every play. I made a few adjustments -- splitting the passes between screens and 0-4 yards short passes... without much improvement in the results (though a bit). But it certainly seems that the opposing teams were gearing up to stop the short pass, and were doing so very effectively. (In a routine game I'm used to, with balanced offenses, we probably complete 75% or more of our screen passes)


So... there's some basic evidence that if you become TOO vanilla with your gameplanning, you're in big trouble.

(I seem to recall some work done a while ago with a team running basically all its running plays into the same spot on the OL, with fairly modest deleterious effects... suggesting that defenses didn't do a great job adjusting to the scheme... this is a different result, apparently.)


Anyone have more to add to the idea of a one-dimensional approach failing (or succeeding) against the computer opponent teams?

[albionmoonlight]

Was your completion percentage better for the first game than for other games? Or for the first half of the first game? Or the first quarter? I imagine that it would not take long for the defense to figure out what you were doing, but if it were a matter of "figuring out," then it would seem that you would be more successful at the beginning of your quest.

The game, of course, could mimic a "figuring out" effect by having some feature that says IF screen pass % > X THEN decrease effectiveness of screen pass. That would have the effect of preventing people from unrealistically calling the same play all of the time, but would seem a lot easier to implement than a dynamic AI that tried to read your tendencies. It would also apply from the first screen pass that you threw.

I know that Jeffrey Boggs' IHOF sacks really diminished when the new patch was implemented to limit the effectiveness of blitzes from one position. And it did not seem to depend on the quarter or my game plan from the previous week--leading me to think that it was something in the game engine that simulates the defense figuring it out, without having the defense actually figure it out.

[albionmoonlight]

dola--

For what little it is worth, I ran an experimental career with FOF2 where I easter egged to get a maxed out team. Then I ran the ball 90% of the time and set passes to 100% long.

In that case, more than any other career that I tried with a maxed out team, I had the best YPC in my running game that I ever had. It seemed that the fact that I threw the ball down the field when I threw it really kept the defense from playing the run--even though I ran the ball 90% of the time. Very anticdotal notes from an old version of the game, but there it is.

[QuikSand]

Was your completion percentage better for the first game than for other games? Or for the first half of the first game? Or the first quarter? I imagine that it would not take long for the defense to figure out what you were doing, but if it were a matter of "figuring out," then it would seem that you would be more successful at the beginning of your quest.

Good question.

My result were awful from the very start. Preseason was a disaster, and from the very beginning the team was terrible in the regular season. In the very first game, we completed maybe a third of our passes for trivial yardage (under 2 yards per attempt, in total).

[Buzzbee]

QS - How many formations were you running? I would imagine that a QB that knows a variety of formations would be critical in this experiment. Running screens from 2 or 3 formations could very quickly become repetitive. Running screens from 8 or 10 formations might effectively 'disguise' it from the AI to a greater extent.

Also, I'm curious if the formation affects the success (or lack thereof) of the screen. Do 3 and 4 WR sets work better since the blockers are already out there? Or are they worse because WR's aren't as good at blocking as tight ends or tackles and such?

Just some food for thought.

[Buzzbee]

Dola - Also, I would imagine that 'getting downfield' would be an important rating for a screen catching RB, in addition to the receiving skills.

[Buzzbee]

Double Dola - Man, so many questions. Also, how were the sack totals? I would also assume that you would get sacked more if you throw a lot of screens. Perhaps the screens were more successful than it appears, but the sacks are pulling down the YPA.

EDIT: Okay, figured I'd edit rather than go for the triple dola. I think I'm pretty sure it was SkyDog who did the experiment involving running 100% over the same spot in the OL. As I recall, the YPC were somewhere around or just below 3.0 YPC when hitting the same hole, and around or above 4.0 YPC when spreading it around.

[QuikSand]

QS - How many formations were you running? I would imagine that a QB that knows a variety of formations would be critical in this experiment. Running screens from 2 or 3 formations could very quickly become repetitive. Running screens from 8 or 10 formations might effectively 'disguise' it from the AI to a greater extent.

Also, I'm curious if the formation affects the success (or lack thereof) of the screen. Do 3 and 4 WR sets work better since the blockers are already out there? Or are they worse because WR's aren't as good at blocking as tight ends or tackles and such?

I was running the five formations that included two tight ends -- and both QBs who saw action knew all five of them.

[QuikSand]

Dola - Also, I would imagine that 'getting downfield' would be an important rating for a screen catching RB, in addition to the receiving skills.

Hmm... I have the opposite impression of that rating, it seems.

[Buzzbee]

Hmm... I have the opposite impression of that rating, it seems.
I guess I was thinking that screen passes are thrown at or behind the line and that YAC would be where the real gains were. It was my impression that 'getting downfield' correlates to yardage after the catch. Thus, it seemed to me like that might be an important rating for a screen pass.

[jbmagic]

what the qb ratings for screen passes?

[Buzzbee]

I recently launched an experimental FOF 2004 career, where I thought it might be interesting to really focus the team on things that I rarely do in my regular FOF careers. I decided to try an offense of all screen passing, and a defense with all DB blitzing -- mostly just for kicks. I built a team with the requisite skills (I think) and got my first season up and running.

I'm thinking his QB's screen passing skills are pretty good.

[Cringer]

I was running the five formations that included two tight ends -- and both QBs who saw action knew all five of them.

I think this could work against you, if those are the only 5 formations you used. Screens can obviously be run for many formations, why not go with more?

The way I see it, if you are always using double tight ends, you are always bringing he defense up and in tight against your offense. And then you choose to throw short when everyone is right there already. Seems like you are working against yourself somewhat.

If you had a trips formation, and those guys are running routes they are taking defenders away from the RB who is going to run the screen. With the two tightends, even if both go out on patterns you still leave a large part of the defense bunched up right around the line of scrimmage where you are throwing the ball.

[Vinatieri for Prez]

Hmm... I have the opposite impression of that rating, it seems.

I used to think it was the deep ball too, but it is in fact YAC ability. After playing the game for years and thinking my burners were the "getting downfield guys" it turns out it is the "big play" rating that determines this after I accidentally stumbled upon in it a scouting report narrative.

See my earlier thread:

http://dynamic.gamespy.com/~fof/forums/showthread.php?t=40521

[QuikSand]

I used to think it was the deep ball too, but it is in fact YAC ability.http://dynamic.gamespy.com/~fof/forums/showthread.php?t=40521

Thanks - I do tend to confuse some of those ratings, and this is a persistent one with me. I think I have even been corrected on this before, but continue to think "getting downfield" means "getting open downfield" rather than "getting downfield after catching a ball." Thanks for re-setting me straight.

[gstelmack]

It would seem like a good way to test this would be to call your own plays and see just how quickly you start getting "The Defense looked very familiar with that play". Having checked game logs, I know that if I go run heavy or pass heavy or lean heavily towards short or long passes, that comes up a lot more often in the latter stages of a game.

[Celeval]

What were the defensive formations looking like? By the fourth quarter, was the opposing team running Cover-2 (or B&R, or M2M) constantly?

[QuikSand]

I neglected to leave on the actual game logs for those games... but if it's of value, I can turn that function on ane run a game or two for experimental purposes. I have dumpd that career as an exercise, but I still have the saved game to use if it helps.

I'll post a full game log later today...