Warrior General Games

Clubber "Pain" Lang - http://goallineblitz.com/game/player.pl?player_id=631916

I just got a new FFUM % AEQ piece for my LB and it's making a huge difference.

His last game - (4 Tkl, 3 FFum) (First game with the new piece)

He's at 40% FFUM chance and he has Power Tackler at 15.

40% AEQ + 45% VA(Power Tackler VA is 3% FFUM x 15 = 45%) = 85% ?

So that means he has a 85% chance to FFUM on every tackle? o.O

That can't be right... can it?

Can someone explain the FFUM ratio/formula break down for me?

I think it's just a percent added on to the roll for forcing a fumble. Something like, if you have a 20 percent chance, it'll be 85% of 20 percent added onto the 20. So in that case, your new percentage would be 37%

It is not a blanketed 85%, it is more of a 85% increase to whatever your chances where to cause it. So if you had an 10% chance in the first place, you now have an 18.5% chance. Similarly, if you had a 50%, you now have about a 85% chance. I am pretty sure this is clear and accurate explanation. Also all math was done in my head, so it may be off a little.

Cool, I was thinking along those lines.

Is there a way I can see my base percentage? I want to get an exact #

Nope, I don't think anyone has access to that formula except for Bort.

Come on Bort! Hook a brotha up!

Originally posted by Robadesigns
Cool, I was thinking along those lines.

Is there a way I can see my base percentage? I want to get an exact #

Base percentage would depend on the other guys stats roll.

Meaning some guys you would have a 10% base others 20% it depends on him I'm pretty sure.

were your tackling tactics set for balanced or power?

Originally posted by Robadesigns
Cool, I was thinking along those lines.

Is there a way I can see my base percentage? I want to get an exact #

Impossible. GLB is generally based on Opposed Rolls. It is going to be your FF chance (which includes your attributes, AEQ bonuses, VAs, your velocity at the time, your tackle quality roll, etc) vs. the other player's chance to not fumble (attributes, AEQ bonuses, off-balance percent, etc.) It'll be different every time.

actually you could come up with a pretty good formula. What you'd need to do is do some sort of regression analysis to come up with bell curves (imo) that represent the average players and use those to figure out what the bonus would correspond to. say if the average FF roll was a 25 with a standard deviation of 4 and the average not fumble roll was 40 with a standard deviation of 5, you could calculate the average expected number of fumbles. then apply the bonus to the FF roll, recalculate the expected # of fumbles, divide by the norm, and I suppose you'd know what the bonus corresponded to.

Actually if you could do this in reverse, holding for constants (which you couldn't do) you could just about reverse engineer the bell curves (assuming they are a normal distribution).

Originally posted by Darren McFadden
actually you could come up with a pretty good formula. What you'd need to do is do some sort of regression analysis to come up with bell curves (imo) that represent the average players and use those to figure out what the bonus would correspond to. say if the average FF roll was a 25 with a standard deviation of 4 and the average not fumble roll was 40 with a standard deviation of 5, you could calculate the average expected number of fumbles. then apply the bonus to the FF roll, recalculate the expected # of fumbles, divide by the norm, and I suppose you'd know what the bonus corresponded to.

Actually if you could do this in reverse, holding for constants (which you couldn't do) you could just about reverse engineer the bell curves (assuming they are a normal distribution).

As I'm sure you're aware, the problem with trying to use regression analysis is that we don't have access to the data we need to account for all of the possible variables. Ball carrier's balance, in-game breath and in-game morale, as examples. Sure, if we had all player's builds, tactics, and extracted speed from the replays using a technique similar to DB's, we could get something rough but I suspect that the other variables play so much of a roll that we'd find significant noise.