WOULD YOU BE BETTER OFF PICKING YOUR PLAYS RANDOMLY?
Smart Football is full of all kinds of brilliance in the strategy-heavy linkdump, but the best bit is at the end discussing whether you'd be better off choosing your plays at random. We have an answer, and it is "It depends." Is Gary Crowton your offensive coordinator? If so, then yes. Yes you are better off going random.
almost 2 years ago
Spencer Hall
21 comments
0 recs |
Email
PrintComments
Charlie Weis has been doing this for years.
Charlie: OK, Jimmy, Heads it’s screen pass, Tails it’s draw play.
Jimmy: What if it lands on its edge?
Charlie: Wildcat.
Brian Kelly says no Spicy Sea Nuggets.
by Ancient Chinese Secret on Jun 21, 2010 12:36 PM EDT reply actions
Game Theory Holla!
Calling all plays randomly is a hopeless strategy.
Calling some plays randomly can benefit you, particularly if you’re up against a coach who’s thinking accurately a level above you. The optimal ratio of nonrandom:random plays is highly contextual, by theoretically one does exist. (And sometimes that ratio means no random plays at all).
Most important and useful is restricting your total available range of plays and adjustments to a handful then picking randomly from that set. This can be extremely valuable if the opposing coach always seems to know what play you’re calling next, but you don’t want to risk an idiot playcall (e.g., an off-tackle run on a 3rd and long with an unfavorable field position).
In a game like poker these strategies can be a great asset to winning high-stakes games. In football , it’s much less valuable. No matter how much we’d like it to be otherwise, playcalling is a greatly overrated factor in football success.
In any case, few coaches have any ideas what the other coach is doing or even what they’re trying to do. Few coaches will say “omg a fire zone blitz is coming up.” Instead, they’ll say, “they keep shifting the safety over here when we line up in this alignment, so we’ll line up in the alignment but run a play that hits the vacated part of the field to punish them.”
The new year approaching, click in. Let’s facelift bar! Open the wardrobe is not yet found love after another the right clothes? So, also waiting for? Immediate action bar!
You lost me in the 2nd word of your post
Too much thinking….
Don't give up, don't ever give up ~ Jim Valvano
" but you don’t want to risk an idiot playcall (e.g., an off-tackle run on a 3rd and long with an unfavorable field position)."
UGA begs to differ.
"It’s not Disneyland, people. Get the hell out of the way." NYC Firefighter
by jokastrength on Jun 21, 2010 12:56 PM EDT up reply actions
Charlie Weis saw your set-up there
and audibled to a screen pass.
"...when the devil says to you: do not drink, answer him: I will drink, and right freely, just because you tell me not to."
— Martin Luther
by Go Big Rev on Jun 21, 2010 2:58 PM EDT up reply actions 1 recs
right- but the central tenet has to be that play calls independent events- they're not, and now we're dealing with the chief problem with Nash equilibrium and the central notions of game theory
I say cluster the playbook based on their situational probability cone position- the cone itself will be highly subjective from team to team, and then randomize within the table those situational probabilities produce. As I posited on the smart football board, you could theoretically improve your probability of winning by randomizing certain aspects, if you limit your exposure to bad play calls you can maximize the benefit of randomizing play calls (defensive confusion). Leveraged playcalling, I’ll call it.
"Voetbal is pas totaal als je wint"- Coach Adun
"The greatest sin is to spurn the gift"- Coach Alistair
We could start a statistically-oriented Davidson football blog
We could call it “Excruciating Analysis for the benefit of John Kilgo”, since he’d be the only one who would read it
(Incidentally, he was our AD when my dad played football in the mid-60’s. Kilgo ranks up there with the Leland Parks and Chalmers Davidsons for legendary status at our dear old alma mater)
The new year approaching, click in. Let’s facelift bar! Open the wardrobe is not yet found love after another the right clothes? So, also waiting for? Immediate action bar!
in a sense . . .
Bill Walsh did a variation of random play calling: having his first 20 plays scripted and (allegedly) never altering the sequence regardless of down, field position, or yards needed for a first down.
It seemed to work out all right for him
I mean, it has a similar method aimed at a similar result as randomization
But going off a script is as far from randomization as possible
The new year approaching, click in. Let’s facelift bar! Open the wardrobe is not yet found love after another the right clothes? So, also waiting for? Immediate action bar!
Randomizing the order of the scripted plays however...
would be a good idea, IMO.
If the script is generated randomly pregame ...
… and varied from game to game, then yes, it is random.
Walsh did have alternate scripts for special situations (ie coming out from your own end zone, third down), but then would go back to his main script once things went back to normal. Like you’re saying, it wasn’t so much about picking 20 plays out of a hat, but finding new ways to piece his playbook together for each game.
agree
The point I was trying to make is that while scripting and randomization are absolute opposites, the result of trying to make your decisionmaking unpredictable is substantially similar (unless, of course, one can find a pattern to the script).
Thats why you randomize the script.
20 plays you are going to run at beginning of game?
Throw 20 numbers in a hat pregame and order them that way.
screen on 2nd and long has a totally different probability than the play on 3rd and short
if you took a bunch of plays that have acceptable ranges of success in multiple situations, scripted then randomized them? maybe that works.
"Voetbal is pas totaal als je wint"- Coach Adun
"The greatest sin is to spurn the gift"- Coach Alistair
context matters
With Crowton, you have to consider the context. He will only go to random playcalling once his current playcalling becomes effective. “hmmmm, we seem to be getting positive yards with this formation and the defense’s adjustments are not stopping us, we better change it STAT.” His favorite switch is to a pass play that takes 10 seconds to develop so that Jefferson can take a sack for a 5 million yard loss. Got to keep them on their toes, don’t want to get stale and score points.
Not as bad as Patrick Nix
That play is working? Lets not run it again, that will fool them.
Which is why Choice didnt touch the ball in the 4th quarter vs Wake (ACCCG).
I thought this was how Coach Giggity called all of his plays. I could of sworn I saw one of those lottery wheel thingys on the sidelines
by D-Macs LoveChile on Jun 21, 2010 5:36 PM EDT reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
