Should dolphins have a statistics/mathematics staffer? | FinHeaven - Miami Dolphins Forums

Should dolphins have a statistics/mathematics staffer?

furball4

furball4

Starter
Joined
Sep 13, 2004
Messages
181
Reaction score
0
One thing that has always bugged me about pro football are the inane decisions made by coaches on a regular basis. Punting on an opponent's 40-yard line when you are down two scores, etc... not that WE have experienced any of those as Dolphins fans, of course (cough, cough, hack).

When I see calls like these, I always wonder why teams don't employ statiticians and/or mathematicians to help them determine what is the highest-percentage call. I'm not talking about the irrelevant historical statistics that announcers throw around all the time to make every game seem exciting and history-making, I mean harder observations such as:

How much clock time is a yard worth?
How much clock time is a down worth?
How much clock time is an opponent's yard worth?
How much clock time is an opponent's down worth?
How many points is a yard worth?
How many points against is an opponent's yard worth?

And conversions between those values. These aren't constants, of course, but something like this has to be done in order to know that you should punt when a drive stalls on your 40, but not punt when it stalls on an opponent's 40. "How many points is a yard worth" comes into play in the red zone, when statistics might show a serious drop-off in points achieved from the 8-yard line v.s. the 9-yard line. My point is that such conversions and estimates are already being made: they have to be. Why not get someone to research them and make sure they are made as logically and accurately as possible?

Also I question the wisdom of playing a game differently at different times. Many coaches talk about getting a lead and then protecting it... but isn't the defense always trying to protect it, and the offense always trying to extend it? If you get a lead and then the offense stops doing anything risky, aren't you changing a good thing? Might they not lose just because they started scoring less? Why not extend the lead as far as possible? If it is because it was so risky, why did you take the risk in the first place?

I'm not saying there is no logic here, just that my impression is that coaches do NOT get these strategies as a result of careful examination. They seem to be weak justifications for particular decisions. For example, on a given Sunday you will probably hear one coach talk about protecting a 14-point lead by changing to a conservative offense, and another talk about how a 14-point lead allowed them to "open up" their offense since they had the defense on the run and tired.

If I were coaching a team, I would want someone who knew how to generate meaninful statistics and incorporate them into strategies, as well as someone who could identify logical fallacies and inconsistencies in game-plan rhetoric. What do you guys think?
 
The one call that bugs me that the Dolphins always do is

The down is 3rd and 10
We are in the 3rd quarter
The play clock is down to 2 seconds

We always take a timeout.

I don't know if it's just me but I would rather be 3rd and 15 and have the timeout
 
Yep, that's exactly the kind of thing I'm talking about. I would want the time-out too. I'd also like some hard numbers to back that decision up.

Crazy thing is that we do that, and *then* punt on their 40 when the drive stalls... we shouldn't lose the time-out unless we are going to take other risks on the drive. Scattering your risks around devalues them, even though it gives a veneer of moderation. You should exhibit complementary degrees of risk-taking within each drive. It may not be a constant, but if not then it should be progressive rather than regressive (i.e., you take more risks the further you get).

By the way, I'm not a mathematician. I just like having good reasons for things.
 
While I think that some of this information could be helpful....I think that it would be unwise to take it to the level that you seem to want to (unless I am misreading you).

This is a game played by humans and there can't be a "right" call for every situation. Coaches are paid pretty well to make the tough calls....and you seem to want to take their judgement out of the equation ond just play the odds.......which aren't true odds at all.

Your one example above that bothers me is your talk of why teams play differently in different situations?? How could you expect them not to?? A team is down 7 points with a minute left and 80 yards to the endzone and they should call their normal mix of runs and short passes....then huddle up after each play?? A team protecting a 14 point lead with 3 minutes left should put the ball in the air, when an incompletion stops the clock, rather than run and make the other team spend their valuable Time outs (I see from your other post that you do see their value)?? You can't truly believe that teams need to play the same at all junctures in the game.

Furthermore, some of the calls that don't seem to make sense to outsiders, are very reasonable if you consider all of the variables. Put Baltimore and Indianapolis in the same 4th and 5 at their opponents 40 yard line with a 7-point defecit and I would think that you get two different calls. If I'm coaching Indy, I likely go for it, trusting my offense more to get 5 yards, than I trust my defense to make them go three and out. Conversely, if I'm coaching the Ravens, I punt w/o a second thought and let my D go out there and shut them down. Add to that your opponents strengths/weaknesses, it becomes impossible to let a statistic point you to the "right" call.
 
Wannstedt quit, I think you'll find a lot of these problems will be taken care of simply because of that fact.

You start to take the human element out of the game, and everything becomes really predictable and boring.
 
I think it is the responsibilty of Head Coach Nick Saban and his Offensive Coordinator to know the mathematics and statistics of whether to go for it on certain plays, and have a list of the best calls for certain situations....I believe most coaches carry these plays on a cheat sheet. I dont think they need to hire a separate person to do this.
 
Your one example above that bothers me is your talk of why teams play differently in different situations?? How could you expect them not to?? A team is down 7 points with a minute left and 80 yards to the endzone and they should call their normal mix of runs and short passes....then huddle up after each play?? A team protecting a 14 point lead with 3 minutes left should put the ball in the air, when an incompletion stops the clock, rather than run and make the other team spend their valuable Time outs (I see from your other post that you do see their value)?? You can't truly believe that teams need to play the same at all junctures in the game.
Click to expand...

No, you are correct. Time management is definitely a good reason to change the style of play. I'm just suggesting that there are also bad reasons, and that having the lead is not in itself a sufficient reason to change the way you play. Also, if you are minutes from the end of the game, sure, go ahead and kill the clock. But if you just have a decent lead in the third quarter, you can't know that killing the clock is going to help you. There's a decent chance you are going to be the team wishing for more time later on. Why not keep your focus on putting more points on the board? More points *always* help. Better field position *always* helps. Anything else you achieve is a two-way street.

Conversely, if I'm coaching the Ravens, I punt w/o a second thought and let my D go out there and shut them down.
Click to expand...

Well let's examine that one. On the face of it, it makes sense. But they can punt it back to you just as far as you punted to them. So even if you stop them in three downs, all you gain is three downs. They get the ball at the 10, go three and out, and punt it back to you at the 40. That's a nice payoff, and you can continue your drive. But chances are they will get at least one first down, maybe two, before the drive stalls (at least these might be the chances... my point is that it does depend on the chances). Which means that they will likely drive you back to your own 40. So let's say that, on average, you are making a decision between taking a 20-yard loss and facing a fourth down. Is twenty yards worth it? I don't know. Not all coaches make these decisions by the seat of their pants, I'm sure, but I'll bet a lot of them do. And of course it does change by team, but that can be worked in.

I'm not suggesting we have a supercomputer on the sidelines calling plays, just that we have somebody review our decisions. Even if the math guy is just there to say "Hey, why did you go for it in that situation?" and evaluate the answer to make sure it makes sense. Humans tend to use very fallacious reasoning when dealing with odds and averages, and I doubt that coaches are exempt. All of the calculations might indeed be summed up by "I trust my defense to improve our position more than my offense", but if you can gain an understanding of the underlying calculations then you can make calls that are right on the line with more accuracy and confidence. You might also start to realize how much a particular weakness in, say, your secondary is forcing your hand at certain decisions, and therefore realize that it is worth correcting.
 
furball4

i agree totally about the taking risks idea and not spreading them around arbitrarily.

Im not sure about the whole stats idea to play a part in in-game decision making, but maybe as a prep week reference. The thing is that most stats dont got far enough or tell you much at all.

Its like those meaningless stats you see on SportsCenter. For example, you'll see something like Herb Nerdler hit .389 with runners in scoring position in the month of may. Thats great, but if you find out he hammered the Devil Rays for six games, sat out for 5 games, and then hit some inconsequential at-bats in blowout games against poor relief, the stat has no real value.

What im saying is that stats in sports have many uncontrollable values. You mentioned success at the 8 yd line as opposed the 9 yd line. its a nice little reference, but you have no idea what defense each play succeded or failed against. nor do you have any clue what defense is coming next.
 
Using statistical analysis to assist coaching is fair idea; and I assume most NFL coaches do use statistics to some degree.
However, this data should only be used as a foundation for decision making.
It's like poker: you can use pot odds, implied odds to support your decision quantitively, but to truely assess a situation, you have to take advantage of the known characteristics of the opposition and their style of play in order to to win consistently.

There are occasions when a coach has to call a bluff, take a risk or defy the odds.
 
Yeah, tons of stats are definitely meaningless. A stat about success at the 8 vs. 9 yard line would probably be as well. Bad example...

One example I'm thinking of is strategy in the game of Scrabble. Obviously it's a completely different situation, but it's an interesting anecdote. For years there were various theories about "rack management"; how to keep the best combinations of letters on the board, what to try to play quickly, etc. Finally (well after Scrabble was being played at a very sophisticated level) someone came up with a quality score for each letter. For example, a Q might be worth -15 because it is hard to use, even though it seems valuable with its face value of 10 points. An E, on the other hand, is worth more than its face value of 1 point. This system allowed players to memorize the quality scores of the letters and manage their racks based on that. Even the most sophisticated player would have a hard time, on instinct alone, deciding whether to play a Q, E, and V or play a V, C and B. It would take a lot of time, and still might be wrong. But nowadays players just add up the quality points and know which play is better (from a rack management standpoint).

The system isn't perfect, but it is such a vast improvement over even the best instincts that pretty much every really serious Scrabble player has come to use the system as, at least, a powerful guideline. In the face of this, there are tons of Scrabble stats that have no application to playing Scrabble, just like in football. I figure the gold is in there somewhere, though!

If you start to look at yardage, field position, downs, time, etc. as resources that can be converted, some stats should rise to the top as being useful. Time can be converted into yardage, but only by consuming downs. However, you get a "discount" on downs when converting in large chunks. (Turn time into yardage at 10yds a pop, and you'll never run out of downs) Field position, as a value, requires yards in order to change. But yards will not necessarily change field position for all purposes. If you are about to punt, being at the opponent's 35 isn't any worse than being at the 30.

One example of how this might play out is in a recognition that you shouldn't necessarily always take what you can get, when you can get it. If on first down a runner breaks through the line and runs 9 yards, then sees that he's going to get tackled in another 2 at most, why not down it at 9.5 yards? His team gets a 2nd down pass attempt a half a yard from a first down, so they can really open it up. If they miss, they have an easy third down conversion. Maybe two of them, if they are in four-down territory. I'm not saying it would always be smart, but if you have a good short-gain running attack then it definitely might be. I've certainly never seen it done, though. If you had a mathematical model for how valuable a down was in relation to yardage, you might find that it was sometimes worthwhile to trade 2 yards for a "2nd and inches" situation.
 
RonnieExpress said:
Wannstedt quit, I think you'll find a lot of these problems will be taken care of simply because of that fact.

You start to take the human element out of the game, and everything becomes really predictable and boring.
Click to expand...

:yeahthat:
 
You must log in or register to reply here.
Back
Top Bottom