Analytics in Coaching

[bobongo]

I shared the Nick Saban example to prove that point. Both he and Cristobal made the exact same decisions (ironically on the same day) but one is a buffoon and the other went unnoticed.

We have a lot less margin to get away with making less than optimum decisions.

What does Cristobal have to do with it? Did we hand off when we could have simply taken a knee?
No way are the two decisions anywhere remotely near the same. One is in a gray area, the other (Cristobal's) is one of the dumbest decisions ever made on a football field.

 

[Thwg777]

What does Cristobal have to do with it? Did we hand off when we could have simply taken a knee?
No way are the two decisions anywhere remotely near the same. One is in a gray area, the other (Cristobal's) is one of the dumbest decisions ever made on a football field.

I think you’re missing my point. I’m not comparing Key’s decision to decline the penalty to Cristobal’s decision to not take a knee.

I’m stating that Saban’s Alabama team did the exact same thing that day that Miami did when the game could have been won with kneel downs. Alabama did not fumble, nor did the lose. No one talks that their decision. But the decision itself was the same, outcomes differ (between Miami and Alabama).

Georgia Tech should not be making less than optimal decisions. We program **** to go to space but cannot fathom using probability tables in football games? I’d have preferred to not be trailing to a middling team by 17 in the 4th quarter but the penalty decision is one tangible inefficiency that we should immediately improve.

 

[MWBATL]

GT got beat on the LOS today.

Period.

The team that played better...won the game.

 

[bobongo]

I think you’re missing my point. I’m not comparing Key’s decision to decline the penalty to Cristobal’s decision to not take a knee.

I’m stating that Saban’s Alabama team did the exact same thing that day that Miami did when the game could have been won with kneel downs. Alabama did not fumble, nor did the lose. No one talks that their decision. But the decision itself was the same, outcomes differ (between Miami and Alabama).

Georgia Tech should not be making less than optimal decisions. We program **** to go to space but cannot fathom using probability tables in football games? I’d have preferred to not be trailing to a middling team by 17 in the 4th quarter but the penalty decision is one tangible inefficiency that we should immediately improve.

Well, I think the decision by Key not being optimal is debatable.

As for not taking a knee, Saban was as dumb as Cristobal if he did the same thing. THAT was a no-brainer, and I don't care who did it or if he got away with it. It's just plain stupid by definition. Take a knee you win, don't take a knee and you may not.

 

[Southern psu fan]

Just take the loss out on Duke and get the win that’ll make all of us happy. We’re still going to have a good season.

 

[roadkill]

Well, I think the decision by Key not being optimal is debatable.

As for not taking a knee, Saban was as dumb as Cristobal if he did the same thing. THAT was a no-brainer, and I don't care who did it or if he got away with it. It's just plain stupid by definition. Take a knee you win, don't take a knee and you may not

It appears that we have gone from "awful", egregious", or "colossal mistake", to simply "not being optimal". That's progress, folks.

 

[Thwg777]

It appears that we have gone from "awful", egregious", or "colossal mistake", to simply "not being optimal". That's progress, folks.

Your contributions to this thread have been as colossally valuable as a center column for an arch.

 

[yeti92]

IMO whether we take the penalty or not was not the issue. Asking the defense to get one stop is easier than asking them to get 2. The bigger wrong decision was not doing a 2nd onsides kick - the defense had only gotten a few (lucky) stops all game, we should have not been putting the game in their hands if we could avoid it. "But the likelihood of recovering two onside kicks in a row is so low!" Well the likelihood our defense was gonna make the necessary stops was just as low if not lower.

 

[Blue&Gold1034]

The awful decision by Key to not accept a penalty to make it second and miserable for Syracuse when we had all three timeouts got me thinking more about analytics. There’s no guarantee what would have happened but the most likely scenario was we get the ball back with one timeout left after they concede the possession.

I’d argue that this decision was a no-brainer. Less egregious than Miami not kneeling it last year to end the game but egregious nonetheless and potentially cost us the game (or at least a chance to tie it).

I wonder what’s stopping us from having assistance from analytics? It would at least show the data / odds for potential options and take some of the emotion out of it. We can build the iron man suit but can’t do this?

I do think Key should've taken the penalty, but analytics doesn't account for the many variables that may influence a play. A player coming out for an injury or substitution the previous play, offensive and defensive coordinator tendencies on certain downs, momentum swings, etc.

 

[MtnWasp]

Even if its two downs. Analytics have everyone over thinking.

If decisions are made based on statistical algorithm, then there is no thinking at all. If coaches make a decision based on following a decision chart, then there is no real accountability "It ain't MY fault, I just followed the chart." If fans have the chart, then they can predict every decision the coach will make.

But the algorithm is blind to novel situations. If there are intangible but relevant parameters in the current situation that are not taken into account in formation of the algorithm, then the analytics don't reliably apply.

It is the difference between human wisdom and creative thought versus the lightning speed of analytics. It is one sided debate these days.

 

[Northeast Stinger]

Your contributions to this thread have been as colossally valuable as a center column for an arch.

Throwing out a reference to a uga campus monument is uncalled for.

 

[Northeast Stinger]

If decisions are made based on statistical algorithm, then there is no thinking at all. If coaches make a decision based on following a decision chart, then there is no real accountability "It ain't MY fault, I just followed the chart." If fans have the chart, then they can predict every decision the coach will make.

But the algorithm is blind to novel situations. If there are intangible but relevant parameters in the current situation that are not taken into account in formation of the algorithm, then the analytics don't reliably apply.

It is the difference between human wisdom and creative thought versus the lightning speed of analytics. It is one sided debate these days.

Tried to say this in another thread but you said it better. As I pointed out, coaches will say you have to know where you are in a game at that moment, not just the analytics which assume all else is equal.

 

[bke1984]

If we push them back the most likely outcome is that they pick up 10-15 on the very next play and we have to burn a timeout. Then we are in virtually the same situation with one less timeout. Our defense was going to concede a 10 yard throw on 2nd and 25. With that being the case it makes sense to just decline and keep the time and the TO. Let it go. It’s not as bad a decision as some of you make it out. This isn’t a video game. It’s real life football. We were going to have to get a stop on 3rd and 10 no matter what.

 

[cpf2001]

Our defense was going to concede a 10 yard throw on 2nd and 25.

Our defense conceded a ten yard throw on third and ten so it’s hard to argue this either way at that point

 

[Jacket05]

I

If the probability of gaining more than 15 on one play is so high as to be relevant than third and 10 is pretty hopeless anyway.

Which is part of why the fourth down play that got stuffed earlier kills me so much.

Both the offense and defense are going to play a 3rd and 10 and 2nd and 22 very differently. Therefore it is not as simple as the probably to get 10 yards on any given play. I don't know the numbers but would argue it is much easier to get 10 plus yards on a second and 22 than it is on the 3rd and 10 because the defense is more willing to give up 10± yd and get into a third and long/medium.

 

[bobongo]

If we push them back the most likely outcome is that they pick up 10-15 on the very next play and we have to burn a timeout. Then we are in virtually the same situation with one less timeout. Our defense was going to concede a 10 yard throw on 2nd and 25. With that being the case it makes sense to just decline and keep the time and the TO. Let it go. It’s not as bad a decision as some of you make it out. This isn’t a video game. It’s real life football. We were going to have to get a stop on 3rd and 10 no matter what.

Agree. It was a lot more about the execution that it was about the decision. Bottom line is we had them third-and-ten and didn't stop them.

 

[awbuzz]

All the D had to do was prevent a first down conversion on 3rd and 10. What are the odds for the that... Pretty high.

The two previous plays before that third down conversion for a complete pass for 0 yd, in an incomplete pass.

We have been giving up chunk plays all game. I'd rather take the gamble that we can make one stop then having them have two opportunities.

 

[cpf2001]

Both the offense and defense are going to play a 3rd and 10 and 2nd and 22 very differently. Therefore it is not as simple as the probably to get 10 yards on any given play. I don't know the numbers but would argue it is much easier to get 10 plus yards on a second and 22 than it is on the 3rd and 10 because the defense is more willing to give up 10± yd and get into a third and long/medium.

If I’m playing D in this situation I would treat the second down as if it were still a third and ten, not as if it were a “normal” second down.

 

[awbuzz]

Making one stop is easier than making two.