Carroll said "doubting fans need to “do their homework”

[BleedGreenNblue]

:49ersmall:

 

[Ad Hawk]

:49ersmall:

:th2thumbs:

 

[HansGruber]

You have proven that teams that have the greater ToP have won their games more times than not. What you have not proven is that they won their games DUE TO ToP, which would make it important. They have accomplished a ToP advantage because they won, they did not win because they had a ToP advantage.

MOCHawk is correct here to apply the "correlation does not imply causation" principle. It is one of those concepts that most people know and if questioned may agree that is logically sound, but do not actually believe. The idea is that A and B may tend to happen at the same time, but that alone does not give you enough information to explain why. A could be causing B, but it is also possible that B is causing A, or C is causing both of them, or they are both simply coincidences.

In this case, it is very likely that teams who are winning a game will start to do well in time of possession in the fourth quarter because of the score differential (B causing A). Knowing this, a superior strategy would be to correlate ToP over the first quarter, first half, and first three quarters of games with the likelihood of winning that game.

First, MOCHawk was not applying the principle of "correlation does not imply causation" (referred to as Corr v Caus from here on out).

"Corr v Caus" is used to establish independence between two seemingly codependent factors in a system. In medicine, we use this principle to ensure that doctors do not misdiagnose an illness by assuming that one symptom is causing another (versus both being the correlated results of some other unknown cause).

We are trying to establish the validity of a SINGLE indicator - not stating that two indicators are causing each other. Unless you are trying to tell me the outcome of a game could be used as a statistic during the game itself. I hope you're not saying that, because that makes my brain hurt.

We have never established a relationship between time of possession and any other statistic. It is a requirement to have established a dependent relationship between two indicators before we can apply Corr v Caus.

Even more important, we have never implied causation. Causation is only being argued by the opponents of TOP as an indicator, such as in your post. It is a common theme I am noticing. Those who oppose TOP as an indicator oppose it because they assume we are saying time of possession CAUSES a win. This was never implied and I have repeatedly stated this is not the case.




With such a simple statistical problem, the only valid means of disproving or discrediting the finding is to challenge the validity of the supporting data itself or question scope (is data extrapolated improperly). No matter how many emotional arguments, logical fallacies, or attempts to twist reality are attempted, the factual data remains the same.

The simple fact is that in 71% of all games played in 2013, the team with the greatest time of possession was the ultimate winner. This is a significant fact, and proves that time of possession is indeed a significant and valid indicator of success.

 

[HansGruber]

Colts had a ToP of 32:55, but lost the game 59-24. Similarly, last year they also had a 22:22 ToP against Detroit, but won that game 35-33. These aren't anomalies, they are regular happenings with ToP.

WRONG. Final scores and statistics are freely available and you can see for yourself. I have analyzed both the 2012 and 2013 seasons and not found one single week where the majority of teams who had the best TOP lost the game.

Your previous estimate of 40% could not be further from the truth, and you are just refusing to accept the truth because you don't like it.

I would agree to disagree but your refusal to accept the facts is troubling and insulting. I took the time to compile and freely provide those facts to you, and instead of saying, "Wow, I was wrong. In fact, when teams win the time of possession battle, they win the game 7 out of 10 times", you continue to argue that this never happens and are now taking your ball and going home. Not very gentlemanly at all, sir.

 

[Lords of Scythia]

"You have proven that teams that have the greater ToP have won their games more times than not. What you have not proven is that they won their games DUE TO ToP, which would make it important. They have accomplished a ToP advantage because they won, they did not win because they had a ToP advantage."

MOC, it would require a statistical analysis to go beyond "correlation is not causation" and thus your position is a fallacious nullity.

 

[Lords of Scythia]

"You have proven that teams that have the greater ToP have won their games more times than not. What you have not proven is that they won their games DUE TO ToP, which would make it important. They have accomplished a ToP advantage because they won, they did not win because they had a ToP advantage."

MOC, it would require a statistical analysis to go beyond "correlation is not causation" and thus your position is a fallacious nullity.

Sorry, I just posted this when I got to his post. I did not realize "Corr v Caus" had been established so thoroughly.

 

[Lords of Scythia]

If you don't do a scientific analysis and prove caus, the corr could be %99 and it still wouldn't prove one is causing the other. The corr could be inverse--the win is causing the TOP because the losing team is passing more and not using any clock.

 

[strohmin]

I think the only two stats that can tell the whole story of a game is the number of punts and number of turnovers in the game. TOP doesn't mean crap in terms of winning compared to these two.

 

[Laloosh]

 

[Cartire]

I think the only two stats that can tell the whole story of a game is the number of punts and number of turnovers in the game. TOP doesn't mean crap in terms of winning compared to these two.

Punts are a factor in TOP. If teams are punting they no longer are gaining TOP. And no one is stating that other things in games are also indicators. This whole argument has simply been about the validity of TOP as an indicator for team success.

 

[Lords of Scythia]

If PC thinks that TOP in itself gets wins, he's living a pipe dream. If you get a lot of TOP and win it's probably because of other factors that are causing both, like getting a lot of first down or turnovers on D. Saying TOP = wins is just too simplistic, as well as wrong.

 

[seedhawk]

Take TOP, and divide it by nymber of offensive plays. Shows tempo, and also if you are just grinding the opponent down.

 

[HansGruber]

If PC thinks that TOP in itself gets wins, he's living a pipe dream. If you get a lot of TOP and win it's probably because of other factors that are causing both, like getting a lot of first down or turnovers on D. Saying TOP = wins is just too simplistic, as well as wrong.

TOP = wins in 7 out of 10 games. Simplistic, perhaps. But it's a fact.

 

[AgentDib]

Even more important, we have never implied causation. Causation is only being argued by the opponents of TOP as an indicator, such as in your post. It is a common theme I am noticing. Those who oppose TOP as an indicator oppose it because they assume we are saying time of possession CAUSES a win. This was never implied and I have repeatedly stated this is not the case.

TOP = wins in 7 out of 10 games. Simplistic, perhaps. But it's a fact.

I can't tell if you are trolling here or not, so I guess the safest thing is to assume that you are.

 

[Lords of Scythia]

If PC thinks that TOP in itself gets wins, he's living a pipe dream. If you get a lot of TOP and win it's probably because of other factors that are causing both, like getting a lot of first down or turnovers on D. Saying TOP = wins is just too simplistic, as well as wrong.

TOP = wins in 7 out of 10 games. Simplistic, perhaps. But it's a fact.

Yes it is a fact. That is a very solid correlation. But it does not mean TOP is CAUSING the wins. There's all kind of stats that correlate to wins. If PC is gameplanning to build up TOP then he is operating on a fallacy. Come on, Hans, you're a scientist. What is the true causal relationship, here? There's a million other factors in the TOP/wins equasion.

The highest priority should be scoring as many points as possible. If PC is passing up a scoring opportunity to try to extend a drive because he's got it in his head that TOP wins games, well then he's on the pipe.

 

[Ad Hawk]

Interesting that we had a lower ToP tonight, even with the short field we gave them a couple times, yet still won. We didn't run it as much tonight, instead taking some shots downfield at opportune moments, and a even more quick-passing game. It was fun to watch!

PC isn't building on a fallacy, Scythia; he knows that ToP is a by-product of his offensive game-plan, thus he can use it as a evaluative metric for what he is trying to do: strong running game, having the ball to kneel at the end of the half/game to control the clock, etc. You're right about there being many factors leading to ToP. All the current Hawk priorities on the field should generally (not always) lead to a greater ToP. This shouldn't be hard to understand.

 

[Lords of Scythia]

That makes sense. You can use stats as tools to measure your performance, but the only one you should be gameplanning to bulid up is the score.