Big Ben wants Steelers to go for 2 after every TD

[mannc]

For each individual game, not the year. You could get all of your 2 point conversions in games that don't end up mattering (blowout loss or win) and miss them all in close games where that is the difference between a win and a loss. 50% for the year, but your record is worse. Obviously that's an extreme example.

 

Lose ONE game because you went for a two point conversion that you didn't have to, and this whole argument goes away. In a 16 game season where every win matters, you just can't take the chance.

This is 100% wrong. This mindset is exactly what analytics is meant to counter. If you are converting 50% of your two point tries and only 90% of your 1-pointers, then you are better off going for two every time, except in the odd situation where a 1-point conversion ices the game. That's just basic math.

[Mr. WEO]

This is 100% wrong. This mindset is exactly what analytics is meant to counter. If you are converting 50% of your two point tries and only 90% of your 1-pointers, then you are better off going for two every time, except in the odd situation where a 1-point conversion ices the game. That's just basic math.

 

 

If the analytics are so sound and the results so predictable and therefore the risk so low, why doesn't a single team do this?

[NoSaint]

 

 

If the analytics are so sound and the results so predictable and therefore the risk so low, why doesn't a single team do this?

He did say "if" in the post. Until a team commits they don't really know the rates they will convert. Otherwise, his point stands that statistically over time you should come out ahead in the win column if you are scoring more points- even if you get a cluster of misses that costs a game across a small sample, you should have that countered by clusters that win 2 games in the long haul (oversimplified but conceptually is there) Edited June 16, 2016 by NoSaint

[mannc]

He did say "if" in the post. Until a team commits they don't really know the rates they will convert. Otherwise, his point stands that statistically over time you should come out ahead in the win column if you are scoring more points- even if you get a cluster of misses that costs a game across a small sample, you should have that countered by clusters that win 2 games in the long haul (oversimplified but conceptually is there)

Exactly. And the reason no one does it is because it's easier to just do what everyone else has done from time immemorial. If a team wins two games by going for two, but blows one when a two pointer fails, the criticism would be merciless (but wrong).

[Mr. WEO]

He did say "if" in the post. Until a team commits they don't really know the rates they will convert. Otherwise, his point stands that statistically over time you should come out ahead in the win column if you are scoring more points- even if you get a cluster of misses that costs a game across a small sample, you should have that countered by clusters that win 2 games in the long haul (oversimplified but conceptually is there)

 

 

Exactly. And the reason no one does it is because it's easier to just do what everyone else has done from time immemorial. If a team wins two games by going for two, but blows one when a two pointer fails, the criticism would be merciless (but wrong).

 

 

But if the analytics, right now (as many have argues, including mannc) support going for 2 all the time, it would make no sense not to commit to that right now. If your analytics tell you it's 50%, why would every team give it a go?

 

If 85% of games are won by a margin of greater than 3 points, the benefit of a few extra points over a season doubtfully will alter a team's record.

[Kirby Jackson]

 

 

 

 

But if the analytics, right now (as many have argues, including mannc) support going for 2 all the time, it would make no sense not to commit to that right now. If your analytics tell you it's 50%, why would every team give it a go?

 

If 85% of games are won by a margin of greater than 3 points, the benefit of a few extra points over a season doubtfully will alter a team's record.

It does support it but for the reasons that you've mentioned it hasn't been done yet. It's going to take someone with job security (like Tomlin or McCarthy) to raise the sample size before the sheep follow. There were 94 attempts league wide last year. That number will continue to grow. Edited June 16, 2016 by Kirby Jackson

[Acantha]

This is 100% wrong. This mindset is exactly what analytics is meant to counter. If you are converting 50% of your two point tries and only 90% of your 1-pointers, then you are better off going for two every time, except in the odd situation where a 1-point conversion ices the game. That's just basic math.

The first part of my post is not wrong...at least for now. After years of data, you can make an argument for season statistics, but right now, it's one game at a time.

 

As for the last part of my post, you're right, analytics is there to overcome that. But human nature is what it is. You watch Rex Ryan go for two for no reason, have that cost a game, and see how crazy the media and fans react. As Kirby said, only a coach with great job security can pull off the testing of that theory.

[mannc]

As I understand it, the numbers do not support going for 2 all the time. I think the success rate on two-point conversions last year was around 46% and for 1-pointers it was around 93%, so it's basically a push. But if you have weak kicker and a strong two-point conversion package, the numbers could swing in favor of going for two almost all the time. But i don't expect any NFL coach to do it, including Belichick.

 

 

 

 

But if the analytics, right now (as many have argues, including mannc) support going for 2 all the time, it would make no sense not to commit to that right now. If your analytics tell you it's 50%, why would every team give it a go?

 

If 85% of games are won by a margin of greater than 3 points, the benefit of a few extra points over a season doubtfully will alter a team's record.

Edited June 16, 2016 by mannc

[NoSaint]

 

 

 

 

But if the analytics, right now (as many have argues, including mannc) support going for 2 all the time, it would make no sense not to commit to that right now. If your analytics tell you it's 50%, why would every team give it a go?

 

If 85% of games are won by a margin of greater than 3 points, the benefit of a few extra points over a season doubtfully will alter a team's record.

you know as well as anyone here that when even a bill belichek does the right statistical call, but it doesnt work he is roasted on monday. no one likes to stand out in the crowd on something that might not pay immediate dividends.

 

also, we dont know if the percentages would hold yet -- using the stats on primarily end of game desperation tries that happen once in awhile vs a team making it a centerpiece of their strategy (and likewise a defense doing so), may have shifts (good or bad).

 

plus as you mention the quality of the kicker, qb, etc... can all play on it. if the pats were STRUGGLING with their kicker but the line was firing on all cylinders i wouldnt be shocked to see them be a team that went this way (likewise the steelers, saints, etc...). teams with coaches that buy into the numbers game and have job security and a franchise qb.

 

as it is, you are seeing incremental growth i believe. how much were 2 pt conversions up this year?

[l< j]

This is 100% wrong. This mindset is exactly what analytics is meant to counter. If you are converting 50% of your two point tries and only 90% of your 1-pointers, then you are better off going for two every time, except in the odd situation where a 1-point conversion ices the game. That's just basic math.

Actually he is 100% right. You are comparing two different sets of outcomes as though they have the same impact on win-loss record. If I make 58% of the 2 pt conversions, that doesn't necessarily result in more games won.

 

Going for it all the time and having success greater than 50% only means the team will have scored more points at the end of the year. It doesn't mean that they success will be evenly distributed across every game. In fact, you can guarantee it won't. So your successful conversions will cluster in games against weaker defenses and your unsuccessful ones will cluster in games against stronger defenses. At the end of the season you will point to the extra points scored across the season and shout "analytics!" But your record won't necessarily be better.

 

kj

[K-9]

Exactly. And the reason no one does it is because it's easier to just do what everyone else has done from time immemorial. If a team wins two games by going for two, but blows one when a two pointer fails, the criticism would be merciless (but wrong).

As I've said, when coaches see a true correlation of success on 2 point conversions to wins, they will adopt the new paradigm. It's why I'm wondering how impactful the Steelers' relative success rate really was in the win column. For now, the old college chart of when to go for 2 reigns.

 

GO BILLS!!!

[Mr. WEO]

 

also, we dont know if the percentages would hold yet -- using the stats on primarily end of game desperation tries that happen once in awhile vs a team making it a centerpiece of their strategy (and likewise a defense doing so), may have shifts (good or bad).

 

 

 

Of course we don't. I have said this in the exact same thread that appeared 6 months ago (with the same participants): http://forums.twobillsdrive.com/topic/184718-no-more-xtra-ptsgo-for-2/page-2

 

The difference in this thread is the some are saying "the analytics" say teams should just go for it all the time and that timid coaches ignore this obvious "math" reality.

 

In actuality, the analytics don't show that every team should go for 2 PAT. Last year teams ranged from 100% failure to 100% success, but the numbers are so small that there can be no meaningful conclusion from them. Posters suggesting otherwise won't admit this.

 

Conversely, the extra point kick success rates are much more tightly grouped (27 of the teams are above 90%) and the sample size is significant.

 

Teams kick the extra point because the exiting data tells them it's a better chance of scoring.

[birdog1960]

 

 

 

 

But if the analytics, right now (as many have argues, including mannc) support going for 2 all the time, it would make no sense not to commit to that right now. If your analytics tell you it's 50%, why would every team give it a go?

 

If 85% of games are won by a margin of greater than 3 points, the benefit of a few extra points over a season doubtfully will alter a team's record.

I don't believe it's that straight forward. if the team scoring a 2 point conversion in a game goes up by 4 instead of 3, it would very likely change the decision making of the opponent making them go for a td rather than an fg with more chance for failure (and ultimately more chance of losing). there are many scenarios where an extra point means more than just the differential of the final score. this is just one example.

 

as evidence, how many times as bills fans have we seen a missed extra point change strategy and outcome?

 

the one thing that I hope everyone can agree upon is that more points for a team are better and overall will lead to more wins. if 2 pt conversions can be definitively shown to produce more points over the long term, they will produce more wins over the long term.

Edited June 16, 2016 by birdog1960

[Mr. WEO]

I don't believe it's that straight forward. if the team scoring a 2 point conversion in a game goes up by 4 instead of 3, it would very likely change the decision making of the opponent making them go for a td rather than an fg with more chance for failure (and ultimately more chance of losing). there are many scenarios where an extra point means more than just the differential of the final score. this is just one example.

 

as evidence, how many times as bills fans have we seen a missed extra point change strategy and outcome?

 

the one thing that I hope everyone can agree upon is that more points for a team are better and overall will lead to more wins. if 2 pt conversions can be definitively shown to produce more points over the long term, they will produce more wins over the long term.

 

 

They missed 6 PATs over 16 games. 4 of them were wins. The losses were both by 8 points. You tell me.

[NorthBuffaloKid]

I don't think it would work. After a couple of games, they would know our tendencies and shut us down.

 

It would give the opposition too good of a look at our red zone offense.

[DasNootz]

What's James Hardy doing now? We could use a big bodied WR for this strategy.

[birdog1960]

 

 

They missed 6 PATs over 16 games. 4 of them were wins. The losses were both by 8 points. You tell me.

not nearly a large enough sample size to be meaningful based on wins. my point re the bills experience was that missed PAT's have frequently change strategy over the many games over many years. it has turned many ho hum games into nail biters often with negative consequences..ie: 1 point is often meaningful to a game.

Edited June 16, 2016 by birdog1960

[NoSaint]

I don't believe it's that straight forward. if the team scoring a 2 point conversion in a game goes up by 4 instead of 3, it would very likely change the decision making of the opponent making them go for a td rather than an fg with more chance for failure (and ultimately more chance of losing). there are many scenarios where an extra point means more than just the differential of the final score. this is just one example.

 

id also assume that most would concede that if scoring to go up 3 or 8 with 2 minutes left, theyd kick even if going all in with 2 pts for the season

 

theres some common sense to it - but coaches are notoriously tight with stuff like this, or 4th down attempts and im sure many other wrinkles through the game.

Edited June 16, 2016 by NoSaint

[birdog1960]

id also assume that most would concede that if scoring to go up 3 or 8 with 2 minutes left, theyd kick even if going all in with 2 pts for the season

 

theres some common sense to it - but coaches are notoriously tight with stuff like this, or 4th down attempts and im sure many other wrinkles through the game.

agreed. but most, if not all scenarios could be anticipated and the move most likely to result in a win chosen by an algorithm. that algorithm is going to result in a large overall increase in 2 point attempts.. the problem will again be going off script. the coaches will need to be consistent with it for the numbers to play out. I don't believe most will be. they'll go by their gut at crunch time when they need the analytics most.

[K-9]

I don't believe it's that straight forward. if the team scoring a 2 point conversion in a game goes up by 4 instead of 3, it would very likely change the decision making of the opponent making them go for a td rather than an fg with more chance for failure (and ultimately more chance of losing). there are many scenarios where an extra point means more than just the differential of the final score. this is just one example.

 

as evidence, how many times as bills fans have we seen a missed extra point change strategy and outcome?

 

the one thing that I hope everyone can agree upon is that more points for a team are better and overall will lead to more wins. if 2 pt conversions can be definitively shown to produce more points over the long term, they will produce more wins over the long term.

And if the team fails and goes up by only 2 instead of 3, would that likely change the decision making as well?

 

GO BILLS!!!