If the move cost you game then it damn sure will cost you your job, those stats are based on a league avg, as mentioned SF could not stop anyone on 4th down giving up 83% that is a far cry from Philly who only gave up 22% meaning the opponent failed more times than not. Now why would I run the risk when punting the ball with a chance of putting them in a hole would benefit the team? I think there are times you should go for it but not based on some meaningless stat sheet. Who you are playing and the situation you are in plays a big part of a coach’s decision
Exactly, there are so many variables and not enough of a sample size to determine from a pure analytic standpoint when you should always go for it. What stat are you using to base these numbers on? You can't use a current season, there is no or very few samples until the season is played out, but by the end of a season those numbers are useless for the next season. You can't use the NFL average, last season teams ranged from 11%-80% success rate on converting 4th downs. Last year only one team went for it more on 4th downs than Philly, Green Bay. GB had a respectable 53% success rate, 1% higher than the Cowboys. GB converted 76% on away games, but only 18% at home. 18%! That's a huge variance, what stat should they use?
You can't really use past stats as a reliable indicator, because teams change from year to year. In 2016, the Cowboys converted 89%, pretty dang good, but it was only 9 attempts. In 2017, they converted 53%, still respectable, but a huge variance. We went for it almost twice as much last year versus the previous season. Maybe we went for it based off the previous years great percentage, but more likely out of need than want. In 2017, we only converted 45% in away games, half the rate from the year before.
What stat could we reliably use to predict success this coming season? Teams change every year, we should be more stable than last year, but there is no way to reliably predict with certainty. Williams should help, but he's a rookie and unproven, we've also lost Hanna and Witten, both were very good blockers and there's no way to predict with any kind of mathematical accuracy how their replacements will fare.
Those stats are a historical representation, but in this case, are dangerous to use to predict our future outcome. Stats and analytics can be useful, but you have to be able to interpret them, what they actually represent, when to use them and more importantly when not to use them. Fans see these analytical breakdowns and holler for their teams to be more aggressive because the generalized numbers say so, without actually knowing what the stats represent. Dolphins fans should scream to never go for it on 4th down unless they absolutely have to, over the last 3 seasons they only average 30%