@daves why you disagree with this post? I thought you even wrote about it once.
Yes, NFL analytics suggest a team should go for the two-point conversion when down 14 points and scoring a touchdown, because it provides a statistically better chance to win the game rather than force overtime. By succeeding on the two-point attempt, the team would then only need one touchdown and an extra point to win. If the two-point attempt fails, a team can still go for two after their next touchdown to tie the game.
I'm a huge proponent of going for two at the first opportunity when down 15 late in the game. You MUST gain the information regarding whether you need one more score or two more scores as early as possible; kicking an XP early and leaving an 8-point deficit for later and HOPING you'll get the 2-point conversion later is an objectively horrible choice.
In the Cards' situation vs. Seattle, a successful 2-point conversion would be irrelevant unless the Cards subsequently prevented a Seahawks FG. A successful conversion would put more pressure on Seattle to score that FG, but I doubt that "more pressure" would affect their odds of success.
So basically in the hypothetical scenario where the Cards prevent the Seattle FG, you'd be trading the odds of winning in overtime (presumably 50-50, though based on how the 4th quarter went, I'd be inclined to give the advantage to the Cards) for the odds of succeeding on the 2-point conversion.
I don't know where you found your numbers, but what I've seen is that success on 2-point conversions is trending downwards, and was far
less than 50% in 2024. Further, the Cards are especially bad at short yardage conversions since Murray cannot or will not sneak the ball, and Conner is out. They had already failed on short yardage earlier in the game. I wouldn't needlessly bet the game on the Cards gaining 2 yards on one play.
From an analytics perspective, if the odds of succeeding on a 2-point conversion were significantly better than 50%, I'd be in favor of doing it almost every time except in specific late game scenarios. Expected value wins in the long run.