A safety is worth less than a touchdown.

But when a team has a safety (since 2000) they go 224-109 (.672)

When a team has at least one touchdown (since 2000) they go 5284-4649-20 (.532)

I don’t know what any of this means. I was inspired to dig into the data because of this post. This doesn’t really answer that specific question, but it touches on the topic.

Edit with some more thoughts:

A few people have made the comment that the good winning percentage for teams with a safety is the result of the fact that good teams get safeties. Therefore, the impact of the safety on the winning percentage is not as causal as the quality of the team getting the safety. I do believe this is a good point and is part of it, how much it factors in, I don’t know. But this made me dig a little bit deeper into the teams that had a safety. I’m just using 2020 data because I don’t have a good way to extract this data for many years.

Those comments seemed to be right, that it is generally good teams that get safeties. In 2020, excluding the games in which they had the safety, NFL teams had a .590 winning percentage. So, yes…safeties are usually from good teams. Comparing that .590 record to games in which teams have a safety (.672 winning percentage), you still see a positive gap between the two (+.082). All else being equal and assuming the WP is due 100% to the quality of the team, you would expect a .590 WP in games they have a safety. But there is a lift of .082 when comparing their overall record and their record in games with a safety. So, I think about half of the success rate in these games is due to the quality of the team that got the safety and about half of it is from the act of getting the safety and the impact it had on the game. Just my thoughts for what it's worth.