'This probably didn't happen... I swear he looked straight at me'

Well fatigue was actually my first and main point! I just want others to come to that conclusion themselves rather than just feeding it to them lol

I can accept that South Africa have a ball in play stat that correlates with a lower fitness/higher strength team, but I don't necessarily buy the argument that one automatically leads to the other. I'd suspect their two stats (high restart numbers low BIPs) likely have separate causes.

Graham made a great point about crescendos. These are what people call momentum swings these days. The build up in fatigue is a momentum swing. The sweeping of the ball down the field in multiple phases is a momentum swing. What is important is that these are far too easily stopped by fake injuries or timely replacements, and that they can happen regularly enough that extending game time (through stopping the clock) becomes irrelevant. It has always been case that to create fatigue play needs to be continuous. What matters is the Work to Rest ratio exceeding 70 secs and still being consistent at the ends of games.

Qualities in bench changes have a different effect, but as their use has become quite adept over time, not so insignificant changes that they should be ignored, I agree. The main problem however is that teams can't dictate the speed of the game, as in, any team can dictate how slow it becomes if they really want to, but the team in possession (they should even have some capability to keep the pace up when not in possession) are too easily foiled when the want to play with a high tempo.

The your question about stats linked to 'incorrect decisions' I've answered more than once above.

In fact it's based on a neat sidestep of the real issue, which is how to reduce the number of pens for minor technical infringements at scrum-time - parading quite wrongly under the banner of 'dominance'.

So you need a model where only clear & obvious offences are pinged, not one where refs are looking to wriggle out of the straitjacket of resets by making a decision.

Hell, I even looked at having Jonah Lomu on the wing as cheating back then, that's how much of a skew in advantage he was.

BTW, just a couple of other things ive picked up on from your tone, Nick's idea promotes more scrums (ie the slowing down of the game into more of a contest), so you should be happy about that one (trade). The other is just a mundane point that I don't know holds any real value to this discussion, but the ball is obviously lively contested at ruck and maul time, it's not after the ref blows his whistle and the 9 is fu#$ing around spinning the ball on his middle finger however.

No reason you cannot be dominant and still use the ball is there?

As you can see from the screenshot in the article, Oli Jager only get separated from his hooker because Jack Boyle is moving around him and Ronan Kelleher is moving in towards him. It's not worth a pen.

Malherbe examples next article! He was going backwards in at least two of those scrums.

Do you view a scrum differently to every other facet? Why's that? And what would you like the scrum to turn into if you implemented some changes?