Ireland Team Talk XII: Farrell's First Fifteen

Smith didn't seem to like him at 13 for whatever reason, picking him at 12 a few times instead. Personally, I think he'd have been a way better pick than Earls at 13 in the 2015 QF. 12. Fitzgerald, 13. Henshaw would have been better than what he went with too, for me. And Jackson ahead of Madigan. Some of the selections for that game are very questionable.

I went back and had a quick look at some attacking stats, ranging from the 2nd test in New Zealand in the 2022 Summer tour, up until and including the QF loss to New Zealand last month. Respectfully, I am not including games against Italy or lower-ranked opposition. I am only looking at games Sexton started.

New Zealand 12 - Ireland 23: Sexton had 27 passes and 2 runs for 30 metres gained.

New Zealand 22 - Ireland 32: Sexton had 12 passes and 1 run for 16 metres gained.

Ireland 19 - South Africa 16: Sexton had 16 passes and 2 runs for 4 metres gained.

Wales 10 - Ireland 34: Sexton had 18 passes and 8 runs for 23 metres gained.

Ireland 32 - France 10: Sexton had 22 passes and 5 runs for 68 metres gained.

Scotland 7 - Ireland 22: Sexton had 27 passes and 5 runs for 20 metres gained.

Ireland 29 - England 16: Sexton had 27 passes and 9 runs for 47 metres gained.

South Africa 8 - Ireland 13: Sexton had 16 passes and 6 runs for 9 metres gained.

Sexton's "pass:run" ratio fell between 3:1 at its most run-heavy (v England in the 6N), and 14:1 at its most pass heavy (first win in NZ). I like this statistic because it normalizes for the amount of times we were actually in possession and not looking to kick.

His metres-gained per carry, for me, is the key. Even into his late thirties, and controlling for outliers, he was good for 4-5m on a carry, which is outstanding. No out-half will come close to this running threat at that age. Especially in tests at that level, and making that many carries per game (robust average of ~4 carries per game).

Defenses absolutely had to take his running threat seriously, and with those runs coming at a robust-average ratio 5:1 across those games, it was a threat that required constant focus by defenses.

Now let's look at the data from the QF loss:

Ireland 24 - New Zealand 28: Sexton had 73 passes and 1 run for 6 metres gained.

Even if you adjust down the 73 passes (at least 37 of them would have occurred in that final drive), this makes 36 passes to 1 run, or a ratio of 36:1. This is two-and-a-half times more than his most pass-heavy game since Summer 2022, and 7 times more than his robust average over that same period. And some might argue that you should not adjust the 73 passes...

But I honestly think you can discount activity in the final drive, and still conclude that Sexton's running threat was practically null for the entire game. We were a pass-heavy and predictable attack. And this policy deviated drastically from our effective performances against top tier opposition in the 18 months previous. Whether it was burn-out, injury, nerves, fatigue, who knows. You will have a hard time convincing me that it was a deliberate tactical decision that they thought would win the game.

Now, you can justly argue that we were nonetheless a Beauden Barrett miracle defensive effort away from reaching a semi-final. So I am not concluding that Sexton's overall performance lost us the game. On another day, we do get that try and probably close out the game. (And probably win the RWC....).

But in the last 18 months, we offered a two-pronged attack and won most games by more than a score (i.e., clear of the influence of these Barrett-like freak moments).

In the game where the data show that we drastically departed from this, we lost.

Note 1: What I mean by "robust average" is taking the exponent of the average of the log transformed data. It controls for outliers, akin to a median, but it's more granular. @jacothelad, I'm sure, can confirm that this is the appropriate measure with sparse and skewed data).

Note 2: Click the scores if you want to look at the underlying data and DYOR (finding matches on that ESPN site is a pain; you're welcome).

Great research, does convince me further that Sexton was not at his best in that last quarter and shovelling on ball. Crowley should have been on, whether he would have made a difference you can argue the toss but I'd wager he'd have ran the ball more.

I'd say that World Cup will haunt those players, we arrived in the best shape we'd been in, suffered few/no major serious injuries throughout and then didn't perform to anywhere near our best when it counted.

Excellent research Neil.

The Gibson Park try was exceptionally soft. Jordan made a terrible defensive read where he doesn't commit to JGP because he is worried about Lowe, despite Lowe being clearly covered by Barrett and falls off the tackle on a scrum half. If he makes that tackle JGB is isolated and Savea is steaming in for the penalty turnover.

Those "sucker punches" aren't a fluke either, they were a direct consequence of NZ selecting superior athletes to Ireland, particularly in the backs. Ireland prioritised system players and cohesion over pace and as with most decisions at the top level you gain and lose based on that selection.

Brilliant post, Neil.

I think one large part of it was that Hansen didn't seem to be 100% following his injury either, and he has become such an important part of our attack that just hampered us even further.