Despite Leinster being 20 point favourites for their 2024-25 Champions Cup semi final, they ultimately lost to a superb Northampton side. There has since been debate on Leinster’s penalty decisions in those last 5 minutes. We’ll dive into them through the lens of our xP metric to investigate which decisions would have enhanced the chances of the winning and ultimately, what would have been easier to execute.
There are so many variables in a rugby match making it so unpredictable between what was only 5 possessions of the ball between the 2 teams. We attempt to narrow down the most common or likely outcome while combining some human intuition alongside our xP metrics for a more rounded view point on this topic.
The Context
The value of winning the game is clear, you’re going to Cardiff for the final. We could assign that a numeric value of “1”. A simple way to look at this would be to give the value of drawing that of half the value of winning, making the assumption that the teams had an equal chance of winning in extra time.
One could argue that Leinster had recovered a 12 point half time deficit and added a lot of experience in the likes of Jordie Barrett and Jack Conan coming off the bench in the 2nd half, that they would have the advantage or momentum in extra time. Compared to Northampton’s less experienced bench.
With the bookies giving Leinster a +20 on the handicap, we could similarly estimate that Leinster would have 0.25 points/per minute advantage in extra time or 5 points over 20 minutes.
We’re going to look at the 2 penalty decisions. 1 with 75.15 on the clock and the other at 79.00.
Starting at 75.15. Using a very basic win probability model, which uses normal distribution (so doesn’t work in scores of 3,5 or 7) and ignores who had possession and where on the pitch, the basic model outputs that Leinster had 13% chance of winning based on their 20+ handicap and Northampton 85% and 2% draw. That basic model obviously has it’s flaws in underestimating the draw but remains an interesting measure to work off.
Leinster do have a penalty from the late tackle on Tommy O’Brien, 11m out from the tryline and 15m in front the touchline on the right hand side. With 4.45 minutes remaining, Leinster will likely get another possession before the 80 minutes and maybe additional possession should there be shorter lengths of possession such as kick tennis.
Also for context, Josh Kemmy the Northampton Blindside Flanker is in the sin bin for the first penalty decision and Alex Coles, is in the sin bin for the 2nd penalty decision.
Penalty Decision 1: 75.15 mins on the clock
Leinster have 4 options here.
- Attempt 3 points
- Kick to the corner
- Take a scrum
- Tap and Go
Decision A: 3 points
There is no value increase due to Northampton having a man in the bin for goal kick itself, but it would increase the chances of Leinster scoring from the subsequent restart.
The chances of the goal kick going over in that location is about 75% in our database but that includes plenty of windy days and harder to kick on grounds than the surface at the Aviva. So we’d estimate that it’s closer to 85% chance of going over here on the calm day in the Aviva.
Should the 85% chance happen and Prendergast makes the kick, then we need to look at the following possessions. There would then be approximately 76.30 to 77.00 left on the clock by the time Northampton kick off, likely giving one final attacking possession to Northampton by them regaining the restart or by Leinster kicking it back to them should they kick deep. On average, the team kicking the restart scores next more often.
We can only hypothesize, but one can imagine Leinster after taking the 3 would be content with going to extra time rather than running from deep and would give the Saints a lineout somewhere around the 10m. Of course Northampton could have kicked short and many other variables but we’ll work off this assumption for this exercise.
An average team would score a try 7% of the time from this possession by the end of the possession and win a penalty about 22% of the time, making the assumption (wrongly or rightly) that they wouldn’t kick the ball. The try of course would win it for Northampton, and the winning a penalty would give them a shot at goal, there are too many permutations on where that penalty would end up so we’ll just use Fin Smith’s 69% from the tee so far in the Champions Cup and add 5% for the nice weather in Dublin that day.
Overall giving Northampton a probability of eventually winning in decision A: 6% + (22%*74%) = 23%.
What about Leinster’s chances of making a turnover at that lineout or in the proceeding phases and scoring after that? Well the chances of the turnover or a penalty win would be 71% from that lineout between two teams of the same level.
The chances of scoring directly from that turnover in that zone would be about 2.5%. Of course a penalty win for Leinster would give them a lineout about the same place on the pitch as Northampton and give them the same % chance of winning from there as Northampton above.
So Leinster’s win chance is estimated to be;
Chance of turnover + try: (71%*2%) = 1.4%
Chance of turnover + pen win + Score from LO (71%*11%*23%) = 1.8%
Northampton win = ~23% 0 value
Draw after 80 = ~73.8% * 0.5 value = 0.367
Leinster win = ~3.2% * 1 value = 0.032
The overall value of Decision A would be 40.1% chance of winning. For the sake of interest, if you said Leinster had a 75% chance of winning extra time, then the decision would be worth 58.6%.
Decision B: Go to the corner
This is the decision Leinster ultimately did make. So here we need to hypothesis less and isolate if the problem for Leinster was the decision itself or the execution.
Probably first of note is that Sam Prendergast’s kick to touch leaves Leinster 6m rather than 5m away, which is significant in terms of getting a maul over the line. The first execution flaw in this sequence.
From here, we can expect an average team to score eventually from this possession about 40% of the time and win another penalty about 17% of the time. Should Leinster score immediately, that would probably give Northampton 2 mins to kick off and score again, plus any time into the red. Should Leinster have taken 3 mins to score, the game would be over.
There are so many permutations, around if Leinster made their conversion etc… so Northampton would only have to make a penalty or drop goal to win rather than draw.
For simplicity, we’ll give Northampton a 3% chance of winning from Leinster scoring eventually.
What actually happens, is that Leinster get turned over by Pollock and Coles. and Northampton have an excellent clearance, giving Leinster that same opportunity described theoretically for the Saints above. Where they finish up with the disallowed try, then won a penalty..
So overall, with Leinster going to the corner on this situation, they gave themselves 3 attacking possessions to score from.
Possession 1 (LO 6m out): 40%
Possession 2 (LO 41m out): 6%
Possession 3 (Tap & Go 5m out): 48%
As they did eventually do, they could have won a penalty and could have attempted 3 points further down the line or went for a drop goal. Potentially from an easier scenario but not in this case.
Leinster kept the ball from their 2nd possession really well, and gained 41m on the possession. An above average outcome.
This is what gave them the 2nd high quality possession to score from and enhances their probability of scoring between the 3 possessions here to 71%. If we were to isolate the chances of scoring either of the first 2 possessions, we’re looking at 43%.
The value of Decision B was actually 71% of a win with 29% of a loss account for the 2 x 5m opportunities. Probably from the point of the decision it would be fair enough to estimate that it was 43% of a win, 7% for draw with an estimation that they would have take the 3 with an easier kick thereafter, 50% of a Northampton win.
So hypothesising, going for 3 would have given a value of somewhere between 40-50% chance of Leinster winning by the end of extra time or 80 mins and going to the corner at the moment of the kick to touch left them with somewhere between 43% and 71%.
Decision 2: 79.00 mins
This decision was a bit more straightforward xP wise. With only 1 shot at it realistically as Northampton would likely not kick the ball back and run the clock down as they eventually did.
From <2m from the touchline, the goal kick chance was about 40% and the tap and go 48% chance.
Draw: 0.5 value x 40% chance of success = 0.2 value.
Win: 1 value x 48% chance of success = 0.48 value.
That’s before we consider the likelihood that Leinster could have won a penalty in a better position to take the 3 from or indeed a drop goal.
Making this decision a no-brainer, but ultimately giving up the ball in contact on the first collision was a considerable error of execution of that decision.
The scrum and lineouts were other options. The scrum minus Coles would have had most likely Chunya Munga going into the 2nd row and Northampton having to make decision on whether to give up a back to solidify the back row or leaving the unattackable blindside without a flanker. They had dropped a back in the previous scrums where they were down to 14. That being said, the numbers still show that scoring from that scrum would be more difficult than the tap. Perhaps with a larger blindside or a scrum with a potential to walkover it would be a worthwhile option.
The lineout was a worse option than scrum or Tap & Go at this juncture.
Conclusion
Breaking it down, the decision was quite tight and evidenced by the diversity on opinion on this, from the commentary box and beyond.
In this article, we tried to divert away from “they lost so decision was bad” line of thinking to advance it a bit more to isolate the decision and the execution as 2 separate but connected items in the outcome.
Overall, this writer’s feeling is that the execution had a much larger impact on the result than the decisions and the value in making different decisions would likely not move the needle hugely on if Leinster had won or not.
The outcome is probably going to justify if your decision was correct or not. That is where we currently sit. Nothing I say will justify the decision we made because the outcome didn’t go our way. Jacques Nienaber – Leinster Senior Coach
All the same it’s always an interesting exercise in investigating these decisions and the psychology that goes into them. Leo Cullen and Nienaber are both quoted as saying that the call didn’t come from the coaches box on either penalty. I personally find this interesting, particularly in contrast to the NFL where Offensive Co-Ordinators call the majority of plays at any time. And they have graphs drawn and analysts dedicated to assisting on 4th down decision making.
