Last season, Munster’s up-and-down nature — week to week, month to month — was broadly predictable if you ignored the scores (always fun when you’re losing) and looked at the one indicator I keep coming back to: net efficiency in the 22. How often you get into the opposition 22 and how well you score when you’re there, set against the reverse — how often you let them into yours, and how well you defend when they arrive.

Across 24 games — URC and Europe — it came out at −0.05 points per entry; basically zero. We got into the 22 slightly more often than the opposition got into ours and came away with slightly less each time. That’s a fair summary of the season, and it’s not all that complimentary, but we know that, at this stage.

The overall figure hides the shape of the season, though, and the shape is the interesting part. I pooled the points and entries over every five-game stretch rather than averaging each game’s net figure. The rolling line spent most of the middle of the season below zero. It bottomed out at −1.01 after the Sharks, the Bulls and Exeter, the point where we were giving away more per visit than we were taking, by a distance. Then it recovered, and we went into the quarter-final in Pretoria at +0.72, the best it had been all season.

Which brings us to Saturday.

Five entries. Four points per entry. All twenty of our points came from those five visits to the 22, and that’s a conversion rate well above the elite line of 2.8–3.0. Net efficiency for the game was +1.64, and the rolling figure barely moved: it’s +0.46.

We still lost.

Glasgow got into our 22 eleven times, and at 2.36 points per entry, they didn’t need to be clinical. They just needed to keep turning up, and eventually, they’d get the reward. In the last 15 minutes, they did just that. Efficiency isn’t our problem right now. We spent most of Saturday in our own half.

It’s one game, and five entries is a small sample to build much on. But it looks a lot like the Scotstoun game last season: six entries to Glasgow’s twelve, better efficiency, and a loss. So the question is whether the rolling line still tells us what it told us last year, or whether it’s missing half of what’s going wrong.

Much of this game boiled down to one thing: Glasgow had too much penalty access to our 22. In a way, they are set up to min-max towards that. They kick long to land the territory, fill the field to contain you and force a kick back, and then kick every single penalty they win down the line — every single one — until you break, or they do.

Get enough entries and, on the balance of averages, you’ll overcome most teams. Give them repeated 22 entries on penalty access, and you’ll eventually get the punishment you asked for. That’s what happened on Saturday, if you boil it down to just the nuts and bolts of what usually wins a game, and what usually loses it.

Two numbers, not one

The short answer is that the rolling line was never meant to tell us everything. It’s a ratio, and ratios don’t care about volume. Five entries at 4.00 and eleven at 2.36 give you a net figure of +1.64. That looks great until you remember the scoreboard. What it can’t show you is that Glasgow got to knock on the door more than twice as often as we did.

So you have to read it next to the entry differential. You can split any result into those two parts. One is what you gained or lost by getting there more or less often than the other side. The other is what you gained or lost by doing more or less with each visit. The two add up exactly to the points margin, so nothing gets lost in between.

On Saturday, the six-entry deficit cost us about 19 points. Our finishing won back about 13. That leaves six, which is the margin by which we lost. Number genius over here.

Scotstoun last season worked out almost the same. The territory cost us about 18, the finishing won back about 10, and we lost by around eight on the derived points. We lost to Glasgow in broadly the same way as we did last season, albeit with them being slightly stronger and us being as weak, or weaker.

I ran the entry differential as a five-game rolling line, pooled the same way as the efficiency line. It tells a different story about the end of last season.

After Benetton, we were getting into the 22 two more times a game than the opposition. Then it slid. By the quarter-final in Pretoria it was −1.8, and after Saturday it’s −3.6. That’s the worst it’s been at any point across the 25 games, even allowing for the change in the new season and all the start up issues that come with that. The previous low was −2.6, which was, of course, after Scotstoun.

So the efficiency line climbing to its season-best +0.72 wasn’t the sign of recovery it looked like. Over those same weeks, we were being pushed further back into our own half, efficiency-wise. The two lines crossed and went in opposite directions.

Last season, whenever the opposition got into our 22 four or more times than we got into theirs, we lost. That happened seven times: Ulster away, Toulon, Glasgow, the Sharks, Connacht away, the Bulls in the quarter-final, and now Glasgow again. We lost all seven. (Toulon is close enough on the derived points that the rounding could move it. The result didn’t.) Leinster away at −3 is the only game all season where we were clearly out-entered and still won.

It’s not that efficiency doesn’t matter. Across the URC season, it was the thing keeping us afloat. Pooled over 19 games, finishing was worth about +2.5 points a game and territory cost about −1.7. But there’s a limit. Once the entry gap gets past three or four, no realistic finishing rate makes up for it. To match Glasgow’s 26 from five visits, we’d have needed 5.2 points per entry. That’s more than a converted try every time.

It isn’t a target so much as it’s a miracle. We won’t sustain it; nobody could. Just ask Ulster last year.

The identity, again

None of this should be a surprise. In the last piece, I said our problem was one of conversion identity. We were, at our best last season, living off scraps: fewer entries, often away from home, turned into points at a high rate. Six of our eight games above 3.0 points per entry came away from home, with fewer entries than the opposition.

On Saturday, it looked a lot like this, but it wasn’t. The scores we had came from set-piece access. When our set piece degraded, so did our scoring efficiency and our ability to halt Glasgow’s.

From here on, the efficiency line gets a partner on the chart: rolling entry differential, pooled over the same five games. If both lines are above zero, we’re winning games. If efficiency is up and entries are down, we get Saturday: a close game that we usually lose. And if both are down, well, you already know how that goes.

What fixes it? Better depth. An 80-minute scrummaging performance. Better discipline. Do that, and the sliders change, and the results change along with them. Those data points point to winning the space between the 22s. If we can do that, the entries will come, and then we see if a new trend line is emerging.