I didn’t state they “were super accurate”. I just reported how well the model fitted the results data.
But I could have reported the mean absolute error, which is basically the forecasting error due to randomness.

Happy to hear about the calls or incidents that went against Penrith if you can list them.

The biggest factor determining Souths’ performance was the head knocks in the 1st half; Murray, Arrow & Gagai.

It forced Souths to operate with one man down for a large chunk of the match and another two that were operating below their best. Murray was clearly not right after that very first tackle. Just because he stayed on the field, does not make the case that he was OK. He played incredibly, but I thought he was a bit of passenger for 20 mins or so after that knock. Souths also lost Gagai for a spell and also I thought he did not play with his usual high energy after that head knock.

When you have those head knocks and lose players, it forces you to burn the petrol of the replacement players earlier than you would like. The end result is the team has less energy to compete at the end of each half. That was a big contributor to Penrith’s dominance of possession and territory.

Three critical bunker calls also all went Penrith’s way.

One, the Luai strip; could have easily been ruled the other way. Two the Gagai high shot; many of these have been penalised recently, including in finals matches. Three, the Reynolds charge down; I’m struggling to recall a similar case where the referee did not rule six again, yet there is a plethora of examples in recent seasons where any players charging down the kicker is deemed to have played at the ball and is 6 again. There is nothing in the rules that allows the bunker official to determine intent. The defender only has to make contact with the ball with his arm for it be classified as a knock-on. You won’t find intent in the rules there.

The combination of head knocks and those three contentious bunker calls was the luck that allowed Penrith to hold out Souths. The fact that Souths managed to get so close to Penrith despite those unlucky breaks, shows how good they really were.

I think if you reversed those head knocks and bunker calls, Souths probably would have won by a lot further than two points.

No mention in your analysis of the 3 crucial head knocks that Souths suffered in the 1st half: Cameron Murray (not HIA’ed but clearly affected), Dane Gagai and Jai Arrow. Three of Souths’ best players taken out by head knocks which clearly handicapped their 1st half performance, yet you don’t manage to identify this, whilst pointing out Souths’ underperformance in the first half. Losing forwards early in the match is massive. It drains you of petrol and disrupts your bench rotation. You either don’t know the game or are biased. I don’t know which one.

Last week Penrith got lucky against Melbourne losing key players early via head knocks. Last night a similar situation occurred. But you don’t seem to be able to join the dots here.

Good article Tim. I agree with your main thrust because as you say, the game has changed significantly in 2021, so historical stats based on points are not relevant (unless you went to the trouble of normalising all the pre-2021 scores based on 2021 scoring rates).

The two best predictors of finals success are 1) the week off and 2) strength of recent form (post-Origin period).

Souths tick both of those boxes. Penrith on the other hand may only just tick the strength of recent form box because I thought they were lucky to get away with the win against Melbourne. Only injuries to key Melbourne forwards in the first 20 mins allowed Penrith to win the match (not to mention Melbourne bombing two tries).

Without going into detail SM, I have found that over the last 10 years, a power rating of form post Origin is a better predictor of GF performance than a whole of season power rating.

To me this makes sense. Teams can be very good going into Origin, falter on the way out and underperform in the finals series.

Conversely, team can underperform in the first half of the season, then improve in the 2nd half in the run into the finals after Origin.

Both of these scenarios would support a post Origin form analysis being superior to a whole of season approach.

It is very hard for a team to be consistently very good throughout the whole season. The trick is to be peaking at the right time. I think Wayne Bennett knows this and it’s why he has Souths playing their best football of the season at the moment.

LBC indicates defence
—
Line breaks, not surprisingly are highly correlated with tries. So, teams with a low average LBC will have good defence. One of the truisms of rugby league is that defence wins matches, more so in the big games where the quality of the teams are higher.
KRM indicates attack and energy
—
Teams with strong KRM’s usually produce strong metres gained in their sets. This then usually translates into more attacking opportunities and in turn tries. KRM also governs a team’s energy level. Teams with a strong KRM means that forwards will have less ground to cover returning upfield after an opposition clearing kick. This will then translate into more energy for attack and defence particularly in the back end of each half as fatigue starts to set-in.
NKM represents territory
—
When I first discovered that net kick metres was strongly correlated with margin of victory I was a bit perplexed. What does how long you kick have to do with how far you win or lose by?
The answer is that NKM is a proxy for territory.
Teams that dominate territory do less long clearing kicks and more short kicks when they are in the opposition’s 20m zone. These teams will tend to produce a negative NKM value.
Conversely, teams that struggle to gain and hold territory have a positive NKM value because they will have to kick long more of the time and short less of the time.

Thanks for all the feedback.

I’m not going to address each individual comment but I would like to respond to those comments that put forward the notion that these statistics are meaningless because it is how teams play on the day that counts and none of this really matters.

There are lots of statistics that don’t really matter; they don’t really tell you anything. These are insignificant results. But there are other statistics that identify clear, consistent and strong trends. These really do matter. They are significant and have a high probability of repeating.

The statistics I have identified here for both the week off factor and strength of post Origin form are significant and strong. The game record and my analysis shows that they are strong predictors for future performances in grand finals.

You can ignore this history if you like, cling to your personal preferences and hope for the best. But that’s poor judgement and those who do not learn from history are doomed to repeat that judgement and be rewarded with what it deserves.

Fair point Barry. My mistake. This paragraph should instead read:
—
In contrast, there have been 18 teams attempt to win the premiership since 2012 with a week off and 5 of those have been successful. That’s a success rate of 27.8%. So teams with the week off have been 10 times more successful than the teams without the week off since 2012.
—
I have asked eds to amend this.

Tim, go watch again and look at Bradstreet punch Coote on the way down in the tackle. Bradstreet punches Coote and Coote punches Bradstreet. It’s all there in the video. As I said, referee Pearce obviously and justifiably saw it as tit for tat, which was the norm in those days. No amount of unsubstantiated conspiracy theories from you will change these facts.

You’ve left out a couple of crucial bits here TB3.

If you watch the replay, Bradstreet comes in with a high, stiff arm to fell Coote. Coot in response, raises his non-ball carrying arm to protect himself and in the process, makes contact with his forearm with Bradstreet’s head, much like JWH has done nearly every week for the past 8 years.

After Coote is felled and lying on the ground, Bradstreet then jumps on him with an elbow to the head. It was this that made Coote angry, get up of the deck and floor Bradstreet.

This all happened right in front of referee Pearce, who obviously saw it as tit for tat and let it go. He could have just as easily marched Bradstreet for his dirty work as he could have for Coote’s punch.

Good question.

I plan to do an amended version of this analysis on all-time try scorers in the NRL. Being a backrow forward with a high try scoring rate you would have to think he would be right up there.

But he has some hot competition like Harold Horder, who scored 102 tries in 86 games; a try scoring rate higher than Ken Irvine.

BMoz came in just outside the top 25 (#26) at a rating of +3.92%.

Small correction; in my explanation above I said:

GMP = Game Minutes Played (total game minutes divided by 4800)

This should have read:

GMP = Game Minutes Played (total game seconds divided by 4800).

Thanks guys,

Following the editor’s guidelines, I didn’t want to bore people with too much data and statistical nerdy stuff. But for those that are interested in this side of things…

The formula
The formula was derived from a multivariate, linear regression on the 2013-20 dataset. The formula extracted from this regression was:

Expected Tries = – 6.627 – (0.0184 * TT) + (0.357 * GMP) + (2.039 * PV)

where:
TT = Team Tries
GMP = Game Minutes Played (total game minutes divided by 4800)
PV = Position Value (a factor derived from the % distribution of tries across positions, with the Interchange position =1)

The Adjusted R Squared value for this formula (a statistical measure on how good the formula fits the data) on the 2-13-20 dataset was 74.70%; a pretty good result considering that the formula includes no skill factors.

Extra results data
I would like to post extra results data, that will enable you to make more sense of the ratings, but the content management system running this site provides no easy way for posters like me to publish this data. I’ll contact the eds about this.

Potential improvements to the model
There are two areas where I think the model can be improved:

1. Adjusting Team Tries to the total number of tries that were scored when the player was on the field.

80-minute players are not affected by this in the current results. But less than 80-minute players are because tries that were scored when they were on the bench are counted as an opportunity against them, when in fact they weren’t (you can’t be expected to score a try when you are on the bench).

2. Using Tackled in Opposition 20 data

I think this would be a strong opportunity variable to use. But until I get the data and test it, I don’t know if that would really be the case or not.

When tries are scored
Daniel, you would first have to calculate the distribution of tries across game time, to see if there were any potential advantages or disadvantages here for players playing less than 80 mins. If there were differences, then you would have to do a significance test to see if these results were just a fluke (ie randomness) or if there was something more to them.

But potentially yes, you could have something to use there to improve the model.

Further updates
If people are interested, I’d be happy to post an updated table of results after each round.