RugbyRebels

Looking at the world rankings before the start of the summer internationals, and the position at the end of it, there's surprisingly little difference in table positions apart from Wales jumping 3 places at the expense of Australia, Scotland and SA (who stay in the same order who drop 1 place each), Fiji & Argentina swap places and Georgia & Tonga swap places. That's about it. Interesting to see how close the USA and Italy now are in points. Wonder who would win a US vs Italy game played in the US?

With SA winning their series, I'm surprised to see them drop. Scotland obviously heavily penalised by that loss to the US wiping out any gains from the annihilation of Argentina

28/05/2018 - before summer tours
Ranking Team Start
1 NEW ZEALAND 93.99
2 IRELAND 89.11
3 ENGLAND 86.23
4 AUSTRALIA 85.49
5 SCOTLAND 83.83
6 SOUTH AFRICA 83.81
7 WALES 83.41
8 FRANCE 79.1
9 ARGENTINA 78.22
10 FIJI 77.93
11 JAPAN 75.66
12 GEORGIA 73.96
13 TONGA 71.87
14 ITALY 71.1
15 USA 69.23


25/06/2018 - end of summer tours
Ranking Team End Change
1 NEW ZEALAND 93.99 0
2 IRELAND 90.12 1.01
3 ENGLAND 86.35 0.12
4 WALES 85.94 2.53
5 AUSTRALIA 84.49 -1
6 SCOTLAND 83.02 -0.81
7 SOUTH AFRICA 82.65 -1.16
8 FRANCE 79.1 0
9 FIJI 76.55 -1.38
10 ARGENTINA 75.54 -2.68
11 JAPAN 75.25 -0.41
12 TONGA 73.84 1.97
13 GEORGIA 73.13 -0.83
14 ITALY 72.56 1.46
15 USA 71.66 2.43

Odd quirk - by losing the first two tests, we gave SA a lot of points and made it so that the rankings thought it was a big accomplishment to win away. If we'd won the first and lost the last two, then we'd've ended up worse off.

I think the US would take Italy on home soil as things stand -they really are starting to get things together now.

Puja

Puja wrote:Odd quirk - by losing the first two tests, we gave SA a lot of points and made it so that the rankings thought it was a big accomplishment to win away. If we'd won the first and lost the last two, then we'd've ended up worse off.

Puja

It was Jones’s plan all along.

Generally I think the ranking system is good but I would suggest one tweak. A 3 test series should be rated as a whole, rather than 3 separate points exchanges. A team that wins a 2 - 1 series because it experimented in a dead rubber should not be rated the same as a team that pulled one out at 1 - 1. They rate RWC matches differently so why not series?

Lizard wrote:Generally I think the ranking system is good but I would suggest one tweak. A 3 test series should be rated as a whole, rather than 3 separate points exchanges. A team that wins a 2 - 1 series because it experimented in a dead rubber should not be rated the same as a team that pulled one out at 1 - 1. They rate RWC matches differently so why not series?

Hmm. The rankings system is always going to be subject to dead rubbers and experimental teams, no matter what though. A chunk of Wales's ranking points gains came from beating SA, despite it being a nonsense game between two second strings. I wouldn't footle with it too much myself.

Puja

Puja wrote:

Lizard wrote:Generally I think the ranking system is good but I would suggest one tweak. A 3 test series should be rated as a whole, rather than 3 separate points exchanges. A team that wins a 2 - 1 series because it experimented in a dead rubber should not be rated the same as a team that pulled one out at 1 - 1. They rate RWC matches differently so why not series?

Hmm. The rankings system is always going to be subject to dead rubbers and experimental teams, no matter what though. A chunk of Wales's ranking points gains came from beating SA, despite it being a nonsense game between two second strings. I wouldn't footle with it too much myself.

Puja

Yeah. Pretty impossible to get a ranking system that's both accurate and relatively simple. This is a good compromise. Some teams will always rank higher than they should and others lower than they should. That's the nature of the game.

Lizard wrote:Generally I think the ranking system is good but I would suggest one tweak. A 3 test series should be rated as a whole, rather than 3 separate points exchanges. A team that wins a 2 - 1 series because it experimented in a dead rubber should not be rated the same as a team that pulled one out at 1 - 1. They rate RWC matches differently so why not series?

I think I've suggested that before on here. For there to be significant differences in the ranking result as a consequence of the sequence of match results seems odd - and I say that knowing that Ireland aced it by getting their loss out first before 2 wins.

Actually Lizard how significant is the difference?

Good question. Let's take the Aust v Ire series as the example.For rankings points purposes, the 2nd and 3rd tests were the same result - a win by 15 or less for Ireland. The difference therefore is which test Australia won:

Actual results (i.e. Aussie win first test):
Before 1st test: Ire 89.11 (2nd); Aus 85.49 (4th)
After 1st test (Au win by <15): Ire 88.05 (2nd); Aus 86.56 (3rd)
After 2nd test (Ire win by <15): Ire 89.20 (2nd); Aus 85.41 (4th)
After 3rd test (Ire win by <15): Ire 90.12 (2nd); Aus 84.48 (5th)

If Aussie had lost tests 1 & 3, but won 2nd test:
Before 1st test: Ire 89.11 (2nd); Aus 85.49 (4th)
After 1st test (Ire win by <15): Ire 90.05 (2nd); Aus 84.55 (5th)
After 2nd test (Au win by <15): Ire 88.80 (2nd); Aus 85.80 (4th)
After 3rd test (Ire win by <15): Ire 89.80 (2nd); Aus 84.80 (5th)

If Aussie had lost tests 1 & 2, but won dead rubber:
Before 1st test: Ire 89.11 (2nd); Aus 85.49 (4th)
After 1st test (Ire win by <15): Ire 90.05 (2nd); Aus 84.55 (5th)
After 2nd test (Ire win by <15): Ire 90.80 (2nd); Aus 83.80 (6th)
After 3rd test (Au win by <15): Ire 89.40 (2nd); Aus 85.20 (5th)

So the differences are not huge but they are there. It does seem counter-intuitive that losing a mere dead rubber in a series you have already won leaves you with fewer rankings points than having dropped one of the first two and having to win the decider to clinch the series. I would argue that forcing a decider but losing should gain you more points than winning a dead rubber. I would also say that at 1-1 in a series, the ranking points should be the same regardless of who won which test.