Performance relative to World Rugby rankings

The World Rugby rankings were developed based on a statistical analysis of all the international games played up to the point they were introduced, with the idea that comparing the ranking scores should provide a good indication of who would win any given match. So I thought it would be interesting to look at the difference in the actual scores relaive to the difference in the teams' rankings.

The results are interestingly consistent.

I can't tabulate on my phone, so I've put match points difference first and then ranking points difference.

Eng - Fiji +24 +8.00 (+11 if allowing for England home advantage)
Fra - Ita +22 +10.59
Ire - Can +43 +19.23
Tga - Geo -7 +6.53
RSA - Jpn -2 +13.09
NZ - Arg +10 +14.5
Sam - USA +9 +4.78
Wal - Uru +45 +22.52 (+25.52 for home advantage)

What I found interesting was that in 5 of the games (including England's if you give them the regulation 3 points for home advantage), the difference in score was roughly twice the difference in ranking. It will be interesting to see if that holds up as the tournament progresses, but it suggests that most of the winning teams performed at par. If that does hold true, it might be a good indicator of the spread to target...

The anomalies are obviously the two upsets, but also NZ and to some extent England and Wales. By the 'twice the ranking difference' rule of thumb, Tonga underperformed by about 20 points, NZ by about 19 (but at least they got the win) and South Africa by a whopping 29 points.

Wales were at home as well, so arguably they underperformed to the tune of about 5 points. England, in contrast, were very slightly ahead of par allowing for home advantage... but where it gets interesting is that they will have home advantage for all their key games. So in the context of this tournament, it makes sense to discount it and just look at the raw difference - because England will always have that home field edge, and the other teams won't, except for Wales against Fiji. Viewed that way, they outperformed by about 8 points - not huge, but enough to make a case that they performed slightly better than most on a weekend when big teams generally didn't excel

??
Hmmmm

Let me read it again. I always sat at the back for maths....

Having reviewed the stats and performed a T test and Correlation Coefficient I've calculated that Wales were absolute Shoite today. No tempo no urgency no ball retention and no hope. Tipuric aside there were 14 statistical anomalies on the field

RubyGuby wrote:Having reviewed the stats and performed a T test and Correlation Coefficient I've calculated that Wales were absolute Shoite today. No tempo no urgency no ball retention and no hope. Tipuric aside there were 14 statistical anomalies on the field

I thought S Williams, Amos were ok

It seems accurate for Wales as we did underperform (FYI I know there are some stalker posters that will take offence to this statement - I'm letting you know in advance that I don't give a sh*t). Given changes disrupt us then yes 5 points is fair, though I expected us to score 60-70 points at least.

Poorfour wrote:The World Rugby rankings were developed based on a statistical analysis of all the international games played up to the point they were introduced, with the idea that comparing the ranking scores should provide a good indication of who would win any given match. So I thought it would be interesting to look at the difference in the actual scores relaive to the difference in the teams' rankings.

The results are interestingly consistent.

I can't tabulate on my phone, so I've put match points difference first and then ranking points difference.

Eng - Fiji +24 +8.00 (+11 if allowing for England home advantage)
Fra - Ita +22 +10.59
Ire - Can +43 +19.23
Tga - Geo -7 +6.53
RSA - Jpn -2 +13.09
NZ - Arg +10 +14.5
Sam - USA +9 +4.78
Wal - Uru +45 +22.52 (+25.52 for home advantage)

What I found interesting was that in 5 of the games (including England's if you give them the regulation 3 points for home advantage), the difference in score was roughly twice the difference in ranking. It will be interesting to see if that holds up as the tournament progresses, but it suggests that most of the winning teams performed at par. If that does hold true, it might be a good indicator of the spread to target...

The anomalies are obviously the two upsets, but also NZ and to some extent England and Wales. By the 'twice the ranking difference' rule of thumb, Tonga underperformed by about 20 points, NZ by about 19 (but at least they got the win) and South Africa by a whopping 29 points.

Wales were at home as well, so arguably they underperformed to the tune of about 5 points. England, in contrast, were very slightly ahead of par allowing for home advantage... but where it gets interesting is that they will have home advantage for all their key games. So in the context of this tournament, it makes sense to discount it and just look at the raw difference - because England will always have that home field edge, and the other teams won't, except for Wales against Fiji. Viewed that way, they outperformed by about 8 points - not huge, but enough to make a case that they performed slightly  better than most on a weekend when big teams generally didn't excel

why are wales getting any home games as its not the host nation ? bias ?

alive555 wrote:

Poorfour wrote:The World Rugby rankings were developed based on a statistical analysis of all the international games played up to the point they were introduced, with the idea that comparing the ranking scores should provide a good indication of who would win any given match. So I thought it would be interesting to look at the difference in the actual scores relaive to the difference in the teams' rankings.

The results are interestingly consistent.

I can't tabulate on my phone, so I've put match points difference first and then ranking points difference.

Eng - Fiji +24 +8.00 (+11 if allowing for England home advantage)
Fra - Ita +22 +10.59
Ire - Can +43 +19.23
Tga - Geo -7 +6.53
RSA - Jpn -2 +13.09
NZ - Arg +10 +14.5
Sam - USA +9 +4.78
Wal - Uru +45 +22.52 (+25.52 for home advantage)

What I found interesting was that in 5 of the games (including England's if you give them the regulation 3 points for home advantage), the difference in score was roughly twice the difference in ranking. It will be interesting to see if that holds up as the tournament progresses, but it suggests that most of the winning teams performed at par. If that does hold true, it might be a good indicator of the spread to target...

The anomalies are obviously the two upsets, but also NZ and to some extent England and Wales. By the 'twice the ranking difference' rule of thumb, Tonga underperformed by about 20 points, NZ by about 19 (but at least they got the win) and South Africa by a whopping 29 points.

Wales were at home as well, so arguably they underperformed to the tune of about 5 points. England, in contrast, were very slightly ahead of par allowing for home advantage... but where it gets interesting is that they will have home advantage for all their key games. So in the context of this tournament, it makes sense to discount it and just look at the raw difference - because England will always have that home field edge, and the other teams won't, except for Wales against Fiji. Viewed that way, they outperformed by about 8 points - not huge, but enough to make a case that they performed slightly  better than most on a weekend when big teams generally didn't excel

why are wales getting any home games as its not the host nation ? bias ?

If Roger Lewis had had his way England v Wales would have been at the Millenium Stadium or whatever it is now.

OK. So the rankings have been updated for the weekend's matches and I have used them to predict the winning margins for the coming week's games... let's see how accurate they are:

AUS 86.67 WIN 19 77.04 FJI
SCO 75.88 DRAW 0 DRAW 76.06 JPN
FRA 81.12 WIN 30 66.28 ROM
NZL 92.89 WIN 62 61.85 NAM
ARG 78.39 WIN 12 72.62 GEO
ENG 88.04 WIN 7 84.63 WAL
RSA 81.15 WIN 10 76.18 SAM
ITA 70.53 WIN 11 65.17 CAN
AUS 86.67 WIN 49 62.11 URU
SCO 75.88 WIN 13 69.32 USA
IRE 84.4 WIN 36 66.28 ROM

The ones that look most likely to vary from the prediction are:
Scotland-Japan - too close to call on paper but can Japan raise their game with only a 4 day turnaround

Argentina-Georgia - I think Argentina have suffered a bit in the rankings for having played the All Blacks, and could win more comfortably than that implies

England-Wales - 6 of the 7 points predicted margin come from England's home advantage. I'm hoping for an England win but it's a very close call.

South Africa-Samoa - The Boks should raise their game for this one. Should.

I am going to put your results into my super bru predictor (with a 7 point swing for a couple of the ones you highlight as being issues)

wolfball wrote:I am going to put your results into my super bru predictor (with a 7 point swing for a couple of the ones you highlight as being issues)

Be interesting to see how it goes. I've not thought of looking at it this way before now but in theory it should give some insights.

To compare with Skybet:

Australia Fiji Your prediction 19 Handicap 26
Scotland Japan 0 18
France Romania 30 46
NZ Namibia 62 80
Argentina Georgia 12 22
England Wales 9 9
SA Samoa 10 22
Italy Canada 11 19

So it seems almost unanimously that the handicap is wider than your prediction. Bet on the favourites to win I guess.

Scottrf wrote:To compare with Skybet:

Australia Fiji Your prediction 19 Handicap 26
Scotland Japan 0 18
France Romania 30 46
NZ Namibia 62 80
Argentina Georgia 12 22
England Wales 9 9
SA Samoa 10 22
Italy Canada 11 19

So it seems almost unanimously that the handicap is wider than your prediction. Bet on the favourites to win I guess.

Thanks for that. We should add the final points difference in as it comes in. It would also be interesting to see what the handicaps were for the weekend's matches