Rugby World Cup 2019 predictions

[bushytop]

With that win ratio it's a wonder why you guys ever think you win

Considering the ratio deficit in terms of population, or even registered players... not to mention finances (though I just did) we're doing aight!

 

[Brigantine]

Interestingly, of the 16 games between Wales and England since 2010, Wales have won six, four of which came when they were ranked lower thank England. So you'd have had more luck predicting a Welsh victory by flipping a coin than using the rankings.

[...]would have achieved the exact same degree of predictive accuracy by simply assuming that England will win every time

This sounded interesting, so I had to check it out... Here are the match results along with England's chance of victory implied by the rankings at the time, factoring in home advantage:
(venue is in bold when home advantage means the lower ranked team is favoured. Score is in bold when the winning margin is high enough (15+) to qualify for the 1.5x weighting on Rankings Points won)

England lost 6 13 v Wales Millennium Stadium 17 Aug 2019 - The rankings said: England behind 1.55 RP + 3 HA -> 27% chance
England won 33 19 v Wales Twickenham 11 Aug 2019 - England behind 3.69 RP - 3 HA -> 47% chance [wrong by 3pp]
England lost 13 21 v Wales Millennium Stadium 23 Feb 2019 - England behind 3 HA - 0.47 RP -> 37% chance
England won 12 6 v Wales Twickenham 10 Feb 2018 - England ahead 7.44 RP + 3 HA -> 100% chance
England won 21 16 v Wales Millennium Stadium 11 Feb 2017 - England ahead 7.47 RP - 3 HA -> 72% chance
England won 27 13 v Wales Twickenham 29 May 2016 - England ahead 0.88 RP + 3 HA -> 69% chance
England won 25 21 v Wales Twickenham 12 Mar 2016 - England ahead 3 HA - 1.80 RP -> 56% chance
England lost 25 28 v Wales Twickenham 26 Sep 2015- England ahead 0.41 RP + 3 HA -> 67% chance [wrong by 17pp]
England won 21 16 v Wales Millennium Stadium 6 Feb 2015 - England ahead 3.21 RP - 3 HA -> 51% chance
England won 29 18 v Wales Twickenham 9 Mar 2014 - England ahead 4.86 RP + 3 HA -> 89% chance
England lost 3 30 v Wales Millennium Stadium 16 Mar 2013 - England ahead 3.52 RP - 3 HA -> 53% chance [wrong by 3pp]
England lost 12 19 v Wales Twickenham 25 Feb 2012 - England ahead 0.88 RP + 3 HA -> 69% chance [wrong by 19pp]
England lost 9 19 v Wales Millennium Stadium 13 Aug 2011 - England behind 3 HA - 1.66 RP -> 43% chance
England won 23 19 v Wales Twickenham 6 Aug 2011 - England ahead 2.93 RP + 3 HA -> 80% chance
England won 26 19 v Wales Millennium Stadium 4 Feb 2011 - England ahead 5.44RP - 3 HA -> 62% chance
England won 30 17 v Wales Twickenham 6 Feb 2010 - England ahead 0.83 RP + 3 HA -> 69% chance

Of the last 16 games, if you had flipped a coin you would be right 8/16 times,
if you had bet on England every time you would be right 10/16 times,
if you had bet on the home team every time, you would be right 10/16 times,
if you had bet on the WRR favoured team, you would be right 12/16 times, (including 4/6 where Wales won).

More importantly, the calibration:

• England had an 80-100% implied chance 3 times. You would expect them to win ~90% (i.e. 2.7) of them, they won 3.
• They had a 60-79% implied chance 6 times. You would expect them to win ~70% (i.e. 4.2) of them, they won 4.
• They had a 40-59% implied chance 5 times. You would expect them to win ~50% (i.e. 2.5) of them, they won 3.
• They had a 20-39% implied chance twice. You would expect them to win ~30% (i.e. 0.6) of them, they won 0.

The rankings did a pretty damn good job for those 16 games if you ask me.

Just for curiosity, I went all the way back to the introduction of the WR Rankings:

England lost 15 23 v Wales Millennium Stadium 14 Feb 2009 - England behind 0.93 RP + 3 HA -> 31% chance
England lost 19 26 v Wales Twickenham 2 Feb 2008 - England ahead 11.40 RP + 3 HA -> 100% chance [big stuff up this one!] <<<<<<<<
England won 62 5 v Wales Twickenham 4 Aug 2007 - England ahead 2.73 RP + 3 HA -> 79% chance
England lost 18 27 v Wales Millennium Stadium 17 Mar 2007 - England ahead 3.44 RP - 3 HA -> 52% chance [wrong by 2pp]
England won 47 13 v Wales Twickenham 4 Feb 2006 - England ahead 0.73 RP + 3 HA -> 69% chance
England lost 9 11 v Wales Millennium Stadium 5 Feb 2005 - England ahead 9.68 RP - 3 HA -> 83% chance [wrong by 33pp]
England won 31 21 v Wales Twickenham 20 Mar 2004 - England ahead 15.78 RP + 3 HA -> 100% chance
England won 28 17 v Wales Brisbane 9 Nov 2003 - England ahead 13.32 RP + 0 HA -> 100% chance

Of all 24 games, if you had flipped a coin you would be right 12/24 times,
if you had bet on England every time you would be right 14/24 times,
if you had bet on the home team every game, you would be right 16/23 times,
if you had bet on the WRR favoured team, you would be right 17/24 times.

More importantly, the calibration:

• England had an 80-100% implied chance 7 times. You would expect them to win ~90% (i.e. 6.3) of them, they won 5.
• They had a 60-79% implied chance 8 times. You would expect them to win ~70% (i.e. 5.6) of them, they won 6.
• They had a 40-59% implied chance 6 times. You would expect them to win ~50% (i.e. 3.0) of them, they won 3.
• They had a 20-39% implied chance 3 times. You would expect them to win ~30% (i.e. 0.9) of them, they won 0.

The only time the rankings look less than perfect, is in that 2008 game. So... what happened on 2 Feb 2008?

 

[bushytop]

The only time the rankings look less than perfect, is in that 2008 game. So... what happened on 2 Feb 2008?

Well, that would be telling... wild night though that's for sure !

 

[Leinster Fan]

But the world rankings (which give a weighting to allow for home advantage) do and they are objective.

In fact they don't, if you factor in home advantage to the rankings they'd have Ireland, England, New Zealand, SA and Australia all down as favourites to beat Wales at home.

 

[bushytop]

In fact they don't, if you factor in home advantage to the rankings they'd have Ireland, England, New Zealand, SA and Australia all down as favourites to beat Wales at home.

I think he's just implying that home and away advantage/disadvantage has always been a factor and despite our away form (which is clearly worse than our home) Wales have still managed to reach the number 1 spot.

There are mitigating factors of course... during the run we played our main NH rivals at home and our SH tour was of Argentina... no one is really getting carried away here, I think @tewdric is just using the fact that we are number 1 (for the time being at least) as a bit of leverage for his banter.

 

[Umaga's Witness]

I thought the same in 2015....I may have been wrong but I wiped the entire incident from my mind.

England did well in 2015, the our up a fight against wales and Australia. They were always going to be dangerous four years later.

England didn't make it to the knock out phase of the last World Cup

yeah England didnt make it to their home World Cup quarterfinals, but they don't remember so we have to remind them.

This sounded interesting, so I had to check it out... Here are the match results along with England's chance of victory implied by the rankings at the time, factoring in home advantage:
(venue is in bold when home advantage means the lower ranked team is favoured. Score is in bold when the winning margin is high enough (15+) to qualify for the 1.5x weighting on Rankings Points won)

England lost 6 13 v Wales Millennium Stadium 17 Aug 2019 - The rankings said: England behind 1.55 RP + 3 HA -> 27% chance
England won 33 19 v Wales Twickenham 11 Aug 2019 - England behind 3.69 RP - 3 HA -> 47% chance [wrong by 3pp]
England lost 13 21 v Wales Millennium Stadium 23 Feb 2019 - England behind 3 HA - 0.47 RP -> 37% chance
England won 12 6 v Wales Twickenham 10 Feb 2018 - England ahead 7.44 RP + 3 HA -> 100% chance
England won 21 16 v Wales Millennium Stadium 11 Feb 2017 - England ahead 7.47 RP - 3 HA -> 72% chance
England won 27 13 v Wales Twickenham 29 May 2016 - England ahead 0.88 RP + 3 HA -> 69% chance
England won 25 21 v Wales Twickenham 12 Mar 2016 - England ahead 3 HA - 1.80 RP -> 56% chance
England lost 25 28 v Wales Twickenham 26 Sep 2015- England ahead 0.41 RP + 3 HA -> 67% chance [wrong by 17pp]
England won 21 16 v Wales Millennium Stadium 6 Feb 2015 - England ahead 3.21 RP - 3 HA -> 51% chance
England won 29 18 v Wales Twickenham 9 Mar 2014 - England ahead 4.86 RP + 3 HA -> 89% chance
England lost 3 30 v Wales Millennium Stadium 16 Mar 2013 - England ahead 3.52 RP - 3 HA -> 53% chance [wrong by 3pp]
England lost 12 19 v Wales Twickenham 25 Feb 2012 - England ahead 0.88 RP + 3 HA -> 69% chance [wrong by 19pp]
England lost 9 19 v Wales Millennium Stadium 13 Aug 2011 - England behind 3 HA - 1.66 RP -> 43% chance
England won 23 19 v Wales Twickenham 6 Aug 2011 - England ahead 2.93 RP + 3 HA -> 80% chance
England won 26 19 v Wales Millennium Stadium 4 Feb 2011 - England ahead 5.44RP - 3 HA -> 62% chance
England won 30 17 v Wales Twickenham 6 Feb 2010 - England ahead 0.83 RP + 3 HA -> 69% chance

Of the last 16 games, if you had flipped a coin you would be right 8/16 times,
if you had bet on England every time you would be right 10/16 times,
if you had bet on the home team every time, you would be right 10/16 times,
if you had bet on the WRR favoured team, you would be right 12/16 times, (including 4/6 where Wales won).

More importantly, the calibration:

• England had an 80-100% implied chance 3 times. You would expect them to win ~90% (i.e. 2.7) of them, they won 3.
• They had a 60-79% implied chance 6 times. You would expect them to win ~70% (i.e. 4.2) of them, they won 4.
• They had a 40-59% implied chance 5 times. You would expect them to win ~50% (i.e. 2.5) of them, they won 3.
• They had a 20-39% implied chance twice. You would expect them to win ~30% (i.e. 0.6) of them, they won 0.

The rankings did a pretty damn good job for those 16 games if you ask me.

Just for curiosity, I went all the way back to the introduction of the WR Rankings:

England lost 15 23 v Wales Millennium Stadium 14 Feb 2009 - England behind 0.93 RP + 3 HA -> 31% chance
England lost 19 26 v Wales Twickenham 2 Feb 2008 - England ahead 11.40 RP + 3 HA -> 100% chance [big stuff up this one!] <<<<<<<<
England won 62 5 v Wales Twickenham 4 Aug 2007 - England ahead 2.73 RP + 3 HA -> 79% chance
England lost 18 27 v Wales Millennium Stadium 17 Mar 2007 - England ahead 3.44 RP - 3 HA -> 52% chance [wrong by 2pp]
England won 47 13 v Wales Twickenham 4 Feb 2006 - England ahead 0.73 RP + 3 HA -> 69% chance
England lost 9 11 v Wales Millennium Stadium 5 Feb 2005 - England ahead 9.68 RP - 3 HA -> 83% chance [wrong by 33pp]
England won 31 21 v Wales Twickenham 20 Mar 2004 - England ahead 15.78 RP + 3 HA -> 100% chance
England won 28 17 v Wales Brisbane 9 Nov 2003 - England ahead 13.32 RP + 0 HA -> 100% chance

Of all 24 games, if you had flipped a coin you would be right 12/24 times,
if you had bet on England every time you would be right 14/24 times,
if you had bet on the home team every game, you would be right 16/23 times,
if you had bet on the WRR favoured team, you would be right 17/24 times.

More importantly, the calibration:

• England had an 80-100% implied chance 7 times. You would expect them to win ~90% (i.e. 6.3) of them, they won 5.
• They had a 60-79% implied chance 8 times. You would expect them to win ~70% (i.e. 5.6) of them, they won 6.
• They had a 40-59% implied chance 6 times. You would expect them to win ~50% (i.e. 3.0) of them, they won 3.
• They had a 20-39% implied chance 3 times. You would expect them to win ~30% (i.e. 0.9) of them, they won 0.

The only time the rankings look less than perfect, is in that 2008 game. So... what happened on 2 Feb 2008?

it was their first game under warren gatland, warren the maligned by English fans. How appropriate

 

[TheOvalBall]

It would always be nice if the Rugby Championship winner meets the Six Nations winner in the final.

 

[FrenchFan]

France are way better than Argentina.

Not Brunel's France

 

[Johnny Storm]

I think he's just implying that home and away advantage/disadvantage has always been a factor and despite our away form (which is clearly worse than our home) Wales have still managed to reach the number 1 spot.

There are mitigating factors of course... during the run we played our main NH rivals at home and our SH tour was of Argentina... no one is really getting carried away here, I think @tewdric is just using the fact that we are number 1 (for the time being at least) as a bit of leverage for his banter.

Actually I think that the number of home games that Wales get each year is something that many Welsh supporters seem to take for granted. More than 50% of their rugby is played at home. In any year there are 6 to 8 test matches played in the Principality stadium and any rugby fan anywhere in Wales could easily make it to every match if they wanted to. Even driving down from North Wales is just a 4 or 5 hour trip down to Cardiff. Swansea, Newport and other cities are much closer and accessible to/from Cardiff by train.

SH teams by contrast seem to tour/travel more and about 60% of their matches are played abroad. The games that are played at home are then divided amongst several cities: Perth/Brisban/Sydney for Australia, Wellington/Auckland/Christchurch for NZ, Cape Town/Durban/Pretoria etc for SA.

I think that this near constant "home advantage" has been a big factor in Wales recent run of success.

 

[bushytop]

Actually I think that the number of home games that Wales get each year is something that many Welsh supporters seem to take for granted. More than 50% of their rugby is played at home. In any year there are 6 to 8 test matches played in the Principality stadium and any rugby fan anywhere in Wales could easily make it to every match if they wanted to. Even driving down from North Wales is just a 4 or 5 hour trip down to Cardiff. Swansea, Newport and other cities are much closer and accessible to/from Cardiff by train.

SH teams by contrast seem to tour/travel more and about 60% of their matches are played abroad. The games that are played at home are then divided amongst several cities: Perth/Brisban/Sydney for Australia, Wellington/Auckland/Christchurch for NZ, Cape Town/Durban/Pretoria etc for SA.

I think that this near constant "home advantage" has been a big factor in Wales recent run of success.

Your somewhat missing the point in so much that home and away is accounted within the rankings formula... therefore you don't actually acquire as many points for a home victory as you would for an away, though you lose more points if you lose at home than you would if you lost away... all adjusted for the ranking of the team that you're playing of course.

In regards to the 'amount' of home games that Wales play. We do 'normally' add an additional 4th autumn international match... So we play four at home in that window. Though we don't play AIs in a world cup year (apart from this year due to Gatland returning to Wales with the Baa Baas)

We normally play 3 matches on a summer tour (excluding world cup years when we play warm ups home and away) so there are three away fixtures.

And then depending on the 6N schedule, so this year for instance we played 2 at home and 3 away.

So it is only really the extra November international that is out of balance (unless you count the world cup warm ups, which you shouldn't as although we aren't touring, we also don't play AIs on a world cup year).

 

[RedruthRFC]

Your somewhat missing the point in so much that home and away is accounted within the rankings formula... therefore you don't actually acquire as many points for a home victory as you would for an away, though you lose more points if you lose at home than you would if you lost away... all adjusted for the ranking of the team that you're playing of course.

Agreed. I suppose you could make a case that home advantage is more significant for sides that play almost all of their matches at the same venue as opposed to the SH teams who play at multiple venues. As you say, the algorithm accounts for home advantage, so shouldn't favour a team that plays every single team at home versus another that plays all of their matches away from home.

 

[dullonien]

Actually I think that the number of home games that Wales get each year is something that many Welsh supporters seem to take for granted. More than 50% of their rugby is played at home. In any year there are 6 to 8 test matches played in the Principality stadium and any rugby fan anywhere in Wales could easily make it to every match if they wanted to. Even driving down from North Wales is just a 4 or 5 hour trip down to Cardiff. Swansea, Newport and other cities are much closer and accessible to/from Cardiff by train.

SH teams by contrast seem to tour/travel more and about 60% of their matches are played abroad. The games that are played at home are then divided amongst several cities: Perth/Brisban/Sydney for Australia, Wellington/Auckland/Christchurch for NZ, Cape Town/Durban/Pretoria etc for SA.

I think that this near constant "home advantage" has been a big factor in Wales recent run of success.

Not convinced a 4 or 5 hour car journey from North Wales to Cardiff makes it easy to attend every international! A night in a hotel would be a must for one thing which aint cheap on match weekends. Other countries who may be much larger either have far better road/train networks, and will often have the option of internal flights to cut the time down to a fraction of the time. Many away games are easier to attend than driving the length of Wales!

 

[FrenchFan]

Are they? France had a **** poor 6n this year bar the Scotland game and regularly drop games to T2 nations. That games a dead heat in my opinion, France at their best will win handy but France at their best is a rare thing.

Agreed. 4mths of preparation 3 friendlies lots of blabla yet les Bleus WC hinges on the one (and first) game v Argentina. Lose and they're out.

 

[Dai Perk]

Hmmm...

I just downloaded the calendars for the Pro14 and the Prem for this coming season. The overlap with the RWC is pretty jarring - should I be pessimistic and think, "Well, the first 6 - 8 weeks of those leagues is gonna suck," or go optimistic and think, "What a great opportunity for some of the younger guys to get a chance to shine, and maybe we will discover the next generation of Internationals"?

 

[bushytop]

Hmmm...

I just downloaded the calendars for the Pro14 and the Prem for this coming season. The overlap with the RWC is pretty jarring - should I be pessimistic and think, "Well, the first 6 - 8 weeks of those leagues is gonna suck," or go optimistic and think, "What a great opportunity for some of the younger guys to get a chance to shine, and maybe we will discover the next generation of Internationals"?

The Pro 14 will be hit harder by this undoubtably but as you say... opportunities abound.

 

[Brigantine]

Not convinced a 4 or 5 hour car journey from North Wales to Cardiff makes it easy to attend every international! A night in a hotel would be a must for one thing which aint cheap on match weekends. Other countries who may be much larger either have far better road/train networks, and will often have the option of internal flights to cut the time down to a fraction of the time. Many away games are easier to attend than driving the length of Wales!

From what I saw in North Wales, the only rugby they care about is the Gogs anyway. They don't fit in like Ulster does. South Wales? May as well be New South Wales.
Any of the Welsh posters on here from the north?

 

[bushytop]

From what I saw in North Wales, the only rugby they care about is the Gogs anyway. They don't fit in like Ulster does. South Wales? May as well be New South Wales.
Any of the Welsh posters on here from the north?

At least they're actually part of the same country though right.

Slightly off topic but there are conflicting opinions about what North Wales could bring to Welsh Rugby (& what Welsh Rugby could do for them). I keep hearing that there are large amounts of gogs that make the trip to the capital for home internationals and living in Cardiff as I do I see the affluent Pontcanna area becoming more and more 'Welsh' (welsh speakers as are a lot of gogs) by the year. I keep saying it but it ain't getting old, in a country of only 3 million(ish) we shouldn't be cutting out a third of it.

 

[Johnny Storm]

Not convinced a 4 or 5 hour car journey from North Wales to Cardiff makes it easy to attend every international! A night in a hotel would be a must for one thing which aint cheap on match weekends. Other countries who may be much larger either have far better road/train networks, and will often have the option of internal flights to cut the time down to a fraction of the time. Many away games are easier to attend than driving the length of Wales!

When I lived in Cardiff, I regularly drove up to Wre

Not convinced a 4 or 5 hour car journey from North Wales to Cardiff makes it easy to attend every international! A night in a hotel would be a must for one thing which aint cheap on match weekends. Other countries who may be much larger either have far better road/train networks, and will often have the option of internal flights to cut the time down to a fraction of the time. Many away games are easier to attend than driving the length of Wales!

I live in Cardiff and I regularly used to drive up north to Wrexham, Colwyn Bay, caernafon etc, do a day's work and then drive back down again. It was a long day, sometimes I stayed over, but mostly I did not. I used to live just off Sophia Gardens and before I had a car I would advertise and rent out my parking space on match days to you guessed it... people from North Wales who would fill a car with 4 and drive down for the match. They would be gone by 6:00pm.

I have to say it again. The Welsh supporters don't know how good they have it when it comes to watching Wales play at home. Driving down from anorth Wales for a day is a lot easier than flying from Perth to Brisbane to watch Australia play a match.

 

[TRF_Olyy]

I hope the WRU don't try and improve their ties with North Wales because it's basically been our extra unofficial academy catchment for the last decade

 

[Paulo]

I think he's just implying that home and away advantage/disadvantage has always been a factor and despite our away form (which is clearly worse than our home) Wales have still managed to reach the number 1 spot.

There are mitigating factors of course... during the run we played our main NH rivals at home and our SH tour was of Argentina... no one is really getting carried away here, I think @tewdric is just using the fact that we are number 1 (for the time being at least) as a bit of leverage for his banter.

Wales and and New Zealand have the relative rise and fall of NH and SH rugby respectively to thank for the current positions. With the rise of Ireland and England, Wales has gotten more rank points for beating them. With the decline of Aus and SA down the rankings comparatively, a loss or a draw to these teams really hurts NZ in ranking points. Without the hiding Aus have NZ in Perth, they would still be No1. Inspite of all of Wales rightful success.

The rankings are objective, as has been pointed out, as they do treat all teams equally and apply weighting's as per home/away, points differential, and rank position as standard. There is perhaps an arbitrary-ness element to what those weighting's are.

They do also make assumptions in a flat manner. Home advantage is home advantage, but not all home games are equal. 15pts in front is the same as 30pts or 40pts. As an example, is 47-26 more difficult than 37-0? Each are treated the same but one is much harder than the other to achieve.

The rank weighting's based on relative position is I think flawed more so now than perhaps in the past. In the past the difference between 1 and 6, was a gulf, now, any top 6-7team could arguably win against another top team. The relative gap is smaller now but the weighting for a win or loss is the same. I think that is not reflective on the level-ness of the top teams.

A solution could be potentially 2 tables for the NH and SH sides, to remove the lack of games elements between teams. This is simple but a bit clunky.

I more sophisticated solution is to add an element of head to head into the algorithm. If a team earns enough points to overtake another team, a head to head check is done, if they played in the last 2 yrs, the ascending team won, then they take the position outright. If they lost, or haven't played in 2yrs, they share the position until either team earns enough points to ascend or descend to another position. In which the head to head check is done again.

In that scenario, Wales and NZ would be first equal, as they haven't played in a while. England would be 3rd. If England earned enough points to over take NZ, Wales takes too spot on their own, and the head to head check is done, Eng and NZ become 2nd equal.

Just my 2cents anyway. Just glad Wales goes into the World Cup as Favourites this time around.