Fiji triumphed today, beating Georgia 45 – 10 in what was an emotional victory not just for the team, but for our ranking model as well.
The stakes of the game were high since the official rankings predicted a Georgian victory, but our model correctly anticipated Fiji being the (significantly) stronger side. Going into this match, the official ranking had Georgia one position above Fiji whereas our ranking had the Fijians ranked higher by a sizeable 5 places.
The 45 –10 margin of victory was one which our system anticipated well, as our ranking table post-match remains unchanged from it’s pre-match state.
Ranking Methods Compared
|Position||Our Rank||Official Rank|
Of the 21 matches played since the tournament began, our ranking model successfully predicted 19 (or 90%) of these outcomes. Given the official ranking was unable to successfully predict the outcome of Fiji vs Georgia, their success rate is now lower at 86%. The two matches which neither ranking method predicted are unsurprisingly the two biggest shocks of the tournament, when Uruguay defeated Fiji and Japan beat Ireland.
|Rugby World Cup||Our Rank Success||Official Rank Success|
|2019 (so far)||90%||86%|
Our ranking has been shown to out-perform the official ranking in terms of modelling expected results on the four previous world cups. So far in this tournament there is no departure from this. More information on our ranking method and testing its validity are presented in our white paper on [Download our whitepaper].
Matches that have contrasting prediction outcomes are exciting in terms of testing the success rates of each ranking method. We’ll endeavor to keep you well informed as to when the next one comes around with insights and commentary.