Dominic Thiem is faced with the task of upsetting Novak Djokovic in the final and Dan Weston gives his thoughts on that possibility...
"Some might argue this key point overperformance as a positive but my research would suggest that this is simply a symptom of variance, with performance on key points almost always reverting back to expectation in the long term (based on service/return points won)"
Thiem's results here unexpected
While Novak Djokovic coming through the bottom half of the draw was expected - he was the heavy tournament favourite prior to and throughout the event - Dominic Thiem's run in the top half was less predictable, with the Austrian fifth seed defeating three top ten players en route to this stage and he will need to beat another one on Sunday if he is to lift his maiden Grand Slam title.
Thiem has twice got to Grand Slam finals but previously they were both in his favourite clay conditions at the French Open (the last two seasons) where he ran into the King of Clay, Rafa Nadal. However, he can't be accused of having an easy run to the final here, although his victory over Alexander Zverev in the semi-final was fairly variance-friendly, winning the last two sets via tiebreaks, having five fewer break point chances and only winning three more points in the match than the German despite winning 3-1 in sets. Zverev also had a better winners/unforced errors count.
Key point overperformance a huge factor for Thiem
In fact, we can make that two matches in a row where Thiem has progressed with similar dynamics. Rafa in the quarter-finals had three more break point chances and lost five more points in a 3-1 loss, and had a better winner/unforced error ratio to boot. This time it was three tiebreaks which got the job done for Thiem and it's difficult to argue that he hasn't been a major beneficiary of taking a high number of key points during these matches.
Some might argue this key point overperformance as a positive but my research would suggest that this is simply a symptom of variance, with performance on key points almost always reverting back to expectation in the long term (based on service/return points won) - it's simply a question of when, not if.
Djokovic boasting far superior tournament data
Certainly, Djokovic has considerably better tournament data in the six matches to get to the final. He's won 4.5% more service points and almost 4% more return points, while also needing to be on court for 61 fewer games - he holds all the statistical aces here.
This is reflected in the market pricing with the Serb priced as a 2/71.29 market favourite, and this is a line which I agree with. Thiem has done reasonably well against Djokovic previously, winning four of their ten head to head clashes on the main tour, but with the exception of a final set tiebreak win over Djokovic in November at the ATP Tour Finals, all his wins were on clay.
On outdoor hard court - a little earlier in their head-to-head series - Thiem was unable to do much at all against the Djokovic serve and I think he's going to have to take his chances again on Sunday. It might well be possible - he's done it in his last two matches after all - but in my view, it's unlikely. The market pricing accurately reflects the implied win chances for the two players.
Follow Dan on Twitter @TennisRatings