Lucic-Baroni unlikely to shock twice
Wednesday's quarter-final action saw a very impressive performance from Serena Williams to oust our fancy, Johanna Konta, while in a huge shock, Mirjana Lucic-Baroni overcame the odds to get past Karolina Pliskova.
Lucic-Baroni's reward is a semi-final with Serena, and the American is available at 1.132/15 to get through to the final. The younger Williams sister's opponent in the final will be known before then, with either Coco Vandeweghe or her older sister, Venus, standing in the way of yet another Australian Open title.
With there being no value on the Williams vs Lucic-Baroni match - my model largely agreed with the market, pricing the American as 1.152/13 favourite - we are going to look at the first match, between Venus and Vandeweghe, with Coco available as the favourite currently, at 1.654/6.
Fast court fallacy
Vandeweghe's rise in this tournament is nothing short of incredible, with the American backed for pennies at 800.0799/1 in the outright betting, and she's now as short as 5.69/2 for outright glory.
It would have been extremely unlikely that she'd have been favourite against Venus if the match was played in the opening round but the markets has got a little excited with her wins over Angelique Kerber and Garbine Muguruza and also with the media commentary suggesting that the courts are very fast at Melbourne Park, with the inference that this suits Vandeweghe's game.
Such a statement is a fallacy, with there being very little deviation between the service hold percentages in the event and the tour means, as well as the aces per game comparisons as well. Both statistics are standard metrics for measuring court speed yet commentators and many observers are oblivious to that fact.
Hold/break stats favour Venus
I'd prefer to focus on hold/break stats on these occasions, with my data suggesting the courts are medium-fast at best, and in the last 12 months on hard courts, Venus has held 66.2% and broken 45.5% (111.7% combined), whereas Vandeweghe's stats are 73.7% and 29.3% (103.0% combined).
Even if we look at 2017 data in isolation, Venus' edge on return is so great that Vandeweghe's serve is negated, and the veteran is a much better returner than both Kerber and Muguruza, and is therefore likely to test the Vandeweghe serve much more.
With all this in mind, it's impossible to make a case for Vandeweghe at these prices, and my model was virtually the reverse of the market, pricing Venus as a 1.66 favourite.
Back Venus Williams to beat Coco Vandeweghe @ 2.506/4
Follow Dan on Twitter @TennisRatings