Although there are quite a few non-Brit seeds this year, the seedings are topped by Nell Miller and Esther Adeshina, who have both enjoyed successful seasons so far. Ellie-Rose Griffiths plays her first junior ITF for 9 months. Full draw at http://www.itftennis.com/juniors/tournaments/tournament/info.aspx?tournamentid=1100036303
I've done a full preview of the draw right here, to help dissect the draw a little better. It will be interesting to see how many of the UTR predictions are correct:
I would be interested to see how Indianna Spink gets on - she is only 13, but a top prospect as is Amarni Banks.
Hannah McColgan is highly regarded and went on the LTA trip to Spain recently. Her big lefty serve will cause problems and I expect her to get through to the QF.
I also rate Olivia Elliott and I'm hoping she will have a decent run.
Sophia Derivan and Nika Zupancic will be dangers, but expecting a Miller v Adeshina final
I've analysed UTR before and found generally an 85% accuracy rate. I'm not linked to them in any way (they re-tweet my stuff occasionally, but that's it) but having formed up a partnership through Britennis, I've since take up some analyses and been very impressed.
It's difficult to judge, sometimes, at this level, because players have played so rarely or so frequently in comparison, and in vastly different competition, but the whole point of UTR is to calculate a rating based on ALL the available results, not just under one organisation.
In terms of the players you mentioned and the UTR comparisons, Olivia Elliott's R1 match should be suitably tight, with only 0.11 separating the players. Amarni Banks actually holds a 0.06 advantage over her 4th seeded Italian R1 opponent, so could snatch a win, while McColgan is rated highly in her section but could run into trouble against Kalaitzis, the Belgian 11th seed, in R2. According to the UTR, Spink will struggle to get many games on the board in her R1 match - she's down by 1.23.
Is 85% the UTR accuracy rate purely for predicting wins ? ( ie their higher rated player winning )
Difficult to know offhand without some sort of analysis and comparison against something else what is actually a good / very good / excellent ( maybe you have done this ? ) How say does that compare against straight rankings where relevant ( it should certainly be a fair bit better given how often we declare players to be under or overrranked and indeed surely much of the raison d'etre of UTR is that it better / more realistically orders players ). You could get the esteemed Joe Bloggs to have a good look at players' records and come up with a winner and assuming Mr Bloggs is into his tennis and a logical sort then he would clearly get well over 50%, indeed so many matches go to form that I would expect him to get pretty well over 70%. When does the % become really good, as I say that I haven't a real feel for that. I suspect 90% plus would certainly do it, and 85% is at least pretty good, just how good ? .....
Anyway all that aside. from what I have read UTR serve a very useful purpose in various areas, and "take UTR with a pinch of salt" seems very dismissive for the clear work that they do on ratings ( and whatever I say on just being hit by a % figure it seems certainly not bad ), and they appear to be very well regarded by many of the colleges.
-- Edited by indiana on Tuesday 12th of April 2016 06:35:09 PM
Here's another article I wrote which could help explain a little further - I did an analysis of Jack Draper's victory in Oslo at a Grade 4 event. 37 out of the 41 matches went as per the UTR.
37 / 41 certainly looks very good, and 12 / 16 not so good ( without context on either re say how many were extremely likely wins or real shocks - it is really how UTR stands up in closer looking matches that is their test ), really need much more data so we come back to your more general 85% on which I just say I'm not totally sure.
Anyway, I will read your article, because this is certainly interesting me, Thanks.
-- Edited by indiana on Tuesday 12th of April 2016 06:33:11 PM