Post by relatedRelated post
Tennis worldwide is coming to a crossroads: with the emergence of the Universal Tennis Rating as a widely accepted method of judging a player’s ability, will the UTR eventually replace the need for national or international ranking systems? Or is a definite ranking necessary for the development of the game? And how accurate is the UTR system really?
Here at Britennis, we often find it difficult to keep up with the varying different levels of the tennis spectrum: events such as the ATP World Tour and the WTA Tour are widely covered in international press, while ITF Futures are a fairly important staple of what this website produces. But when you dig a little deeper, into the ITF Junior circuit, the AEGON British Tour, or the Tennis Europe events, being able to judge a player’s ability, or the result of a specific match, is severely impacted due to the varying different backgrounds each player has come from.
For example, a couple of weeks ago, Paul Jubb played Jack Draper in the final of a ITF Juniors Grade 4 event in Norway. Jubb, 16, plays the majority of his competitive matches at national level in the UK, and as such, his ITF ranking was significantly lower than the rest of the main draw field. Jubb, therefore, was required to qualify for the event. In the case of Draper, 14, he had played most of his recent competitive matches on the Tennis Europe tour, meaning his ITF ranking was also significantly lower than his ability suggested. Both players came through a field of higher ranked players with arguably a lower level of tennis ability to reach the final, simply because the ITF rankings only take into consideration results on their own tour, and not results elsewhere.
Would it not be fairer to seed players based on an overarching system that takes into consideration ALL the results that a player has recorded, rather than just those on one specific circuit? That’s where the Universal Tennis Rating is beginning to make an impact.
First and foremost, the Universal Tennis Rating is used by American universities to assess the ability of the players they are potentially recruiting. How is an American college coach expected to be able to understand all the national rankings, ratings and tournament results given the standard of tennis may be much higher in some countries in comparison to others? Craig Lambert, writing for Universal Tennis recently, writes:
“Global recruiting is still a road with some bumps. It is hard enough to pin down the actual potential of juniors who’ve played in high-school and USTA-sanctioned events. But making a reliable assessment on a 16-year-old girl from, say, Cambodia, can be rather perplexing. A key stumbling block is the plethora of different tennis rating systems used around the world. There are often even multiple ranking systems within one country. Each system has its own formulas and metrics, resulting in tennis’s Tower of Babel. This makes the college coach’s task somewhat like that of a United Nations translator expected to understand the languages of all the different national delegations. Not possible.”
Since its introduction, the Universal Tennis Rating has become a key factor in college recruitment in America, which arguably has the best and most coveted collegiate tennis program in the world, and is allowing players from an ever-increasing amount of countries to be fairly assessed and recruited, no matter what their previous playing experience. Audra Cohen, the Head Coach of Women’s Tennis at the University of North Florida, claims that “Universal Tennis is the most useful recruiting tool we have for evaluating the level of each recruit,” while the UTR system is being used to evaluate players at worldwide tennis academies as well: Scott Trielby, the Recruiting Director at the IMG/Bollettieri Academy, says that “Our players hail from over 50 countries, so it allows us to fairly evaluate everyone. It provides our college-bound players a fair ranking since they are playing multiple circuits.”
So how does the Universal Tennis Rating actually work? How does it ‘level the playing field’, as it so often suggests? Lisa Stone, writing for Parenting Aces, explains it best:
“Universal Tennis (UTR) publishes ratings based solely on actual matches played. They look at a player’s 30 most recent singles match results (doubles are not included), apply their proprietary algorithm, then rate the player on a scale from 1-16.5 to provide a snapshot of where a particular player is in comparison to other players in a given week. Gender is not a consideration. Neither is age nor country of origin. All players world-wide are rated together on the same scale. Only matches that are actually played are included. Walkovers or defaults are not counted. And, UTR pulls match results from a wide variety of sources including USTA junior tournaments, USTA adult tournaments, high school matches, ITF tournaments, ITA tournaments, and college dual matches among others. According to the UTR guiding principles, any two players within a 1.0 rating differential should have a competitive match, and if a player rated more than 1.0 below the opponent wins the match, that is considered an upset.”
In a nutshell: the UTR combines the last 30 results of any given player, male or female, and provides a Rating based on these results. In order to assess the accuracy of the Universal Tennis Rating, we decided to carry out a little experiment: using the same tournament mentioned above, the Grade 4 event in Norway where Jack Draper defeated Paul Jubb in the final, we decided to compare the actual seedings of each player with their UTRs.
Our study found that of the 31 matches that were played in the competition, the UTR Ratings were correct in 26 of them, with one match-up having two players with identical UTRs. Of the four that were incorrect, the differences in the UTRs were as follows: 0.16, 0.12, 0.03 and 0.01. Universal Tennis suggests that a difference of 1.0 is necessary to be considered an ‘upset’; these differences are clearly marginal.
Let’s firstly analyse a couple of the matches to demonstrate the difference between the ITF Ranking used for the seedings, and the Universal Tennis Rating:
Round 1:  Christian Linge (NOR) vs [Q] Emilien Voisin (FRA)
Ranked #450 and seeded third at the event, Linge’s seeding would have suggested he should reach the semi-finals at least. However, Frenchman Voisin, who had come through qualifying to reach the main draw, produced a ‘shock’ in the first round by defeating the Norwegian 6-1 6-2 and eventually progressing as far as the semi-finals. So what happened? Linge’s UTR is 10.49, whereas Voisin’s is 12.04. Therefore, according to the Universal Tennis system, had the result gone by seeding, then it would have been a pretty major UTR upset. The difference between the two players is this: Linge’s higher ITF Ranking is partially because since the start of 2015, the Norwegian has played 23 tournaments at Grade 3, 4, or 5 level. Voisin has played just four. The Frenchman, however, has generally beaten better quality players in those four tournaments, hence his higher UTR. So that’s one reason the UTR is benefitting some players: those who are unable to play so often internationally are still receiving equal amounts of credit as those players who ‘spam’ tournaments on the Junior circuit in order to falsely boost their ranking.
Round 2: Jack Draper (GBR) vs  Gilbert Jaeger (SWE)
This was the match that had the biggest UTR ‘upset’ in the tournament: Jack Draper, holding a UTR of 12.61, defeated fourth seed Gilbert Jaeger, holding a UTR of 12.78, 6-4 6-0 in the second round. Draper would, incidentally, overcome another UTR deficit in the quarter-finals against Anton Matusevich (UTR 12.73), meaning that en route to the final he defeated the two players with the highest UTRs in the tournament – a worthy champion. As mentioned above, Universal Tennis suggests that UTR differences of less than 1.0 lead to ‘competitive’ matches – another part of their algorithm. According to their website, “Universal Tennis considers a “competitive” match to be any match in which the losing player wins at least one game more than half the minimum number of games that are required to win a match. In a common best-2-sets-out-of-3 format, that’s seven games.” Perhaps the noteworthy thing about this particular match is that it wouldn’t have been classed as competitive, with Jaeger only winning 4 games. But perhaps this can be written off as Draper simply being in excellent form that week?
How would this event have proceeded if the seedings had been given out based on the Universal Tennis Rating, rather than the ITF Junior Rankings? Below are the top eight players in the tournament, based on their UTR, how far they progressed in the tournament, and who they were beaten by. It is worth noting that the top five players according to the UTR were in the bottom half of the draw, and the ITF Rank given is the ranking each player had on the week of the tournament:
|Seeding||Player||Nat.||ITF Rank||UTR||Result||Beaten by?|
|4||Gilbert Jaeger||SWE||457||12.78||R2||J. Draper|
|Anton Matusevich||GBR||1002||12.73||QF||J. Draper|
|5||Jacob Eskeland||NOR||458||12.60||R1||A. Matusevich|
|Johannes Abrahamsen||NOR||568||12.48||R2||A. Matusevich|
|Q||Paul Jubb||GBR||1594||12.45||RU||J. Draper|
|1||Carlo Donato||ITA||408||12.35||R2||P. Jubb|
|Viktor Thoresson||SWE||954||12.11||QF||P. Jubb|
Several things emerge from this table. Firstly, the second and fourth best players according to the UTR met in the first round, with Matusevich taking out fifth seed Eskeland. Similarly, the two best players in the top half of the draw, Jubb and Donato, met in the second round, with one of them having to exit the competition earlier than someone like Britain’s Jeremy Gschwendtner, the seventh seed, who with a UTR of 11.75 had the lowest UTR of those who reached the quarter-final. If the competition had been seeded according to UTR, the quarter-finals may have consisted of the eight players above and led to four highly competitive matches: the seven players in the top eight who didn’t win the title were all beaten by other members of the top eight.
Are the ITF Junior rankings therefore useless? And what about the ATP and WTA Rankings, which, generally speaking, are reflective of the UTRs of the players involved? I spoke to Universal Tennis about their opinions on the benefits of the UTR compared to the worldwide ranking systems, and it all boiled down to the one decisive factor that everyone agrees on: that of a common denominator.
“Rankings and ratings are different. We don’t do rankings. What does differentiate us from other tennis federation ratings, is that we provide credit from many different tennis pathways and circuits, and we are becoming more global each day. Think of us as becoming the international golf handicap for tennis, or a tennis passport for players who have the desire to play outside of their tennis community. Or Tennis’s Metric System.
Tennis Canada, who have begun to input all of their national results into the Universal Tennis Rating system, agrees:
“The reason why UTR is so useful to college coaches is because it pulls data from other rating tools. UTR then provides a universal scale determining the playing level of a particular athlete and this rating system has now become the common denominator whereby players can compare themselves to other tennis athletes and as well the top tennis players in the world, such as: Roger Federer, Rafael Nadal, Maria Sharapova and Serena Williams.”
So why mention this now? What does this have to do with British Tennis? Well, yesterday, Universal Tennis inputted over 1,100 match results from LTA tournaments, predominantly British Tour events, into the system, finally providing British athletes with similarly accurate Universal Tennis Ratings as their Canadian and American counterparts – something which is crucial in finding colleges for British youngsters. Tennis Smart recently posted this on their Facebook page:
“Well done to Tennis Canada for leading the way with Universal Tennis Rating, sending 100% of their junior results to the UTR team. We really need the same to be done for the British Juniors, as the lack of results are causing the UTRs for all British Juniors to be inaccurate & underrated. This then makes placement harder, and places even more importance on the player assessment of Tennis Smart. With more and more college coaches making decisions based on a player’s UTR, it has become a little more challenging when placing a British Junior who only plays British tournaments.”
So, we decided on another experiment. We decided to compare the UTR with the current LTA Rankings for the top five girls born in 1999, 2000 and 2001, to see the major differences between the two. Here were our results, with rankings accurate at the time of writing:
|Name (born 1999)||LTA Rank||ITF Junior Rank||UTR|
|Name (born 2000)||LTA Rank||ITF Junior Rank||UTR|
|Name (born 2001)||LTA Rank||ITF Junior Rank||UTR|
Generally speaking, the LTA Rank seems to correspond more to the UTR than the ITF Ranking. Why is this? Well, the LTA Ranking not only gives points based on tournament results on a national level, but also takes into consideration results from results at ITF and Tennis Europe events. The difference between the two being that the LTA Rankings takes the best six results in singles and doubles over the course of the year; the UTR simply takes the results of the last 30 matches. So which system is more accurate?
Let’s take a couple of examples: 2001-born pairing Sasha Hill and Esther Adeshina are ranked roughly 30 places apart on the LTA Rankings, with Hill being higher ranked. However, the UTR shows Adeshina at an overwhelming advantage. Indeed, the pair met at the Nike Invitational Winter National Tour Finals at the start of February. Hill was seeded fourth, whereas Adeshina was unseeded. Adeshina won 7-5 6-0, eventually going on to win the competition – coincidentally, she had the highest UTR in the draw, ahead of top seed Olivia Peet (UTR 9.34), whom she defeated in the second round.
A second example: last week, 2000-born Nell Miller defeated 2001-born Gemma Heath in the final of the Grade 4 event in Sweden. According to the LTA Rankings, Miller is well ahead of Heath in terms of ability, with a whopping 47-place gap between the two. The ITF Rankings have it the opposite way around, with Miller at #551 and Heath inside the top 300. But according to their UTRs, they’re very evenly matched: Miller (UTR 10.11) defeated Heath (UTR 9.98) in a very tight final 1-6 7-6(2) 6-1, reflecting the marginal difference between the two players in terms of ability, a difference which neither the LTA or the ITF Rankings would suggest.
So what’s the next step? How should tennis associations, and, as we’re a British Tennis website, how should the LTA be implementing the Universal Tennis Rating? Should they get rid of rankings altogether?
Well, maybe. If we were to further analyse the UTR on the professional circuit, we would find that, for the most part, the top 10 players on the ATP Rankings are also the top 10 UTR-rated players. Being the World Number One would mean less if it were based on a rating, rather than a ranking. But at junior level, the distinction is less clear cut. Generally speaking, juniors are assessed based on their ITF Junior Ranking, but that’s not particularly fair in some cases: youngsters who do not have the financial backing to play events across the world will naturally play more national competitions and as such gain less exposure, but may still have an equally good or better game than some players ranked in the top 100 of the junior rankings. Perhaps the UTR should release a comprehensive list of junior UTRs, so that youngsters are accurately able to assess their progress in relation to the rest of the field. Perhaps junior tournaments should be seeded according to UTRs?
The best thing about UTRs is that they transcend age groups – a 14-year-old can have a similar UTR as a 25-year-old or a 45-year-old, and given the chance to play against each other, both parties will learn from a competitive match of tennis. At national level, players are rigidly grouped by age, but the fact is, 14-year-old Jack Draper is head and shoulders above the majority of the rest of his age group. Why shouldn’t he play against older players at a similar level to earn more useful experience, rather than thrashing a local 14-year-old 6-0 6-0 and gaining nothing? Tournaments across the US and Canada are already using UTRs to produce more competitive fields and more beneficial match-ups for all. At the very least, the LTA should start doing something in a similar vein.
About Mark Gregory
25-year-old University of St Andrews graduate with a rather insane passion for British Tennis. Boston United fan (don't ask). Favourite tennis player: the Brummie bunch - Dan Evans and Lloyd Glasspool.