Regional Rating Differences

soldierIt is interesting to see the differences in tennis ratings by region of the country. One would think a 4.5 is a 4.5; but that is not always the case.

Even at my ripe old age, if I am accurately rated as a 4.5 tennis player in Florida, in New England I would only be considered a 4.0 player. Is that because they interpret the standards differently here? No, according to a friend who is exactly even with me and rated 4.0, it is because of league play.

He feels that stronger, younger players – who should be rated 5.0 – don’t have good leagues of their own to participate in; so they get “rated down” to play in 4.5 leagues. So just by comparison to them, the standard 4.5s get pushed down to being 4.0s.

And that doesn’t even take into consideration the standing argument that AGE IS A FACTOR in ratings… that a 35 year old 4.5 rated player is just not the same as a 65 year old 4.5 rated player.

What is your situation in other sections of the country?

If you are not on my “new posting alert email list” and want to be (I promise, no other uses of your email address!), just drop me a note at George@seniortennisandfitness.com

8 thoughts on “Regional Rating Differences

  1. The NTRP ratings system is completely broken.
    Early ratings have just been published in New England and I have been bumped up to 4.5.
    Only one other teammate, on my 4.0 adult team, was bumped, despite the fact that most of the players (who are in their 20’s and 30’s) are stronger players than I (at age 74).
    The entire system needs to be scrapped and replaced with a fairer, more accurate, system.

    Michael – we need Spike to redo his system! tks, george

  2. Hi George!
    I find the same thing is true in Michigan. At our club, Deer Lake Racquet, in Clarkston,Mi. the 4.0 players would be playing at a 4.5 level in Florida. First of all, we have many more young players than we have in Florida, especially in the Collier, Lee County area. I believe that this is one of the main differences.
    As a week 4.0, I played in the adult 18 and over for the past several years with hardly a win. I did not play this season since I felt that I was not contributing to the team.

  3. Here are my observations after playing in the USTA/Southwest Section for a year:

    – In the USTA/Southwest section everyone plays League Tennis, mainly doubles & based on ratings. Adult aged-based singles tournaments are offered by the USTA in a few locations, but outside the major cities (e.g. Phoenix), they rarely get any entries above the 55 or 60 age divisions.

    – Participating players are definitely rated lower than they would have been had they been playing in NE. Our team from Sedona won the District, then went to the Sectionals (18 yrs & over, 4.0 rating) in Albuquerque, NM. In my opinion, a number of the top players I observed clearly would have been rated 4.5 or higher back in NE. And I concur with the view that a 25 year old player who is rated 4.0 is not the same as a 65 year old 4.0 player.

    Bruce – thanks. and, i miss you on tuesdays in NH! geo

  4. I consider myself a true 4.5 player from the rating system back in the 80’s.

    In collier county i have had my ass handed to me in league and tournament play by ex collegiate players all under 35 in the 4.5 divisions.
    I always matched up well everywhere Else i played/went against 4.5 players. I think the other comments are spot on

    Marc

  5. In NE Ohio I am a 74 yr old 4.0 player with good speed on the court. What I notice
    mostly about the 60 and over men’s doubles players is their lack of mobility . Many who were once 4.0 doubles players are now really 3.0 and 3.5 because they no longer can take those quick steps to place themselves in the optimal hitting position. Instead they have to reach and most have problems returning serve. Many also think they can still cover the court as they did 20 years ago and play too close to the net and cannot make it back for lobs. Others have become baseliners and even retreat after hitting a good shot and become easy drop shot victims. My advice for the over 60 player: when your partner is serving stay closer to the service line than the net. In rallies get off the baseline and join your partner up around service line.

    Jerry – I agree… as mobility goes, so goes the rating. tks, george

  6. Interesting topic. I am in the Bay Area and play in two leagues – SF and East Bay. There is a huge difference between the low 4.5s that just got bumped from 4.0 and the many players who play 5.0s for one season and seemingly purposely lose 4-5 matches and come back down and go 11-1 or 10-2 at 4.5s. These guys are usually well under 40 years of age and have some collegiate Tennis in their past. They game the system and there are many managers that recruit and harbor these sandbaggers…

    Yesterday – I was in a 9.5 combo match. My partner a sold 5.0 and I played two of these very high 4.5s with 5.0 experience players. We pulled out a third set tie-break. One of the other players was every bit the 5.0 player my partner was….

    The districts and sectionals are usually filled with perenial stacked teams filled with such players.

    Solution – Perhaps they should split 4.5 and have a 4.25 league as well….

    Mike

    Rambo – a friend of mine in FL (Spike Gonzales) was one of the architect’s of the system and said it was designed to be done in tenths (ie 4.0, 4.1, 4.2, etc). TKS. geo

  7. George, the ratings system currently uses a computer algorithm to measure head to head play but it (a) can be easily “gamed” and (b) tends to retain people in rating categories lower than where they may actually belong, especially with doubles, because, strangely enough, it has become overly objective and is not subjective enough.

    To appreciate this, you need to understand a little of how the algorithm works. Using a computer, the system keeps track of ratings to the one hundredth of a point and is “dynamic” in the sense that the numbers regularly change based on a player’s scores playing against other opponents. Each rating category is actually a range and is not a fixed number. So, a 4.5 player is really someone with a 4.01 to 4.50 dynamic range. It is not a fixed number. The system predicts the most likely score of a match based on the relative position of the dynamic ranges of the opponents prior to the match being played using tenth of a point gradations. For example, if one player has a 0.1 edge over the other player – say 4.50 versus 4.40 – the most likely score is predicted to be 6-4, 6-4 for the player with the slightly higher dynamic rating. Carrying this further, this means that a “true” 4.5 (i.e., having a 4.50 dynamic rating) should beat a “true” 4.0 (i.e., having a 4.00 dynamic rating) 6-0, 6-0 pretty much every time they play.

    Therefore, to “game” the system, all that a player trying to sandbag needs to do is to be aware of the predicted scoring differentials and to keep it close, or even to drop a set now and then. So, an actual 5.0 player who is sandbagging down on a 4.5 team cannot win sets 6-0, 6-0 against his 4.5 opponents in singles and expect to stay on a 4.5 team. He needs to win, say, 7-6, 7-5 or even drop a set now and then – especially when he plays against a weaker (e.g., 4.01) rather than a stronger (e.g., 4.49) 4.5 level opponent. Most 5.0 players are certainly good enough that they can “rig” a match so it seems they are playing unusually close against an opponent but still winning the match.

    But if the 5.0 player playing on a 4.5 team in the previous example only kept it close but still managed to win all of his matches in a season, the computer algorithm may still catch this and bump him up to 5.0 at the end of the season. So, to further “game” the system, he might not play that many singles matches but, instead, be paired with some of the weaker (e.g., 4.01 – 4.24) players on his 4.5 team in doubles. The computer algorithm for figuring out dynamic ratings with doubles is more complicated than with singles but, as I understand it, it uses a formula that essentially averages the dynamic ratings of the doubles players. However, this fails to take into account the reality of the adage that a doubles team is only as strong as the weaker player – that is, taking an average of dynamic ratings is not a very good indicator or predictor of the actual outcome of a doubles match because, almost always, the weaker player is going to pull down the level of play for the team as a whole than what the stronger player, acting alone, is able to control. So, if the sandbagging player mainly only plays doubles and avoids playing or only rarely plays singles, he runs much less risk of getting bumped up because the actual scores (and outcomes, including the occasional unexpected loss) of his doubles matches with weaker partners are almost always going to be less favorable than what the computer system “predicts” they should be based on mathematically averaging of the doubles pairings.

    This, in turn, has the further negative effect of retaining players where they are who may actually not want to sandbag but who would prefer to “move up” in the computer ratings to a higher level where they can improve and face stronger competition – especially if all they ever play is doubles and even more so if they regularly get paired with weaker partners themselves because they are perceived as being able to “carry” weaker partners in matches. So, not only do you have a greater facility for sandbaggers to play “down” under the current system, but you also make it harder for players who genuinely want to improve their ratings to do so, especially if they start out at with a lower rating team and mainly play doubles with the weaker players as partners.

    In the old days, before the USTA moved to the present computer based dynamic rating system, it used to have roving ratings officials, especially at the district and sectional playoffs, who would more subjectively (but I think more accurately) identify players who should be bumped up, and these officials would have the power to do that on the spot. Nowadays, I don’t see these officials around any longer, which I think is a mistake. While comparing head to head match scores using a computer algorithm may be a decent way to keep track of the vast majority of league players, it is certainly not foolproof and can, and does, allow outlier situations to exist that should not happen. It is in this respect that I think the system has probably now become TOO objective and could stand to allow a little more of the older subjective system to come back in to act as a fail safe for the situations described above.

    Marty – Great explanation! tks , geo

  8. George, here are some other pet peeves that I have with the current system that I offer in case Spike may have some influence in pushing USTA to fix it:

    1). The regional differences that you and others mention between New England and Florida can be just the opposite in other areas of the country. For example, while a 4.5 in Florida may be a 4.0 in New England, there are 4.5’s that I have played against from California who would destroy many 5.0 players near me, in Middle States. Rambo’s recent post about solid 5.0’s in the San Francisco area who sandbag down to 4.5 seems to bear me out. Either there are just fewer 5.0’s in Middle States generally, which is certainly possible given the climate difference, or sandbagging is more prevalent in Northern California, which is also possible.

    2). It is indeed correct that a 60 year old 4.5 is not the same as a 25 year old 4.5, but I don’t think it is age alone that is the major difference. Age is just an indicator. It is mainly foot speed, reaction time and general athleticism that account for the differences. If you take the time to read them, you will see that the current rating criteria are too heavily weighted toward stroke production and do not adequately take into account other, more physical factors. To illustrate: A 60 year old 4.0 player who has played all his life is going to have some nice looking, even textbook strokes. He can probably put the ball pretty much anywhere he wants on the court and can hit varying degrees of pace and spin. He also has every shot in the book – that his, he can volley, hit overheads, has a decent serve, can hit half volleys, can lob, etc. But put that player up against a 25 year old 3.5 player who is new to the game, plays a lot, is aggressive and fast on the court, but who can only hit relatively hard groundstrokes (i.e., he cannot volley, has a weak serve, an inconsistent overhead, etc.), and my bet is on the younger player winning 3 out of 4 times just because he is faster, can chase things down, and can stay out there all day. Indeed, in several seasons of 18 and over league play, I have seen this very thing happen regularly with singles matches. Although it would probably be too complicated to administer, the ideal system should have differentiated ratings that take into account age (and, by implication, foot speed, reaction time and general athleticism) in addition to the current, stroke production oriented ratings factors. So, a 4.0 or 4.5 would still be a 4.0 or 4.5, but it would be different if you were between 18 to 39, between 40 and 59, and 60 and above. (These are my suggested age groups only; they could be tweaked if only the USTA would buy into the concept.)

    3). If the USTA truly wanted to make the system more accurate and fair, it would also publish different ratings for singles and doubles. Many players are decidedly stronger in one of these categories. I am certainly a good example of this. I will be the first to admit that my singles game is uneven, at best. (This is the main reason why I have been playing so many tournaments in singles lately; I want to get more consistent and improve.) I regularly beat solid 4.5 players, but I also lose matches that I personally have no idea how I could lose to supposedly “weak” 4.0’s. But the opposite is true with my doubles game. I just feel more confident playing doubles and, as a result, I almost always play doubles better than singles. Indeed, I have played tournament and other matches in doubles over the years where I have been the only player on the court who was NOT a 5.0 or better player and, while I was certainly out of my comfort zone hitting with and against these better players, I still managed to hold my own, and even win. I don’t think I am alone in this. Many players feel they are stronger in either singles or doubles. Ideally, the rating system should reflect these differences.

    4). At least in some sections – Middle States being what I am most familiar with – the current computer based ratings system only tracks play in the same sex USTA leagues (that is, it leaves out mixed doubles) and it does not track matches played in non-NTRP tournaments. So, if you play a 4.5 NTRP tournament, my understanding is the system will count your results and adjust your dynamic ratings score accordingly. But if you play, say, a 55 and over “open” tournament but every opponent that you face, and beat, happens also to have a 4.5 rating, the system will not count any of those matches for purposes of further adjusting your dynamic rating. This seems to me to be nonsensical. I can appreciate that maybe one reason for this is the USTA is trying to encourage people to play more tournaments by not putting their league ratings at risk, but if the whole goal of even having the ratings is the promotion of accuracy, then all matches played against players who actually have ratings with which to compare should be included.

    Some of the issues that I have identified above have been talked about and criticized for years. However, I don’t think the USTA has any real interest in fixing the rating system because, as bad as it is, it is better than not having a system at all and the USTA probably feels it is “good enough” to use with organized league play, which is its main purpose. This is too bad because, with some tweaking, I think the system could be made much better and remove many of the frustrations that I and others have with it.

    Marty – i really think your idea of different singles and doubles rankings is a good one! tks geo

Comments are closed.