One thing about improving your rating

I’ve been asked a few times today about ratings and dynamic ratings and Early Start ratings and how to get your ratings up or explain why they go down and yadda yadda yadda. People, I can barely hit a forehand over the net three times in a row, and when it comes to computers, I’m writing this on an old Commodore 64 with my internet hardwired in via a proxy using native IP protocols with systemic DNS subsections routed through an 802.1 vortex.

(Totally made that last sentence up and have no idea what any of it means, but you get the point)

However, I went to Kevin Schmidt and his terrific and favoritist tennis stats website (bookmark it!!!) and found this great Q-n-A I highly recommend reading, especially the comments. I have pulled out the more interesting ratings parts:

Q21: So if you’ve figured it out, how does it work?
A: I’m not going to disclose everything I’ve researched, but the basic algorithm works like this. The computer looks at the dynamic ratings of the players prior to a match and based on them establishes an expected result. If one player/pairing does better than expected, their dynamic rating(s) will go up while the opponent’s will go down. How much the actual result differs from the expected result determines how much the ratings change. There is also some averaging done with prior dynamic ratings so that your current rating incorporates the results from not just your most recent match, but several of your most recent matches.
Q22: Does it not matter if I win or lose then?
A: Strictly speaking, yes. It is possible that the expected result is that you will lose 6-2,6-2 and so if you lose 6-4,6-4, you did better than expected and your rating will go up. Similarly, you could be expected to win 6-2,6-2 and instead win 6-4,6-4 and your rating could go down.
Q23: I want to improve my rating, how can I do it?
A: Simply put, you need to do better than the computer expects you to do. The best way to do this is to practice, take lessons, and work on your game including improving your strokes, fitness, and mental aspects of the game.

Q24: But now you just sound like my local tennis pro trying to get me to pay for lessons, is there nothing else?
A: Ok, there are some things to consider when trying to “impress” the computer. The key thing is in doubles, it is not just your rating but the rating of your partner that matters, and of course the rating of your opponents. But most of us can’t pick our opponents but do have some influence on who we play with.
It may seem counter intuitive, but in order to give yourself the largest opportunity to improve your rating, you actually want to play with a lower rated partner rather than a higher rated one. By doing this, the computer will expect less of you so there is more upside. Now, you still need to go out and do better than expected, so picking a partner that may be lower rated but is improving and perhaps playing “above their rating” would be ideal. Similarly, you’d want to avoid playing with a high rated player that is struggling or perhaps playing with a niggling injury as they would be playing “below their rating” and it may be hard to meet let alone exceed the computer’s expectation.

Q25: I beat a Benchmark player, that will help my rating more, right?
A: Not exactly. As noted above, a B rated player achieves that designation for playing in the playoffs,which may make you think that means that B also means they are the “best” at their level. This is somewhat true, but there are exceptions and you cannot make the generalization that B means anything more than the player making playoffs in the previous year.

For example, say a 3.5 team goes to playoffs and they have a 12 man roster that includes a bunch of players with dynamic ratings above 3.3, but their bottom four players are at 3.2 or below. In local playoffs, they were confident of winning against a certain team and they played two of their “worst” players on court 3. At sectionals, they wrapped up their sub-flight early and could afford to play their other two “worst” players on court 3. Thus, all four of these players have played in playoffs and get a B rating, specifically a 3.5B, even though their dynamic ratings are below 3.2, in the lower part of the range for 3.5s. Further, their four best players get bumped up to 4.0 at year-end as a result of having ratings between 3.5 and 3.7, and they get the 4.0B ratings. Their other four players end up being 3.5B rated with ratings between 3.3 and 3.5.

The result is that all 12 players are B rated, but only four are actually the “best” at their level, the other eight are actually in the lower half of the range for their level. So one could actually make the argument that B rated players are generally weaker within their level than C rated players given this scenario which probably isn’t that uncommon, especially if more of the players were to be bumped up.

Q26: You mentioned early start ratings before, what are these?
A: Some sections or district elect to start play in year X leagues in the prior year. This is usually done for court scheduling purposes. If you were to try and play 18 & over, 40 & over, 55 & over, 18 & over Mixed, 40 & over Mixed, etc. all at the same time, there simply wouldn’t be enough courts. So for example in my section, we start our Mixed leagues in the Fall of the year before. These leagues are considered “Early Start” leagues.
Since these leagues are actually for the following year, it doesn’t seem fair for a player who has improved and is going to be bumped up at year-end to get to play in this league at their old (lower) rating, so the section will publish early start ratings usually in the July or August timeframe that will be used for the early start leagues. The early start ratings are based on the current dynamic rating and can be considered a good indication of if a player will be bumped up or down. For example, a 3.5 may have a really good year in their Spring Adult league and get their dynamic rating above 3.50. They would get an early start rating of 4.0 and have to play in the Mixed league as a 4.0 rather than a 3.5 even though year-end ratings haven’t come out yet.

Q27: Can a player be bumped up on the early start list but go back to their old level at year-end?

A: Yes, this is possible and I’ve seen it happen numerous times. And it can go the other way too, a player isn’t on the early start bump list but then is at year-end. It can happen for at least a couple reasons.
First, the player may play more matches after the early start ratings come out and if these cause their rating to go up or down, this could result in them not being bumped up/down at year-end. These matches could be from subsequent rounds of playoffs that they play in or another league that their section chooses to include in year-end ratings. If a section includes tournament results, these could also affect one’s year-end rating and make it different than the early start rating.
Second, as noted above, year-end calculations incorporate some extra calculations to normalize ratings across the sections. At a very high level, you can think of it this way. If the team that goes to Nationals from your section does very well, that may be an indication that your section was stronger than your ratings indicated, and the good results from that team will cascade back to the players that played against them within the section, district, and area that they came from. This can in turn “pull up” the ratings of those players. However, if that team does poorly in playoffs, the effect can be that the ratings of the players that played them would go down.
Q30: I went undefeated all year but lost in playoffs and wasn’t bumped up, shouldn’t my undefeated regular season have gotten me bumped up?
A: Not necessarily.

First, going undefeated does not guarantee your rating will actually go up as it is possible you played low rated players in getting those wins and/or the scores may have been very close, and is the rating of the opponents and the scores, not just wins and losses, that determines your dynamic rating.

Second and more importantly, if your dynamic rating had gotten above the threshold to be bumped up prior to playoffs, it is entirely possible that the playoff loss would have dropped you below the threshold. This is for two reasons. A) Your rating is affected more by your recent matches than older matches. B) Playoff matches are given extra weight in the year-end calculations. So that playoff loss carries a bit of a double whammy and is likely the reason you didn’t get bumped up.

Q39: How does playing mixed doubles affect my rating?
A: The short answer is “it probably doesn’t”.  The longer answer is that mixed results don’t count if you also play in a men’s/women’s league during the same year, specifically you’d need to play in at least three men’s/women’s matches in order to get a year-end C rating in which case your mixed matches aren’t used at all in determining your rating.

If you don’t play at least three men’s/women’s matches but do play at least three mixed matches, then you may get a Mixed-exclusive or “M” rating at year-end.  I say “may” because if you had a C rating at the end of the prior year, my experience is that it will carry over to the next year even if you only played mixed.

Advertisements

About johncotey

John was a sportswriter and columnist for the St. Petersburg/Tampa Bay Times for 25 years, before being pushed from the Titantic, and hopes all of you non-subscribers are happy reading your free news on the internets while his journalism brethren suffer. Not that he's not bitter or anything. His real true joys, however, are his wife, kids and tennis, though not necessarily in that order. Unless his wife is reading this. Then DEFINITELY in that order.
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s