June 2, 2025

Tip of the Week
Showing Frustration or Confidence.

USATT Initial Ratings Problems
For many years there’s been a serious bug in the USATT ratings software, which I’ve periodically pointed out to USATT people and here in my blog. Specifically, it often gives out nonsense ratings for initial ratings, especially if the player wins or loses all of their matches in their first tournament. This is unfair to these players, who find themselves way over-rated (and so can’t play in rating events where they might be competitive) or under-rated (making it unfair for others who have to play them, especially in rating events), and messes up seeding. This didn’t use to be a problem. (Here is the Guide to USATT Ratings, which explains how USATT ratings are processed, including initial ratings.)

There are four ways to judge an initial rating: based on wins and losses; based on games won and lost; based on scores; and based on the tournament director’s judgement. It used to be that the software took all four into account. (Tournament directors were given an optional field to estimate ratings for new players, and if they did so, it was factored into the equation. I don’t know if they still do this, but that’s not what I’m writing about here.) From what I’m told, a number of years ago a member of the Ratings Committee objected, saying that scores and game results shouldn’t be used as there’s always the chance a stronger player may let the weaker player score more points than he should, out of mercy. And so they were taken out of the algorithm – leading to the problem. Even worse, the algorithm that used game and point scores for initial rating was also apparently deleted and lost to eternity.

It is sometimes true that a strong player lets a weaker player score more than they should, but scores are a good indicator of a player’s level after a few matches. (Plus, if you require, say, at least five tournament matches to achieve a rating, then inaccurate ones are averaged out.) But using that data almost always gave a more accurate initial rating than just winning or losing. For example, if a player loses to a 1500 player at 1,1,1, it’s fairly obvious the player is probably under 1000 level, while if he loses 11-9 in the fifth, it’s fairly obvious that he’s probably close to 1500 level himself. But now the software looks at these two as the same, which is silly.

Here are three examples of nonsensical initial ratings. (I’m not giving out the names publicly, to protect the innocent players who had nothing to do with these mistaken initial ratings.)

  • A woman played two matches in her first tournament, losing to a 2125 player at 3,3,1, and to a 2115 player at 6,5,6. Based on this, she’s probably under 1500, perhaps under 1000. But her initial rating after this? 2090! How can you lose to a 2125 player at 3,3,1, and have the software judge you to be almost as good, at 2090? By ignoring data! After this was pointed out to USATT Headquarters (and after playing a tournament with the 2090 rating where she lost a series of matches to players under 1000), she was manually readjusted to 900, and is currently 881, which is roughly accurate.
  • An 8-year-old went 0-8 in his first tournament, losing every match 0-3. Against players ranging from 1645 to 1862 he lost at 4,5,2,1,4,1,4,5,4,3,5,3,2,10,4,3,3,2,6,3,2,8,10,7. Based on this, he’s probably about 1000 level. Instead, he was given an initial rating of 1601, which was silly. (He lost to the 1645 at 4,5,2, for example.) After this was pointed out, he was manually readjusted to 1200. But for a week or so, this beginner was #1 in Under 9 in the country!!! (He thought it was pretty funny. In tournaments afterwards he quickly dropped to under 1000.)
  • A 7-year-old played exactly two matches in his first tournament. He lost to a 663 player at 1,4,7, and to a 1361 player at 1,1,2. Based on that, he probably should get an initial rating of maybe 300 or 400, at most. But the computer gave him a rating of 655, only 8 points lower than the player he’d averaged 4 points a game against. After this was pointed out to USATT Headquarters, he was manually readjusted to 400. (He went under 400 in his next tournament.)

A player should also play a minimum number of matches before achieving a USATT rating – perhaps five. Here’s an example. Another player played only two matches in his first tournament, beating players rated 648 and 541 at 8,6,9 and 3,3,4. From this, he was given a 660 rating, which was silly since he’d beaten the 648 player at 8,6,9 and the 541 player at 3,3,4. From just these two matches, you can’t really accurate rate him, but he should be at least be given a rating considerably higher than the 648 he beat 3-0, and way ahead of the 541 he beat so easily. In his second tournament, he played four matches, again winning them all 3-0, against players ranging from 1013 to 1592 – and came out 1276, which clearly wasn’t accurate. (He beat a 1329 at 4,3,3, though his match with the 1592 seemed close at 10,9,5.) If they’d waited until he’d played at least five matches (after his second tournament), the scores would more accurately show him at be perhaps at least 1800 level. In his fourth tournament he finally faced real competition, and came out at 2135, which turned out to be accurate. (All four of these tournaments came within a three-month period.)

There are many such examples – alas, I haven’t kept track of them. I remember another case of a foreign player going undefeated in his first tournament, beating 1800 players at around 3 points a game – and getting a high 1800 rating when the scores showed he was much, much stronger. (He turned out to be a 2500 player.)

Currently, the USATT solution is to ignore the problem, but if a tournament director complains, then they manually readjust, as they did with two of the above examples. My suggestions are as follows:

  1. Adjust the ratings algorithm to take games and scores into account for a player’s initial rating. USATT has a Technology Committee that should be able to do this. Or bring back the Ratings Committee, which used to deal with such issues. Or just hire a programmer with a math background. (I have a math degree and a minor in computer science, and many years ago I could have done this. Alas, I’m a bit rusty now.)
  2. Encourage tournament directors to give estimates of ratings for new players.
  3. Require a minimum number of tournament matches to achieve a rating – perhaps five.

Cataract Surgery
I had the surgery on my left eye last Wednesday. (I had it on my right eye on Mar. 25.) All seems well so far, but as before, I have to go about ten days without any heavy exercise, no lifting of anything over ten pounds, and wearing heavy sunglasses anytime I go outside. I see the eye doctor against this Friday, and if all goes well, I’ll be able to start coaching and practicing against next Saturday. One small issue – the left eye doesn’t seem to be as sharp as the right eye. I was astonished when my right eye went from 20-150 to 20-15 – making it way above average, and allowing me to read small print from a distance. But the left eye isn’t as sharp. I’m guessing it’s 20-20, which is average. If all goes well, from June 7-27, I train for the Nationals, then I fly out on June 28. Alas, I’ll have to be careful – I’m still having shoulder and right knee issues, though the injury to my side seems to be better, and (quick roll call) most of the rest of me seems to be working. When they bother me in practice, I stop. If they bother me at the Nationals, I’ll make do, unless the arm or leg actually fall off.

RIP Yaroslav “Bob” Kaminsky
Here’s the obituary - he died on Thursday, age 89. He may have been my first informal coach when I first showed up at the New Carrollton Table Tennis Club in early 1976, though I’m not sure – Jim Mossberg and Yvonne Kronlage also gave me some informal instruction. He had a huge influence on my game as the first table tennis between top players I ever saw was Bob (then rated about 1900, shakehands with short pips on both sides) against Herb Horton (a chopper rated about 2000, with antispin on both sides), where Bob would just hit over and over and Herb would chop them all back – long rallies. From this, I realized you could hit anything, and so I became an all-out forehand hitter from that. (It was only later I learned about looping.) As I improved, I also got to play against Bob many times. Bob's wife, Barbara Kaminsky, is a member of the US Table Tennis Hall of Fame. 

Tableball, and What If We Took Sport Names LITERALLY?
Here’s the video (73 sec) – go to 47 seconds to see 6 seconds of literal TABLE tennis. One of the great tragedies of our sport is when they adopted the name “Table Tennis” in 1926, since Ping-Pong was trademarked by Parker Brothers. There’s baseball, football, basketball, racquetball, handball, volleyball, dodgeball, and others with names that show what the sport is about – so why didn’t they adopt the obvious name, Tableball!!! Table tennis makes us look subordinate to tennis. (Of course, they just play court table tennis.)

ITTF Election Shenanigans
The headlines almost tell the story. Read on.

USATT National Junior Ranking Tournament II Kicks Off in Charlotte, NC Charlotte, NC
Here’s the USATT article by Melanie Griffin. The Trials end today. Here are complete results. Click on Draws and use dropdown menu to see results for each event.

Butterfly Training Tips

How to Feed the Ball to Advanced-Level Students for Practice
Here’s the video (4:03) from Ti Long.

Best Training Drills to Boost Your Reflexes!
Here’s the video (11 sec) from Pingispågarna – I may have to copy this one!

Short Backspin Pendulum Serve
Here’s the video (3:26) with Cheyanne Chen from Pong Space.

New from PingSkills
Ask the Coach

The Redemption Of Hugo Calderano
Here’s the video (6:38) from Taco Backhand.

Major League Table Tennis
Follow the action! (Here are news items over the last month.)

New from Steve Hopkins/Butterfly

New From ITTF

We Challenged Table Tennis Legend Tiago Apolonia!
Here’s the video (9:41) from Table Tennis Daily! Apolonia of Portugal is world #94, formerly #13.

Backhand to Backhand – on a Skinny Table!
Here’s the video (17 sec) – could you do this?

Table Tennis Musical Bouncing Dance
Here’s the video (17 sec)!

Kitten versus Rhino Ping-Pong Championships
Here’s the video (37 sec)!

Cute Puppy Wants to Play Ping-Pong
Here’s the video (41 sec)!

Non-Table Tennis – Short Stories
On Thursday I sold “Thirty-Five Genie Heads on a Wall” (8,400 words) to Black Cat Weekly, my third sale to them and 236th overall. While being bullied at the beach, teenager Robbie is sucked into a Coke bottle. So begins his life as a genie, including classes at Genie University. His owner is a brutish king on a planet far from Earth – and the squirming, tormented heads of the king’s previous 35 genies are on display on a wall. It’ll come out in a few months.

“Cemetery Whale” (3,700 words) came out in the Tiny Terrors anthology from Graveside Press. A lonely music-loving autistic killer whale ghost haunts a graveyard, and interacts and battles with a local girl and a ghosthunter with a deadly ectoplasmic harpoon.

“Thank You, Miss Kittykat!” (900 words) came out in the Amazing Stories: Best of 2024 anthology. In a future where we've genetically enhanced the intelligence of dogs and cats and given them opposable thumbs, there's a war (humans and dogs allied against the cats), the cats win, and humans and dogs are now pets while cats rule.

***
Send us your own coaching news!