Wednesday, July 19, 2023

Wimbledon 2023 and the GOAT debate

The 2023 Wimbledon ended when Carlos Alcaraz beat Novak Djokovic 3:2 in Men’s Singles Final on July 15.  Alcaraz was not in the match during the first set but reemerged afterwards.  Djokovic was visibly tired at the end of the second set and the whole third set, allowing Alcaraz to take charge, and the outcome was never really in doubt.  No matter how great Djokovic is, you can be 25 only once, and currently Djokovic’s best weapon, his mental advantage over the two generations of players after him, is no more against Alcaraz.  Alcaraz’s 60+ winners are twice of Djokovic’s.  Djokovic fought possibly as hard and well as he could under the conditions, but the contest was not as close as the score indicated.  

This brings back the old question: is Djokovic the GOAT now that he has collected 23 GS and probably more before he quits, and will Alcaraz be the GOAT?   

There is a loose GOAT view, and there is a strict GOAT view (Sept 13, 2021 blog).  In the loose view, we can always have a GOAT up to date, and annotate a new one whenever the timing is right.  In the strict view, GOAT must have staying power, must withstand the tests of time; sure, GOAT can change, but the (new) GOAT must have done something significantly different and better than everybody among peers and in history.  So the strict view will rather have no GOAT than having to crown new GOATs every 5 years.  The bar is high.

 In the strict GOAT definition, therefore, if one has to have a GOAT in men’s tennis, it will be the big 3 rather than pinpointing only one.  Considering the careers of the big 3, their GS numbers and the head-to-head records are the products of luck and age differences more than anything else such as skills.  Federer is unlucky because Nadal is left-handed.  If Nadal were right-handed, even with the same forehand topspin and athleticism, Federer would have beaten that version of Nadal.  Nadal is unlucky because he is sandwiched between Federer and Djokovic, two of his equals.  Djokovic is the luckiest of the three because he has nobody after him, and he is younger than the other two.  While Djokovic’s head-to-head records are better, they are barely above 50%, and much is due to Federer and Nadal getting old.  Same with their GS numbers.  In other words, Djokovic does no better than Federer and Nadal, the differences merely by chance.

A more objective comparison is looking at other sports.  For baseball Bath Ruth is the GOAT, any doubts are manufactured.  Track and field or at least sprints, Usain Bolt.  For the reasons, just see how much different Bath Ruth and Usain Bolt distinguished themselves among their peers, beating everybody by “miles”, and their records still in conversations today.  American football, Tom Brady.  Badminton, Lin Dan, practically the big 2 or even big 3 in tennis combined.  Were one of the big 3 getting close to 40 GS, there will be no debate in men’s tennis.  Even women’s tennis, Steffi Graf is still more the GOAT than Serena Williams.  Serena gets one more GS, but played much longer than Graf because she had the records in mind, whereas Graf retired from tennis when she lost interest and had no more goals.  But Graf was still OK to play, having lost the Wimbledon Final as essentially her last match.  Were Graf dragged on like Williams, 25 GS would be easy.  

Basketball has murky territories: Wilt for personal stats, Jordan, LeBron, or someone else?  Then in men’s table tennis, is Ma Long the GOAT?  In terms of wins and records, Ma Long is clearly above anybody else, but is it enough for distinction?  Head-to-head, check.  Winning streaks and in important matches, check.  Two Olympics Golds, Ma Long all alone.  Three straight World Championships?  One person had done it in modern era, and Wang Liqin has also three, albeit no consecutively.  From a strict GOAT standpoint, Ma Long is borderline.  If he wins one more Olympics Gold or World Championship, he will be GOAT since there is little chance another later person will match that.  But Ma Long winning another big one is not going to happen.  Have to settle that there is no GOAT before him; if there is one after him, Ma Long has set a very high bar.

A curious side issue is that why tennis players like the big 3 in their mid 30s still dominate younger players, yet great badminton and table tennis players cannot.  Even Lin Dan was no GOAT caliber when he was over 30.  An answer is the serves.  In tennis you can serve big and score an easy point, saving you a lot of energy.  The relative long breaks in games and points help, too.  But serves are a nonissue in badminton, as you win by rallying, which requires stamina, “old” guys lagging behind young guns.  Further, you don’t have much time between points to rest.  Table tennis has the no hidden serve rule, eliminating much the advantage from the old days.  And even more than badminton and tennis, table tennis demands fast reflexes, youth always trumps aged.  Of the big 3 in tennis, Federer and Djokovic have good serves, which is why they have aged better than Nadal.  

Back to the question: will Alcaraz be the GOAT?  The mere fact that Alcaraz enters the discussion is astounding.  After the big 3 it is hard to imagine anyone will ever be as good, but so quickly came Alcaraz since 2022, who clearly combines the best of the big 3, some areas even better.  His both wings are as balanced as Djokovic, but he runs as willingly and fast and his forehand as strong as Nadal.  Then even at full speed he hits shots as naturally as Federer. So much power and so much touch and feel at the same time.  Alcaraz even has the best drop shots and the best lobs in the game.  You see good drop shots, and you see good lobs, but never from the same player.  Until Alcaraz.

The answer is wait 10 years and we will tell.  A lot can happen, like injuries, etc.  At this point he needs more experience, like knowing when not to run down every ball.  But by 2033 if Alcaraz is close to 30 GS he will sure be the single person GOAT.  

 

Monday, May 22, 2023

Sudirman Cup 2023: China gambled and survived, and technical analyses

Sudirman Cup has had a history of over 30 years, but the 2023 version (Suzhou, China, May 14-21) is one of the most dramatic if not THE most dramatic.  Two reigning Olympics champions got injured and retired during quarterfinal matches, completely ending their teams’ hope.  And the host China survived tough tests, first against Indonesia, then barely escaped Japan, then almost fell into the same hole against South Korea in the Final.  Much is due to the the first match, mixed doubles, and the Chinese choice of mixed double pairs.  Consequently, Chinese public opinion railed against the coaches; the coaches are not infallible, but the prevailing thought is wrong.

Ironically, China’s mixed doubles are considered China’s strongest point.  Zheng Siwei/Huang Yaqiong are the best in the world and ranked No. 1 almost continuously for the past 5 years.   The second pair Feng Yanzhe/Huang Dongping (Huang being the reigning Olympics champion) has a good record against all the major opponents not named Zheng/Huang.  The third pair is young but with a lot of potential and just won the Asian Championship a few week back.  So what can possibly go wrong?

The answer is that Zheng/Huang in 2023 are not as dominant as in 2022, either due to age or a drop in forms.  During 2022 they were winning most matches 2:0, leaving opponents with little chance.  But so far the short 2023 season has seen a lot of 2:1 and even comebacks.  No doubt everybody senses that, including the Chinese coaches.  In the group stage against Singapore HYKT/TWHJ, the first set Zheng/Huang were down 7:11, 9:15, 18:20, before winning 22:20.  The second set was 21:19, not a cake walk either.  HYKT/TWHJ were “winning” by fast exchanges at the net and quick smashes whenever possible.  While it was just a group stage match, it foretold things to come.  

The same strategy was employed even better by the Indonesian pair RR/WGE during the quarterfinals, as WGE was much taller than Huang and dominated the net more.  Zheng/Huang had no chance in the first set, 13:21, and was down 9:13, 11:17, 14:19 in the second.  Then Zheng/Huang hit fast and flat to RR/WGE, inducing many errors into the net.  After losing the second set, RR/WGE lost their energy and quickly the third set.  

The win is a must for China because Indonesia is the weakest in mixed doubles.  Had Indonesia stole the first point, the whole thing could go 3:2 either way.  The second match, men’s singles, Shi Yuqi vs Ginting, was still crucial.  Shi has an excellent record against Ginting, but Indonesia picked Ginting because: 1. their last two meetings were 1:1; 2. Ginting is in good shape, just winning the Asian Championship; 3. the other Indonesian Christie plays better against Shi historically, but had lost in the group stage, so is probably not at his best.  Ginting was leading the first set throughout, 15:9, 18:12, and 20:18, before running out of gas.  There is a reason why Shi plays well against Ginting: both are attacking players, but Shi, being much taller, has a wider reach and better defense.  Ginting tried to neutralize it by smashing at Shi’s body, which worked a couple of times in the first set, but no more as the match progressed.  After Shi’s win, the women’s singles ran the course naturally.   

Even though Zheng/Huang won, it sounded a warning bell for China, and the coaches must have realized that Zheng/Huang are no longer automatic and durable.  Hence the pick of Feng/Huang against YK/YN of Japan.  The Chinese public opinion denounces it as a wrong decision, but it is only a hindsight after Feng/Huang lost.  The fact that Zheng/Huang are not at their 2022 peak and Feng/Huang had beaten YK/YN twice comfortably made it an easy and sensible choice.  

In the match it was clear that Feng/Huang could have prevailed if Feng had been calmer.  In the first set Feng/Huang was leading most of the time, caught up by YK/YN at the end, and Feng served out to give YK/YN a soon-converted set point.  The second set Feng/Huang were leading all the way and won.  The third set they were leading even more, 12:6, 17:11, 19:15, 20:18.  Then Feng served out again at his match point.  Feng’s hands were visibly shaken on TV and legs gave away left and right, so the loss was entirely predictable even before the last shuttle dropped.  But in an alternate universe, Zheng/Huang could still have lost to Japan, then Feng/Huang would perform even worse against Korea, an even tougher opponent.

Three hours later, with Japan leading 2:1 came the most dramatic, everyone-talking-about match of the tournament: men’s doubles, Liu/Ou vs Hoki/Kobayashi.  The first two sets were evenly contested.  The third set Hoki/Kobayashi had the clear upper hand, leading 18:12, 19:13, 20:16.  One more point, Japan in the Final.  Then Liu/Ou scored the next 6 points to achieve the most improbable comeback ever in Sudirman Cup history.  Did Hoki/Kobayashi choke?  Not necessarily.  Sure, they lost 6 straight points, including 4 match points, but they didn’t give away any silly errors; the credit should go to Liu/Ou, who managed to make zero mistakes in fast exchanges during those 6 crucial points.  Contrast that to Feng/Huang in the mixed doubles in their own match points.  If Hoki/Kobayashi surrendered a win, so did Feng/Huang 4 hours earlier.  Still, the loss by Hoki/Kobayashi was more devastating, deflating the Japanese team and making the women’s doubles a nonissue.

So far so good for the Chinese coaches, despite the most scary drama to the Final.  But their choice for the Final was the real problem, again notwithstanding the result.  

Zheng/Huang to play was a no-brainer.  They were rested and the best there is.  Nobody would pick Feng/Huang or anybody else.   But Shi for men’s singles was a gamble.  Chen Yufei for women’s singles was also questionable, albeit justifiable.

Shi had played in the quarterfinal, coming back to beat Ginting.  He played an even longer match in the semi, coming back to beat Naraoka, a mere 18 hours ago.  All maximum pressure matches.  Shi is the MVP of the Sudirman Cup 2023, even discounting the Final.  But Shi has never been a stamina guy.  While a tired Shi is still better than Lee YG of Korea, the Chinese coaches were running an unnecessary risk.  It showed.  The second set saw Lee much stronger than the first set, leading 7:1 early, and Shi visibly tiring.  Had Lee stole the second set, he could have won the third set as well.  China should have picked Li Shifeng, the reigning All-England champion.  Li’s playing style is very consistent, nothing flashy, but makes few mistakes, a good match and favorite against any Korea men’s singles, the only weak spot for Korea.  Li is no newbie either, having played in team events in 2021 and 2022 and performed competently at crucial stages.  

Of course Shi did win, making the selection question moot.  But why gambled when you had a sure hand?  Especially considering what happened earlier, the mixed doubles match, Zheng/Huang vs Seo/Chae.  After Seo/Chae won the see-saw first set, Zheng/Huang were clearly in shock, 3:9 right away in the second set.  Credit to them to right the ship fast and mount another comeback.  Still down 16:19 and 19:20, before winning the next 3 points to avert defeat.  The third set Seo/Chae were eerily like RR/WGE of Indonesia.

So what is the moral here?  First, based on their performance during the week, Zheng/Huang were no sure bet against YK/YN either.  Second, had Zheng/Huang lost to YK/YN and China still went into the Final, China would surely pick Feng/Huang, who could lose to Seo/Chae as well.  Third, Zheng/Huang are world No 1 for a reason: you have to play over 100% to beat them, and if you let the opportunity slip by, you will have no energy sustaining the third set.  But fourth, had China lost the mixed doubles, Shi Yuqi would have played his third straight full pressure match in 3 days,  No guarantee he would have enough energy or play better than Li Shifeng, especially when Lee YG would have been fully motivated.    

The third match, Chen Yufei was a surprise pick against An SY.  An SY is the best player in 2023, and Chen just lost meekly to Yamaguchi 18 hours earlier.  China could have selected He Bingjiao, but the coaches probably reasoned like this: 1. Anybody is an underdog against An SY, Chen or He. 2. Every match is a new match, and past performance is no guarantee of future results.  Case in point, Yamaguchi already lost twice in the tournament, but still beat Chen easily yesterday. 3. The fact that Chen lost fast actually meant she was still physically fresh. 4. Chen matches up OK against An.  In terms of playing styles they are the same, rallying players, and both read each other’s shots well.  Chen has the advantage of being taller and covering more space.  An has the advantage of being younger and becoming faster.  In recent matches An evolves by attacking more and winning.  So the strategies used and executed would determine the outcome.

This match turned out to be quite telling in badminton thoughts.  Chen arguably played the best and fastest in her life.   Chen was a pure rallying player when she first came out.  Then she improved attacking in 2019, which enabled her to threaten the best players.  Yet in 2021 the Tokyo Olympics saw her reverting to the old defensive style in the Final and winning the Gold.  She has then maintained the same passive mode ever since, losing more and more to the top opponents.  But in the match against An she was attacking again, at the earliest chances, even starting from services and service returns.  Such a strategy was so apparent and consistent that Chen used it even at the last five points of the match.

In comparison, An was passive until about the last 10 points of the match, by then it was too little too late.  Not certain why, because An’s rise in 2023 coincides with her more frequent smashes.  Maybe Chen’s attacking first made it hard for An to attack.  Maybe it was her strategy to tire out Chen before attacking in the third set.  Or maybe An didn’t play to her full potential because of the pressure with Korea down 0:2?  Regardless, one can argue that Chen played her best, and An did not.

One last point: why did Chen’s attack win, while Tai TY lost to An in the quarterfinal, considering Tai is a stronger attacking player than Chen?  One explanation obviously is every match is a new match.  And expectation and strategies differ among players.  But another reason may be that Chen’s and Tai’s attacks are different.  Chen’s attacks are mostly smashes.  Tai uses tight dropshots at the net coupled with smashes.  Tai’s deceptive net shots are less a threat to An because An is willing to run.   When you drop a netshot from the back court, if your opponent retrieves it, she will likely return with a net shot, and you must rush to net from the back, spending a lot of energy, so Tai naturally makes more errors at the net, which happened at the last points in her Olympics Final and again in the quarterfinal against An.  In contrast, at least in this match, Chen attacked early, which saved her energy, and she intercepted the shuttlecock early and was able to smash often from mid court, which was more effective.  In other occasions she pulled An to one edge of the court and then smashed/dropped wide cross court, another scoring formula.  In sum, Chen’s win was due to her proactive strategy and An’s inability to adjust.  It will be quite interesting to see whether Chen keeps it up, or returns to the old, boring style in the future.

Sunday, February 19, 2023

ChatGPT: close, but no cigar

If there is one buzz word in the past three months, it must be ChatGPT.  This is the fastest, actively downloaded application ever, 1 million in 5 days, and 100 million in 2 months.  There have been AI tools, such as AI drawing, or even AI in photography for years now, but ChatGPT is different because it needs no training and is as simple as searching with google, while yielding answers more detailed, organized, and useful than google, in virtually every area.  Due to this unique nature, ChatGPT is going to be one of the watershed moments for AI in human history.

There have been a lot of hypes about ChatGPT, most justifiable.  Many educational institutions have now banned students using ChatGPT, since it will do their homework for them, often times quite adequately.  ChatGPT can give you answers to most if not all subjects, and those answers are at least plausible, in an intelligent and coherent manner.  Thus, people are crying that certain professions will lost all their jobs.  But this obviously depends on how good ChatGPT answers are.

At this moment, it is clear that ChatGPT is very good or proficient enough pertaining to questions with well defined rules or knowledge.  For example, coding, with well defined grammars and structures, and excel, which has finite formula and programming.  ChatGPT can also give you believable recipes, CVs, business plans, and steps on getting better at tennis, since there has been years and years of evolving yet consistent information available.

But ChatGPT is far from capable of doing a few other things, mostly involving creativity and new scientific knowledge.  ChatGPT may write you stories, even mimicking certain authors’ styles, which may be good enough for homework assignments or just amusement purposes, but the contents aren’t going to be anything special or usable for any novel publication.  Worse, if you ask scientific questions, not those kinds in the textbooks, but in active research areas, in hope for some literature synthesis and AI perspective, you will be utterly disappointed.  ChatGPT answers, language still presentable, are just a sequence of words that are generalized for similar question/answers but likely completely useless or wrong scientifically.  ChatGPT even made up references to support its answers: when I go to check its references, maybe 90% are nowhere to be found.  Some fake references don’t even support the claims by ChatGPT.  But more importantly, one has to wonder where ChatGPT gets these references from and how?  Does ChatGPT know how to cheat too?

So ChatGPT is not yet ready for prime time, and the current product is already showing worrying signs.  It will get better progressively, but whether it can truly overcome the above problems remains to be seen.  Another issue is: as there are fake news, there will be fake AI as well, “fake” in the sense that AI can be manipulated to be reliable on certain subjects but deceiving on others.  Fake AI will give fake answers that fool at least many people without the ability to know better or simply wanting to believe because AI looks so real!  A scenario is ChatGPT dominates the field, but it gives wrong answers while people act on them.  The other scenario is there are multiple AIs that feed diverging groups of users with different answers in certain subjects.  The same thing is already playing out daily, e.g., in the traditional and social media, and it is certain AI will be another, future divider.  How will AI such as ChatGPT will impact the world?  It won’t be tidy for sure.  

Sunday, December 25, 2022

COVID-19 Consensuses in China

China faced a tough time in Dec 2019/Jan 2020 and again three year later.  The difference is that for the first one people didn’t know what to expect, but for the second one the public do, even when the loss in 2022/2023 will be much higher. 

Unlike my Nov 1, 2022 blog posited when there was no consensus on DZ at the time, in less than 2 months’ time a few consensuses have taken hold in China.  The first one is that DZ did save lives.  The contrast between Oct 2022 and Dec 2022 couldn’t have been starker in terms of infections and deaths.  But the second one is that opening up is the only way going forward.  DZ was too costly.  Even DZ supporters, a significant number, who oppose the current policy, admit that once it is opened, it is not possible to return to DZ.  The third consensus is that most people can recover at home, hence the fear factor is much diminished.  The fourth consensus is that many will die, but that is the cost of moving beyond the pandemic.

The fourth consensus is the reality nonetheless vexing the public.  Should or could the government have done anything, or done better, to prepare for the shift from DZ or to lessen the impact?  The only one definitive is that the government should have pushed vaccination harder (Dec 2, 2022 and Dec 10, 2022 blogs).  Other than this is speculative.

The biggest complaint right now it is the government moved away from DZ too quickly, without a smooth transition to opening up.  It does seem like it, but the actionable question is: what exactly can the government do to make the transition smooth besides vaccination?  There have been debates on this in the media.  A school of thought is that you have either DZ or full relaxation, but no middle ground.  If DZ is a 9, what the rest of world is doing is 1, then the current China is 2, but there is no 5 or 6, because 5/6 is not feasible to begin with.  In other words, little can the government do to satisfy those who want the safety of DZ minus the restrictions.  Just have to bite the bullet.  Short-term pain is better than long-term. 

The other position is that we should have tried 5/6.  Even if it doesn’t work, the weeks and months of trying it will soften the blow going into 2.  5/6 presumably means very limited lockdowns, less frequent RNA testing, health code requirements going into crowded places, etc, which my previous blog had hoped for.  We will never know the answer now as it was never attempted.  Still, DZ being ineffective against Omicron didn’t stop the government from doing it for most of 2022, then why did it go straight to 2 without ever trying 5/6?  

The last position is that the government failed to prepare the necessary medicines to handle the massive infections, such as ibuprofen, etc.  But this kind of situation happens all the time around the world, and no country or government can avoid it.  According to a famous US general and President, plans are useless, planning is essential.  If you want everything prepared well before you do anything, which is subjective and easily turned into circular thinking, you will never do anything since you can never prepare everything well.  China is a big country with 1.4 billion people, so bound to have shortage somewhere sometime no matter what.  The bottleneck is seldom the production but the distribution and panic buying.  Indeed, the first two weeks saw the most empty shelves, which in late Dec is far less an issue. 

Even though we now have vaccines and Omicron is less deadly, a situation better than 2020 and 2021, everything happens all at once, family members are sick altogether, creating a new danger.  While most people stay home, even 0.5% going to the hospitals will squeeze the system, leading to unnecessary deaths.  If you didn't have Omicron before you went to the hospital, once you did, you would have it.  The ventilation systems in Chinese buildings could be upgraded to avoid long distance infections, perhaps another area the government should have done better.  At this rate China appears to aim for 2-3 months to go through what the rest of the world went through in the last 2-3 years in order to move past COVID-19.  The exact infection and death numbers are impossible to come by because they are too fragmented and localized, so have to wait for months and check the excess death data for 2022 and 2023.  Local cities and provinces are giving very high numbers which are only surveys and estimates.  People use home Ag tests and don’t report their results, or have the symptoms but don’t do Ag tests, so it is impossible to tell if all with the cold/flu symptoms are Omicron infections.  This winter is cold in much of China, so a busy common cold/flu season as well, which would infect 200 million or more people.  Worse, China doesn’t have a tradition of tracking cold/flu.

There is abundance of information about COVID-19 experiences and death announcements in the Chinese media, which helps foster the consensuses and educate the public.  The only thing missing in almost all cases is that it would have been better and more informative if their vaccinated status is shared.  So far of all the dozens if not >100 death reports, I have seen only one ever mentioning vaccination status (not vaccinated).  For this round of infections vaccination is unfortunately too late for many, but 6 month from now when the immunity wanes, it will help to vaccinate and boost sooner, again.  How China recovers from these tough 2-3 months will be closely watched.