Hacker News Reader

Comment by devindotcom | Hacker News Reader

I love watching extreme outliers in play - its one of the main reasons I watch. Guys like Messi, Jordan, Ali are so high above the standard (and often so different) that they elevate the whole sport just by taking part.

View on HN

The cricket player Donald Bradman [1] is the outlier among outliers. Even compared to greats like Michael Jordan, the statistical significance of his achievements almost defy belief.

[1] http://en.wikipedia.org/wiki/Donald_Bradman

If you're willing to branch out beyond physical sports, Marion Tinsley [1] is perhaps even stronger, losing only 7 games in his 45-year career of checkers.

[1] http://en.wikipedia.org/wiki/Marion_Tinsley

I don't think checkers is comparable, since there's no luck component in checkers. In a physical sport with random winds/ground conditions/ball imperfections/bat imperfections/etc, you have to be that much better than your competitor to beat them consistently (and that margin varies by sport, where baseball, for example, has many more random factors than tennis).

This is a false conception people have about sports like checkers, chess, go, etc. It depends on how you think about luck. Sure, if you're an omniscient being who knows the entire decision tree for a given move in [checkers/chess/go/etc...], there's no luck, in the Monte Carlo sense. But, except for simplified positions, nobody is fully aware of the consequences of their move in these games. This "fog of war" is a good analogue for the random winds/ground conditions/ball imperfections that you mentioned in other sports. It's often why great players in these games sometimes lose to far worse players.

Checkers is a solved game. It proven possible to always not lose a game if you play correctly from the beginning. As such high level games feature many draws and few wins. Marion Tinsley also didn't compete in the World Championships for 20 years in the middle of his 45 year streak, during which time Walter Hellman was world champion for 17-odd years and only lost 7 games at the World Championships during that time.

Marion Tinsley may be the best checkers player of all time, but his achievements are nowhere near unbelievable given the nature of the sport and achievements that are similar to his. Statistically, Donald Bradman is much more of an outlier when compared to other people who have played his sport.

That's an interesting perspective. Thanks.

I'm not sure I agree with you

physical sports sometimes come with sustainable competive advantages

until he gets old, LeBron is always going exceptionally big, and athletic

in contrast, checkers exists on a much more fragile competive advantage

You should check out this [1] titled "The most remarkable graph in the history of sport" to get an idea of just how much of an outlier Bradman is. Read the whole thing, it's brief, but a quick excerpt:

   To understand how Bradman’s 99.94 average compares with other batsmen, 
   consider that a typical topflight batsman has an average in the range 45 to 
   55. Batsmen with averages above 55 are once-in-a-generation phenomena who 
   dominate the entire game. After Bradman, the second highest average in 
   history [2] belongs to South African Graeme Pollock, with 60.97, and the 
   third highest to West Indian George Headley, with 60.83.

[1] http://michaelnielsen.org/blog/the-most-remarkable-graph-in-...

And it would have been an average of 100, if he hadn't got out for a duck (no score), in his last game.

Yep, all he needed was a single 4 and he would have had an average of 100. Impressive guy, the idol of basically every kid whos ever played cricket.

Prior to the 1932-33 Ashes Tour(1) his test average was 112.29 after averaging over 200 per innings against South Africa and dropped to a mid career low of 89.55 as a result of the English tactics. Very impressive.

(1) http://en.wikipedia.org/wiki/Bodyline

Agreed, but I'd put Heather McKay up there with The Don.

https://en.wikipedia.org/wiki/Heather_McKay

> When she retired in 1981 at the age of 40, McKay had gone nearly 20 years undefeated (with the only two defeats to her name occurring at the beginning of her career).

Interestingly, both come from Australia.

Great to see a cricket comment.

I wonder how he would've fared against the present day Akhtars, Akhrams, Donalds, Steyns, Warnes or Muralitharans.

Would older greats have stood out as much faced with the technological and biological advances of the present day?

possibly apocryphal story, but...

During the recent radio coverage of the current Ashes tour, between England and Australia, I heard a great quote about Sir Donald Bradman, the famous Australian cricket player. It went as follows: In the 1980s, England's fast bowler Bob Willis had the fortune to meet Sir Donald Bradman, legend of cricket. Bradman had been the dominant player from the 1930s and 1940s and the Australian had amassed stats that are unlikely ever to be bettered. Willis was keen to get The Don's view on what he might have averaged if he had played in the modern game. Bradman looked at Willis and replied that he reckoned that he would have averaged about fifty runs per game. Willis expressed surprise, having thought that the great man might have expected to have averaged more. Bradman looked Willis in the eye and then responded along the lines of "Well, I am in my seventies now!".

from http://www.sportinglife360.com/index.php/donald-bradman-cric...

I have watched on TV Ian Chappell narrate this same story, as an incident between him and Bradman when they met at a party celebrating Bradman turning 90. Very likely apocryphal.

Bradman survived the 'Bodyline' (http://en.wikipedia.org/wiki/Bodyline) series where the bowlers (pitchers in American parlance) literally targeted the body to injure the players and to hamper Brandman's prowess. It had the desired effect (Brandman only averaged 56 for the series as opposed to his career average of 99.94) and resulted in multiple injuries.

I think he would do just fine today.

As long as he gets the advantages of modern nutrition, fitness conditioning, and analysis, then I'm pretty sure Bradman would tower over modern day players.

Don't forget, in Bradman's time, there were no covered pitches, leg side theory was legal, batmen didn't wear helmets or much in the way of protection (and bowlers weren't that much slower than they are now).

Sure, but that is more than a hundred years ago. Back then the sport was not nearly as competitive.

I've always wondered about what it is that kicks someone up from great to "how is that even possible?". For example Mozart being asked to remove a ring because his playing was so good an audience thought the ring was a talisman. It's one thing to be so good people think you cheat. It's another level when people think you are magic.

Clearly opportunity, talent, and work play their parts, but there have been many people that have all three and don't turn out a Gauss, or a Newton, or a Pele.

Actually, thinking about it a bit more I think there is a natural explanation. If measured ability follows a normal distribution then the farther out you go the more spread out (on average) individuals become. For example, draw a billion samples from the right half of a normal distribution with mean 0 and stddev 1. There will be more drawn between 4 and 5 than between 5 and 6. The extreme outliers will be in some sense more lonely.

Edit: here are the most extreme values from a run of a billion draw with the distance to the next smallest value.

5.967211966 0.056293894

5.910918072 0.043174783

5.867743289 0.081026984

5.786716305 0.043416671

5.743299634 0.094533593

5.648766041 0.000088773

5.648737164 0.005932669

5.642804495 0.011669773

5.631134722 0.003206244

5.627928478 0.004643796

5.623284683 0.013723654

5.609561028 0.014653298

5.59490773 0.018714748

5.576192982 0.001690759

5.574502223 0.031026791

5.543475432 0.002331827

5.541143605 0.022237834

5.51890577 0.004242771

5.514662999 0.009347961

5.505315038 0.004298492

5.501016547 0.02271157

"Ability" isn't a single variable either. There's a combination between natural talent and hard, determined work. For instance, many people thought Ricardo Quaresma and Cristiano Ronaldo had similar amounts of natural talent as young players, but Quaresma wasn't as determined as Ronaldo to become the greatest player in the world. Conversely, Michael Jordan's contemporary Charles Barkley didn't really have Jordan's natural talent, but made up for it by working exceptionally hard. So you have at least two (and possibly more) variables, all of which have to reach extreme outliers together, to produce an athlete like Messi, Jordan, or Gretzky.

One thing that might play into it is that actual talent distributions are leptokurtic, ie they have a fat tail, while our internal model is more of a normal distribution, or maybe even some kind of truncated normal based on what we have previously seen. So these outliers that occur naturally in the leptokurtic have a much lower perceived likelihood.

PS: the bit about talent distributions being leptokurtic is pure conjecture, no idea if it is true. Seems like something Nassim Taleb would say.

This is an amazing comment. The only way it could have been funnier is if you had made up the word "leptokurtic."

You should get Malcolm Gladwell's people on the phone and set up an interview.

If you look into that wave of great central european composers most of them were third generation successful musicians/composers from big families. Natural selection in action. I'm sure genes played into it.

We have the rare opportunity of watching two top 10 forwards of all time at their prime at the same damn time. One of those sport rivalries to tell your grandchildren about.

Similar thing in tennis with Federer and Nadal.

Nadal's stats on clay is incredible. He's lost just 11 matches and won 270 odd in the last 9 years.

I think an even better example would be Nadal v Djokovic on the 2012 Australian Open final. According to Nadal the best game he ever played and consolidated Djokovic as one of the greats. And you could see how one lifted the other up to do better and be better. Such an amazing game.

http://en.wikipedia.org/wiki/2012_Australian_Open_%E2%80%93_...

That game was such a amazing thing to watch, I would be hesitant to assign a winner or loser to that game, I would say both won the game.

I remember at the presentation ceremony the players were finding it difficult to stand due exhaustion and Nadal started his speech[Post match] with a 'Good morning' since it lasted so late into the night.

What about Nick Kyrgios defeat of Nadal today at Wimbledon?

Quote from Kyrgios before the game:

"Between us, we have 14 Grand Slams, so it should be a great game".

The guy's got a great future.

Well, one man can only dominate for so long.

The most amazing part about them is their ability to adapt with age.

The Michael Jordan of his first 3 years in the NBA was about attacking the basket with an insanely fast first step. The Michael Jordan of his second trio of championships didn't have that level of quickness or burst anymore, but he developed a post game along with a fadeaway jump shot that was virtually impossible to guard. The man had a 20+ point per game average after returning from retirement for the second time at age 40. I don't think the average person understands how quick and athletic NBA players are. A 40 year old man being able to play that game at that high a level with a terrible team around him for 82 games a season is insane.

I wonder if these kind of stats are available for players of past eras. I would be fascinating to see what Pelé or Maradona would look like compared to Messi !

It is true, though, that you only really become a football legend when you win the World Cup (or show some insane talent at the World Cup). Pelé, Maradona, Zidane, Ronaldo et al. all won world cups. And Pelé is considered the greatest of all time precisely because he won 3 world cups out of the 4 he participated in.

Very much doubt it - it's all about http://www.prozonesports.com/subsector/football/

Pretty impressive kit.

I don't understand the point of this link. Is this related to the comment you replied to or is simply spam?

Another player to add to this list is the hockey player Pavel Datsyuk. IMHO he is the most underrated player of this generation, although a perennial all-star and stanley cup champion. The way he handles the puck and dekes people out is unreal. . . He is unique on this list as he is not a dominant superstar, but he is an unstoppable outlier.

It's the same in tennis... Rafa and Federer have both been titans, in a class of their own, for years, but sport generally is just getting more and more professional, competitive and extreme as we hone our ability to analyse performance.

Another outlier: Sébastien Loeb (paired with Daniel Elena) has been in a league above every one in rally. His abilities extend way beyond rally: you could throw him anything with four wheels and he'll pull something great out of it (see his team's 2nd place at Le Mans in 2006)

The unfathomable thing is that in France, basically no one cares about his (and others) success, while everyone raves about the national men's† football team.

† women's team apparently performs significantly better on the same period but it's a blackout too. I know why I'm not into football, the whole thing is pervasively biased in so many ways...

I watched rally for a few years, but lost interest after Loeb won his third straight WC. There was absolutely no suspense, everyone knew Loeb would win most of the rallies barring the winter courses.

As for Women's football, there's a reason it's not really appreciated and that is the fact that the skill level is significantly lower. Far fewer women play, which results in less competition and less highly talented individuals. Another problem is that the goals are sized for men, while the women keepers statistically speaking are smaller and can cover a smaller area, which means it's far easier to score with cheap shots. It all makes for rather dull spectating, sadly.

Not sure why that's unfathomable.

Canada is nuts about hockey. Americans as a whole are indifferent at best, yet we share a border. Mexicans are crazy for soccer.

If the US had the world's greatest soccer player, the majority of the country wouldn't care. Soccer is a sport that is played by about 30 million Americans a year, but it's a blip on the map at the collegiate and pro level here. People simply enjoy different sports, and the local culture reinforces it. Jim Brown was probably the greatest lacrosse player who ever lived, but his career was a footnote because the sport he played professionally was football, and he was an outlier at it.

Not true that no one cares about the success, but football is just much more popular than rally. Motor sports are incredibly expensive to practice and you can't expect them to have the popularity of cheap "universal" sports like football.

Usain Bolt is an example of this. I remember being totally shocked when he ran the 100m in 9.59 [1]. Check out this awesome graph from the "Men's 100 m progression" wikipedia page.

[1]https://www.youtube.com/watch?v=3nbjhpcZ9_g [2]http://upload.wikimedia.org/wikipedia/commons/6/62/World_rec...

I think one thing that's interesting is how this relates to things like pay equality.

People throw their arms up and say "How the hell can a CEO pay themselves 1000 times what the lowest paid staff gets paid"

But that CEO may be a "Messi".

Some people really are 1,000x more productive/as talented as the average person. And so should be compensated accordingly.

Besides the obvious inequality, and unlike Messi, ceos rig the game.

If the company performs incredibly well, they receive dynastic levels of wealth. If they completely fuck up... they receive dynastic levels of wealth. If they're essentially fired for cause after 5 months... they receive enough wealth to retire ($7.3m for 130 days including weekends) [1]. If an already stunningly wealth ceo submits fraudulent expense reports trying to get in the pants of one of his employees (who doesn't seem to have been much besides eye candy), demonstrating at bare minimum incredibly poor judgement along with the possibility of sexual harassment, he's given dynastic levels of wealth ($40m) and merely made to resign. [2] In short, there's no performance requirement.

[1] http://money.cnn.com/2012/05/14/technology/yahoo-ceo-no-seve...

[2] http://www.businessinsider.com/backlash-against-hewlett-pack...

AFAIK every one of those statements can be replacing the CEO with a sports star and everything is still true

In short, there's no performance requirement?

Sounds nice except when you consider that:

1. Even Leo Messi isn't "1000x more productive/as talented as the average [footballer]". Messi is a difference maker to be sure, and one might even say he counts for two or three men on the pitch, but even an extreme outlier like him only, perhaps, double or triples the statistics of another very good player. 1000x? No, simply no.

2. All of the recent studies coming out demonstrating that highly paid CEOs have no correlation with high performance, or worse either a correlation or possible contributor to poor performance, would say differently.

3. While most CEOs are certainly talented, many simply achieve their position thanks to connections and networking, and are not necessarily superhumans compared to their employees.

So no, this really says nothing about pay equality.

1000x is a bit much, but double/triple production is hugely significant when you realize that teams are constrained to only playing 11 players at once. Its not like they can make up his production with 2 average players playing together.

Nah, people have been able to block even Piero if they gang up on him at once this world cup. Messi himself was stiffled by Iran until the last few minutes of the game.

That's how most of the teams have been able to counter Messi so far; whenever he gets the ball, make sure two or three guys are on him. That he's still scored four goals despite that is a testament to his prodigious skill, but its certainly been effective at cutting that number down.

You're sort of on the right track with this (and I'm not sure why you're getting downvoted). The difference in talent of the top 250 or so CEOs is actually very small, but the huge size of modern companies means that even a small difference in CEO talent makes a huge difference in dollars.

Gabaix and Landier [0] run a counterfactual (based on their model) and suggest that "if firm number 250 could, at no extra salary cost, replace for a year its CEO by the best CEO in the economy, its market capitalization would go up by only 0.016%. ... this talent difference implies that the pay of CEO number 1 exceeds that of CEO number 250 by 530%. Substantial firm size leads to the economics of superstars, translating small differences in abil- ity into very large differences in pay."

[0] http://qje.oxfordjournals.org/content/123/1/49.abstract

250 doesn't sound like a large enough statistical sample to eke out even the most trivial performance differences in a single person when you consider the size of the companies being discussed.

High pay isn't really the issue so much as spiraling executive pay while other wages are stagnating and the economy is screwed. Also, if tech was run like modern football, the engineers salaries would be far higher. Tech is more like the old days of football before the players got more control over their contracts.

Definitely. Sports star salaries aren't keeping up with growing gate receipts, concession sales, TV DEALS, etc. Especially when you at the short careers and how injury can derail things. Owners/club execs are laughing all the way to the bank.

It's almost more like a guy going broke trying to invest his life savings on a stock tip than a CEO. You fail to ascend to superstardom? You end up with no money, no roots since you've moved around 30 times in five years, few employable skills outside of being physically fit, but wait... you're not physically fit because your body is broken from years of abuse.

I don't know if ya'll are watching the world cup right now but Messi was stifled by Iran, and Nigeria just recently. Belgium is coming up next so....

Argentina would be doing far better with a team that play together better than with Lionel Messi at the helm, and Messi himself said so.

Statistically, high paid CEOs are worth less than low paid ones. There's a negative correlation between CEO pay and stock performance.

Here's a thorough analysis that argues that Dennis Rodman was not just the best rebounder of all time, but that he dominated his position more than any other basketball player dominated theirs.

http://skepticalsports.com/?p=112

Reply on news.ycombinator.com