After Millions of Dollars, Microsoft Bing is Just as Smart as…Las Vegas [When Data Fails]

At Kellogg, we learned that people in aggregate tend to be quite correct (for example, say you have a random amount of jelly beans inside a big jar. Ask people to guess the amount of beans. When you average all the guesses, it will come out quite close to the real number, even if the real number is large and random, like 1,724).

According to How Microsoft got so good at predicting who will win NFL games, Microsoft Bing is an awesome prediction guru of human intelligence, machine learning, and big data:

Bing Predicts is run by a team of about a half dozen people out of Microsoft’s Redmond, Washington headquarters. It uses machine learning and analyses big data on the web to predict the outcomes of reality TV shows, elections, sporting events, and more.

How Microsoft got so good at predicting who will win NFL games

In 2014, Bing was 67% accurate predicting NFL winners.

In all, the Bing Predicts model considers hundreds of these different signals, or data points, for each event, like an election or game, Sun said.

So far this year[2015 to game three], Bing is about 60% accurate in predicting NFL matchups.

Sounds great, right? However, my first thought was, who cares about winners? I can’t bet on winners, this is why the spread exists, to create (theoretical) 50/50 bets that bookies can make stable revenue from.

My next question is, in this awesome model built from millions of dollars in labor and computing power, are the prediction results better, hopefully at a statistically significant (p = .05) level, than information I could get free from a public resource? How little can I spend to get reasonably close results to aid in my for-profit wagering?

Let’s look at Las Vegas betting spreads.

Booking Odds

According to Inpredictable.com and its 2013 article Is the NFL Betting Market Getting More Efficient?, the answer is NO, Bing’s modeling is no better than me looking at the latest odds online.

From 1989 to 2013, Las Vegas favorites were correct 66.8% of the time. With a sample size of 15 years, and looking at the chart above, I can say that Vegas is pretty good.

1 signal – Vegas odds – versus hundreds of signals – Microsoft Bing = the same result.

Great work, Microsoft.

The Best of City Slam Chicago [Photos]

Ha and I went to City Slam Chicago a couple of weekends ago. It was held at Seward Park, which was also the setting for this Uncle Drew / Kyrie Irving / Pepsi commercial below:

I won’t spoil the winner for you (it will be aired on TV in late July), but here are some of the best photos Ha took of the event, with judges like Doug McDermott of the Chicago Bulls judging.

One last photo of the park with downtown Chicago in the background:

10 Things to Learn from Soccernomics (by Simon Kuper)

imageSoccernomics: Why England Loses, Why Spain, Germany, and Brazil Win, and Why the U.S., Japan, Australia, and Even Iraq Are Destined to Become the Kings of the World’s Most Popular Sport by Simon Kuper is basically the Freakonomics for Soccer. It looks at how the game is played but also the business of the sport and has many many many good things to note. Ideally, this article would be fifty things to learn, but here are ten things to learn from Soccernomics:

  1. the most dangerous corner was the inswinger to the near post. The beauty of the inswinger was that it sent the ball straight into the danger zone.
  2. In short, the more you pay your players in wages, the higher you will finish, but what you pay for them in transfer fees doesn’t seem to make much difference. While the market for players’ wages is pretty efficient—the better a player, the more he earns—the transfer market is inefficient. Much of the time, clubs buy the wrong players.
  3. A new manager wastes money on transfers; don’t let him.        Use the wisdom of crowds.        Stars of recent World Cups or European championships are overvalued; ignore them.        Certain nationalities are overvalued.
  4. Older players are overvalued.        Center forwards are overvalued; goalkeepers are undervalued.        Gentlemen prefer blonds: identify and abandon “sight-based prejudices.”        The best time to buy a player is when he is in his early twenties.        Sell any player when another club offers more than he is worth.        Replace your best players even before you sell them.        Buy players with personal problems, and then help them deal with their problems.        Help your players relocate.
  5. It was Rupert Murdoch who went to English clubs and suggested putting them on satellite TV; the clubs would never have thought of going to him. In fact, the clubs often fought against new moneymaking schemes. Until 1982 they refused to allow any league games to be shown live on TV, fearing that it might deter fans from coming to the stadium. They couldn’t grasp that games on television meant both free money and free advertising. There is now a good deal of research into the question of how many fans are lost when a game is shown on TV. Almost all the evidence shows that the number is tiny, and that the gate revenue that would be lost is usually well below the amount that would be made from selling extra matches for television coverage
  6. Let’s compare the breakeven rule to the salary cap, widely used in American major-league sports. A salary cap generally restricts teams’ spending on players to a fixed percentage of average club income (55 percent, say). UEFA’s breakeven rule restricts player spending to the level of club income defined as “relevant.” The key difference is that the US cap is the same for all clubs, whereas breakeven caps each club at the level of its own resources—a lot for the larger clubs, not so much for the smaller clubs. So while the American salary cap encourages competitive balance between clubs (it stops the Dallas Cowboys from spending infinitely more than the Jacksonville Jaguars), UEFA’s breakeven rule cements inequality by making it harder for smaller clubs to compete with the aristocrats. This hardly seems fair, unless by “fair” you mean the idea that big clubs should be protected from competition from upstarts.
  7. “Two thirds of goals come from possessions won in the final third of the field,” he lectured. The great sin, to him, was losing the ball near your own goal.
  8. “Soccer” was the most common name for the game in Britain from the 1890s until the 1970s. … Soccer conquered the world so fast largely because the British gentleman was such an attractive ideal.
  9. Daniele Tognaccini, longtime chief athletics coach at AC Milan’s “Milan Lab,” probably the most sophisticated medical outfit in soccer, explains what happens when a player has to play sixty tough games a year: “The performance is not optimal. The risk of injury is very high. We can say the risk of injury during one game, after one week’s training, is 10 percent. If you play after two days, the risk rises by 30 or 40 percent. If you are playing four or five games consecutively without the right recovery, the risk of injury is incredible.
  10. Kevin Alavy, the managing director of the Futures Sport + Entertainment consultancy whom we met earlier in the book, says that often the Premier League enters a new territory by offering itself on free-to-air TV. That’s what it did in eastern Europe, across the former Soviet Union, and more recently in Indonesia. Until a few years ago, hardly any of the 247 million Indonesians knew soccer. But given the chance to watch top-class English games on TV for free, many began watching. In recent years, says Alavy, “typically Indonesia has been the number one market by TV audiences for the Premier League.” Once Indonesians were hooked, the Premier League gave them a new TV rights deal that offered them far less free soccer.

Stephen Curry Team USA Jersey Review (AliExpress, Aimee Smith)

To reward myself for a recent 2nd place (but a cash prize!) finish at an innovation competition at Kellogg, I decided to buy myself a fake jersey. Normally I wouldn’t do this, at least not knowingly (damn you eBay!), but I wanted to look at the process of buying things from AliExpress (an Alibaba company), which helps Chinese vendors sell directly to international consumers. Prior to this, I had been curious about the quality of fake jerseys, and who made the best ones. You can find plenty of reviews of fake jerseys on YouTube, but sub-Reddits like http://www.reddit.com/r/sportsjerseys and http://www.reddit.com/r/basketballjerseys/ are also very useful.

I ended up deciding on a Stephen Curry Nike FIBA 2014 jersey, primarily because as far as I know, it was never sold to the public as Curry is signed to Under Armour. For the seller (there are a number of them), I chose “Aimee Smith”, who has great reviews both online and within the AliExpress storefront – as of my writing this post, she had received feedback from 10,299 people with 99.4% reporting positive transactions. This means about 10,237 people liked “her”, while just 62 people did not. I, for what it’s worth, also liked her.

Before going into the jersey , I’ll talk a bit more about the AliExpress experience:

https://i0.wp.com/style.aliunicorn.com/wimg/buyer/single/standard-definition-logo.png

  • Security: AliExpress is very easy to use – you do not need to worry about credit card security issues, or any other negative fears you may hold about buying from some “random” site in China. AliExpress (and Alibaba) is no random company – they have been doing this a long time, and you will see that in your shopping. AliExpress accepts all major credit cards (American Express, Visa, Mastercard) and also has a clear buyer protection policy. The website is no harder to use (and is likely easier) than any other American site that you like.
  • Responsiveness: It is very easy to ask sellers questions, and Aimee Smith in particular is very responsive. She answered all my questions in less than a day and often, within minutes (keep in mind the different time zones). On eBay, whether you get any responses at all is random based on seller. From my experience, eBay sellers response properly less than 50% of time.
  • Reviews: If you are fairly comfortable buying things on eBay or Amazon 3rd Party sellers, you will be fine on AliExpress. What I especially like about the service is that you can review the specific product from that seller. Thus, I could see what others felt about the specific Curry jersey I was buying from Aimee Smith. This makes sense for AliExpress since their products are not one-off goods (e.g. I only have one to sell, like my limited edition baseball card). This wouldn’t make as much sense for many small volume eBay sellers but it would for some and it would definitely make sense for many Amazon sellers.
  • Shipping / Tracking:  Assuming you get a good seller, AliExpress also does a great job of letting you know the status after you order. Even though you can check your order details on the site, however, the site never sends a detailed receipt via email

If you’re interested in AliExpress, I definitely say try it without fear, but do make sure you check for sellers with strong feedback before doing so, just as you would (I would hope) with purchases from eBay or Amazon 3rd party vendors.

On to the jersey!

I am happy with it. No complaints considering the price and fact that I cannot get (a major incentive to buy fake jerseys is when an authentic version is unavailable) a real one. The Curry jersey cost less than $22 shipped, and I received it in about 2.5 weeks (coming from China, after all) after ordering. If you are interested in fake jerseys, I would not hesitate to get one from Aimee Smith, and all the feedback online I have seen agree.

While I have not worn the jersey to play basketball in, the material is really soft. I am sure it does not have Dri-Fit or any other moisture-wicking technology built-in, but the jersey is light and I could see myself wearing it. In this sense, if you are normally someone who wears t-shirts for athletic wear, I would recommend this as an alternative. I bought a size small (I am 5’6, 140 lbs.) and I feel it was correctly sized.

To get into the details, let’s do a comparison of images from Getty Images,

455063814 455055924a Paul George authentic jersey auction on eBay,

and Nike Store images.clip_image0014clip_image001And for reference, my Stephen Curry Aimee Smith USA Jersey:

I imagine it might be tricky to compare all these images in this kind of vertical-line view, so I’ll summarize what I see as best as I can:

  1. For the USA lettering on the front, the authentic jerseys are flat (almost like a screen print integrated into the jersey material) and the borders around the letters are dark. On the Aimee Smith version, the borders are red, and the lettering is stitched.
  2. I believe the USA badge (on right chest of player above Nike symbol) on the authentics is printed on the jersey, while it is embroidered on the Aimee Smith.
  3. The placement and size of the “4” on the front match Getty images fairly well. There is no FIBA patch on the left clavicle, but some fake sellers have it.

Moving to the back and other components. This time, I will show Getty first, then the Aimee Smith and Paul George authentic alternating different parts.

455160982imageimageimageimageOther than the “4” on the back being too high relative to the Curry name, the back looks pretty good as well. While I could not tell how accurate the lettering was based on the Getty Images photo, comparing the Aimee Smith Curry to the Paul George authentic reveals that the “R” looks pretty close. Again, the Aimee Smith is more of a Swingman jersey in which all letters and numbers are stitched, which is not the case with the authentic. The jock tag at the bottom front of the jersey is much different between the two jerseys, which is also true of the collar tag. Nonetheless, if you had nothing to compare either with, it would be difficult to say that one of them appears fake.

As I mentioned before, the quality of the jersey is excellent – it probably looks no worse than an authentic Obsidian Warriors Swingman alternate jersey I bought from the NBA store a few months ago for 4 times the price. It could have easily sold as an authentic Nike USA swingman and it is likely better than an authentic replica Andre Iguodala Nike USA jersey that I own. I hope this helps, but feel free to ask me questions!

10 Things to Learn from Game Changer: How the English Premier League Came to Dominate the World (Mihir Bose) [Soccer, Sports Business]

imageContinuing my research into soccer-business (see my previous post, 10 Things to Learn from Sounders FC), I recently finished Mihir Bose’s Game Changer: How the English Premier League Came to Dominate the World. Although the book may discuss the TV contract negotiations / creation of the Premier League in the early 90’s a bit too much, there is a lot of great insight into the modern game (from a business perspective).

Here are 10 things that I would like to highlight from the book:

  1. All the major English clubs have targeted these fans and Chelsea have been particularly active trying to target disillusioned fans of other clubs who are not doing so well. These fans follow success and do switch loyalties.
  2. Indians probably watch more live Premier League matches, including the 3pm Saturday kick-offs, than most do in Britain.
  3. When the Premier League was founded in 1992, La Liga in Spain and Serie A in Italy were the dominant European leagues, secure in their own homelands and in the wider world. Italian football had even invaded England’s football scene, being broadcast every Sunday afternoon on Channel 4. But in the last 20 years all that has changed.
  4. In another curious reversal of the American experience, football embraced segregation in order to cope with fan violence. The Americans spent much of the 1950s and 1960s trying to eliminate segregation based on colour from their society. English football in the 1970s decreed that fans could only watch if there was strict segregation between fans of rival teams. For all the changes that have since occurred in football, this separation of home and away fans still exists, with grounds having large signs directing them away from each other. And even in new stadiums such as the Emirates, Arsenal’s ground, it is made clear even in the executive box areas that fans should not be wearing the colours of the visiting teams.
  5. Against this background, it is hardly surprising that English football failed to attract non-white fans. This failure persisted even throughout the 1990s when English football made strenuous efforts to oppose racism. The most prominent initiative, Let’s Kick Racism Out of Football, was launched jointly by the PFA, the Commission for Racial Equality and the Football Trust, and within a year all but one of the professional league clubs in England and Wales had signed up to its 10-point plan. This effort was supplemented by multiple individual initiatives from clubs, fanzines and community groups. However, in 2001 the FA Premier League’s national fan survey found that only 0.8 per cent of “active top-level fans” were Black British or British Asian. This represented a rise of only 0.1 per cent since the previous survey in 1997, and compared to a total minority ethnic representation of 13 per cent in the UK population. The same survey found that 7 per cent of all Premier League fans had reported witnessing racism against other fans and no fewer than 27 per cent had reported racism displayed against players at matches.
  6. Hilly’s great idea was to take the score and the time and put it on the screen. I remember thinking, “Oh fuck — why have we not thought of that before?” It had never been done. Anywhere in the world. Can you believe it, only 20 years ago when you were watching football, you’d switch on and you didn’t know who was playing, you didn’t know the time, and you didn’t know the score.
  7. The Champions League was born! The new marketing concept was both innovative and commercially adapted to the changing market conditions. Each sponsor would receive exclusivity in its product area, not only in the stadium as was previously done, but also on TV, with commercial airtime spots and programme sponsorship. By linking stadium advertising together with on-air sponsorship, it became almost impossible for non-sponsors to associate with the competition. The three pillars of stadium advertising, commercial airtime, and programme sponsorship generated a “multiplying media effect” that offered new levels of recognition to the sponsors. A “less is more” approach was taken and a maximum of eight international sponsors was decided upon. The sponsor package included four stadium advertising boards, ticket allocations, and identification on TV interview backdrops and in the VIP and press areas. Each of the sponsor ticket holders was also invited to specially arranged hospitality suites before and after the matches.
  8. In the years since 1995 the Bosman ruling has led to other changes in the transfer regulations. It led to transfer windows allowing player transfers only twice in a season, once at the start and once in the middle. But most significantly, it greatly increased player power. The court ruling meant that footballers were now free to move when their contracts expired. And this in turn paved the way for footballers to earn multi-million-pound salaries. Sport could no longer be exempt from EU competition rules and had to be treated like any other business. The net effect was that unless a club arranged a transfer before the player entered the last year of his contract he was free to move at the end of it. This tilted power decisively in favour of players and away from clubs. Now players, particularly high-profile stars, were masters of their own destinies. And as free-agent players they could suddenly demand huge signing-on fees and salaries on the basis that the club they were joining did not have to pay anything in transfer fees. Football clubs were powerless to prevent their best players from leaving at the end of their current deals. Conversely, players under contract could demand bigger, better and longer deals — because the threat of being able to leave for free, especially if they would otherwise command high transfer fees, was something clubs could not ignore.
  9. In his very first season Abramovich spent £111.3 million on transfers. Not only was this more than anyone had ever spent before in English football, but the Russian dramatically changed the terms of trade. He paid the full transfer fee at the timing of signing the player. This broke with the usual convention of fees being spread over a period of time. Cash on the nail proved a lifeline for clubs facing cash-flow problems. Indeed, it was immensely beneficial to clubs such as West Ham who were then under financial strain, as the then club chairman, Terry Brown, acknowledged. Since 2008 and the purchase of Manchester City by Abu Dhabi United Group, Chelsea have been challenged and even overtaken in transfer spending. In 2010–11 the club spent £141 million on players, the third successive season they had spent more than £100 million, double that of Manchester United and comfortably ahead of Chelsea at £91 million.
  10. While the 20 Premier League clubs had an income of £2.3 billion, the remaining 72 clubs in professional football in England between them had an income of under £700 million. Two of them, Port Vale and Portsmouth, are in administration and 13 of the clubs in these three divisions are classified by financial experts as in distress, meaning that they have serious court actions against them, including winding-up petitions and high court writs, or have been issued with striking off notices for late filing of accounts or have county court judgments against them.