Here’s a timely example: traditionally, when you have two successive quarters of negative GDP growth, you’re in recession. We have a quarterly GDP time series for Australia, and we want to know how many recessions they’ve been having down under. How do we count these events?

Of course you can always write a loop, but it’s clumsy. You can use the “embed” function, that’s much better in a lot of cases. Or you could use regexps. Here’s how.

First, our Australian GDP data, courtesy of the expsmooth package:

plot(ausgdp, main = "Australian GDP per capita", ylab = "dollars", xlab = "Year")

(caveat: the series is per capita GDP, not absolute GDP, so my recession count might be off).

Regexps find patterns in character strings. They run the Web, pretty much, so googling “regexp tutorial” should turn up about a million results. Patterns are specified using a mini-language, for example “1.1” will match all strings of length three of the form “1 followed by anything followed by 1”. In R you can use gregexpr to find matches:

txt = "54131x1blahblah"
gregexpr("1.1", txt)

## [[1]]
## [1] 3
## attr(,"match.length")
## [1] 3
## attr(,"useBytes")
## [1] TRUE

The great advantage of regexps over other techniques is that they can match patterns of variable length. In the case of recessions that’s exactly what we want.

We turn our GDP time series into a sequence of 1s and 0s, where 1 means GDP increased in that quarter:

delta = (sign(diff(ausgdp)) == 1) + 0
head(delta)

## [1] 1 0 0 1 1 1

regexps work on character strings, so we paste everything together:

ds = do.call(paste0, as.list(c(delta)))
ds

## [1] "1001111110001101111111000111111110111111100000111111111110011111111111111100000011111111111111111111111111"

(does anyone know a more elegant way to do this conversion?)

The pattern we’re looking for is two consecutive zeros or more: the appropriate regexp is “00+”, which matches patterns that have at least two zeros.
By default the regular expression engine will find the maximal match. What this means is that in the string “1100011”, it will consider that there is only one match: the substring “000”, and not two successive “00” matches. That’s not always the behaviour we want, but in the case of recessions it’s exactly what we need.
gregexpr will return all the matches, along with the lengths of the matched segments (that’s how long recessions lasted):

matches = gregexpr("00+", ds, perl = T)[[1]]
matches

## [1]  2 10 23 42 58 75
## attr(,"match.length")
## [1] 2 3 3 5 2 6
## attr(,"useBytes")
## [1] TRUE

We can use these matches to highlight recessions. The following code outputs a list of indices in the ausgdp time series that correspond to a recession:

m.length = attr(matches,"match.length")
recessions = sapply(1:length(matches),function(ind) matches[ind]+0:(m.length[ind]))

and we can now highlight the corresponding regions in the plot:

hl = function(inds) lines(time(ausgdp)[inds], ausgdp[inds], col = "red", lwd = 3)
plot(ausgdp, main = "Australian GDP per capita", ylab = "dollars", xlab = "Year")
tmp = sapply(recessions, hl)  #Used for side-effects only

That’s it. Things get a bit trickier when you have patterns that might be overlapping, for example when you’re interested in extracting all patterns of the form “1.1”. See that question on stack overflow for a solution involving look-ahead.

[Note: updated following comment by DataMeister. Fixed variable name]

     Some statistics that defensive players might excel in are rebounds, steals, and blocks. Let’s see how big ‘ole G’s defensive stats look compared to the rest of the league leaders.

     Dwight Howard led the league with 12.4 rebounds per game. With the 23^rd highest average in the league, Marc gathered in 7.8 boards per game. Stunning. Let’s also mention that his Grizz teammate, Zach Randolph, was 4^th in the league by grabbing 11.2 rebounds each game.

     Let’s compare another statistic. The NBA season is 82 games. Marc Gasol displayed great stamina by playing in 80 games for the Grizzlies. After an 80 game season, Gasol had collected a total of 80 steals. His point guard teammate, Mike Conley, played in the same amount of games (80), but he finished the season with, a league best, 176 steals. He doubled Marc’s amount.

     Last but not least, blocks absolutely must be compared. Two of the front-runners for this year’s Defensive Player of the Year award led the league in blocks. Serge Ibaka (3 bpg) and Larry Sanders (2.8 bpg) were ranked first and second in the category. Then, you scroll through a list of really good defensive players’ names and finally come across Marc Gasol, who ranked 12^th in the league by blocking 1.7 shots per game.

     When you see a Marc Gasol highlight, it’s usually his polished, rainbow swish at the top of the key from 20 feet with the shot clock going off, that was modeled after Dirk. Then, you keep SportsCenter on for 20 more minutes to check out the top 10 plays of the day from the entire sports world. Somewhere in that top 10 is Larry Sanders running down the court to swat an easy layup by one of the game’s best scorers (which he has done again and again). You don’t remember the 9 points that Larry just dropped, you only remember the ball flying 6 rows into the stands because of a remarkable, defensive play.

     Serge Ibaka isn’t the Thunder’s most consistent scorer. But he is the league’s most efficient defender. I mean, for the love of good marketing, he even has the nickname to intimidate some of the league’s better post- scorers, Iblocka. How can you be rewarded with a defensive-minded nickname and have some of the best numbers in the league, but still not get the award? Gasol doesn’t even have a nickname yet. Or maybe it’s ‘Ole G.

     I mentioned some of Marc Gasol’s teammates in some of our comparisons earlier (Conley, Z-Bo). I don’t even think any of those guys are the best defenders on the Memphis Grizzlies squad. I strongly believe that Tony Allen is the game’s grittiest and in-your-face player (on both ends of the court). And Tayshaun Prince may be old, but his length allows him to still be a good defender against some of the game’s best. Tony Allen and Tayshaun are matched up with Kevin Durant, Lebron James, Paul George, Carmelo Anthony, Andre Iguodala, each night. They are the guys that have the toughest task for the Grizz.

     Oh, but there’s that big guy in the middle. As soon as Mike Conley gets beat and the lane is wide open, there’s only Marc Gasol. Marc now has to play 2 guys. Marc can step up to stop the ball and that forces CP3 to either attempt the floater or he has to throw up the alley-oop to Blake Griffin. And Gasol looks like a genius cause he defends Chris Paul and Blake Griffin to perfection. As soon as Tony Allen collapses to help Gasol, CP3 has nowhere to go with the ball, so he finds Matt Barnes in the corner to shoot up a brick.   

     Marc Gasol won the Defensive Player of the Year award for his presence, not his play.

Statistics is being a bitch again and now I need to do an assignment on it. This far, it looks easy. But I’ll have to do it to actually know.

Biology is just year 12 all over again, doing experiments with pepsin and shit. We need to write a report on it and submit it to them. Shouldn’t be too hard either.

Chemistry on the other hand… let’s just say I haven’t been doing the tutorial questions for about 3 weeks. Lot of catching up to do.

*sigh* 6 weeks later though, after all the exams are finished, I can finally take a deep breath and relax. But until then, the Stress Index is staying strong and will soon hit an all time high.

Everything else is working out fine. I’m just going to leave the question of relationships out the window. I don’t have time for it. Hell, I don’t even have time to play WoW, let alone anything else.

Until next week.

The info is about Australian society. Sorry to all our international readers. But who knows, it might still be interesting!

Last 13 winners recorded an RPR of 116+ and had won at 12f+

Only two in the field match this:

Cavalryman
Joshua Tree

Both match up pretty well – 7 wins and 6 wins, 30 and 25 runs, RPR 124 and 118, both won a Group 1, won over 16f / 15f.

Preference would be for Cavalryman, especially has he doesn’t have a penalty, and that is further confirmed when looking at their history. It seems Joshua Tree has one big target each year – the Canadian International at Woodbine in October. The last 3 years he has won twice and finished 2nd in the other.

Cavalryman’s record when running in fields of 10 or less, outside Group 1 company, and with less than 10 weeks break is 6 wins and 2 2nds from 8 runs.

Selection:

Cavalryman

This season has had its share of amazing footballing moments: from van Persie‘s free-kick against Manchester City to Joey Barton‘s insult to Zlatan. Here are some of this season’s best goals, moments, funnies (and embarrassments):

• Lewandowski’s super hattrick against Real Madrid
• Watford x Leicester’s amazing finish to the game
• Barcelona’s 0-7 trashing against Bayern Munich
• Ivanovic’s winner against Benfica in the Europa League final
• Sir Alex Ferguson’s last Guard of Honor
• Suarez goes cannibal and bites Ivanovic
• Wigan wins the F.A. Cup against Man. City
• Arsenal 7-5 Reading
• Big Sam’s Mona Lisa
• Frank Lampard breaks Chelsea’s all-time scoring record
• Three amazing goals in an amazing match. Spurs 2-1 Lyon
• Suarez goal x Newcastle!
• Borussia Dortmund x Malaga, Dortmund with two late goals and a win
• Bradford defeating Aston Villa in the league cup
• Hazard kicks a Swansea ballboy…
• Hulk destroys Liverpool’s net
• David Luiz with a world class goal
• … and an another goal from the Brazilian
• Celtic beat Barcelona on their 125th anniversary
• Stevie G gets it right against Manchester City
• David Luiz smiling after a tackle from Rafael
• Rafa Benitez booed by his own fans
• And of course: Jonathan Walter’s worst career game

Have any more links you want to share? Comment!

Sultan Aldori

The Mariners were supposed to be in for Dee Gordon this winter, and were willing to pay. The Dodgers, however, were anything but motivated sellers. Maybe now we know why the Ms were so hard up: Hardball Talk points out, that Mariners’ shortstops are providing less offensive production than NL pitchers. Not a few NL pitchers who can handle the bat, NL pitchers as a class. Seriously:

http://hardballtalk.nbcsports.com/2013/05/16/mariners-shortstop-are-hitting-worse-than-nl-pitchers/

The brutal truth, about a quarter of the way through the season:

NL pitchers, hitting: .124, .323 OPS

Mariners shortstops, hitting: .121, .318 OPS

The article further speculates that the Mariners might call up prospect Nick Franklin (.982 OPS at AAA). Franklin’s MLEs from 2012 graded out at 10-10-.240. The growth so far appears to be on the power side this year. Super-2 says not before mid June.

Ah what the hell. We’ll piss off their mothers, but let’s close out this post by naming and shaming the guilty. Brendan Ryan has started 30 games at short so far this year for Seattle. Robert Andino has started the other 10.

P.s…I think its measuring tape and body fat callipers and the dreaded photographic evidence next week….trying to do them once a month….wish I was a whizz on photoshop

Body Fat % = 30 (-1.3%) OMG!!!!!

Weight (lbs) = 152 (+2lb) Nearly fainted when I saw that on the scales, but stayed calm and realised its all muscle and water

Water % = 49.3 (+1%)

Muscle Mass (lbs) = 101 (+4lb) OMFG!!!!  Yep!! Gained 4lbs of muscle – YAY!!!!!!!!! DDDDD

Metabolic Age = 36 (-2 years) Gawd I feel like a bloody teenager

Basal Metabolic Rate = 1441 (=41 kcals) I can eat more

Bone Mass = 5.4 (+0.2) Excellent result, really pleased

Visceral Fat = 4 (no change)

#beinglazy

            While ranking players in the NFL according to statistics is difficult since games aren’t played on paper, I have endeavored to do so. I decided to rank a few of the NFL’s top active players who I find to be versatile. Versatility to me was being able to perform a variety of roles at a certain position. For quarterbacks this is basically limited to passing and rushing performance. However, for wide receivers and running backs there are a few more areas to excel in. These positions can run, receive, and return kickoffs and punts, which is exactly how I tried to quantify them.

First off I ranked a few quarterbacks. My list is of quarterbacks who I found capable of throwing and running the ball efficiently such as Cam Newton, Russell Wilson, Aaron Rodgers, RGIII, and Colin Kaepernick. I compiled a spreadsheet and for quarterbacks I used passing yards and touchdowns and rushing yards and touchdowns. To determine the versatility rating for each player I created a simple equation involving these statistics. The primary statistics, passing yards and touchdowns, were left alone since any decent quarterback should be able to do these well. (Note: I did slightly weight passing touchdowns since they are less frequent than yards). Now the secondary statistics or versatility statistics as I like to call them were rushing yards and touchdowns. These statistics I decided to weight since they are more important to defining the versatility rating – not all quarterbacks can run efficiently. So the equation turned out like this -

Versatility Rating= (pass yards + 10(pass touchdowns) + 1.5(rushing yards) + 20(rushing touchdowns))

(games played)

The division by ‘games played’ is to make sure that people with massive pass yardage because they have played more seasons.

I expected Cam Newton (despite somewhat of a sophomore slump), Aaron Rodgers, or RGIII to have the highest versatility rating. However when it came down to crunching the numbers, Colin Kaepernick had the highest versatility rating at 381.64 (see spreadsheet below for comparison)

To calculate versatility for running backs and wide receivers I had to incorporate the kick and punt return statistics, since many of these players perform these roles. The formula for versatility was much the same as was used for quarterbacks only with punt and kick return yards added in as weighted statistics (see spreadsheet below for exact formulas).

I predicted that Percy Harvin would have the highest versatility of any wide receivers analyzed and that Darren Sproles would have the highest versatility of runningbacks.

It turned out that Randall Cobb of the packers had the highest versatility rating of wide receivers (180.75) and Darren Sproles of the Saints had the highest of the running backs and of any player studied. (196.60).

Using these formulas it is easy to calculate the versatility of any player in the NFL. Wide receiver and running back positions are actually quite comparable because they use the same stats, but comparing them to the quarterback potions would be more difficult. These ratings I’ve created are only meant to be compared against other ratings, there is no perfect rating or rating that is the cutoff for good or bad (just better or worse). Also it’s necessary to keep in mind that these rating are based on stats and stats alone! Stats can mean nothing when a player is on the field in a particular game, matched up with a certain team. This study is just a good comparison, and could be useful in trade evaluations, and even fantasy leagues. I hope all of you found this interesting and unique, a fresh taste at how we look at players in the NFL! If you have any questions or want me to add a player to this list please comment and/ or email me at a.snider85@gmail.com.

NFL versatility spreadsheet

Every day, the site provides analysis about what you wrote. For this day (Saturday, May 11, precisely), I think the main character’s feelings and concerns are pretty accurate given that she’s trying to solve a murder that happened in a bank. At first, I thought it was strange that “affectionate” was the second strongest emotion, but my protagonist likes her two co-investigators in this scene, so I think that might be coming through.

Commanders Faced Over the First 20 Games

After positive recaps and one more clouded one, we now start dipping into the purely negative observations. Let’s be honest: it wasn’t all roses for the Blue Jackets in 2013. In fact, while it’s mentally easy to separate the late-season run and the early-season struggles the two are a part of a single whole. And perhaps even more frightening? The post-trade-deadline team Fenwick Close was still sub-par, a meager .4696 only 19th in the league. Yes, things were better than the ghastly outset, but the improved record was primarily due to the guy in net.

So who or what is to blame for all the shots against? The answer lies in both ineffective forwards and porous blueline. I’ll get to the former in a future post. Today I’m going to look at the two most detrimental defensemen on the Blue Jackets team: the ever-popular and constantly awful Jack Johnson, and the praised-but-not-quite-deserving rookie Dalton Prout.

I’m sure neither of these picks will be well received, but humor me while I try to justify my choices.

Let’s begin with Jack Johnson. The thought among many hockey fans is that Johnson is somehow good at hockey. This is something patently false (or at least not defensible by any logical means) but the idea remains. I’ve written before just how poor Johnson is and I’ve linked to more than a few articles that tear apart Johnson’s contribution to any team (both in terms of shot differential and goal production). However, the warm fuzzies continue toward #7 so I’ll continue ranting away until either he shows improvement or there is more unanimity in the idea that Jack Johnson sucks.

Of course, this leads me to ask: why do people believe Jack is so good or so valuable? I assume it must be his top-10 draft pick status, his role as a much-vaunted Puck Moving Defenseman, his presumed status as an offensive threat, his ability to throw big checks, or this year’s ability to become the marathon man and eat 30+ minutes of ice time in some games. Unfortunately, these positive traits are either inaccurate or don’t necessarily represent a beneficial impact to a hockey team.

Starting with the baseline: Jack Johnson is not good at preventing goal scoring opportunities. Johnson’s net shot differential per 60 minutes of ice time was -9.45, second worst among defensmen on the team with 20+ games played. Relative to the team, it was a -8.1, worst on the team. So Johnson played huge minutes and was still crushed by opposition, allowing more shots than almost any other d-man on the Blue Jacket team.

But maybe you don’t like shot differential. Maybe you think goal impacts are more important. Here, Johnson also fails. Instead of being an offensive powerhouse, Jack’s on ice goals for per 20 minutes was 3rd worst among Columbus d-men. From that same data set, Johnson was 3rd worst at goals against per 20 minutes. He wasn’t helping the offense and he was hurting the defense. That’s not good.

And what are we to make of Johnson’s logging huge minutes? It’s simply a terrible decision. While he is physically capable of skating at top speeds for 30+ minutes, his poor positioning and defensive inability more than make up for that and make Jack’s presence on the ice a dangerous prospect for the club. In fact, it’s troubling that Todd Richards viewed Johnson as the go-to choice when other d-men were injured.

We can take a more detailed look at Johnson’s impact by checking out how players performed with #7 and without him (a WOWY table). Using David Johnson’s awesome Hockey Analysis website, we can build and look at WOWYs between Jack and other CBJ d-men as seen below:

Jack Johnson WOWY for CBJ defensemen at 5-on-5, data from stats.hockeyanalysis.com

The differences column on the right show just how negative an impact Johnson was when paired with other players. Of course, you could argue Moore, Goloubef, and Nikitin didn’t have enough ice time with Jack to make significant conclusions, but they’re included for data completion.

In summary: Jack Johnson is not good at preventing shots, he’s not good at generating offense, and he’s not good at preventing goals. I’ll get to some league-wide context for this in a bit. But next up…

Dalton Prout. He has been praised as a stay-at-home d-man revelation for the club and supporters have pointed to his (seemingly) impressive +/- that leads the team at +15. And there’s hope for him as he’s only 23. Unfortunately, +/- is misleading, particularly for Mr. Prout.

His +/- was earned playing in front of a 5-on-5 save % of .974. If your eyes haven’t popped out, please reread that again. .974!!! Unless you believe that Dalton Prout is the world’s first ever significant shot quality reducer (and you shouldn’t), that’s all on Bob. Furthermore, that’s absurdly ridiculous luck for a single player to have. So if the +/- and his goals against figures are misleading due to the brick wall behind him, what can we make of Prout’s shot differential information?

Unfortunately, he was nearly as bad as Jack Johnson. His on ice shot differential was -11.39 per 60 minutes (worst on the team), and relative to the team was -7.7 per 60 minutes (2nd worst on the team). If it wasn’t for Sergei Bobrovsky stopping almost literally everything while Prout was on the ice, that +/- would have been horrifying.

Thankfully Prout does have two redeeming qualities that Jack Johnson lacks. The first is his age: Prout is only 23 and was only in his first season-ish of NHL hockey. There is plenty of room to improve due to age and experience. The second quality is his offensive generation. At 5-on-5, the CBJ goals for with Prout on ice was actually higher than with Johnson on. If Prout can grow into his NHL role and continue to improve the scoring, he could well be a strong blueliner in the future.

So neither Johnson nor Prout was a very positive influence on the Blue Jackets this year. To give their negativity even greater perspective, we give one final look at their shot differential performance in the context of the whole NHL. Johnson and Prout’s shot differential were 27th and 19th worst among all NHL defensemen with 20+ games played (that’s out of 210 eligible players). Johnson’s and Prout’s performances considered relative to their team (Corsi Rel) were 32nd and 36th worst in the NHL over those 210 d-men.

That’s pretty bad and it makes picking them as the CBJ 2013 worst defensmen an easy task.

Photo via ESPN.com, by Victor Decolongon/Getty Images

Welcome to our weekly delve into local real estate market segments.  Today we are going to take a look at Condo Prices in the Glebe and Old Ottawa South area of town.

Right now (May 14th 2013) there are 37 condos available for sale and 1 under conditional contract.  Ranging from $259,000 for a bachelor apartment right up to $2,995,000 for a two story, 3 bed/4bath condo loft with glass wall overlooking the canal.  As you can see, there are a lot of variety in the type of condos in this community.

So far this year 9 condos have sold (about 2/month) for an average price of $477,611 in 43 days.  The sales have ranged from $271,000 up to $717,500, and condos have sold the day they came out or take up to 176 days to sell.

Last year in the same date range, 18 Condos had sold (about 4/month) for an average price of $473,400 in 69 days.  So while, there has been a decline by half in the number of sales this year, the ones that did sell, sold quicker and for more money.

For the whole of 2012, 43 condos sold (again about 4/month) for an average price of $510,803 in 52 days, ranging from $1.5 Million down to…get this $213,000 for a condo town house, on Riverdale, by the Rideau River.

I speculate, given the very consistent nature of last years sales, and a quick look at the last 24 months of sales (still about 4/month) that a large segment of condo buyers in the Glebe are buying new constructions.  Which could mean some great deals coming up down the line, be sure to keep an eye out for them.

This Spring’s box office blockbusters, brought to you by L. Ron Hubbard!

This spring’s release of Will Smith and son’s After Earth & Tom Cruise’s Oblivion poses an important question: how is M. Night Shyamalan still employed when will audiences reach their saturation point towards post-apocalyptic movies? Between these two films, the latest zombie craze (Walking Dead, Revolution, Warm Bodies, World War Z) and the Hunger Games franchise, soon Hollywood has to run out of reasons why the entire human race, minus Earth’s most beautiful people, has been wiped out. Possibly scarier than the loss of civilization is the thought of Hollywood without the stars of the two films above. Will Smith (44) and Tom Cruise (50) are arguably the two biggest movie stars alive today, but despite what their physical trainers would tell you, both are well past their action-star primes. They draw obvious comparisons: both are Scientologists with well-documented personal lives, both have helmed huge box office franchises, and both had notable roles where they pranced around shirtless.

Who’s the bigger star? The numbers paint an interesting picture. I tried to develop a crude way to measure an actor’s “Star Power” a few posts ago, when I took on the cast of The Avengers. The formula boiled down to three parts:

Quality of movies x Box office success x Social Media following = Star Power

Taking a look at the first portion, here are each of their films’ Rotten Tomatoes Scores, in order of age the actors were at the time:

(click to expand)

(click to expand)

Tom Cruise boasts a 64% RT average, while Will Smith’s only averages 55%, in sixteen less films. Somehow, after thirty (!) years, Risky Business is still Tom Cruise’s highest rated film at 98%, beating his recent surprise critical hit, Ghost Protocol. Meanwhile, Smith thrives in that 60-70% “entertaining, but flawed” category, with seven of his biggest films (Independence Day, I,Robot, Pursuit of Happyness, Ali, Hitch, MIB3, I Am Legend) falling in that range and only the first Men in Black and his first two tiny films rating over 80%. By comparison, Cruise features twelve films over 80%. Even ignoring the imperfect numbers from Rotten Tomatoes, Cruise has 10 Oscar/Golden Globe nominations with 3 wins under his belt, while Smith only has 4 nominations (six if you count Fresh Prince). Cruise has the clear advantage in the quality of his resume.

30 years and 300 parodies later, Risky Business is still Tom Cruise’s most critically acclaimed film. (picture via JustJared)

Edge: Cruise

Here are their Opening Weekend Box Office dollars, adjusted for inflation (data from BoxOfficeMojo), along with a trend line for each actor:

(click to expand)

Agent J strikes back, beating Ethan Hunt with a $45.8M to $28.2M average opening weekend gross, adjusted for inflation. MIB, Independence Day & I Am Legend all had adjusted opening weekends of over $85M, a number Tom Cruise only bested once with MI:2. What’s interesting is how audiences turned out when these two had similar movies released around the same time. In June 1996, hot off Jerry Maguire momentum, Mission Impossible grossed a fantastic $81.6M- only to be defeated two months later by nearly $10M by Independence Day, Smith’s first real blockbuster. Both movies even had the same exact Rotten Tomatoes score of 61%. Perhaps the starkest contrast came in June 2002, when Steven Spielberg’s sci-fi thriller Minority Report rode its rave reviews (92%) to a nice opening weekend raking in $48M. A month later, the highly anticipating and majorly disappointing (39%) sequel to Men in Black, buoyed by Smith’s stardom, enjoyed a whopping $72M opening weekend. After Earth vs. Oblivion will be the best case study yet – same genre, same overdominating CGI, and same poster font. Based on sheer numbers so far, Will Smith is a box office giant.

Edge: Will Smith

To measure Social Media presence, albeit crudely, I took the number of Facebook likes for each movie’s individual Facebook page (some of their earlier films didn’t have one). Franchises (Mission Impossible, MIB & Bad Boys) were only counted once for these charts, since those either share a Facebook page or have inflated numbers due to multiple pages.

(click to expand)

(click to expand)

Cruise averages 12M Facebook likes over 28 films, but Smith has a comfortable advantage, averaging 17M over 16 films. Shockingly, Smith outpaces Cruise on their personal Facebook pages 45M to just 3.8M for Cruise. Smith does have the Fresh-Prince and his brief music career to boost his numbers, but why do  almost 12x more people “like” Smith? My guess is the  Les Grossman-sized gap is due to the Smith family’s carefully cultivated public image. Barely any buzz has come out of the rumored conflict between he and wife Jada Pinkett Smith, and the whole Scientology thing is never held against them the way it’s held against Cruise. On top of that, Will Smith has meticulously chosen roles that support his good-guy image- most recently, he turned down Django Unchained because, on top of it tackling risque subject matter, the role of Django wasn’t the lead. In 1999 he turned down a smaller sci-fi flick written by some unknowns in favor of  another $170M-budget Barry Sonnenfeld blockbuster that featured the uh, hilarious, scene below. That role he passed up? Neo from The Matrix.

On the other hand, the negatives in Cruise’s personal life have been well documented- from his outburst on Oprah to his public divorce with Katie Holmes and the ensuing hijinx. All this has hurt the public’s perception of him and definitely have an effect on his “likes”. In addition, career-wise, Cruise, unlike Smith, hasn’t been afraid to take chances. Look at the chart above: his highest rated movies are straight dramas, some of which are extremely controversial- and thus have the least “likes”. He’s played the manipulative brother of an autistic savant (Rain Man), an orgy participant (Eyes Wide Shut), and a paralyzed war vet that invented a familiar high school game (Born on the Fourth of July). Meanwhile, his lower-rating, more accessible films - Mission Impossible, Top Gun -  have the most likes. Nearly all of Will Smith’s movies are mass-audience friendly, which explains his huge advantage in both Box Office dollars and Facebook likes.

Edge: Will Smith

The formula below was applied to each movie individually, and then averaged and totaled below.

Drumroll please…

In sixteen less movies, Will Smith still blows Tom Cruise out of the water in Star Power total, and handily eclipses him in average.

In 2012 Forbes released their annual list of highest-paid actors- Cruise ranked #1 earning $75M, while Smith ranked #8 raking in “only” $30M. Similar to my previous Avengers article, if you divide their salary by each actors’ Star Power, you get a multiplier. The lower this number, the better – meaning  the studio is extracting more of this “Star Power” out of their investment. That number is 543 for Tom Cruise, and only 53 for Will Smith. Not since the days of Michael Jordan has someone with a salary of $30M been called underpaid, but Will Smith over his lengthy yet selective career has been just that.

Wondering how I  found the self-control to write an entire article about Tom Cruise without making a short joke? Have some matchups you want to see? Tweet me @kramerkram, leave a comment, or just vote below.