Valencia in Transition

By:

Using STATS Playing Styles and Advanced Metrics to Show How Marcelino’s Valencia are Employing an Effective Counter Attack to Return to Spanish Relevance

Marcelino couldn’t contain his excitement and pulled up short with a hamstring problem after Simone Zaza scored a late winner against Real Sociedad in late September. The injury was of little concern to his club for at least a triad of reasons.

  1. Marcelino, 52, is Valencia’s manager, and no part of a professional football match has hinged on the performance of his legs since 1994.
  2. It was overshadowed by the fact that it gave Valencia yet another result to sway detractors amassed over the past two seasons back in their favour.
  3. This was hardly an exceptional occurrence. Marcelino, as a manager, once injured himself taking a seat for a presser.

It’s unsurprising to see Marcelino competing for the celebratory spotlight with his players in his first season with the club, though he’s said he recognises the need to tone down his touchline merriment. Given the past two campaigns, it’s of greater note that Valencia have had so many such opportunities through their first eight matches with the excitable man in their coach’s box.

Using STATS advanced metrics, we can show his players might need to reciprocate and go crazy for their new manager once in a while, even if he’s being particularly unpredictable when determining whether he’ll include each of them in his starting XI. Eight different lineups in as many fixtures might seem as erratic as the manager’s touchline fervor, but there appears to be something of a method behind it. We’ll get back to that in great detail on team and individual levels.

First, a bit on why Valencia’s strong start matters.

It wasn’t as long ago as it may seem that clubs outside of Madrid and Barcelona won the Spanish top flight. The last were Valencia in 2003-04 and ’01-02, and before that Deportivo La Coruña in 1999-2000. Modest success hasn’t eluded Valencia since. European football had been an expected part of the gig at Mestalla until recently. But there’s no arguing the past two campaigns in which supporters have endured successive 12th-place finishes – their first in the bottom half since 1987-88.

Los Che now find themselves second in the table as one of three unbeatens in what’s arguably Europe’s top league. The other two – Barcelona and Atlético Madrid – have reached at least the quarterfinals of the Champions League for the past four years.

So how do Valencia find themselves back up in the early-season mix for direct Champions League qualification? It’s not for a lack of competition. Quite the opposite, in fact, and it could be argued Valencia’s early-season fixtures have been as demanding as any Spanish club. Four of their eight league matches have come against clubs playing European football. That injury-inducing match came away to Sociedad a week before the same 3-2 scoreline played out in less-exciting fashion at home against Athletic Bilbao, Sociedad’s Europa League contemporary. But those victories bury the lede of quality draws in Marcelino’s second and third matches with the club.

Valencia left the Bernabau with a 2-2 result against the reigning European and Spanish champions after holding a lead into the 83rd minute, then followed the first international break with a scoreless home draw with Atlético.

Most recently was Los Che’s chaotic 6-3 win Sunday at Real Betis, who remain in the top half of the table.

So what’s changed from a season ago? A bit of everything. The manager, of course. Players. Players’ efficiency. Method – and that’s where we’ll begin by calling upon STATS Playing Styles before going more granular with advanced individual metrics.

Last season, Valencia went through four different managerial periods and three different bosses – Pako Ayestarán until Sept. 20, club ambassador and longtime centre back Voro González for the next eight days, Cesare Prandelli from Sept. 28-Dec. 30, and the ever-present stopgap Voro again until the end of the season. In terms of style, that unsurprisingly amounted to very little differentiating from La Liga averages:

Valencia’s 2016-17 playing styles measured against La Liga averages (0%).

They played more of a fast-tempo game than much of the league, but they didn’t sustain threat when doing so and were rather blasé in all other areas. What followed was a minus-9 goal difference for their worst mark since ’07-08 (-14) when Los Che finished 10th.

Through eight matches this term, there’s still not some overhaul of telling possession-based attacking styles that typically signify a dominant club – they rank 16th in possession at 45.3 percent, which is lower than last season (48.3 percent). It’s unsurprisingly a drastic departure from other top teams in the table. Barcelona are first (61.2 percent) and Real Madrid are second (60.6). But there is order to how Valencia score goals. It’s frequently about transition:

Valencia’s 2017-18 playing styles through eight La Liga matches measured against league averages (0%).

That plus-55 percent counter attacking style against the league average leads La Liga – yes, ahead of even counter masters Real Madrid (+37 in second). Among the top-five European leagues, only Benevento in Italy are countering at a higher percentage against their league average. Anyone familiar with the Italian table then jumps to a logical question: Why are Valencia succeeding and Benevento the clear-cut worst side in Serie A with eight losses and a minus-17 goal difference?

The answer is probably that counters ending with a striker tripping over the ball don’t mean much. The Serie A newcomers have had 59 possessions with a counter attacking value of at least 50 percent, and it’s amounted to two goals. And they spend far more time defending, as is evident by their overall style, so they’re not exactly creating chances in other tactically sound ways:

Benevento’s 2017-18 playing styles through eight Serie A matches measured against league averages (0%).

Valencia, meanwhile, are effective in their counter – more effective even than Real Madrid. Among Los Blancos 48 possessions on which their counter attacking style is of a value of at least 50 percent, they’ve scored once. Valencia have 53 such possessions and four goals after scoring eight goals off the counter all of last season.

All of this must originate somewhere, and that’s where counter attack regains come in. Last season, Valencia were 13th in La Liga in regains to begin a counter attack (154). That trailed leaders Real Madrid by 70, so nearly two per match. Their counter attack distance covered (8,705 metres) – made up of the total of counter attack distance carried and counter attack distance passed – ranked 14th. This season, they’re first in regains (53) and distance (3,171 metres) – more than a third of the way to reaching last season’s marks.

There are more reasons transition works for one club and doesn’t for another. To measure the efficacy of the correlating defensive and midfield play, we’ve got to get beyond simple sums. We showed last week how Kevin De Bruyne has been one of Europe’s most dominant offensive players, despite having a comparatively limited direct involvement with goals and assists. We did this with STATS’ ball movement points. BMP is a metric that considers every involvement a player has in a possession to credit or discredit decisions with the ball and reward creativity. It’s what football minds could always see but never quantify. It goes beyond expected assists by looking at the full chain of passes and weighs the probability of that pass leading to a shot later in the play. Passing points generate expected shot points, so if a player generates one BMP, he’s generated passes to lead to – or defend – one shot.

Yeah, that’s ambitious. So how is this done? The process is measured and assigned objective value using machine learning and massive amounts of historical league data to express the level of threat or wastefulness that can be attributed to a player. It’s broken down into categories of offensive and defensive as well as positive and negative with net values telling the more conclusive story.

There’s dBMP+, which measures how many created chances a defender prevents – breaking up attacks in important situations. There’s dBMP-, which measures liabilities in possession – giving the ball away in dangerous areas. Combine that for net dBMP. While Benevento sit in the bottom half of Italy with a 0.13 dBMP rating, Valencia (0.27) lead Spain. So we’d previously established with playing styles that Benevento are spending a lot of time defending, and dBMP helps us show that they’re not making great decisions with the ball when doing so. Valencia might not be the most attack-minded club in Spain, but they’re at least effective in their own half. That might not matter quite as much for ball-dominant clubs such as Barcelona. It absolutely does for sides that have to pick their attacking moments judiciously.

So on the pitch, who specifically is to reward for executing the system Marcelino seems to be implementing?

We’ll start with the sexy goal-scoring numbers from a striker mired in that special brand of Italian sorrow this time last year for his happenings with club and country.

Zaza scored six goals in 20 matches in his time with Valencia last season and has passed that already this season with seven and three winners. With six goals in his last four matches, he seems a healthy distance from his Euro 2016 penalty miss for Italy and his disappointing spell with West Ham United. The numbers back that up with the 26-year-old ranking in the top five in Europe’s top-five leagues in finishing with an expected goal differential of plus-3.5 among a pretty elite group a season after posting a minus-1.9 xGD. Notice than in Spain, he jumped this past weekend ahead of even a guy named Messi:

As we noted before with ball movement points, midfield play has a lot to do with Valencia’s success, and that’s true on an individual level as well. Bringing in on-loan Geoffrey Kondogbia from Inter Milan as a central presence might have displaced 20-year-old Carlos Soler some after the latter became a mainstay in the middle last season, but it seems to be working out for Marcelino. Kondogbia, who’s attracting attention from top Premier League clubs, ranks second among all midfielders in Europe’s top-five leagues in dBMP, and he’s one of three to really distinguish himself from the pack:

Valencia don’t quite make the same use of the corresponding playmaker guiding a dangerous attack at the other end. Their top-ranked player in oBMP among the top-five leagues is Dani Parejo tied for 29th, but when filtering that down to only La Liga, it’s good enough for fifth among a star-studded top 10. It’s rather impressive when considering the opportunities and surrounding creatives much of the rest of this list has to work off of:

(Graphics by Stephan van Niekerk)

Bored yet? OK, let’s talk goals again. We can’t forget about Rodrigo, who scored five goals in 19 La Liga matches last season and was an objectively mediocre finisher with a -0.6 xGD. He was with Spain as they wrapped up qualifying last week for reasons that go beyond David Villa nearing 36 years old. Rodrigo has scored in five straight matches for the club and also got one in his start against Albania on Oct. 6. Although none for Valencia have been match-winners, he has compiled early-season efficiency (+1.6 xGD) to show he’s not exactly feasting on scraps.

Finally, goalkeeping. Neto, who spent the past few seasons behind Gianluigi Buffon at Juventus, has a +2.1 expected save differential, which is calculated by subtracting expected saves from saves to show how a keeper is performing against league averages. That mark ranks sixth in the division and, you guessed it, is better than his former mentor Gigi (+0.7). It’s not quite the level of Pau López (+6.2), Jan Oblak (+6.0) and Guaita (+5.3), but the Valencia keeper is still going above and beyond from time to time. It’s also important here to consider that Valencia aren’t leaning on him to continually bail them out in an unsustainable way.

So Valencia have a manager pushing for some consistency in style, and he has players making it happen at various levels that we’re now equipped to properly measure. That’s what it takes to earn 18 points in Spain through eight matches, three of which have been draws and another three being one-goal victories. But this is La Liga, home to the two most dominant clubs in the world in recent years. Recall Real Madrid’s Spanish-record unbeaten run of 40 in all competitions ending in January. Barcelona notched 39 in 2015-16. Is it right for a club that’s won domestic titles of its own to make much of this just yet?

Given the circumstances of the past two seasons, it somehow feels right for Marcelino to stick with those celebrations.

The Daring Process of Kevin De Bruyne

By:

How STATS Ball Movement Points Show Manchester City’s Midfielder Has Been the Most Dangerous Creator in Europe This Season

Kevin De Bruyne’s 67th-minute strike in a 1-0 win at Stamford Bridge wasn’t just an aesthetically impressive example of fast-tempo football – it was arguably the most important goal to be scored in England so far this season. It distanced Manchester City from title holders Chelsea entering the international break and gave the 26-year-old his first Premier League goal of the campaign against the club that sent him to Wolfsburg.

That finish, as pretty as it was, isn’t the main reason we should praise De Bruyne’s efforts for the Premier League leaders.

Even in that six-match goalless stretch to start the season, you weren’t going to hear many supporters complaining about De Bruyne’s form the way one might if Sergio Aguero went through such a drought. Nor will you in the week following his inclusion in the Ballon d’Or 30-man list, from which fellow City creative David Silva is somehow once again conspicuously absent.

De Bruyne has been directly involved in four of Manchester City’s 22 Premier League goals. His goal – which featured the Messi-esque combination of build up and finishing inclusions from a pretty one-touch layoff to Gabriel Jesus before getting it right back on the run for a two-touch 20-yard left-footer to the upper corner – and three assists works out to an 18 percent involvement from the goal-and-assist perspective. It’s not a particularly high rate, and on a personal level, it’d be his lowest in three seasons with City.

But that’s precisely why KDB’s start to the 2017-18 season can act as one of the most relevant contemporary examples of why there need to be better metrics in football. Now settled into a deeper central position after the tinkering Pep Guardiola went through with his creatives in his first season in charge, De Bruyne is still generating all sorts of threat. He’s always been the type of creative midfielder who gains praise for his passing ability and his field awareness. He makes work difficult for defensive players with his precision, and his ball movement is rarely lacking ambition. It’s now measurable with ball movement points, which reward the process in a way traditional binary metrics such as successful passes fall short.

And, through seven matches, BMP shows the Belgian has evolved into the most dangerous playmaker in Europe with his globally overlooked club teammate not far behind.

First, a quick rundown on BMP, which we’ve used a few times before when discussing transfers and how key players are contributing to the top-five European leagues. BMP is a metric that considers every involvement a player has in a possession to credit or discredit decisions with the ball and reward creativity. It’s what football minds could always see but never quantify. It goes beyond expected assists by looking at the full chain of passes and weighs the probability of that pass leading to a shot later in the play. Passing points generate expected shot points, so if a player generates one BMP, he’s generated passes to lead to – or defend – one shot.

Yeah, that’s ambitious. So how is this done? The process is measured and assigned objective value using machine learning and massive amounts of historical league data to express the level of threat or wastefulness that can be attributed to a player. It’s broken down into categories of offensive and defensive as well as positive and negative (oBMP+, oBMP-, dBMP+ dBMP-) with net values telling the more conclusive story.

Got it? Good. Back to KDB.

First, the basics. In 2015-16, his seven goals and nine assists among Man City’s 71 goals amounted to 22.5 percent involvement. Last season, his six goals and 18 assists out of 80 Man City league tallies amounted to a 30 percent involvement. At that level, this season’s inclusion looks like a regression.

It’s not. While Guardiola experimented last season, De Bruyne’s 7.62 oBMP still managed to place second in England behind Mesut Özil (9.00). The season before, Manuel Pellegrini’s last, his 4.14 oBMP ranked 16th in England, and that came behind teammates Yaya Toure (4.82 in eighth), Silva (4.60, 10th) and Fernandinho (4.32, 13th). Özil (10.95) led then as well. For the sake of comparison, La Liga’s leaders last season were Lionel Messi (7.52) and Toni Kroos (5.97).

Onto the current term. It’s of course very early in the season and this is a tight pack, but now consider this season’s oBMP rankings across Europe’s top-five divisions in league play. Of the top 20, 11 are midfielders, and plenty don’t have the goals or assists to display their true value:

(Graphics by Stephan van Niekerk)

De Bruyne is leading, but he’s also the only player to separate himself from the next best player by any considerable sum. Project that 2.29 mark out over a 38-game season, and he’s looking at a 12.44 oBMP that cruises past his own marks from the past two seasons and even Özil’s. And right behind him is his teammate, who again hasn’t been given any love by the Ballon d’Or brass despite consistently creating on elite levels among the world class.

It follows that these players must have a pretty impressive oBMP+, meaning they are ambitious and effective with ball circulation in the attack – they find the channels and play a mean through ball, or they consistently deliver that low, bending cross that makes centre backs trip themselves. It also follows that they may have a considerable oBMP- because of the number of chances they have to craft opportunity, but they’ve got to limit that to exist as a leading creative player. For example, Alexis Sanchez ranked fourth in oBMP+ last season (10.66) but first in oBMP- (-5.01), so his wastefulness drops his net oBMP (5.65), which was not only a considerable margin behind teammate Özil but also Arsenal midfielder Granit Xhaka (5.92).

Now keep in mind BMP does not take into account finishing, which STATS quantifies by calculating expected goal differential (subtracting expected goals from actual goals). That’s where players such as Radamel Falcao (plus-5.8), Paulo Dybala (+4.7), Ciro Immobile (+3.5), again Messi (+3.4) and Mathew Leckie (+3.1) have distinguished themselves this season.

Maybe you can see where this is going. We’re going to write more in the coming weeks and take this a step further by quantifying the value of overall offensive contribution. But when you focus solely on the process of getting the ball to those finishing players in positions to succeed, right now there’s been no better creative than De Bruyne.

Is Jet Motion Sweeping the NFL?

By:

Jon Gruden consistently makes bold claims during Monday Night Football broadcasts while using his fiery personality to drive his points home. Not surprisingly, his emphatic delivery often prompts viewers to poke around for holes in those boasts.

So when Gruden claimed the jet motion offense has been on the rise across the NFL during last week’s Monday night matchup between the Chicago Bears and Minnesota Vikings, it became only natural to do some digging. After all, a play like the jet sweep still feels a bit gimmicky and unusual.

As it turns out, Gruden was right all along.

According to STATS X-Info data, use of the jet motion offense is up 36 percent league-wide from last season. Using jet motion simply means an in-motion player is heading toward the quarterback at the snap. The ball does not have to be handed off to that player – usually a wide receiver – meaning fakes and screens or even regular handoffs to the running back still end up being results from jet motion formations.

The jet sweep is the play most associated with that type of motion. Here’s an example from the New England-New Orleans matchup in Week 2. Brandin Cooks runs for a big gain to help set up Tom Brady’s 19-yard touchdown pass to Rex Burkhead.

The jet motion the Patriots used on the Cooks run no longer is that uncommon. NFL teams used jet motion an average of 1.71 percent of the time in 2016, but that average is up to 2.33 percent this year. That might not seem like a big jump on the surface, but take a look at the chart below. Only 11 teams are using jet motion less on average than last year, and even then the percentage points are not that far off.

Jet Motion Percentage - NFL Teams

According to STATS X-Info data
Team20162017
Arizona0.801.89
Atlanta1.512.61
Baltimore2.593.94
Buffalo1.850.33
Carolina3.246.01
Chicago0.310.78
Cincinnati0.870.99
Cleveland0.000.30
Dallas3.962.39
Denver0.202.04
Detroit2.492.17
Green Bay0.002.03
Houston0.500.89
Indianapolis1.081.29
Jacksonville2.010.00
Kansas City7.115.21
Los Angeles Chargers0.360.66
Los Angeles Rams4.938.32
Miami1.120.86
Minnesota3.642.52
New England2.285.34
New Orleans4.185.65
New York Giants0.441.99
New York Jets0.461.78
Oakland0.270.00
Philadelphia2.071.47
Pittsburgh0.630.61
Seattle1.812.09
San Francisco2.521.80
Tampa Bay0.254.67
Tennessee0.504.26
Washington1.180.41

Some numbers jump out immediately. Green Bay is using jet motion a decent amount after not using it at all in 2016. Denver’s percentage is up considerably from last season, and New England is using jet motion almost double the amount of plays on average.

Last season, Washington used jet motion on 1.18 percent of its plays under then-offensive coordinator Sean McVay. That’s what makes the Los Angeles Rams’ league-leading 8.32 percent jet motion use so stunning in the 31-year-old McVay’s first year as the Rams’ head coach this season.

The presence of the versatile Tavon Austin likely has plenty to do with that significant jump under McVay’s watch. Originally, McVay admitted having trouble implementing Austin into L.A.’s scheme other than to return punts – something he no longer will be doing after losing his third fumble of the season on a muffed punt in last week’s 16-10 loss to Seattle. He muffed another, but the Rams recovered.

But Austin has added a dangerous dimension to the Rams’ offense, which ranks fifth in the NFL averaging 382 yards per game after finishing dead last in total offense last season. STATS X-Info data reveals Austin has lined up in four different positions on his 60 snaps – 23 at slot receiver, 19 at running back, 16 at outside receiver and two at tight end.

“I might not be producing that much with the ball in my hands, but my fakes, my jet sweeps, it’s doing numbers, and that’s the main thing about it,” Austin said earlier this season.

Todd Gurley explained that Austin’s use in the jet motion as a decoy has helped him drastically improve in the running game. Gurley is averaging 4.1 yards per carry compared to 3.2 last year, and his seven touchdowns (four rushing, three receiving) have already surpassed his total scores from 2016.

Quarterback Jared Goff also is succeeding after a very difficult seven games as a rookie last year, completing 61 percent of his passes with seven touchdowns and three interceptions – two of which came just last week against Seattle’s tough defense.

The scheme implementing jet motion – specifically Austin’s role in it – by McVay and offensive coordinator Matt LaFleur has provided more space and opportunities for each of the Rams’ main offensive players. And even the threat of a jet motion play creates other chances.

Here’s Austin’s 27-yard touchdown run from last week, when Austin lined up at running back and remained stationary until the snap.

“A lot of times you talk about keeping your gap integrity as a defense, and it’s predicated on where guys are aligned,” McVay said earlier. “When you’re flying a guy fast across the field, it causes some conflicts in your run fits and guys get out of gaps, or they’re looking at it or you might regulate some different things that you’re doing. You don’t know if it’s coming, and there’s some complements off that.”

Scoring is down in 2017 – teams have combined for 368 offensive touchdowns compared to 389 through five weeks last season – as defenses adjust to tired schemes coming from recycled offensive coordinators. And to McVay’s point, running an offense that keeps multiple defensive players thinking is paramount to moving the ball with regularity.

Installing more jet motion appears to be a major reason for the Rams’ 3-2 start after they went a dismal 4-12 last season, when their solid defense performed even better than it has this year. It’s only logical to attribute the addition of jet motion to the basis for the Rams’ turnaround, and other teams have implemented it more often.

Looks like Chucky was right after all.

The Rugby Championship Roundup 6

By:

South Africa vs. New Zealand – Newlands Stadium: 7 October 2017

The public begged for the Boks to come back after the lopsided last meeting, and boy did they deliver. New Zealand spoke highly of South Africa and stood firm that the 57-0 drubbing in New Zealand three weeks ago was not a true reflection of the match. New Zealand still managed to win, though they were matched pound for pound in the 80 – no, 90-minute match.

It was that type of game where no one would ever want it to end. On the stroke of halftime, neither team stopped play and in the end the match went into halftime with the clock at 49:57, 24 minutes with the ball in play and the score standing at SA 3-8 NZ. The second half continued to be a massive brawl with the Boks scoring first to grasp the lead back to 10-8. At the end of the match, one could say that the New Zealanders kept their composure better and won off of the Boks errors and one piece of brilliance from the smallest player on the pitch in Damian McKenzie.

With the score at 17-15 and 11 minutes to play, McKenzie cut through the Bok defence, leaving Damian de Allende in no man’s land to score arguably the try of the match. The game was all but over at that time, but when the Boks had to chase eight points after de Allende conceded a penalty for a late charge and was subsequently given a red card, the valiant effort was just one point too far.

It was the first time in the Rugby Championship that a team matched the All Blacks with running metres and had the world champs defending more than ever, having to make 299 tackles. Still they managed to somehow scrape the win and show the world of rugby why they are the No. 1 team. South Africa and New Zealand produced arguably the best test match of the 2017 season, which leaves both teams in a very good space for the end-of-year tour.

Argentina vs. Australia – Estadio Malvinas: 7 October 2017

With the Championship already concluded as New Zealand wrapped up the cup the previous week, there was still the slug out for outright second spot on the log. With South Africa losing to New Zealand, Australia only had to secure victory to ensure they finish second overall, and that was exactly what they did.

With halftime approaching, Argentina had the opportunity to take the lead with a penalty on the half-way line. They chose the line-out option but failed to capitalized, and the teams went into the break deadlocked at 13-all. Australia came out firing in the second half and wrestled to take the lead, only to last for all but four minutes as Argentina smashed back to make it 20-all with 30 minutes to play. Australia, however, stepped up the pace and took the next 17 points unanswered to eventually break Los Pumas spirit and take the match, 20-37.

Argentina will reflect on a disappointing campaign in 2017 and will hope to build towards the end-of-year tour to try and turn their season around. Australia, however, will go into the end-of-year tour with a lot more confidence and will surely be a strong competitor in all matches.

The Professor vs. Washington’s Quick Learners

By:

STATS TVL data and Video Solutions break down Kyle Hendricks’ start against Bryce Harper, Daniel Murphy and the Nationals in Game 1 of the NLDS

Most envision Dusty Baker clenching his jaw in anger so hard that he snaps his trademark toothpick whenever he’s presented with any sort of advanced metric. In a sport where literally everything is counted, Baker has a reputation for being an old-school manager who trusts his gut far more than he’ll ever trust baseball’s version of a mathematician.

It seems fitting then that Kyle Hendricks, nicknamed The Professor with a degree in economics from Dartmouth who crunches numbers for fun, will be on the mound for the Chicago Cubs in Game 1 of the NLDS trying to shut down Baker’s Washington Nationals.

Ironically, STATS’ TVL algorithm mostly sides with Washington over Hendricks for the opener of the best-of-five series.

TVL tracks pitch type (T), velocity (V) and location (L) for each MLB pitcher and records the data into categories such as usage percentage of a specific pitch, the average velocity of each pitch type and the percentage a batter hits the ball on the ground against that pitch. The data is broken down further to show opponents’ batting average, slugging percentage, swing percentage and swing-and-miss percentage each time a specific pitch is thrown.

A pitcher’s TVL then can be pitted against a hitter’s success when facing specific pitches to project how the hitter would fare versus a particular pitcher, and unfortunately for Hendricks, he projects to struggle against the meat of the Nationals’ order.

Remember the last time Daniel Murphy faced the Cubs in the postseason? He homered in each game of the New York Mets’ 2015 NLCS sweep, including him tattooing a Hendricks two-seam fastball in Game 3. Murphy is entering this series on a tear as well, and that doesn’t even include his success against Hendricks the last time they faced off.

Murphy finished the regular season on a seven-game hitting streak, going 12 for 26 with a homer, two doubles and a triple while also walking four times. He also went 10 for 19 over his last five home games. That might not bode well for Hendricks, who served up two homers to Murphy back on Aug. 4.

Murphy crushed Hendricks’ 71-mph curveball that he hung over the plate in the first inning of the Nationals’ 4-2 victory. Hendricks then threw five straight fastballs over Murphy’s next two at-bats, and after flying out, Murphy hit a four-seamer on the outer half for an opposite-field homer.

Have a look at how each approached those at-bats using STATS Video Solutions (SVS):

 

Those previous blasts might be a bit misleading given that Murphy is 4 for 15 (.267) including the postseason in his career against Hendricks, but that success – and the TVL data – shows Hendricks needs to be careful with his pitch selection against Murphy’s power potential.

Murphy’s homer off Hendricks’ four-seamer isn’t all that surprising when knowing left-handers hit .319 against Hendricks’ four-seam. Hendricks has retired Murphy on mostly changeups and curveballs, and Hendricks will want to locate his change accordingly given that Murphy projects to hit .233 against that pitch, compared to solid averages against Hendricks’ four-seamer (.409), two-seamer (.383) and curve (.359).

Here’s how Murphy and other Nationals hitters are projected to fare against Hendricks’ pitches, according to STATS TVL data:

Judging by that data, Bryce Harper is going to be waiting for a Hendricks fastball to pounce on after he went 3 for 18 in five games since returning from a hyperextended left knee that kept him out over a month. Hendricks stayed true to the TVL data in the only game he faced Harper this season, throwing mostly curves and changeups. Harper lined out to third, hit a weak grounder to short that he beat out for an infield single, then popped out.

Harper’s projected low averages against Hendricks’ curve and changeup could bode well for Hendricks again. Hendricks’ changeup is his second most-used pitch at 27.9 percent, and that goes up to 29.6 percent against left-handed hitters. Those lefties hit .197 and slugged just .256 against it while whiffing on it 35.7 percent of the time – Hendricks’ highest whiff rate for any of his pitches.

However, Harper is notorious for taking advantage of pitchers’ mistakes, and Hendricks certainly can hang a pitch in the heart of the zone from time to time. Last season, despite the numbers favoring Hendricks, Harper ripped a hanging changeup for a solid single on the second change he saw in the at-bat.

Hendricks by no means is a power pitcher, and his average velocity thrown to Harper in their 15 matchups is 85.2 mph. That’s probably why Harper tends to be late even when Hendricks gets his fastball to 88 or 89 mph in their matchups.

But Harper’s ability to hit for power to all fields makes him a threat no matter the location or speed of the pitch. The one time Harper has gone deep off Hendricks – May 26, 2015 – Harper didn’t even think he got enough of it, taking an 89-mph fastball over the left-field wall. Watch his reaction after contact.

Perhaps the most interesting TVL projections are Ryan Zimmerman’s. The Nationals’ veteran first baseman is 0 for 9 with two strikeouts in his career against Hendricks but has solid projections when facing Hendricks’ two-seamer, curve and change.

That’s especially interesting when considering Hendricks has retired Zimmerman with a changeup five times in their matchups. But Zimmerman is entering Game 1 having going 9 for 11 in his last four home games with a pair of two-homer contests. So maybe things are ready to turn around for Zimmerman in his matchups with Hendricks.

The same could go for Anthony Rendon, who finished the season on a five-game hitting streak and hit safely in his last 14 at Nationals Park. Rendon fell to 2 for 11 with a walk and three strikeouts in his career off Hendricks after facing the Cubs starter Aug. 4.

Hendricks took advantage of Rendon’s .196 projected average against his changeup, throwing Rendon four straight changes while striking out Rendon in the first inning. Hendricks then went after Rendon with two-seamers in the next two at-bats – a walk and a lineout to second. Rendon has a .348 projected average against that pitch, so Hendricks will want to be careful about repeating that sequence.

The middle of Washington’s order – Murphy (.307), Harper (.294), Zimmerman (.335) and Rendon (.293) – all have pretty solid projected overall averages against Hendricks, who will have to mix up his pitches, locate and remember what worked in the past if he hopes to tame the red-hot Nationals hitters and help the Cubs strike first in the series.

Trouble Could Find Him

By:

STATS TVL Data Shows Why Cleveland Indians Manager Terry Francona’s Decision to Start Trevor Bauer against the New York Yankees in Game 1 of the ALDS May Backfire

It’s bold.

The decision to bypass an adequately rested ace in Game 1 of a five-game series comes with the hope that postgame praise of an intrepid call will follow. For Terry Francona, though, there’s reason to think there’ll only be more questions concerning the decision to start Trevor Bauer over Corey Kluber.

Some of the surface-level data supports the Cleveland manager’s decision. Dive a bit deeper into STATS TVL data, and it looks more like the Indians might need to put up some serious offense to not fall into an early hole in their ALDS matchup against the New York Yankees.

TVL tracks pitch type (T), velocity (V) and location (L) for each MLB pitcher and records the data into categories such as usage percentage of a specific pitch, the average velocity of each pitch type and the percentage a batter hits the ball on the ground against that pitch. The data is broken down further to show opponents’ batting average, slugging percentage, swing percentage and swing-and-miss percentage each time a specific pitch is thrown.

A pitcher’s TVL then can be pitted against a hitter’s success when facing specific pitches to project how the hitter would fare versus a particular pitcher, which is what we’re going to use here to give some insight into the specific Bauer-Yankees matchups that’ll take place tonight.

Francona’s reasoning is centered around the series including two days off, so Kluber, the AL Cy Young frontrunner, would still be starting a decisive Game 5 and doing so on a normal cycle. The manager said the routine was important to Kluber, and there wasn’t another viable option to line it up that way. Additionally, if they win in four, then Kluber lines up nicely for Game 1 of the ALCS. But there’s also something to be said for avoiding a 1-0 hole in a five-game series, and few will argue which pitcher gives you a better chance there.

Now, the surface-level data. Both pitchers have been great in two starts against the Yankees this season. Kluber has probably been a bit better, going 2-0 with a 1.59 ERA and .105 opponent batting average and .381 OPS while chewing up 17 innings. Career: 5-1 with a 1.80 in seven starts with five straight dominant wins.

Bauer went 2-0 with a 1.38 ERA in 13 innings, winning both home and away in August with a .229 OBA and .661 OPS. His career numbers against the Yankees aren’t nearly as strong, and they roughed him up as recently as last season with present Yankees Brett Gardner, Didi Gregorius, Jacoby Ellsbury and Chase Headley all notching two-hit nights and an RBI each to tag Bauer with all five runs in a 5-4 final. So he’s not Yankee-proof.

Such individual matchups, and specifically how the New York lineup projects to hit against Bauer, is where the decision becomes particularly teeth-clenching from a predictive standpoint. Using TVL, the Yankees lineup projects to have a .289 average and .525 slugging percentage against him. Get even more granular, and six batters project to hit at least .303 against Bauer’s entire arsenal of pitches:

There’s a lot of projected doom there, but it doesn’t mean there aren’t ways for Bauer to work through it. Let’s now look into how he might best handle the order and exploit the occasional projected weakness to maintain his manager’s reputation. Bauer’s second-half pitch selection shows how he evolved over the course of the season – just not always for the better.

In the first half, Bauer’s four-seam fastball and curveball accounted for 67.1 percent of his pitches. Those are still his most-used pitches, accounting for 68.5 percent in the second, but the effectiveness of the four-seamer has actually fallen off. In the first half, he threw it 39.0 percent of the time with a .269 OBA and .433 slugging. In the second, 36.6 percent usage with .295 and .476 splits. Return to the chart above, and that doesn’t bode well against the likely Nos. 1-5 in the New York lineup. He’ll also want to be selective with his curve against a few of those bats.

The more positive change for Bauer arrives in that he was in the first half making far more use of a two-seamer (15.3 percent vs. 6.5 in the second half) and cutter (10.7 vs. 5.0). Opponents hit .333 against the cutter with a .704 slugging percentage in the first half. They were at .289 and .566 against the two-seamer – still impressive numbers.

Bauer has in part replaced them with an effective slider, which probably bodes well against the middle of the Yankees’ lineup. But he still only threw that pitch 11.6 percent of the time in the second half. Is that enough to maintain what’s been an impressive 13 innings against the Yankees, or is the bottom about to fall out?

Gregorius is 4 for 9 with a home run and double in the last two seasons against Bauer. Gardner is also 4 for 9 in that time. Todd Frazier is 6 for 14 with a home run and double. Bauer is yet to retire Aaron Judge – 1 for 1 with two walks – and from the looks of it, he should have very little confidence throwing the young slugger either of his go-to pitches.

Those numbers are from Bauer’s top two seasons in the bigs. Many of them also happen to be in line with TVL. Regardless, the 26-year-old’s manager has expressed plenty of confidence in him.

Again, it’s bold.

NL Wild Card Preview: Who Can (J.D. Martinez) and Can’t (Paul Goldschmidt) Hit Jon Gray’s Slider Using STATS TVL Data and Video Solution

By:

J.D. Martinez injected power into Arizona’s lineup almost immediately upon his arrival. Gone were the days of Paul Goldschmidt being the only reliable home run threat, as the Diamondbacks homered an average of every 22.4 at-bats after acquiring Martinez compared to every 27.6 prior.

But according to STATS TVL data, only one of those sluggers seems to have a feel for the breaking pitches of Jon Gray, who will try to stymie the heart of Arizona’s lineup for the Colorado Rockies in Wednesday night’s NL Wild Card showdown.

TVL tracks pitch type (T), velocity (V) and location (L) for each MLB pitcher and records the data into categories such as usage percentage of a specific pitch, the average velocity of each pitch type and the percentage a batter hits the ball on the ground against that pitch. The data is broken down further to show opponents’ batting average, slugging percentage, swing percentage and swing-and-miss percentage each time a specific pitch is thrown. A pitcher’s TVL then can be pitted against a hitter’s success when facing specific pitches to project how the hitter would fare versus a particular pitcher.

Martinez hit 29 of his 45 homers after making his Diamondbacks debut July 19 and led MLB homering an average of every 8.0 at-bats in that time, way up from his 12.5 AB-to-HR ratio with Detroit. His Arizona surge includes going 11 for 23 with three homers and four doubles in six games against the Rockies, who trailed the D-Backs by one-half game prior to the Martinez trade before finishing six games back.

Martinez homered twice off Gray in six at-bats against him. Not much explanation is needed for Martinez hitting one off a Gray four-seam fastball, given that STATS’ TVL data projects Martinez to hit .448 against Gray’s most-used pitch. But how is it that Martinez blasted a slider out of the park when he’s projected to hit just .201 off Gray’s slider?

The homer shown above is exported from STATS Video Solution. It was the first time Martinez and Gray faced each other and the homer came on a 2-2 count, one of two sliders Martinez hit out of the park with the count reading as such this season. Three times Martinez crushed 0-1 sliders for homers. Overall, despite hitting .221 against sliders one the season, Martinez homered nine times against that pitch, the second-most he hit off a specific pitch behind the 18 fastballs he hit out.

Martinez’s success against Gray’s slider might be an anomaly considering opponents hit just .164 and had a 35.3 swing-and-miss percentage when facing Gray’s slider. Martinez’s homer was one of only two Gray allowed when throwing that pitch all season.

Still, it seems wise that Gray threw only one other slider to Martinez in their six times facing each other this season – on a 3-0 pitch that Martinez was expected to take and did. Instead, Gray has fed Martinez a heavy dose of curveballs, pitches that were thrown to complete Gray’s two strikeouts of Martinez on the season.

Below is Martinez’s second at-bat against Gray on Sept. 12 following the homer earlier in the game. He also struck out Martinez on a curve Sept. 2 – not all that surprising considering Martinez is projected to hit just .191 against that pitch when Gray throws it.

Whether or not Gray decides to try his luck with a slider against Martinez on Wednesday night remains to be seen, but you can be sure Goldschmidt will see that pitch plenty hitting in front of Martinez.

Goldschmidt enters this winner-take-all contest 0 for 9 with five strikeouts against Gray this season and 0 for 11 in his career overall. Gray, who threw 57.2 percent of his pitches as fastballs this season, almost exclusively tossed breaking balls to Goldschmidt. The two faced off three times June 20, and Gray threw three fastballs, nine sliders and a changeup to Goldschmidt, who struck out swinging on a slider low and away in the dirt in all three at-bats.

Goldschmidt saw 17 sliders in his nine at-bats against Gray and struck out four times on that pitch, with the other coming on a changeup – a pitch Gray threw 25 times all season, seven to Goldschmidt.

Watch the at-bat the last time Gray and Goldschmidt faced each other in the sixth inning Sept. 12. Goldschmidt saw two fastballs in five pitches. The first was nowhere near the strike zone, and Goldschmidt appeared taken aback by the second before striking out.

And while Goldschmidt is projected to hit .461 against Gray’s 4-seam fastball, he’s likely to see very few – if any – from Gray on Wednesday. Take a look at his projected averages against Gray’s other pitches in the graphic below, as well as how Gray is projected to fare against other Diamondbacks hitters.

Maybe it’s just a matter of Goldschmidt trying to do too much the pitch that Gray gives him rather than making quality contact, which is something Rockies’ MVP candidate Charlie Blackmon learned heading into another matchup with D-Backs starter Zack Greinke.

Blackmon won the NL batting title with a .331 average and tied with teammate Nolan Arenado for third with 37 homers, but he went just 3 for 16 (.188) with six strikeouts against Greinke. The strategy for Greinke is to stay to the outer half of the plate against Blackmon and let Blackmon get himself into trouble trying to pull the ball. When he goes toward the inner half on Blackmon, the ball dips into the dirt and the Rockies’ slugger chases.

Two of Blackmon’s three hits off Greinke went to the opposite field when he went with the pitch, and you can watch how Blackmon approached both of those at-bats below.

 

Of the 13 times Greinke retired Blackmon, six times Blackmon attempted to pull it and the out was made on the right side of the field. He grounded out to second four times, lined to right and flew out to center. Blackmon struck out six more times, with five coming on low breaking pitches. Only when Greinke missed his spot with a fastball, as noted in the screenshot of the SVS interface below, did Blackmon have success pulling the ball.

Blackmon projects solid averages against most of Greinke’s pitches – .433 against the four-seam, .410 against the two-seam, .307 versus the curve and .303 against the slider. But Blackmon’s success Wednesday likely will depend on how he approaches the location of those pitches.

STATS NBA 2017-18: An Advanced Look Into The Offseason Shakeup

By:

The ground has most certainly shaken in the East with the most unlikely of trades between conference frontrunners.

Kyrie Irving and Gordon Hayward have landed in Boston to join forces with the latter’s former coach. LeBron James has been dealt an unlikely top-three scorer.

Transcontinentally, 2016-17’s DIYers James Harden and Russell Westbrook have been given some serious help. Or is it the other way around: Chris Paul, Paul George and Carmelo Anthony still don’t have a true last name among them, but are they the main beneficiaries?

Yeah, there’s a Jimmy Butler-Tom Thibodeau reunion on in Minnesota, but the Timberwolves also brought in a point guard who may better suit their bigs.

Lob City is no more, but what it ultimately comes down to is everyone’s still chasing Dub City. Come along with STATS and the most reliable and fastest NBA data available as we take a deep diagnostic dive to weave select stories of how the pursuit may go down in 2017-18.

The Rugby Championship Round 5 Roundup

By:

South Africa vs. Australia – Free State Stadium: 30 September 2017

Another so-close-yet-so-far case for the Springboks and the Wallabies as they played out to their second draw of the season against one another. With South Africa desperate to put away the demons of their last match against New Zealand, they were left to rue their missed chances on a day that should have kept their title challenge alive.

The Springboks came out of the blocks firing and had the Wallabies under pressure with an expansive style of play by running from deep and trying to break the defensive line from the first minute. Australia, however, stuck to their guns and did not give an inch on defence with Michael Hooper ever impressive at the breakdown halting the Bok charge. Making close to 200 tackles in a match where the Springboks accumulated the most running metres of their campaign, the Wallabies showed true grit and determination not to lose the match.

With a bit of hair pulling and players charging one another off the ball, all seemed to have been left out on the field as the Boks had the opportunity in the 79th minute to seal the deal. Once again, as is the story of their season, they were unable to hit the nail in the coffin, and both teams will have to reflect on what could have been.

Australia and South Africa will now fight for second position as the Wallabies travel to Argentina to face a wounded Los Pumas, whereas South Africa will have to pull their weight if they are to inflict any sort of revenge on the All Blacks.

Argentina vs. New Zealand – Jose Amalfitani Stadium: 30 September 2017

Unstoppable, unbeaten and ever impressive. These are the words and phrases to describe the first 40 minutes of the All Blacks against Los Pumas. Retaining yet another Rugby Championship title, the world champs showed their true class in a 30-minute blitz where they outplayed and outscored the Argentine side at will.

The All Blacks went almost a point per minute, putting Argentina to the sword in the first 30 minutes and racing to a 26-point lead to effectively win the Rugby Championship. Argentina did put up a fight in the second half, conceding only one try for the next 40 minutes. The All Blacks overcame a disjointed and error-ridden second half by defending like Trojans and never gave a sniff at a comeback to Los Pumas.

Argentina will now aim to get their first victory of the Rugby Championship when they tackle a desperate Wallabies side, and the mighty All Blacks will want to reap the rewards of a new management strategy and inflict even more pain on the Springboks.

What Not to Throw Aaron Judge (And Other Less Obvious Matchup Stories Assessing the Yankees-Twins AL Wild Card with STATS TVL Data)

By:

Ervin Santana missed with a slider, then threw Aaron Judge a 94-mph fastball up in the strike zone and not nearly far enough away in a 2-1 count. A predictable conclusion followed: The Yankees slugger hit the ball over the right-center field fence for a 1-0 lead in a game the Yankees won 2-1.

It’s everything the Minnesota Twins will want to avoid tonight as they try to make their first postseason appearance since 2010 stretch beyond one game.

It happened in the bottom of the first with one out on Sept. 18. It was the first time the two ever faced each other, but that’s no longer a passable excuse. There was still probably little need for Santana to throw Judge that pitch in that situation, according to STATS TVL data.

TVL tracks pitch type (T), velocity (V) and location (L) for each MLB pitcher and records the data into categories such as usage percentage of a specific pitch, the average velocity of each pitch type and the percentage a batter hits the ball on the ground against that pitch. The data is broken down further to show opponents’ batting average, slugging percentage, swing percentage and swing-and-miss percentage each time a specific pitch is thrown. A pitcher’s TVL then can be pitted against a hitter’s success when facing specific pitches to project how the hitter would fare versus a particular pitcher, which is what we’re going to use here to give some insight into this winner-take-all wild-card game.

Let’s first revisit that Santana-Judge sequence using video exported from the At-Bat Viewer in STATS Video Solution:

In the next at-bat, Santana threw Judge four straight sliders – a pitch he threw right-handed hitters 48.5 percent of the time in comparison to the four-seamer at 34.6 percent. The at-bat begins with a runner on first before Brett Gardner advanced to scoring position on a ball in the dirt, and the outcome was considerably different with a strikeout on a slider out of the zone:

It’s one thing to be able to analyze these matchups after the fact, but with TVL, we can go a step further and project pitch-specific matchup outcomes. Consider the Santana-Judge numbers, and it’s easy to see Santana should be incredibly selective with when he brings the heat, though that varies throughout the New York lineup:

Aside from Judge’s contrast, the number that stands out here is Greg Bird’s suspiciously low projected average against Santana’s four-seamer. He’s 2 for 5 with two home runs in their career matchups, and one of those came off a four-seamer. But that was back in 2015. Bird missed all of last season with a torn labrum and most of this season after undergoing foot surgery. He’s since shown little ability to catch up to that same pitch and is 3 for 38 off right-handed pitchers’ four-seamers this season, while left-handed batters have hit .173 against Santana’s four-seamer. Bird is still able to mash a changeup, which was the Santana pitch that accounted for that other homer, so the Twins should have an idea of how this matchup has evolved over two seasons, despite the two seeing little of each other again in that time.

There are similar insights to be had when assessing Luis Severino against the Twins, despite the right-hander not facing any of his Tuesday night opponents in more than two at-bats. For example, Severino may want to consider setting up Jorge Polanco and Eddie Rosario differently than Brian Dozier given the ability of the first two to track the slider:

It doesn’t take quite that level of data to show the Twins have lost their last 12 playoff games or that nine have come against the Yankees. But digging that deep might help Minnesota end its plight in another game where one pitch could prove far more costly than it did two weeks ago.