Data

Data

Data Rant: Nathaniel Clyne

June 29, 2015 Tags: , Data 14 comments
featured image

Towards the end of last season the press was replete with articles citing a “senior source inside Old Trafford” as evidence that Manchester United was beginning preliminary negotiations with Nathaniel Clyne and his club Southampton. Yet, Ed Woodward has done little to counter Liverpool’s acquisition of the full-back this week. Read More

Data Rant: in search of goals

May 16, 2015 Tags: , Data No comments
featured image

Manchester United has qualified for next season’s Champions League and a good result against Arsenal – together with any further slip-on on the London side’s part – may even see the Reds sneak directly into Europe’s premier club competition. United must finish third to avoid a Champions League play-off against potentially difficult opposition next August. Read More

Data Rant: Long ball Louis

March 17, 2015 Tags: , Data 11 comments
featured image

Perhaps a detailed statistical analysis of Louis van Gaal’s long ball football is not necessary – it is plain that the strategy has not worked this season. Manchester United’s performance against Tottenham Hostpur last Sunday may have been excellent, but the game saw United attempt a long ball 12.3 per cent of the time* – the fourth highest figure this season. It is in this context that a study remains useful. Read More

Data Rant: Moyes vs Van Gaal

January 18, 2015 Tags: , , Data 13 comments
featured image

Manchester United’s victory against Premier League strugglers QPR increased the Reds’ total to 40 points this season. It was United’s 22nd game of a mixed campaign. Much has been made of the total after 21 games – just 37 points – the same amount David Moyes’ side had amassed at the same time last season. So who has really done better – Moyes or Louis Van Gaal? Read More

Data Rant: Ashley Young

December 30, 2014 Tags: , Data 4 comments
featured image

Louis van Gaal has transformed Ashley Young from stuttering winger to a functioning wing-back this season. The 29-year-old has not set the world alight in the new role, but his presence has served a useful purpose amid an ongoing injury crisis at Old Trafford, with new signing Luke Shaw absent for most of the campaign.

If nothing else, Young has always been diligent and success in the wing-back position is in part dictated by athleticism; a hard-working winger makes for a good substitute in the role. Another industrious winger, Antonio Valencia, has also been used as a make-shift wing-back this season.

Young has taken to the new role with some gusto. There is no denying, however, that his performance has been judged leniently given the player’s previous mediocrity. The Manchester United defence is yet to display any consistent solidity this season. It is easy to mask incompetence in a sea of ineptitude. At United, simply doing the job is never enough.

Using Squawka’s index score, which totals a player’s (or team’s) contributions over each game, we investigate the relationship between Young’s performance and United’s overall performance.

Figure-1

In Figure 1, above, the correlation is strong enough to suggest that Young is making a solid contribution to United’s cause. After all, the Englishman is trusted to marshal the left flank on his own in Van Gaal’s narrow 3-4-1-2 system. Another possible interpretation is that the team is carrying Young; nine other Reds on the field playing well enough to take the heat off the 29-year-old.

A make-shift defender naturally lacks inherent defensive qualities and in United’s three-man defence, the left-most centre-back must be dominant enough to allow Young to concentrate on attacking. If the theory is right, Young’s Squawka ratings will be highly correlated with his centre-back partner’s.

Figure-2

In Figure 2, above, we discover this is not the case. That is, Young’s good run has not been due to Tyler Blackett or Jonny Evans mopping up after the Englishman. If anything, Young has been exemplary in defence – his numbers, particularly interceptions, dwarf those of a proper left-back in Luke Shaw.

What, then, about Young’s midfield partner?

Figure-3

In Figure 3, above, the trend is stronger than that between Young and his central defensive partner’s performances. This makes sense – it is the left-sided central midfielder, not the central defender, who is in the immediate vicinity of a left wing-back. Note – United’s game vs. Newcastle was excluded due to distorting effect of Wayne Rooney’s goals on his score in that game.

Indeed, Young is heavily dependent on central midfield. If Young is allowing the midfield to prosper then there is no reason why his partner central defender has not prospered too.

Curiously, Young’s first two games this season were very poor. Young recorded his third lowest Squawka score against Arsenal. These three games share a commonality in that the left central midfielder was a dedicated holding midfielder. In all other games, his central midfield partner – Angel di Maria, Juan Mata or Rooney – was the more attacking of the central midfield duo.

Young’s poor performance in his first two matches at left wing-back may have been caused by acclimatisation to the new role, but United was up and running by the game at the Emirates. So what about the third game of the season against Burnley in which he had the third best game so far?

It may be harsh, although not unfair, to describe Young as a mediocre talent. There is two seasons worth of evidence. The former Aston Villa man may have the pace and athleticism, but lacks the technique to navigate the attacking third without relying on speed. As a wing-back, Young has an extra 15 yards to accelerate into. The Englishman has enjoyed plenty of space to gather momentum this season.

Young may have a head start as wing-back, but he isn’t the kind of player to beat a man on his own. Playing next to a holding midfielder, the onus is entirely on the former England international to make ground. With a central midfielder such as Di Maria running with the ball and pushing forward, Young faces an easier task.

United’s recent game against Tottenham Hotspur is indicative. Young enjoyed his second worst game of the season according to Squawka. Jonny Evans, and later Shaw, had underwhelming games too, but Rooney struggled badly, misplacing 25 per cent of his passes. Young has not somehow reinvented himself as a wing-back, rather Van Gaal’s introduction of a proper box-to-box midfielder – more specifically deploying one near Young – has played to Young’s strengths.

There is little doubt that Shaw will walk into the team ahead of Young, fitness permitting, in the games ahead. Young’s deployment should be seen as Van Gaal’s attempt at making the best of what he has rather than a true renaissance – recall that Marcos Rojo has always been picked ahead of Young on the left. This analysis suggests that Young requires a highly specific system to function. The suspicion is that Young does not have the talent to warrant such treatment.

All data from Squawka
Assumptions dictating linear regression have not been held strict

Data Rant: will United make the Champions League?

November 3, 2014 Tags: , , Data 12 comments
featured image

Each week, Louis van Gaal appears to break a new record: the “worst x games in Manchester United history”. This time round the lowest points total at this stage since 1986 – in the corresponding week that Ron Atkinson was sacked 28 years ago. Indeed, the David Moyes experience has tinted everyone’s outlook though Sir Alex Ferguson’s remarkable rise after a lean first few years should not be forgotten just yet.

With that said, United’s loss to City on Sunday all but conceded the title. Barring a catastrophic meltdown on Chelsea’s part, United’s 21st English title will have to wait. Worse, for United, reaching the top four spot is under genuine threat.

United is only four points behind Arsenal in fourth and the gap may be bridged to just one should the Reds prevail against the London side in two week’s time. Moreover, while Southampton’s point haul so far is amazing, at the risk of being patronising, the Saints’ thin squad will surely suffer as the season goes one, dragging the side down at some point.

In that context the situation may appear somewhat manageable, but it is worth remembering that there are still 28 games to go – continuation of the current form will see United missing Europe for two seasons in a row.

After all, as Data Rant previously noted the number of “poor streaks” is a good indicator of the final points tally. A poor streak is two or more consecutive draws or losses and “a team may have a poor patch of form – even a pretty severe one – but it’s fine as long as the team can get back to winning ways and stay on that path.”

Figure-1

The past two seasons suggests that side’s in the top four have suffered at most five poor streaks during a campaign, though realistically only three are allowed. United has already suffered two by this definition and there is little margin for error if history has any weight.

Figure-2

Data Rant also analysed the relationship between the final point tally and points from the first five games of 2012/13 and 2013/14. The relationship was weak – after all, five games is only 13 per cent of a season. Let’s see if we can judge a season after a quarter of it has passed.

Figure-3

Quite obviously the fit of trend line has tightened. The predicted final tally based on 13 points at this stage is just 51 – nowhere near the top four. Normally, a mid-70 final point total is required to qualify for the Champions League.

Points Total

The standard deviation here is 13.9, assuming normal distribution, meaning there is about a 4.2 per cent probability of United gathering 75 points come May, according to Data Rant’s calculations.

Points Total

Excellence is no longer enough – only consistent excellence will do. A busy winter transfer window is a must though this may very well be a job too far even for the “Iron Tulip.”

All data from Squawka
Assumptions dictating linear regression have not been held strict

Data Rant: United’s use of width

October 28, 2014 Tags: Data 5 comments
featured image

Perhaps a push for the Premier League title is a step too far this season, but a place in the top four appears an achievable goal after Sunday’s draw with Chelsea – it leaves Manchester United only three points off the coveted fourth spot. Not least because Southampton and West Ham United are unlikely to sustain their excellent form, while Arsenal and Liverpool are nearing the point of permanent self-destruction.

The normally rampant Chelsea side was comfortably held by United at Old Trafford on Sunday, despite the Reds requiring a late equaliser, with seven yellow cards speaking to how desperately the Londoners defended at times. Indeed, much of the tactical battle was won by United manager Louis van Gaal. Cesc Fabregas was silenced by Maroune Fellaini, while Daley Blind and Juan Mata pressed high up the pitch forcing Chelsea to play long more often than is typical.

More interesting, perhaps, was Robin van Persie’s role, with the Dutchman dropping deep to create space for Fellaini to exploit. In fact, Van Persie, who started the season poorly, had an excellent game, which was capped by that dramatic late goal.

Tactics

Previously, this column has identified the importance of shots per game and take on attempts as two key indicators of a striker’s goal tally. The Dutch striker’s shots per game ratio has dropped sharply from 2012/13, and so has interceptions per game.

Players stationed deeper are more likely to intercept opposition passes so the figure is a good indicator of a player’s position on the pitch. The numbers suggest that Van Persie’s isolation upfield has been causing the Netherlands captain to shoot less.

Table 1

Van Persie has lost the pace of yore and perhaps the aging process has also caused the 31- year-old to become more of a poacher. With no ‘big man’ to play off, Van Persie needs runners around him to ‘ghost’ into the box. As the most advanced player, Van Persie has to hold up the ball and lead the line – something that the former Arsenal player can no longer do as effectively. Van Persie enjoyed a fine game against Chelsea, ironically, because he was further away from goal.

Curiously, United frequently resorted to directing the play wide against the Londoners. True, Angel Di Maria is the most creative player in the squad, while the youthful Adnan Januzaj offers genuine pace, but a more central approach could theoretically have been more fruitful. With Fellaini providing a physical presence, Van Persie might have made more out of through balls than aerial battles. And, terrible recent form notwithstanding, Mata was also marginalised as result of the strategy. Note that United’s reliance on the flanks has been almost Moyesian of late.

Take for a moment the following thought experiment. With David de Gea initiating a United attack the Spaniard has an option to go wide or central. Choose, for example, Di Maria on the left and the Argentinean can cross, or pass to his central midfield colleague, with the latter more likely to attempt a through-ball. In our example, had de Gea picked out Ander Herrera the new Old Trafford arrival can attack the opposition box or release Rafael da Silva down right for the Brazilian to cross.

Picture 2

This decision is a matter of risks and rewards. A sideways pass is relatively safe – connecting typically more than 75 per cent of the time. Threading the ball through to a forward is hard, but much more likely to yield dividends in terms of chances created than crossing into the box, especially given that the Reds don’t have a towering striker. Assume, respectively, success rates of around 15 and 10 per cent.

Suppose further that United players are indifferent between passing sideways and vertically – about half the time each. Logic dictates that the Reds are slightly better off taking the central route by 0.7 percentage points.

Despite the meagre difference, the figure is significant. Given the amount of creative talent in the middle, and Van Gaal’s tendency this season to field a number 10, the central approach will be far more productive in creating chances than the theoretical 15 per cent. Further, with United lacking the brawn to make the most of crosses the supposed success rate of 10 per cent is likely a very generous estimate. Therefore the gap between success through the middle or from wide areas is likely to be much bigger.

So why has United been so focused on playing down the flanks?

Mata and Wayne Rooney’s failure to shine at number 10 could be an issue, but a much more likely reason is the Reds’ leaky defence. When United concedes possession in central areas, the opposition has the entire width of the pitch through which to conduct play. Lose the ball out wide, however, and United can contain the danger to that flank and adjacent areas – sweeping passes to the opposite flank is more likely to end up in the stands than a good attacking situation. From a defensive point of view, keeping the ball wide is the safer option.

Put succinctly, van Gaal has likely sacrificed some attacking flair for defensive solidity. United has been playing this way all season – it follows that Van Gaal has been concerned with United’s defence from day one.

The famous Latin phrase Quad Erat Demonstrandum is rarely used to conclude a proof on the grounds of pretension. Mathematicians, instead, draw a little square to end proof – and square passes are the keys in unlocking the van Gaalacticos.

A runner down the left flank will benefit by having a midfield to offload the ball to. In the middle, a player must always have safe options to relieve pressure in order to progress with United’s defensive structure intact. As the newly assembled United squad assimilate to playing with each other, there should be more central – therefore more creative – approaches fostered.

Van Gaal has talked the talk of upholding United’s attacking tradition. When will he start walking the walk? It has already been three months.

Data Rant: the first five games

September 25, 2014 Tags: , Data 1 comment
featured image

It is some surprise that under Louis van Gaal Manchester United has made a worse start to the season than under David Moyes last year. It is surely the calibre of the Dutchman that has kept discontent at a minimum. With so much investment in the summer, United’s minimum goal is to finish in the Premier League top four, while the lack of League Cup and Champions League games might even allow the Reds a shot at the title.

Still, after just five games, United has drifted to eight points behind Chelsea and five off Arsenal. With 33 games remaining these points can be made up, but there is no denying that a win and two draws out of five games against relative minnows is a very poor start.

Club’s that suffer a bad run at the beginning of the season often attribute the malaise to a ‘squad gelling,’ while Sir Alex Ferguson often cited his side as an example of “late starters.” But what about teams that make a great start? Does this contribute to a strong season as a whole, with the side continuing to win. Or do teams’ great starts unravel as the season marches on and injuries take toll? Data Rant takes a look.

Figure 1, below, charts the total points gained in a season versus points gathered in the first five games – we look at the teams involved in past two Premier League campaigns. There is little to suggest that the final point tally can be predicted from early season form.

Figure-1-Points-vs-Five-Games

In 2012/13, Chelsea enjoyed the best start by gathering 13 points from five games yet ended up third. Liverpool had a comparatively bad start, ending up with two points in five games – the same as Reading and Q.P.R. Yet, the Anfield side finished  seventh while Reading and Q.P.R were relegated. Newcastle United, on the other hand, enjoyed a relatively good beginning to the campaign, gaining eight points, but barely escaped relegation. Tottenham Hotspur, who managed the same eight points, almost beat Arsenal to Champions League qualification.

In the following season, 2013/14, Arsenal and Tottenham won two more points in first five games than Manchester City, Liverpool and Chelsea yet finished fourth and sixth. Sunderland only mustered one point, but survived relegation despite Norwich City, Fulham and Cardiff City winning four, four and five points at the start.

Any team can have an off day; a well prepared minnow can also take a bigger side by surprise and win on its merits. The goal for every team is to not only minimise losses and draws, but also to make sure that it bounces back from any surprise.

In this Data Rant a poor streak is defined as two or more consecutive draws or losses and its length as the number of these consecutive draws or losses. In figure 2, below, there is stunning correlation between total points and the aggregate of poor streaks. It is also totally predictable – the team that loses and draws the least will win the title.

Still, this result warrants some discussion – consider last season’s champions Manchester City. The club had two poor streaks during the campaign, each with a loss and a draw. Apart from those runs, City had four ‘isolated’ losses and three ‘isolated’ draws – more lost points in isolation than through poor form.

Figure-2-Points-vs-Aggregate-Poor-Streaks

Liverpool, the runners-up, narrowly missed the title in 2013/14 with only one more draw than City, but the Merseysiders had two more poor streaks than the Etihad side. Arguably, resilience won the title for City.

In figure 3 we look directly at the relationship between points and the average length of poor streaks. At first, there is little to speak of, but a trend emerges, in Figure 4, when we remove the outliers such as Reading.

Figure-3-Points-vs-Ave-Length-of-Streak

 

Figure-4-Points-vs-Ave-Length-of-Streak-Outliers

Notice that a seemingly clear outlier in Chelsea, 12/13, has not been removed. The London side shall serve as a litmus test, of sorts, to determine if either the length or number of poor streaks has a discernible effect on the final points tally. The trend is much stronger in the latter.

Figure-5-Points-vs-Number-of-streaks

Chelsea managed to reduce the average length of a poor streak length by 56 per cent over the two seasons yet only gained a 9.3 per cent increase in total points. In short, a team may have a poor patch of form – even a pretty severe one – but it’s fine as long as the team can get back to winning ways and stay on that path.

Chelsea-points

In 2012/13 and 2013/14 the top four sides, at most, suffered five poor streaks (Arsenal 12/13). United has already totted up one such streak this season and is on the cusp of another. It is worth noting that Arsenal beat Tottenham to fourth in 2012/13 by one point, while Liverpool, 2nd, and United, 7th, both suffered four poor streaks. Given that, Van Gaal can afford only two more bad runs this season. A title run remains feasible – though there is little room for mistakes . Then again, title runs tend to be that way.

All data from Squawka
Assumptions dictating linear regression have not been held strict

Data Rant: Danny Welbeck

September 3, 2014 Tags: , Data 11 comments
featured image

Danny Welbeck’s departure to Arsenal is peculiar in the sense that the London club was arguably in greater need of Radamel Falcao than Manchester United. As is, United has upgraded the forward line while Arsenal has acquired a Premier League quality forward to weather the loss of Olivier Giroud. Win-win, perhaps, but Welbeck’s transfer is far more nuanced than squad betterment alone. It leaves the question of just how successful Welbeck can be at Arsenal?

A local lad leaving his boyhood club has understandably stoked an emotional response from many Reds. On the pitch, though, the England international has divided opinions. Welbeck’s athleticism is obvious for all, but his technique and, in particular, finishing have been often questioned – the former United striker has never broken the 20 goal-a-season barrier.

There is a popular counter argument though: that Welbeck has often been shunted into roles that are not his best, either deep or out wide, to “do a job.” Welbeck will, goes the argument, fulfil his goalscoring potential once he is given a run as the main forward. Daniel Sturridge has flourished since his move to Liverpool and has often been cited as an example supporting this thesis.

Welbeck, at least until Giroud returns from injury, will be deployed up-front at the Emirates – the Longsight-born striker will have a solid series of games in his favoured position, injuries permitting. It is a role he may keep should goals come for Arsenal.

Yet, Wayne Rooney’s rapid decline and Robin van Persie’s age may also necessitate reinforcement at United beyond Falcao in the coming years. With neither Rooney nor van Persie always injury-free, Welbeck’s sale could very well go down as a strategic mistake should the Englishman become the forward some believe him to be.

Data Rant looks to statistical analysis to see if this scenario is possible. We have used 2013/14 season data and look at forwards who scored more than 10 goals in the Premier League.

First, we investigate whether number of shots per match has any relation to goals scored. After all, Welbeck’s stats should improved as a regular first team starter as he will have more opportunities to strike at goal.

Figure-1

Figure 1, above, makes perfect sense – more shots equals more goals, although the question of why remains. Next, we look if shot accuracy has any relationship with goals scored. After all, more shots on target should lead to more goals.

Figure-2

In figure 2 there is no trend to speak of. This also makes sense – after all, one goal in 10 shots on target and one goal after nine off-target means essentially the same thing. In figure 3, below, we look at correlation between shot conversion rate and goals scored.

Figure-3

Again there is little correlation. If we look deeper at the data there is little variation in accuracy or conversion (Emmanuel Adebayor’s freakish 33.3 per cent conversion rate aside). Does this mean that there is little to finishing apart from getting into the box and taking a chance? Figure 4 demonstrates that this theory does not hold up.

Figure-4

So far we have neglected the role of provider: the team must do just as much work to feed the striker for a goal to be scored. To see how important chance creation is in goals scored we look at number of assists each striker’s team recorded in 2013/14.

Figure-5

In figure 5, above, there is a fairly strong relationship, which is obvious. What isn’t obvious is the fact that the number of goals scored is almost purely dependent upon the number of shots attempted. Amongst the élite, there seems to be little difference in ‘finishing’ and scoring principally comes down to chance creation and frequency of goal attempts.

One thing that the data, at least the publicly available kind, doesn’t capture is movement. But if off-the-ball movement was so important then the role of creator would be greater than the data appears to demonstrate – after all, off-the-ball runs have to be found.

Figure-6

In figure 6 there is a very strong relationship between take on attempts per game and goals. More important than having creative teammates is a striker’s ability to create his own chances.

The data indicates that finishing as a technique is overrated. Instead a striker should be judged by his ability to take as many shots as possible. That ability is predicated somewhat by his team’s creative prowess, but more importantly by a striker’s ability to beat his marker on his own. This conjecture is supported by the struggles that some ‘fox-in-the-box’ players, such as Javier Hernández, face in modern game.

It is difficult to predict Welbeck’s future given that he has played out of position for much of his time at United. Welbeck’s conversion rate of 25 per cent and shot accuracy of 53 per cent do suggest that enough raw finishing ability is available to the Gunners. What is clear from the analysis, however, is that he must shoot and dribble more often. This should come with more game time and greater confidence. There is every chance that United will come to regret the move.

All data from Squawka
Assumptions dictating linear regression have not been held strict

Data Rant: box-to-box midfielders

July 12, 2014 Tags: Data 15 comments
featured image

The soon-to-depart midfielder Anderson was converted from a typical advanced playmaker to an ‘eight’ by Sir Alex Ferguson – the rationale being that the Brazilian has all the attributes of a great box-to-box midfielder. Anderson has not lived up to a promising first season at Old Trafford after joining in 2007, leaving Manchester United short of attacking thrust from central midfield for the past seven years.

Ferguson could afford to get by on a functional midfield partly due to Michael Carrick and Paul Scholes’ excellent passing, but mostly due to a reliance on high-flying wingers. Cristiano Ronaldo’s sheer explosiveness helped mask a pedestrian midfield, while Antonio Valencia, for a while at least, consistently provided dangerous crosses from the right.

That Valencia has suffered two poor seasons in a row, while Ashley Young and Nani face an uncertain future at Old Trafford, says something about the degradation in that department. Luke Shaw and Rafael are exciting young fullbacks, but they are limited by their position on the pitch. Yet, having thoroughly figured out 5-3-2 at the World Cup, Louis Van Gaal might even do away with wingers entirely if he is to get the best out of his United squad.

Central midfielders, therefore, must break the lines to provide attacking impetus from midfield; Ander Herrera has been recruited for that task. The question of whether United actually needs such a player remains though. After all, direct football, such as that employed by Netherlands in Brazil this summer, tends to bypass the middle entirely.

In this Data Rant we look at the best dribbler from each Premier League side’s central midfield then see if their skills have any correlation with points the player’s team earned last season. In figure 1, below, there seems to be little relationship between the two. That is, having a good dribbler in the engine room is not strongly related to winning games.

Figure-1: Dribbles vs Points

Notice, however, that six points are clearly distinguishable from the rest in upper-middle areas – those belong to the top six sides. With these clubs removed, in figure 2, the correlation is stronger. In addition, United is now an outlier – the Reds are the best of the bad bunch and the worst of the good.

Figure-2: Dribbles vs Points

The fact that United’s best midfield dribbler is Tom Cleverley, and that the Reds had the worst such player in the Premier League is, in a sense, interesting. From a purely box-to-box point of view, the Reds should have ranked lower in the table. Put simply, there are other factors that make up for the lack of an all action midfielder.

In figure 3, below, the top four lie above the blue line. It is that line van Gaal’s United must aim for in the coming season. Notice the horizontal spread between these clubs. There are significant gaps between City, Chelsea, Liverpool and Arsenal’s best midfield runners, despite little differences in points. This is another piece of evidence that on-the-ball running is not an important a factor in winning football matches.

Figure-3:  Dribbles vs Points

This conclusion runs against intuition, of course. When a skilful midfielder is pinned down, he can work his way out of trouble, enabling his team to create an overload or attacking opportunity. Yaya Touré, for example, would not have scored twenty goals had he not been an exceptional dribbler. Meanwhile, stronger sides have to regularly break down defiant defenses. While holding on to that ball could also earn the team that last-minute set piece, it is relevant to note that any player with the ball can accomplish this feat.

Finally, in figure 4, there is a significant relationship between points and number of take-ons attempted by each Premier League club. United is about 200 take-ons short of the top four.

Figure-4:  Take Ons vs Points

Juan Mata chalked up 52 in his last full season at Chelsea, while Herrera supplied 73 for Athletic in 2013/14. Provided the pair performs to that level in the coming season United should dribble its way into a stronger position in the Premier League table.

This bodes well for Ander’s impact at Old Trafford. United was deficient in many things last season, but at the very least, David Moyes was short an Ander Herrera.

All data from Squawka
A brief note on methodology:
1) All categories are weighted equally
2) Each figure has been adjusted relative to the ‘best’ in each category
3) Assumptions dictating linear regression have not been held strict