We have completed three weeks in the small open online course (SOOC), Observing and Analysing Performance in Sport.

It has been a relatively quiet week for online exchanges. It has given me time to reflect on the format of the SOOC and explore ideas with my fellow facilitators.

I have written daily wraps for the course.

We have 477 enrollments for the course. I am hopeful that the asynchronous, non-linear format of the SOOC enables enrollments at any time. I feel that there is nothing to miss and everything to gain through enrollment.

The quiet week has led me to think about peripheral participation and how an invisible community of practice can flourish.

I am clear that we are a SOOC rather than a massive open online course (MOOC). Following Alexis Madrigal’s lead I have been thinking about how SOOCs can mobilise the power of the dark social to grow their connections.

I have been reflecting too on the role of facilitators in SOOCs . I am extremely fortunate to have five colleagues who have shared the workload of the first three weeks of the course. This has made it possible to have a twenty-four hour service should anyone have teachable or sharing moments.

I am keen to extend the SOOC exchanges beyond a single language and during the week have been thinking about the Tower of Babel and an alternative … polylingual diversity … with nodes of sharing. The visitors to the SOOC from eighty countries could make this possible.

Photo Credit

David asks for directions

Position: 12th

Played: 12   Won: 3   Drawn: 5   Lost: 4   Goals Scored: 13   Goals Conceded: 17

Top Goalscorer: Demba Ba (8)

Form (Last 6): LDWDLL

Formation for last game (against Swansea)

INDIVIDUALS

GOALKEEPER

DEFENDER

Williamson has made the most aerial duels and clearances in the Newcastle team. 

• He also has the highest win % in the Newcastle team for Ground Duels (82%).
• Rickie Lambert will have a tough game up against Williamson, with a Centre Back capable of winning the ball in the air and on the ground.
• Other Players: Cabaye has made the most interceptions in the team with (25). However, he is injured. Gutierrez has made (18), with both Tiote and Santon on (15) interceptions.

MIDFIELD

• Ben Arfa is the most creative Newcastle player, creating (24) chances for his teammates. He also has (2) assists and (2) goals to his name.
• Out of the (30) shots on goal (22) have been with his left foot. Therefore, can Saints try and force him onto his right foot which will help minimize the chance of him scoring.
• Ben Arfa also has the most successful final third pass with (66%) pass accuracy, this shows he’s looking for the penetrative ball into the front men. Notably, 26% of his passes are played backwards and only 22% are played forward.
• He has also had the most ground duels in the team (201). But with a 46% success rate shows he’s not the strongest. He has also lost the most possession for Newcastle (243) and been dispossessed (40) times which is also the most. Therefore, if Saints can get close enough to force a mistake to make him think twice about a decision this should help win back possession of the ball. However, if you give him enough room he will take you on (105 dribbles) and with a 58% success rate he will most likely get past you.

ATTACKER

• Ba is very much an all-round centre forward, capable of taking players on, winning aerial balls and dangerous with the ball inside the penalty box.
• Ba has had (50) shots on goal with a (20%) conversion rate. (Goal Chart Below)
• Interesting enough out of his (265) passes (32%) are played to his right. Therefore, he is the player most likely to get Ben Arfa on the ball. If possible, one of the Saints midfielders needs to position themselves in a way where they can’t get the ball out wide to Ben Arfa; therefore forcing the ball out the other way or losing possession.

• Looking at the Demba Ba Goal Chart, we can clearly see the bottom left corner is where Ba likes to put it and where he’s become successful at scoring from.

TEAM

• Looking at the team statistics we can see it’s fairly similar in aerial and ground duels.
• Saints are known for winning back possession by interceptions, therefore they just need to carry on as they’ve done before.
• Interesting enough the Newcastle defence has been dribbled past (72) times which is a lot more than Saints (55). Therefore, players such as Adam Lallana, Gaston Ramirez and Jason Puncheon should look to take on the Newcastle defence. Saints could even look at getting the full backs on the overlap which could force a 2 v 1 situation on the Newcastle full backs.

• The attacking stats are very similar, if we break down the shots stats, Newcastle on average has a shot every 8 minutes, whereas Saints have a shot every 10 minutes. Both teams have a shooting accuracy of (39%). However, Saints are more clinical in front of goal with a (14%) shot conversion rate, whereas Newcastle is (10%)
• One of the stand out stats is the amount of dribbles attempted by Newcastle (228), this shows us their attacking style of play were they like to break forward at speed and take on players. Therefore,  Lallana and Puncheon will need to carry on their form from the QPR game where they were excellent at recovery runs supporting the full backs and forcing the opposition to a mistake or making them play inside to stifle there attempted attack.

• Notably the direction of passes are very similar, with Saints playing (5) passes every minute shows they play at a higher temp than their opponents. As some of you are aware I did my last article on ‘Saints Direct Approach’ where Saints have become more direct in the last few games. However, that is nothing compared to Newcastle massive amount of (791) long balls so far this season. Expect long balls onto the front man.

•  One of Nigel Adkins key points is possession of the ball. If you don’t have the ball you can’t attack the opposition; therefore you need to look after it and be productive with it. Saints don’t give the ball away a lot and don’t get dispossessed as much as some other teams in the Premier League. Notably, when Saints don’t have the ball they press the opposition forcing them into mistakes, last week was a prime example when Schneiderlin and Cork worked well as a pairing winning the ball back countless of times.

More Analysis!!

As we can see Newcastle tend to concede goals during the middle period of the first half, (4) goals in a period between 11-30 minutes. Usually this is the point when the game has settled down from a frantic start; therefore this shows that they may settle down a bit too much causing them to be careless on the ball which has led to more shots on goal. Notably, they concede a lot of goals late on in the game.

However, if we look at goals scored..

They have scored 5/13 goals in the last 10 minutes of the game. Therefore, Saints need to make sure they are fully concentrating limiting the amount of chances that they offer Newcastle. I thought Steven Davis was excellent when he joined the match late on against QPR, hassling the players on the ball forcing errors and winning back possession, depending on the score line I expect the same sort of impact.

RED- GOALS SCORED

BLUE- GOALS CONCEDED

Above is a line graph showing when Newcastle score and concede.

Out of all the reports I’ve done so far, Newcastle United are the first team to show a real weakness of conceding goals from a header; 6/17 goals coming off the head. Therefore, Saints from set pieces and open play should look to get balls into the box looking for key headerers of the ball, such as Rickie Lambert and Jose Fonte.

On the other hand, the majority of the goals scored have come from a shot. 12/13 shots came from within the Penalty Area. Both headers scored were from Demba Ba, which shows despite the amount of goals conceded from the head, they can still score if the ball is put in the right areas.

Key Points

• Hatem Ben Arfa is their most creative player, take him as much out of the game as possible and the chances Newcastle will create will be halved.

• Newcastle are in a tough patch at the moment with results not going their way, they have also lost one of their key players in Cabaye. They are also playing in the Europa League; so on paper this is a great time to play them. However, as we all know we can never underestimate the opposition.

• Due to the amount of long balls played by Newcastle and the person receiving it being good in the air, Saints should look to force pressure on the goalkeeper or defenders forcing them to lose accuracy in the kick long. Therefore, giving Saints a better chance of retaining possession of the ball.

ENJOY THE GAME!!

Source: EPL Index (stats)

The abstracted infrastructure, intrinsic feature of cloud computing, has the side effect that customers have no way to truly know the type of computers their applications are running on. Due to the absence of successful mechanisms to enforce performance isolation the performance measurement of an instance many not be constant nor reproducible. There is a need for some sort of agreement between the cloud provider and the customer ensuring that the price paid for a specific amount of resources will lead to a performance that fulfills the user’s expectations, rather than availability of service, referring to the amount of time the service is up and responsive. The author suggests the future SLAs should take on the problem of specifying performance, in such a way that the customer the knows limits between which its application performance may vary and reference parameters such as throughput, response time, and availability should be used for designing the performance SLAs.

REFERENCES:

Victor Delgado. Exploring the limits of cloud computing. Masters thesis, KTH Information and Communication Technology, Stockholm, Sweden, October 2010.

PRESENTATION: Exploring the Limits of Cloud Computing

REFERENCE:

Lei Wang; Jianfeng Zhan; Weisong Shi; Yi Liang; , “In Cloud, Can Scientific Communities Benefit from the Economies of Scale?,” Parallel and Distributed Systems, IEEE Transactions on, vol.23, no.2, pp.296-303, Feb. 2012. Download

PRESENTATION: In Cloud, Can Scientific Communities Benefit from the Economies of Scale?

We started with 279 enrollments on the eve of the course and at the end of Day 14 we had 420 enrolled.

I have been writing daily wraps for the course. They can be found here. I have found it remarkable that for each day’s wrap, by the time I had completed writing we had new enrollments.

My aim has been to provide a summary of activity during the day for very busy performance analysts. I have started to add some additional discussion points too.

My experience of the course has been one of fascination. I have really enjoyed the pathways participants have taken through this non-linear course. I have been delighted that there is a willingness to share and explore ideas.

We have four groups of Module facilitators. Each of whom has engaged with participants during the first two weeks. Information about the facilitators can be found in this Welcome post. All activities have been overseen by Adam Brimo at OpenLearning.

We have had lots of discussions about Observing Performance. There was some exciting discussions about Augmented Reality in the first week.

Our experiment with Open Badges is on-going. I see one of my roles as being an advocate for these badges. Adam Brimo has made Karma Open Badges available. These Karma Badges recognise contributions to discussion forums.

I am looking forward to week three in the course. My colleagues, Darrell Cobner and Adam Cullinane, at Cardiff Met facilitate discussions overnight Australia time. Here in Australia Mark Upton has worked with me during European night time.

Photo Credit

Banksy? (Markus Trapp, CC BY-NC-SA 2.0)

At the beginning of the season Saints received many plaudits for the way they played their football. However, it’s only worth playing good football if you are picking up points at the same time. One thing I’ve noticed over the last few games Saints have tended to go more direct to the front man; whilst playing the long ball more often they are still playing their fast attacking inter-play across the ground which is a pleasure to watch at times. I’ve put together a few graphs showing whether playing the long ball is creating more chances and shots on goal for Saints.

Looking at the graph above, we can clearly see a rise in long balls over the games being played this Premier League season. Notably, we beat Aston Villa quiet easily and in doing so we played 58 long balls, the 2^nd most amount of long balls we’ve played this season. The last few games we have certainly played a lot better as a team, I also feel the reason why we’ve been more direct is because Rickie Lambert has become more isolated when we’ve tried playing a high tempo passing across the ground. By playing at a high tempo across the ground will also lead to teams looking to intercept the ball to regain possession and due to Saints now playing against World Class players as soon as they intercept passes the chances are the ball will end up in the net. Therefore, I believe by mixing up our style of play we become less predictable in what we will do with the ball.

The graph above shows the Shots on Goal Saints have had this season (includes Shots on Target, Off Target and Blocked Shots). Looking at the graph it’s harder to tell if playing the long ball has had an effect on shots on goal. However, the Swansea and QPR game we can certainly see an increase in shots on goal (16 and 20). Saints first win of the season against Aston Villa also saw 16 shots on goal. So maybe this more direct approach is suiting Saints attacking style of play.

Looking at this graph it is nice to see Saints are creating a lot of chances for themselves, if we were losing games and not creating chances I would be a lot worried but this is very much a positive part of Saints game. Similar to the Shots graph there is a great progression in the last few games against Swansea and QPR. Realistically, I have probably done this report a little early, but nonetheless we can look back in a few games time to see if this progression is still on the way up and whether this change in style of play has worked.

The line graph above shows all three graphs put together.

Red Line- Long Passes

Blue Line- Shots on Goal

Green Line- Chances Created

This graph reiterates what I have previously mentioned. But looking at this graph we can see more clearly that the long ball has had an influence in improving our creativity and shots on goal.

Finally, I am now going to look at the man who receives most of the balls in the air. Rickie Lambert.

Blue- Shots on Goal

Red- Chances Created

One of Rickie’s great strengths is holding the ball up and playing others in, if we look at the Fulham game he may have had a few shots on goal but he didn’t create anything, in that game Saints played only 40 long balls which was their 2^nd least amount of long balls so far this season. As a result he was substituted in the 67^th minute. However, if we look at the last few games against Swansea and QPR where we’ve seen a rise in long balls he has created a total of 9 chances as well as having 5 shots on goal where 1 went in. Whenever I see Rickie at the far post when a ball has been knocked in there is a great sense of belief that something will happen, with him either going for goal or knocking it back to players such as Adam Lallana, Gaston Ramirez and Jason Puncheon. This worked very well in yesterday’s match against QPR.

Overall, this shift in playing style has certainly helped Saints perform better on the pitch. It’s helped get more influential players on the ball and in doing so has meant Saints are creating more and in turn that is providing more shots on goal. Nigel Adkins has done well in changing this, but saying all this he will no doubt want to be playing more of the ball on the ground in the future as Saints look to play the Southampton Way.

Sources: EPL Index (stats)

Merilyn Childs was one of the speakers at the conference. Her talk was titled Not business as usual: MOOCs, Badges, OERs & global personal learning activism. Merilyn has a copy of her presentation on SlideShare.

My conversation with Alan about the conference was one of the prompts I received during the day about performance analysis and connected communities.

• I started the day looking at and writing about what had happened overnight in #OAPS101.
• I followed up on a conversation about coach development with a colleague from a sports institute.
• I had the good fortune to participate in the Performance Analysis unit at the University and was able to introduce Alexis Lebedew to the group.
• Alexis told me about his brother’s Facebook page At Home on the Court and Huy’s  Volleyball Blog and Newsletter
• I exchanged some ideas about observation with a coach.
• I met with a colleague from the Australian Paralympic Committee to discuss an open source approach to sharing resources.
• Ended the day at the University discussing Learning and Teaching in the context of  #OAPS101 and racing over to see the end of a football coaching session to talk with a coach.

The day underscored for me the diversity and delight of being involved in performance analysis.

Positon: 20^th

Played: 11   Won: 0   Drawn: 4   Lost: 7   Goals Scored: 8   Goals Conceded: 20

Top Goalscorer: Bobby Zamora (3)

Formation for last game (against Stoke City)

Note: Ryan Nelson pulled out of International Friendly with Injury, may not feature. Park Ji-Sung shouldn’t be back from injury in time for Saints game.

INDIVIDUALS

Attacking: Due to QPR attacking style of play in theory they have a very potent attack. Junior Hoilett, Esteban Granero and Djbril Cisse all players that like to have a pop on goal; however none of them can match the trigger happy feet of Adel Taarabt who has had the (30) shots on goal, (10) on target. Notably 27 of them have been on his right foot; therefore if Saints can we must try and force him onto his left foot. Interesting that so far this season (5) QPR players have missed two clear cut chances each. We should not take them for granted despite their poor start to the season; they are still very talented footballers.

Defensive and Discipline: Like Saints, QPR have also let their fair share of goals in. Looking at some stats Estaban Granero has a tackle win percentage of 91% and also has gone in for the most ground duels (114= 47% win). The defence itself has the centre back partnership of Ryan Nelson and Anton Ferdinand who have won 61% and 62% of their aerial duels. QPR have made (3) errors which have led to goals, it’s up to Saints to force that mistake by pressuring the QPR backline. Granero pops up again but for his discipline this time, having been booked (4) times this season. He has also won (22) fouls but also lost (24).

Creativity: Esteban Grenero is obviously central to how QPR play, this time creating the most chances for the team (23). Of these (23), (12) of them have been from open play and (10) from set-pieces. Adel Taarabt has also created (19) chances but (17) of them from open play. Also notable creators are Park Ji-Sung (14) and Jose Bosingwa at fullback with (6), this shows us that they like to get their fullbacks forward looking to get involved with as many attacks as possible. Due, to this players like Adam Lallana and Jason Puncheon will need to make sure they are supporting their fullbacks in case of a QPR overlap. Junior Hoilett has attempted the most dribbles with (33) but with a success rate of 39%. However, there are 8 players in the QPR team with over (10) attempts at dribbling. This further backs up there attacking style of play where they are looking to take players on.

Passing and Possession: Diakite has a pass completion rate of 90% which is the highest of any of his fellow teammates. Adel Taarabt has made (113) successful final third passes; this shows us he’s looking to penetrate the opposition defences by playing in their fast attackers in Cisse and Hoilett. Taarabt has also successfully made (2/3) accurate through balls, with these balls being played in behind Saints centre backs Jose Fonte and Maya Yoshida need to be aware of the forward runs that the attackers are making, the one thing QPR aren’t lacking in is pace. So the centre backs are going to have to think smart if we are looking to prevent QPR from receiving these balls. The centre backs for QPR, have made the most interceptions with 37 in total between Nelson and Ferdinand. So if Rickie is playing up top, he’s going to have to make sure he meets the ball instead of the ball meeting him, otherwise Saints will lose possession very quickly which could lead to a QPR counter attack.

TEAM

The table above shows us a few interesting things, the physicality of the QPR team shows us they win a lot of the ball on the ground, however in the air they seem to lose the majority of balls. The interception stat shows us that Saints best bet in winning back possession is through intercepting slack passes from the QPR team.

I found these stats quiet interesting, in a way it shows us why both teams are at the foot of the Premier League. Both teams losing possession the same amount, also nearly the same amount won back in the defensive third and if we look at the pass completion rate it’s pretty much the same as well. However, Saints have won more in the midfield and attacking third, if they can keep that up for the match this weekend this could be the deciding factor in who wins the tie.

As mentioned before in this report, I said how QPR like to attack with pace and put the opposition under pressure. When you look at Saints attacking line-up of Lallana, Ramirez, Puncheon and Lambert they are all capable of creating something out of nothing, but I wouldn’t say they have the natural pace to beat a player. Instead they think smart and find the space in order to get time on the ball and look to drive on. However, they aren’t like QPR who have Cisse, Hoilett and Taarabt who all have pace to burn. The stats back this up with an incredible amount of dribbles so far this season, but it also shows if they are forced out wide they don’t have a very good crossing accuracy, whereas Saints are very dangerous when a ball is put in the box. Possible something the team could exploit.

Again very interesting how both teams have had around the same amount of shots, but it shows the difference in how clinical Saints have been compared to their rivals who have only mustered up a mere 7% chance conversion rate which is never going to be enough to survive in the Premier League. Not saying Saints is enough for survival because it probably isn’t, but I feel more confident in our attacking players than QPR attack.

More Analysis!!!

The graph above shows us when QPR are conceding the majority of their goals, notable trends is the first 20 minutes of the game. In that period they have conceded (6) goals which shows us that they start very slowly. If we can get an early goal this could replicate the frustration that QPR have suffered so far this season. They also tend to concede goals in the final 10 minutes of the game, (4) goals. This again will prove to be an edgy time for the home fans if both teams are drawing or winning by 1. Hopefully, their edginess can transfer onto the players performance.

This pie chart shows us how their goals are being conceded, with the majority coming from shots on goal (14). Out of the (20) goals they have conceded (18) of them have been from inside the penalty box with just the solitary (2) from outside the box.

Looking at when they score their goals, (4) of them come just before and after half-time. Saints concede a few of their goals before half-time so they will need to keep working hard to make sure that doesn’t happen.

I find this particular interesting, the fact that out of the (8) goals (3) have come from outside the penalty box. Therefore, Saints need to make sure they close down the space in order to prevent them from shooting.

Looking at this we can see that QPR receive a lot of bookings, notably receiving (8) cards in the last 10 minutes of the game. Interesting enough they have received (6) cards in the 51- 60 minute stage in the game, if we look at the Goals Conceded we can see they have conceded (3) goals in that period. Therefore, I put it down to frustration because the goals scored have effectively killed off any comeback QPR might thought they had.

Of these cautions, 18 were Yellow Cards (including one 2^nd Yellow) and 2 were red cards. As I have just mentioned I felt that QPR get frustrated easily and you can see with 10 cards for Unsporting Behaviour. Therefore, Saints need to get under QPR skin and force them to lose their cool and by doing that the best way is to score a goal. I also believe this is one of the main reasons why QPR haven’t achieved what they have set out to do due to the discipline of the side.

Key Points

• Esteban Granero is the key to the team. Tackling, Passing, Dribbling and Shooting all come with impressive stats. Morgan Schneiderlin or Jack Cork will have to be on their game to prevent him from having an impact on QPR style of play.

• Counter Attack: I can see QPR will look to exploit Saints on the counter with their pace. Attacking players will have to make sure they get back and support the Saints defence.

• Discipline is QPR major weakness. Keeping hold of possession and scoring goals has led the QPR team to become frustrated in past games.

• Do NOT underestimate this team, they have struggled and their morale will be low but on paper this team should be a lot further up the table.

Enjoy The Game!!

Source: EPL Index (stats)

You may not have heard of the term Sabermetrics, you may have heard of the term Moneyball, or the book, or the Oscar nominated movie with the same name. If you haven’t it is not surprising as it deals with the typically American obsession of Baseball and Baseball statistics. It is a shame that it is not that well known on this side of the Atlantic as it does raise a question that would strike fear into the heart of every CIO…..are you using the right yardstick to measure your business performance?

Back in 2002, it was considered in the world of Baseball the team that spends the most wins the most. At that time arguably the most famous Baseball team in the world, the New York Yankees was spending over $130 million a year on the team players salaries. By the normal measures of that day the most valuable players were the best performing, and yet they were out performed by the Oakland Athletics team with spend of only $40 million. What happened at Oakland in 2002 changed baseball forever and many of the assumptions of baseball player performance.

In 2001/2 Billy Beane the manager of Oakland Athletics could not spend the same money as some of the bigger teams, and lost some of his most valuable players, who transferred to other teams. Restricted by budget his limitation spawned some innovation, and a large chuck of revelation.

At the time players were measured on ‘Batting Averages’, ‘Speed’, ‘Home Runs’ and other more intangible factors. What Billy and his staff found, was that these measures were not a true indicator of player performance, they were 19^th century statistic measures left over from when the game was first created. What he discovered was that players that had been perceived as less valuable as ‘They throw funny’, ‘Can’t run well’, ‘He can field, but can’t bat’, was that their underlying data showed that they ‘Get on base’ or in more English terms ‘Completed a run’.

With this insight and information he put together a team of what he considered performers, but for a far lower cost. Suffice to say his reputation was questioned and the fans shouted and media speculated wildly that Oakland would fail, with its team of has-beens, second and third choices.

The results were not immediately obvious and Oakland Athletics did not start well and it seemed that the data was proving false, but then after a historic run of 20 games undefeated, beating a record held since the 1930’s, then winning the America League West division with a record of 103 wins and 59 losses, one or two people started to take notice. Most dismissed it as fluke, Billy had been just plain lucky.

At the end of the 2002 season, Billy was approached by the Boston Red Sox, who saw what Billy was doing and saw that his analytical, evidence-based model was the future of the game. They offered him a salary of $12.5 million dollars which would have made him the highest paid manager in history. ‘Would’ is the correct word, he turned it down and stayed with Oakland, maybe showing that loyalty does have a price.

Two years later after embracing the method and analysis Billy had used, the Boston Red Sox won their first world series since 1918.

Since then as other teams have started using the Moneyball model, the advantage that the first teams who used the model has reduced. Unlike the business world, in the world of Baseball the player statistics are freely available to everyone, causing a democratic levelling of information. The race is now to a better model, a better analysis of the data to predict performance. Player talent scouting is now more analytical, and moved away from the old measures, and the old instincts.

What we should take from this tale is when turning data into information, we have to consider carefully how to define the measure of performance, and if we are using out-dated or outmoded concepts to drive our business.

Could your company be unknowingly competing with another, perhaps with their very own own Billy Beane?

You may have already read our posts on our work with the English Institute of sport (the organisation responsible to the training of the British Olympic team of Athletes), the principles of which are very similar to that of system applied by Billy Beane. Whereby, in implementing a BI solution to help improve their existing Athletes’ Services Database, EIS subsequently became equipped with the skills, technologies, applications and practices to help its business acquire a better understanding of its resources, whilst also providing the organisation with the ability to target and improve the performance of international athletes. As well as providing historical, current, and predictive views of business operations, EIS now has a platform to allow better decision making.

As Jon discusses, the principles of data analysis and effective decision making stretch way beyond the realms of sport, and as competition is a major driver for success in sport, so it is in business. Successful information management strategy, through the adoption of effective tools to assist in the critical decision making process is a must if business’s are to maintain a competitive edge and ensue success. 

For more information on our work with EIS or any of our other projects and solutions, please do not hesitate to get in touch.

(Vocaroo Summary of this post)

I have really enjoyed Week 1 of the small open online course Observing and Analysing Performance in Sport.

There have been some fascinating exchanges particularly about Seeing and Observing and Decision Making.

The numbers of participants enrolled in the course has increased this week. Shortly before the course started we had 155 and now it is 374. This means it is still a small open online course. I have discussed the approach taken in the course in a number of Clyde Street posts and in this Day 1 post on OpenLearning.

I have been keen to offer Open Badges for the course.

Conversations in the first week have prompted me to think about:

• Feedforward
• Performances of Understanding
• Personal Learning Environments

Will Oldham’s post Analysts as Educators acted as a catalyst for my thoughts. In a post that synthesises a variety of ideas, Will concludes that:

We must take be confident enough in our skills and abilities that we are able to provide assurance to those who require it that we can add value to established coaching setups and processes, that we’re not in the business of coaching revolutions, but simply the development of athletes and coaches.

I commented on his post and suggest that the value we add is as educational technologists. I should have added that wherever possible bring an interdisciplinary understanding to performance. I think this requires a sensitivity to a narrative of performance that is customised to athletes and coaches.

My three big issues for the first week:

Feedforward

What if performance analysts decide to share the world as it might be?

I think feedfoward gives us the opportunity to do this. In Peter Dowrick’s words “subjects see themselves not so much as they were but as they might be“.

Performances of Understanding

Last year I wrote about Sam Stosur’s victory in the US Open Tennis. In that post I noted that “I am fascinated by the process by which athletes prepare to perform. I am fascinated too by the realisation of the readiness to perform in actual performance”. My ideas about performances of understanding help me think about this readiness.

A decade ago I followed a Harvard University course online, Teaching for Understanding Using New Technologies. In that course performances of understanding were important indicators. Such performances:

… require students to go beyond the information given to create something new by reshaping, expanding, extrapolating from, applying, and building on what they already know. The best performances of understanding help students both develop demonstrate their understanding.

Personal Learning Environments

I am hopeful that many of the participants in the Observing and Analysing Performance in Sport course will write about their experiences as performance analysts or their interests in performance analysis. There is so much experience to share.

I am keen to find out about personal learning environments. My participation in the CCK08 course opened my eyes to the benefits of sharing openly. I have written about personal learning environments since that course. I see personalisation as one of the ways to offer a service to athletes and coaches.

Into Week 2

I am looking forward to Week 2 of #OAPS101. I am hopeful that this post might stimulate discussion that goes beyond the content of the course and helps address some second order questions about performance analysis.

Photo Credit

I received the photograph in this blog post from a friend. I have been unable to find a source for it. I am keen to learn whether it is a Creative Commons Licensed image.

Fellow student and analyst Chris Fry raised the point that, during BBC footage over the weekend, live performance analysis was publicised through shots of the Welsh rugby team’s support team, including analyst Rhys Long, frequented millions of screens across the nation. In isolation this can only be seen as a positive for our discipline – positive publicity for the practical applications available to sporting institutions. However, when we consider how often clips of the coaching teams on BBC’s flagship football programme, Match of the Day, includes analysts being actively involved in the discussions and live performance reviews we can only be disappointed. For a sport with such financial clout, football seems to fall short in the level of performance analysis on show.

This isn’t to say that no analysis takes place during these games, the presence of Pro-Zone cameras and operatives alone is enough to counter this point, alongside the increasingly popular MCFC Analytics website, but can we be sure that all clubs are engaging in the discipline to this level and if not why not? Contributions to explain this varied, but a general theme of a cultural difference having some part to play seemed to be present. The “old guard” are seen as less than embracing when it comes to young disciplines, whilst the “young pretenders” aren’t in the positions to be making a significant difference in applied practice on a broader scale. There will always be exceptions to prove the rule, however this seems to be an underlying trend. Whilst I agree that there will always be cultural differences not only from sport to sport but even within individual sports, I struggle to see this as an acceptable reason for a potential lack of analysis.

Recent participation in a Short Open Online Course entitled ‘Observing and Analysing Performance in Sport‘ (#OAPS101 for the tweeters amongst you) has allowed me to be privy to a number of viewpoints from across the globe on many topics, including whether or not technology in performance analysis is still seen as threatening or confusing by coaches. The point was quickly considered likely but not reason enough for analysis to be avoided – it must be down to the analysts and those who understand and appreciate the discipline to educate those who do not. We must work as part of the various coaching and performance communities in order to drive forward the levels of excellence achieved in both.

Sir Clive Woodward is a public advocate of performance analysis in sport

Sir Clive Woodward’s article in this week’s Sunday Times Sport section revealed the level to which the RFU, driven by himself, bought into performance analysis as a discipline over a decade ago, using the publicising of this investment as a tool of intimidation ahead of the 2003 Rugby World Cup, and we all know what happened Down Under. Following success with England Rugby, Woodward attempted to cross the cultural boundaries previously mentioned and develop a successful performance structure at Southampton FC. For whatever reasons comparative levels of success weren’t achieved – whether or not this was down to a reluctance from those ‘upstairs’ to fully buy into the methodology of Woodward remains a mystery but, especially given his subsequent success with the British Olympic Committee, one can only assume this played a part.

Analysts simply can’t afford to expect people to throw jobs at them – the relative youth and public obscurity of the discipline combined with stubborn potential clients mean analysts must be educators as well as practitioners. We must be confident enough in our skills and abilities that we are able to provide assurance to those who require it that we can add value to established coaching setups and processes, that we’re not in the business of coaching revolutions, but simply the development of athletes and coaches.

Next up on my individual analysis is Japanese International Maya Yoshida. It’s been a bit of a bumpy ride for Yoshida since his move from VVV Venlo in the Dutch league. Saints needed a new and improved defender to join the squad, so there was a lot of pressure put on Yoshida from the off.

First up was Arsenal, having only met up with the team a few days earlier he was thrown straight into the squad for the trip up to London. No one would’ve expected him to come on as a first half substitute due to the injury Jos Hooiveld sustained. Nerves were certainly there and due to not knowing how the team play he was at fault for Gervinhos first goal, Maya stepped up believing the other defenders would. However, Gervinho ran in behind Yoshdia who then finished well.

Since then Saints have still conceded far too many goals, this has been mostly down to individual errors and lapses of concentration. Yoshida would’ve hoped for an easier start, but it’s not been nice to him. Switching positions to cover for injury, constant changes to starting line-up and getting used to a new team it has certainly been a learning curve for Maya. Hopefully it will prove to be worthwhile in the future.

Let’s have a look at some stats!

Passing

I believe that his passing accuracy is at a pretty good level; out of all the defenders I would say Yoshida is the most positive. He likes to get the ball forward as soon as possible and with 43% passes played forward that backs that statement up. He has also attempted 2 dribbles of which both have been successful. His passing accuracy in the defensive zone is 91%, attacking zone 68% and in the final third its 59%.

Tackling

Looking at his defensive part of the game, we can see his tackling and ground duels are above average for the league. I would say his aerial duels are a weak point in his game, if the opposition attacker is particularly big he tends to struggle in the air which does prove to be a problem as the 2^nd ball is usually played in behind with fast wingers or strikers looking to pick up on the flick-ons. I was particularly impressed with yesterday’s game against Swansea. I thought he was very good, the majority of the balls crossed were blocked by Maya, and he looked a very calm and organised centre back making the Swansea attack very frustrated. Maya reads the game very well and with 16 interceptions to his name proves that he strives on picking up on slack passes or slow minded attackers. On average he wins a challenge every 10.5 minutes. I know a lot of you are thinking his error cost us 3 points against Swansea, but was it his entire fault? If we break the goal down Jack Cork potentially could’ve played the ball forwards or sideways twice before the ball ended up back to the goalkeeper, Gazzaniga probably should’ve played the ball long instead of giving Maya an awful ball to control. It’s a shame it happened as the whole defence were much improved from previous games, yet football is a results business and the defence have to take responsibility.

Here’s a comparison with fellow defender Jose Fonte.

Comparison

Looking at the defending aspect; Fonte has an excellent tackling win % but his ground duels is particular poor. I believe a lot of this is down to Fontes lack of pace, which will most likely coincide with Yoshida 4 last man challenges. As Maya is quiet quick across the ground.

POSSESSION	MAYA YOSHIDA	JOSE FONTE
Loss of Possession	4	0
Possession Won	47	39
Minutes per Possession Won	15	25
Dispossessed	3	4

Possession stats are very interesting, Maya may have lost more but he wins a lot of the ball back. With the Premier League being one of the most brutal leagues in the world winning back possession is massive, if you don’t have the ball you don’t score. With a 10 minute difference in winning back possession Yoshida is clearly a key man in terms of breaking down the opposition in the defensive third.

 Overall, I believe we are slowly seeing the better side of Maya Yoshida. As every game passes I feel more confident with him in defence, the 2^nd Half against Spurs and 98% of the Swansea game really showed what he’s about. Let’s hope he continues to progress and help Saints propel up the table!

COYR!

Together As One

Sources: EPL Index/ SaintsFC Pics

This morning I enrolled in a small open online course (SOOC), headed by Keith Lyons, amongst other notable contributors to the discipline of Performance Analysis. The course is entirely voluntary, with a 24 hour forum being active throughout the course of a month during which participants are encouraged to engage in and complete five modules. Such an environment lends itself to maximum learning being inextricably linked to maximum engagement simply through it’s very nature – a potentially time expensive commitment.

Establishing the position of a five-person performance analysis team within an established University sports team has proven taxing in a number of ways – logistically and temporally for starters, so why take on further commitments and how can one expect to manage both in a way which detracts from neither?

I’ll address the initial part of this conundrum first – why tax yourself beyond the necessary? Recent discussions on twitter, around performance analysis, have highlighted the disgruntlement with which many graduates, employees and, indeed, employers view unpaid internships currently available for graduates and non-graduates alike. The general consensus amongst those opposed to many of the current vacancies is that the job requires an individual with a sound grounding in the discipline who requires experience alongside a mentoring scheme capable of assisting in the development of individuals from novices through to experienced analysts. The question which must be asked remains that of how much of this ‘sound grounding’ can be gained simply through University studies alone? However much one applies oneself, surely the student who looks beyond the books and lecture theatres is likely to prosper in a jobs market seeking experienced individuals capable of pushing themselves beyond expectations.

So how does one manage additional commitments beyond a course requiring ‘x’ number of independent study hours on top of weekly lectures and assignments? I feel that the answer to this lies in 5 key areas:

1) Applying a scheme of prioritising and acting accordingly

This doesn’t necessarily mean saying unequivocally that A is always more important than B and therefore will always come first – certain aspects of a task or commitment may appear less critical yet be an aspect with which you are least confident, is most time consuming or requires the most independent thought, how does this affect it’s place within your priorities?

2) Understanding your personal strengths & weaknesses in the required areas

Consider conducting a SWOT analysis of yourself – the results may surprise you. They may not. Either way, this provides the basis from which your self-reflection can achieve new depths which, in turn, affects the previous point of prioritising.

3) Appreciating the difference between a weakness and a limit

You may currently be far more comfortable producing data than developing it with data visualisation skills – this is a weakness. Having 2 hours available to develop your data visualisation skills to the level currently held by an experienced analyst is pushing limits – can you understand and appreciate when a situation moves from challenging to out of control and, more importantly, are you able to address it accordingly?

4) Your ability to identify and work within a team

Very rarely is an analyst or student entirely alone in their exploits. Whether it be a head coach, a lecturer or even a peer, there is always a ‘team mate’ available for advice, direction or simply an alternative viewpoint. How many of these are you aware of and how many do you use?

5) The collective strength of you with your support network

Beyond the professional team is the personal support network. It’s vital that individuals are aware of and utilise their support networks efficiently – this may not mean the same thing to you as the analyst sat on the next laptop along. The individuality of a support network is one of it’s key features and something which cannot be overemphasised. Take time to understand and appreciate yours.

So, time management – the ability of an individual to control the amount of time exerted on an activity or group of activities. I hope this gives you some insight into the way in which individuals can overcome seemingly excessive commitments in order to get ahead.

The course goes live later this morning.

I think we have a remarkable month ahead of us.

Starting Out

Introduction

Today is the first day of our small open online course, Observing and Analysing Performance in Sport.

Welcome.

Thank you for enrolling on the course and being part of an open approach to sharing ideas and practices.

This is going to be a wonderful learning journey for me and I hope for you too.

I have been writing about the ideas that underpin the SOOC in my Clyde Street blog. For the duration of this course I am going to use the OpenLearning blog facility as my primary blog. This way I hope to share updates and summaries with you directly.

I have posted a video welcome here. There is a written welcome here too. … and an audio welcome (using Vocaroo). Video, text and audio will be important characteristics of this course.

The Course

The characteristics of this course are:

• It is free.
• It is open.
• It is self-paced.
• It is non-linear.
• It is an introduction to observation and analysis.
• It shares content from a number of authors.
• It offers open badges to acknowledge your involvement and engagement in the course.
• Content is shared under a CC BY 3.0 license unless otherwise indicated (see for example the licenses of the three images used in this post).
• It will flourish with discussion and sharing.
• It will change throughout the course.
• It will remain on the public internet as a resource for discovery and sharing.

This Week: Week 1

We recommend that you:

Look at the Connecting and Sharing resources in the About the Course Module.

Spend some time finding your way around the OpenLearning platform (including the Need Help? page) for this course.

How about …

Visiting the Getting Started Activity.

Looking at the Introduction: Connecting and Sharing.

Then checking out Connectivism.

We realise that this is a personal learning journey for all of us. One of my hopes is that we share these journeys through the open sharing of our personal learning environments. You might find the Personal Learning Environments resource of interest

Modules

This course has five modules. All five of them are available today. Other content will be added subsequently and we will let you know about any changes to content.

Our idea is for you to follow your interests.

We have set aside a month to share and discuss ideas.

I am hoping to monitor discussions and deal with any requests for clarification. Please let me know if I can be of help.

To make this a 24 hour a day experience, Darrell Cobner and his colleagues at Cardiff Met will monitor the course whilst Australia sleeps.

We do see enormous opportunities for conversations and exchange. We do not have any formal webinars planned and we are not using Universal Time stamps for our work.

We realise that what we are offering is fallible and that it is constructed by all of us.

We hope you are going to enjoy the course and I am keen to learn if I can be of any service to you as you work your way through the course. I am just an email away either through OpenLearning or at this email address.

The Whisperings Within The Course

It is fascinating to find other locations where the approaches used in the course are explored.

For example, Uberveillance‘s Ethos statement is:

This website provides the general public, the community and those afforded free and unconditional access to the world wide web, access to information regarding the concept and understandings of Uberveillance. This website provides invited Authors an opportunity to directly contribute to the development of the website whilst maintaining an open dialogue with the general public.

Uberveillance’s objective is to provide open comment that is:

• relevant to the topic – is not selling nor spamming
• conversational – pictures and addresses the audience
• coherent – maintains consistency
• respectful – provides reading enjoyment, challenge and value
• generative – builds trust and connection

The spirit of this  #OAPS101 course is invitational, voluntary and supportive. It aims to build trust and connections through participation. It seeks to do so through public sharing.

Photo Credits

Dawn Platform (Isuru Senevi, CC BY-NC-ND 2.0)

The Road Ahead is Shrouded in Mist (Joysaphine, CC By-NC 2.0)

Share @Mehanata Bulgarian Bar, NYC (o.blaat, CC BY-NC-SA 2.0)

 The flexibility and scalability of commercial cloud infrastructures make them an attractive application migration target, however, due to the associated complexity, it is difficult to make newly migrated applications run efficiently in clouds. For example, while clouds are good candidates for supplementary system platforms during occasional overload of Internet applications (e.g., electronic commerce), reports on Amazon EC2 consistently mention that network latency may affect overall system performance considerably. Such issues are compounded by dependencies among system components as requests are passed among the tiers, which are characteristic of real scientific applications. Despite of the scalability and cost benefits that clouds provide, the tradeoff between guaranteed performance (e.g., bounded response time) and economical efficiency (e.g., high utilization for sustained loads) remains a serious challenge for mission-critical applications.

 Scientific Computing:

 Cloud computing has seen tremendous growth, particularly for small database workloads and commercial web applications. The on-demand “scale by credit” model creates a flexible and cost-effective means to access compute resources. For these reasons, the scientific computing community has shown increasing interest in exploring cloud computing. However, the underlying implementation and performance of clouds are very different from those at traditional supercomputing centers. Scientific computing requires an ever-increasing number of resources to deliver results for ever- growing problem sizes in a reasonable time frame. It is therefore critical to evaluate the performance of High Performance Computing (HPC), High Throughput Computing (HTC) and Many-Task Computing (MTC) applications in today’s cloud environments to understand the tradeoffs inherent in migrating to the cloud.

 The availability of the large, virtualized pools of compute resources that cloud computing offers raises the possibility of a new compute paradigm for scientific research with many advantages. For research groups, cloud computing provides convenient access to reliable, high performance clusters and storage, without the need to purchase and maintain sophisticated hardware. For developers, virtualization allows scientific codes to be optimized and pre-installed on machine images, facilitating control over the computational environment. However, due to multi-tenant cloud architecture, there is always a performance decrease due to co-located virtual machines running resource-intensive tasks. The drop in performance is slight for CPU and memory intensive workload and very significant for disk and network I/O intensive workloads. Also, the variability of the measurements increases dramatically in every case when there is high background load.

Thus, we conclude that though the current cloud computing services they may form a good solution for the scientists who need resources instantly and temporarily, but a significant performance improvement is needed for clouds to be a viable alternative to grids/cluster based scientific computing.

PRESENTATION: Performance Analysis of Cloud for Scientific Computing Workloads

REFERENCES:

A. Iosup, S. Ostermann, N. Yigitbasi, R. Prodan, T. Fahringer, and D. Epema, “Performance Analysis of Cloud Computing Services for Many-Tasks Scientific Computing,” IEEE Trans. Parallel Distrib. Syst., vol. 22, pp. 931–945, June 2011. Download

As of this morning we have 134 participants enrolled. I am delighted that so many people are interested in this small open online course (SOOC).

It is intended to be an exploration in open sharing and connecting. I like the idea that this course will be sustained by intrinsic motivation.

We have tried to make participation as easy as possible. Although nominally the course will run through November, all the content for the course will be available when we go live on 5 November.

We think it will be important for participants to look at the Connecting and Sharing Module first in order to appreciate the functionality of the OpenLearning platform but even this is negotiable. Thereafter the course welcomes non-linear participation in the SOOC.

We have no webinars planned in universal time and I see three ways we will connect and share:

1. Discussion forums

2. Video posts

3. Participation portfolios within or beyond the OpenLearning blog and wiki options

We aim to manage the discussion forum as a 24 hour activity each day. Colleagues here in Australia and in the United Kingdom will monitor the discussions. I aim to follow as much of the discussion as I can throughout the course.

I am hopeful that the month will be a great time for produsers … using and producing resources to share under a Creative Commons license.

We have two Open Badges to offer in this course.

Enrolled

* You enrolled for the course.
* You hose a personal path through the material.
* You submitted a summary statement about your experience of the course.

Participant

* You enrolled for the course.
* You chose a personal path through the material.
* You used the OpenLearning tools to contribute to discussion or to comment (karma).
* You developed an e-portfolio to record your involvement in the course and to reflect on your experience of the course.

We hope the distinction between enrollment and participation is reasonable. We think that the participation mode will enable a community of practice to flourish.

Earlier this month I wrote about the kind of atmosphere that we hoped would pervade the course.

Stephen Downes provided me with a timely reminder about this in one of today’s OLDaily posts. In it he observes:

If you were to review my writing on MOOCs and similar phenomena you would see me most frequently refer to (what we would call) ‘students’ as ‘participants’. The term ‘participant’ to me most accurately represents the relation between MOOC and an individual person – they are not ‘students’ because that implies studying and the master-student relationship, which are antithetical to MOOCs. Nor either are they referred to (much) as ‘learners’, as this suggests that learning is the dominant paradigm at work here. In fact, the logic of MOOCs is not the logic of learning, but rather, of participation, and that’s why I use the word.

This is why we use the term participant too in #OAPS101.

Photo Credit

Fireworks

The course is an introduction to observing and analysing performance in sport and is being hosted by OpenLearning.

We have a team of twelve authors working on content for five topics in this course. You can find out more about the course here.

Collectively the authors of the course recognise that we have an opportunity to explore some engaging content and learn more about connecting with colleagues interested in the observation and analysis of sport.

I have been keeping a scrapbook of links that exemplify our approach to the course. You can find them at SOOC 001.

To date registration for the course has been by word of mouth. This morning there were 110 registered members of the course. I have been blogging about the course for some time but I thought I would use this post as an open invitation to Observing and Analysing Performance in Sport.

Each day I find some reassuring posts about the approach we are taking.

I am hopeful that it will be an event in the way that Stephen Downes and Dave Cormier describe.

It would be delightful if the course exemplified approaches to learning that Will Richardson advocates.

… and Jabiz’s enthusiasm for blogging.

We would be delighted if you would like to join us in November for the course (#OAPS101).

Photo Credit

Bus Stop Crowd

After playing around with the data, I developed the following visualizations graphics that map City’s key passes and attempts. These events are linked, as an attempt always follows a key pass.

The visualisations need no explanation, illustrating where the passes that leads to attempts came from.

Key passes

Attempts

“Beware the left-wing”, that should be the warning that managers of teams playing Manchester City should be giving to their players

The visualisations clearly show City’s preference for attacking from the left half of the pitch. With only three of Manchester City’s 14 shots coming from the right, Bolton’s defenders in that part of the pitch must have enjoyed an easy afternoon.

So the message to teams playing Manchester City is clear – seek to neutralise the Champion’s attacks down the left side of the pitch.”

Position: 5th

Played:8  Won:4  Drawn:2   Lost:2   Goals Scored:15   Goals Conceded:12

Top Goalscorer: Jermain Defoe (5)

Top Assists: Aaron Lennon, Jermain Defoe and Moussa Dembele (2)

Formation vs Chelsea (last match)

Note: Gareth Bale, Emmanuel Adebayor and Moussa Dembele all expected to return for the Saints game.

INDIVIDUALS 

Attacking: Spurs have a very potent attack that can cause problems to all sorts of teams; there top scorer Jermain Defoe on average has a shot every 22.53 minutes with a shot accuracy of 43%. If Saints give him any time or space around the penalty box they will be sure to pay the penalty. With Spurs playing with a front three, they have a very attacking style of play looking to put the opposition under threat by forcing a mistake and catching them on the counter attack. On one side they have Gareth Bale who averages 28.08 mins per shot and has a shot accuracy of 55% which is very high for any player.

Defensive and Discipline: So far this season they have conceded 12 goals which is pretty high for a team of their quality. Both fullbacks in Walker and Vertonghen like to get forward; notably Vertonghen goal against Manchester United shows his strength and determination in the final third of the pitch. In total Gareth Bale and Moussa Dembele have won the most fouls than any other player with (9). On the other hand Sandro has lost the most fouls with (11). Oddly enough Defoe has handballed 4 times more than any other player in the squad so far this season; hopefully the referee will keep a close eye on that.

Creativity: This is something that Saints are going to have to be aware of due to the amount of chances Spurs create. Aaron Lennon is the most creative player in the squad having created (22) chances, of which all of them have been created in open play. This makes him the 5^th most creative player in the Premier League, sitting in 9^th place is Gareth Bale who has created (20) chances, of which 60% have been from open play and 40% from set pieces. Lennon and Bale have also attempted the most dribbles (22) but Lennon has a far superior success rate of 55% compared to Bale’s 32% the lowest in the team.

Passing and Possession: Looking at some of the individual passing stats, they are very impressive. Overall, the lowest in the team (excluding goalkeeper) is Defoe with a Passing Accuracy of 75%. In particular, Kyle Walker has made the most passes in the team with 375 which is interesting as he is a fullback (although very attacking minded) his passing accuracy is 84%. The centre back partnerships of Gallas and Caulker have a passing accuracy of 92% and 90%. Looking at the possession, Sandro has won back possession of the ball 53 times (most of any Spurs player) this is at an average of 12.36 minutes per possession won. He also loses possession of the ball at an average of 7.71 minutes. Sandro is Spurs’ main enforcer by winning back possession in the middle third of the pitch 35 times; this is where Tottenham look to break once he has recovered the ball. Despite the threat of Bale, he has lost the most amount of possession in the Spurs side (151) times. Both full backs in Walker and Vertonghen have lost the ball the 2^nd and 3^rd most (142 and 110). As I’ve stated before this shows there role in pushing forward and putting the opposition on the back foot.

Tackling: Sandro has made the most tackles (31) at a win % of 71%; he also has the most attempted ground duels (70) at a win % of 63%. Jan Vertonghen has won the most aerial duels with (19) at a win % of 66%. Looking at the tackling breakdown; Sandro has made the most challenges at (121), each challenge averages at 21.13 minutes. However, minutes per tackle won averages at 29.77 minutes.

TEAM

Looking at the table above, there’s not too much difference between the two sides. Looking at the aerial duels in particular Saints should win the majority of balls that go up; however with Spurs playing Defoe at the top I can’t see to many balls getting knocked up to him which is probably why there aerial stats are lower. However, Adebayor could be set for a return to the team, therefore there may be a few more direct balls.

This table is looking at the passing comparisons of each team, again very similar in the accuracy of the pass although Saints look to play at a higher tempo with more passes per minute. Looking at where they pass it Spurs look to play it out wide a lot more, this hardly surprises me when you have players like Bale and Lennon who are capable of taking on players and putting the ball in dangerous situations.

As stated above in the passing comparison, you can see how they like to take players on with over half more attempted dribbles than Saints. By doing so this is creating more chances for the Spurs team. Whereas, Saints who are good going forward aren’t creating anywhere near the amount that Spurs are managing. However, all is not lost if we have a look at the table below.

Due to Spurs style of play they will always create opportunities for themselves, but there chance conversion I believe is not the greatest for a team of their quality. Saints have a good conversion rate for the chances that they have created, but they will have to make sure they put that to use when we play Spurs on Sunday.

Key Points

• Pace, the pace of the team as a whole is quiet frightening.
• Fullbacks getting forward, midfielders will have to track back otherwise the opportunities we will present them will not be nice to see.
• Sandro, the underrated player of the team. Stopping him play will halve the amount of times they try to win back possession off us; he is also the main passing distributor to the wingers. So cutting off the supply will cause frustration amongst the team.
• Take our chances when they come are way.

Enjoy The Game!!

Source: EPL Index (stats)

Saints fans have been taking to social networks in their droves with reasons why we’re not picking up points. Some have already had enough with Nigel Adkins and would like him replaced, others believe it’s to do with the defence and some have mentioned Saints tough start.

So I thought I would have a look to see if I can find any trends or patterns that have cost us points.

Let’s start with the goals that Saints have conceded, in total we have conceded 24 goals in 8 games so far. For any team that is not good enough and for us to stay in the division this will have to improve which I believe everybody knows that. So when in the match are we conceding? Looking at the graph below which shows goals conceded we can clearly see that we concede the majority of our goals between (31-40mins) and (81-90+mins). For me this shouts out two things, concentration and fatigue levels, now this is sadly stats I don’t have due to the clubs confidentiality to releasing fitness details. But this for me is the main reasons why we are conceding goals, the team are working so hard at pressing the opposition during the start and middle phases of each half as well as keeping hold of possession that the last 10mins in each half are causing Saints to make errors, in total 5 errors which have led to goals and chances have been made so far this season.

The goals conceded have mostly come from open play (21) with (3) coming from set-pieces.

(22) Of the goals have come from a shot on goal, with (2) of the goals coming off the head.

So the next part I am going to look at is whether the substitutes have played a part in why we haven’t picked up points.

Looking at the bar chart we can see that the majority of Saints subs come on in the final third of the match with the most (8) coming in between 71-80 minute. The aim being to either bring on a sub to get Saints back in the game; for example Man City with Rickie Lambert and Steve Davis getting on and scoring or bringing on a fresh pair of legs to keep the tempo and pressing up to its high levels. However, as I’ve mentioned before due to this work rate the Saint’s players are putting in you can’t replace 11 players for the last 20 minutes; therefore you see tired and sloppy play filter in to the final minutes, such as being out of position, not pressing when you’re supposed to and giving possession away which no doubt puts the fans on edge. We can all say they are professional footballers who should be able to deal with the extremities of football, but this is a huge factor. As much as we would like to see Rickie Lambert stay on the pitch for the full match, this is almost impossible for a player like Rickie. Yes, he brings a goal threat and creates a buzz about the place. But times are moving on and the physical demands he is putting himself through will not let him play every minute, this could prevent him playing in future games due to injury or fatigue. Saying that if he was fit and ready to go he should’ve started against West Ham. Some changes may have cost us points such as Manchester United when Nigel Adkins took off Jason Puncheon, Adam Lallana and Rickie Lambert when realistically one of them should’ve at least stayed on. Against Fulham, some people believe the change of Lambert for Guly cost us the win, but how can that be when the goals conceded were irrelevant to the position Guly was in. Below is a chart on how many times Saints players have been substituted on and off the pitch.

Now to look at the defensive side of the team, below is a table showing some us some of their % stats.

Here’s a comparison with three other defenders from the Premier League.

When you compare the Saints defenders with three other defenders from the Premier League we can see that their isn’t too much difference, some maybe surprised at how decent Jose Fonte stats actually are. Overall, the passing accuracy is at a very good level all round which suits in with Saints playing style. So some maybe thinking, why are we still conceding late goals? The answer I believe is back to the concentration levels, due to the demands of the Premier League a lot of mistakes are made late on. If we look at the game against Manchester United below we can see some of these errors; I think these errors are purely down to concentration and fatigue levels in the players.

Robin Van Persie hat-trick vs Saints http://www.youtube.com/watch?v=gFBZlfj2enI

These three goals have summed up a lot of the season, 1^st Goal straight away Fox needs to get tighter to Valencia (one of the best crosses in the game) then Clyne manages to slip up allowing a world class finisher in Van Persie to slot away. The 2^nd Goal shows that there was no support out wide which allowed the cross to come in, where Ferdinand managed to get a free header, then the ball came off the post and the first player to react was Van Persie. For the 3^rd Goal Fonte sliced the ball out for a corner and from that resulting corner Fonte forgot to mark his zonal area which allowed Van Persie to score his 3^rd goal. Looking at this infuriates all of us, because we outplayed Man Utd, just three silly errors which could’ve been avoided cost Saints points, this seems to be the trend this season and something that needs to be cut out in order for us to push on up the table and pick up points.

I guess this can also be influenced by the fact that Saints are playing against a lot of world class players where they are always looking for errors and due to the fact Saints lack a lot of Premier League experience in the back five it proves to make them wobbly when it comes to the final minutes of the game. There have been rumours today of Michael Dawson from Spurs coming in which would be the ideal sort of defender Saints are looking for, it’ll be interesting to see if the rumour still has legs by January.

So what can be done about it? I believe the players should keep striving to be the best of their ability and look for ways to cut out their own mistakes, that maybe a fairly open and poor view. But it’s the simple things that can get the best out of them. In the long run, an experienced defender maybe the solution but were going to have to deal with the present. As long as they keep working hard in training and put as much as that back on the pitch then Saints can only get better and help win over some of the fans who think it’s all doom and gloom.

There’s still a long way to go and there’s plenty of time to improve, we have shown against teams like Manchester United and City we can mix it with the best and I fully believe we can start moving up the gears.

Thanks for reading; I aim to bring in my next scout report for Tottenham by the end of the week!

Sources: EPL Index (stats)

This is my second analysis of the 2011/12 EPL Opta dataset provided by the MCFC project. In my previous one, I looked at Pass direction in order to refute the Fink Tank analysis on Long balls.  In the process, I produced some graphs which actually give further support to my conclusion, and are a clear demonstration of the power of visual analysis.

Let’s start with the data…

Looking at this table is not easy to spot major differences in Pass direction between the clubs.  But if we plot this data (see graph below), and we fit an orthogonal straight line (red line) to it, we can now easily notice that this line is (or almost) vertical for some clubs, and (almost) horizontal for others.

The line position indicates that there is a major (significant) difference between the two sets of clubs.  Those with a vertical line show a predilection for sideways passing (left/right), while the others for forward passing (back/forward).  The graph does not give a clear-cut answer for all clubs (four points are too few to produce a well-defined line in such a small graph), and we can see that Chelsea is borderline.  In fact, a discriminant analysis of this data shows that Chelsea, Liverpool, Everton, and Fulham also belong to the sideways set.  More importantly, what the graph show clearly is that the top four clubs (with Swansea and Wigan) all tend to pass the ball sideways, rather than forward.

To further emphasise the power of visualisations to highlight irregularities in the data, below is a radar (polar cap?) chart of the same data.  The clubs are in order of Forward passes. Some interesting information can be gathered just by looking at it, for example:

1. Stoke makes the lowest number of passes.
2. All clubs make an almost equal number of left/right passes.
2. Manchester United is unique in making an almost equal number of left/right/forward passes.

The authors of an open online course, Current/Future State of Higher Education (CFHE12), observe in Week 2 of that course:

New technologies alter the relationship between learners and educators. The mediating role of higher education – selecting and brokering access to learning materials and ideas – is changing with the rise of the internet and constant connectivity.

Observing and Analysing Performance in Sport is seeking to explore these changes and to do so with a short, introductory course hosted on the OpenLearning platform. The course starts on 5 November and runs for five weeks.

Open Online Courses

There is lots of discussion about open online courses at the moment. In Australia, for example, The Conversation is running a series of articles this month. I enjoy reading Tony Hirst’s perspectives from the UK. Recently he observed:

Another of the things that I’ve been pondering is the role that “content” may or not play a role in this open course thing. Certainly, where participants are encouraged to discover and share resources, or where instructors seek to construct courses around “found resources”, an approach espoused by the OU’s new postgraduate strategy, it seems to me that there is an opportunity to contribute to the wider open learning idea by producing resources that can be “found”. For resources to be available as found resources, we need the following:

1. Somebody needs to have already created them…
2. They need to be discoverable by whoever is doing the finding
3. They need to be appropriately licensed (if we have to go through a painful rights clearnance and rights payment model, the cost benefits of drawing on and freely reusing those resources are severely curtailed).

Tony points to the approaches taken in cMooCs and ds106. Like Tony I have found these two approaches to open learning liberating.

(cMOOC) offerings inspired by the likes of George Siemens, Stephen Downes, et al., where a looser curriculum based around a set of topics and initially suggested resources is used to bootstrap a set of loosely co-ordinated personal learning journeys: learners are encouraged to discover, share and create resources and feed them into the course network in a far more organic way than the didactic, rigidly structured approach taken by the xMOOC platforms. The cMOOC style also offers the possibility of breaking down subject disciplines through accepting shared resources contributed because they are relevant to the topic being explored, rather than because they are part of the canon for a particular discipline.

The course without boundaries approach of Jim Groom’s ds106, as recently aided and abetted by Alan Levine, also softens the edges of a traditionally offered course with its problem based syllabus and open assignment bank (particpants are encouraged to submit their own assignment ideas) and turns learning into something of a lifestyle choice…

I am keen to be part of that discussion and to do so through the practice of a small course committed to open sharing and flourishing. I take Jim Groom’s point about open architecture very seriously:

We have to broadly experiment with and come to terms with how we design an open architecture that provides for a coherent personal digital archive.

I have enjoyed reading about A Domain of One’s Own at the University of Mary Washington:

A Domain of One’s Own is a new pilot project from the University of Mary Washington and a collaborative effort between the Division of Teaching and Learning Technologies and the Office of Information Technology Services. This pilot will give 400 students and faculty their own domain name and web space to install a portfolio of work or map to existing systems. Content from coursework in the pilot will be aggregated here, as well as an exploration of aggregating the work students are creating directly to umw.edu. We believe this pilot project will give students the flexibility to build out their “e-portfolio” using a variety of software and approaches in a space that gives them the power to easily migrate and transport that data when they graduate. We’re excited about the possibilities and will continue to update everyone on the project right here at umwdomains.com as the pilot unfolds.

John Udell has written about this too and the importance of cloud opportunities “where students could, and would, take care of their own services, relying on standards to interoperate with the institutions they’d serially associate with during their careers and lives”.

Leigh Blackall makes some important points about principles too in regard to: connected and constructed learning, open access, free content reuse, international, cross cultural and collaborative engagement, transparent processes and open documentation, peer to peer assessment and acknowledgement of people breaking conceptual ground in the lobbying and development of open and networked practice. Leigh notes Teemu Leinonen’s 2008 approach in Composing free and open online educational resources:

The course readings and the assignments in this course will familiarize participants with the main concepts related to open education resources and to the historical and philosophical ideas behind them. The participants will also do their own projects where they will learn to create and participate in projects producing free and open educational resources.

I admire Stephen Downes’ work immensely. Last week Stephen shared a presentation on the Connective Learning Environment. In that presentation he gave an insight into the development of CCK08 and subsequent open online courses. Stephen links to Dave Cormier’s video What is a MOOC?

Invitation

I hope that the background information I have shared here gives a feel for the epistemological foundations for Observing and Analysing Performance in Sport.

This post is an invitation to participate in the course and to do so with the voluntary assumption of risk that it will be a messy experience. With OpenLearning’s help I think we will be able to adapt to an emerging course.

I have cleared my diary for November to optimise my opportunity to be part of the course. We are still writing the content for the course but I sense that the range of contributors will make it a very distinctive experience.

Photo Credits

Orchestra

Riffing

However, analysis of Opta data from the MCFC Analytics project shows that the short ball game is the way to success.

The data

Opta classifies Passes in various categories.  For this analysis we have looked at what Opta specifies as Long balls as well as Long/Short Passes.  Opta also specifies Successful (ON Target) and Unsuccessful Passes/Long Balls.

Long balls

Using this data immediately highlights some inconsistencies with Fink’s data.  Opta supports Finks analysis that Stoke “is the club who plays the largest number of long balls”, and moreover, has the highest ON target (success) rate.  Also Opta confirms (in general) Fink’s claim that “the less good clubs use a greater proportion of Long balls”.

However, while the basic Fink hypothesis stands up to scrutiny, the numbers themselves don’t.  Fink then goes on to state, for example, that Long Balls make up 2% of Stoke’s passes.  Wrong! 15% of the Potters passes are long balls – quite a difference!  Not really relevant, but note that Stoke made also the lowest number of passes (nearly half those of the top clubs).   Also according to Fink’s table (Hit and myth…) Arsenal played only 78 long balls last season (0.38% of their 20594 passes), that is about 2 every match.  These are 608 says Opta, of which 124 ON target – quite a difference!

* http://angryofislington.com/2012/09/10/long-balls-are-more-effective-than-short-passes-apparently/)

Winning strategies

In the next stage of his analysis Fink compares Long balls to Short passes, “to judge which one was the more successful techniques” i.e. to get the ball from the penalty area to the opposition one.

Opta provides data on Pass direction: Back, FWD, Left, Right, as well by Pitch location: dividing the pitch in three area (DEF third, MID third, FWD third).  A problem with using the Pass direction data is that includes Crosses and Corners.  But these Events are around 7% of the data – too few to skew the results of the analysis.

Below are the results of our statistical analysis of OPTA Pass direction data, with the relevant (to Fink’s claim) comments:

Tree 1.                                                                     Tree 2.

Tree 1. shows that there is a significant difference in Pass direction among teams.  The teams on the left (mainly the top ones) favour passes to the sides (Left/Right); those on the right (mainly the bottom one) favour FWD passes.  The percentage of Back passes is almost equal.

The difference is startling. The top teams make significantly fewer FWD passes (only 34%) compared to the teams in the bottom half (40%).

This difference is even more clearly shown in Tree 2. Here, we have ignored Back passes, and grouped Left/Right ones. It is now clearer that the successful teams prefer to move the ball sideways (61% vs. 53%).

We have further analyzed the Opta data to look at the FWD passesfrom players in different positions (ignoring forwards, who make very few Long passes), and comparing it to their eventual league position.  The results are shown in the tables below. (Note that FWD passes include Long Balls).

FWD passes by Position

It is clear that the poor teams are those that rely more on FWD passes. The three relegated teams were actually pretty good at using them.

But the teams that were successful in the League were actually pretty poor at playing FWD. Champions Manchester City’s defenders (DEF), for example, found that only just over 41% of their FWD passes were successful, yet relegated Bolton’s defenders were 16% points more successful.

These tables confirm the details of what already shown by the two graphic trees: top teams make significantly fewer long, forward passes. Since their lack of success with the Long ball has already been established, they must have a higher success rate with Short passes to overcome the effects of their poor FWD pass technique.  Otherwise, how could they be top?

The top teams know that hoofing the ball forward isn’t the answer to winning games. They know it is less accurate, they have (buy?) fewer forwards suited to this type of game, and they have the better players who can pass the ball accurately.

The good teams also know that they can carry the ball to the edge of the opposition penalty area more often by moving it sideways. Hoofing is NOT essential to top teams who have the skill to move the ball up the field in a more controlled way – sideways is their winning way!

upc_dump

There are several tools that can utilize GASP to analyze UPC applications. The most dumb tool is upc_dump, which is included into Berkeley UPC compiler. It doesn’t actually analyze application, it simply dumps all tracing information. Here is how you compile UPC application with upc_dump:

> /opt/bupc-runtime-2.8.0/opt_inst/bin/upcc -network=udp --inst-toolname=dump /opt/bupc-runtime-2.8.0/opt_inst/bin/gasp-dump/gaspu.upc -L/opt/bupc-runtime-2.8.0/opt_inst/bin/gasp-dump -lgasp-dump -T=4 --inst bin_file

First of all you need to compile your code with instrumented version of UPC from opt_inst subdirectory. I described how to compile opt_inst earlier in this post. With --inst-toolname you provide the name of instrumentation tool (it seems that it doesn’t matter what you specify here). Then you compile your application along with gaspu.upc. Usually developers of performance analysis tools put GASP related call-back implementations into a library and all upcalls to UPC code into a separate source file (such as gaspu.upc for upc_dump), since library is compiled with C compiler and cannot include UPC code. Then you provide path to the library with -L flag and name of the library with -l flag. The last --inst flag instructs the compiler to instrument all UPC library functions. --inst-functions also instruments all user-defined functions. Additionally, --inst-local allows to instrument all local accesses to shared memory within the thread.

upc_trace

upc_trace is another tool distributed with Berkeley UPC, but it has some performance analysis functionality. Unlike upc_dump you don’t need to manually specify all flags to the compiler. What you need to do is to compile an opt_trace subbuild. And upcrun application with the -trace flag. When execution is completed you will have several trace files which you pass on to upc_trace tool. Performance analysis data is represented as output text file. You can control which types of events are gathered using GASNET_TRACEMASK environment variable.

Parallel Performance Wizard

PPW is probably the one performance analysis tool for UPC with rich functionality. PPW installation is simple:

# ./configure --prefix=/home/fred/ppw-2.6 --with-upc=/home/fred/bupc-runtime-2.10.0
# make
# make install

Then to perform simple tests use PPW wrappers for upcc and upcrun:

> ppwupcc -network=udp --inst-functions -T=32 upc_code.c
> UPC_NODES=”n1 n2 n3 n4″ ppwrun --output=upc_code.par upcrun -n 32 -nodes=4 bin_file

PPW has GUI where you pass the .par file and see what happened in the application.

I already described basic Berkeley UPC compiler installation here. So now lets go deeper in details.

Backend Compilers

Basically UPC compiler is a translator from UPC language to C. After translation is done, backend C compiler is invoked to actually compile the code. On Linux clusters GCC is used by default, if you have Intel, Sun or any other high performance compiler installed, then use CC and CXX flags in UPC runtime configure step:

./configure CC=icc CXX=icpc --prefix=/opt/bupc-runtime-2.12.1-icc
./configure CC=suncc CXX=sunCC --prefix=/opt/bupc-runtime-2.10.0-suncc

Optional UPC builds

By default Berkeley UPC is installed in two configurations: debug (with GASnet assertions enabled and debugging info compiled in) and opt (optimized version for everyday use). You will see debug and opt subdirectories in your working UPC runtime build. But you can install additional versions of runtime for other uses.

Berkeley UPC has integrated tracing facility. If you upcrun application with the -trace flag, tracing data is collected and you can analyze it with upc_trace utility. Tracing build can be compiled by using opt_trace multiconf option:

./configure --prefix=/opt/bupc-runtime-2.12.1 --with-multiconf=+opt_trace

Berkeley UPC has integrated callbacks (called GASP) for third-party instrumenting utilities. Instrumentation allows developers of performance analysis tools to gather all sorts of information about UPC program execution. Like functions called, their arguments, etc. If you want to develop your own UPC performance analysis tool you can use this feature during development and instruct users to build opt_trace version of UPC to be able to use your tool later.

./configure --prefix=/opt/bupc-runtime-2.12.1 --with-multiconf=+opt_inst

You can debug UPC applications with dbg build, if you are a developer and use instrumented build of UPC and need to debug it, then build a dbg_inst version. There was a dbg_inst.patch (find link below) to add dbg_inst functionality to UPC, but it’s already integrated into compiler as far as I remember.

./configure --prefix=/opt/bupc-runtime-2.12.1 --with-multiconf=+dbg_inst

There was also another bug which broke dbg_inst in 2.12.1 (which was originally implemented in 2.10.0) with the following errors:

/root/install/berkeley_upc-2.12.1/gasnet/gasnet_trace.c: In function ‘gasneti_trace_finish’:

/root/install/berkeley_upc-2.12.1/gasnet/gasnet_trace.c:988: error: ‘gasneti_mallocreport_filename’ undeclared (first use in this function)

/root/install/berkeley_upc-2.12.1/gasnet/gasnet_trace.c:988: error: (Each undeclared identifier is reported only once

/root/install/berkeley_upc-2.12.1/gasnet/gasnet_trace.c:988: error: for each function it appears in.)

To resolve this issue apply mallocreport.patch00 (find link below). But if you use recent Berkeley UPC build you won’t see this bug.

Block size

If you work with huge matrixes and want to distribute them in large chunks of consecutive rows, then you will run into UPC limitation of block size. UPC pack pointer representation into one 64 bit integer. By default 34 bits are allocated for memory address, 10 bits for threads and 20 bits for phase (or block size). 2^20 is basically a 1048576 elements which is a very small number. You can redistribute bits with --with-sptr-packed-bits=value value=’phase,thread,addr’ configure option, but then you will either have small address space or small number of threads.

Another option is to use --enable-sptr-struct configure flag which changes shared pointer representation from int to struct. It will increase block size to 2^(32-1) which is 2147483647. But it could also be too small if you conduct performance measurement and need to run your code for 1 thread. Then the whole matrix is a one huge block. 50000×50000 matrix is already hit the limit.

If 2^(32-1) is not enough, then the last option for you is to use row distributed algorithm instead of row-block distributed.

POSIX shared memory problems with InfiniBand

UPC support two one-node inter-thread shared memory communication types: POSIX shared memory and SYSV shared memory. POSIX is configured by default. If you want to register large amounts of shared memory with many PSHM processes using --shared-heap key you can see errors like these:

*** FATAL ERROR: Unexpected error Bad address (rc=1 errno=14) when registering the segment

NOTICE: Before reporting bugs, run with GASNET_BACKTRACE=1 in the environment to generate a backtrace.

*** Caught a fatal signal: SIGABRT(6) on node 29/32

To solve this problem reinstall runtime using the following options:

./configure --prefix=/opt/bupc-runtime-2.12.1 --enable-pshm --disable-pshm-posix --enable-pshm-sysv

Bug when building translator

For some vendor-build GCC releases, like Red Hat, older versions of translator fail to compile with error like:

/usr/bin/ld: ipl_summarize_util.o: relocation R_X86_64_PC32 against `Phi_To_Idx_Map’ can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: final link failed: Bad value
collect2: ld returned 1 exit status

It’s a bug number 2202 in UPC Bugzilla and is described here. Solution and patch are described in post 17. Find copy of patch below.

UPC I/O support for large files

UPC have parallel I/O extension. In version 2.14.0 and earlier by default UPC I/O supported files 2GB in length. It led to upc_all_fread_shared() returning -1 “Invalid argument” for data above the 2GB limit. To change defaults from 2^(32-1) bits size to 2^(64-1) use BUPC_IO_64 variable during runtime configure step:

./configure CC=”gcc -DBUPC_IO_64″ CXX=”g++ -DBUPC_IO_64″ --prefix=/opt/bupc-runtime-2.12.1

Replace GCC with your own compiler.

SUN compiler issues

If you run into an error (I had it in version 2.10.0):

“/home/fred/install/berkeley_upc-2.10.0/upcr_profile.c”, line 36: left operand must be modifiable lvalue: op “=”
cc: acomp failed for /home/fred/install/berkeley_upc-2.10.0/upcr_globfiles.c

Apply patch sun_const_field.patch00 (find link below).  Additional info can be found in Berkeley UPC Bugzilla, bug number 2696.

Another bug (not an error, but an annoying warning) looks like numerous warnings throughout compilation:

“/home/fred/install/berkeley_upc-2.10.0/upcr_atomic.h”, line 876: warning: result of paste undefined and not portable: 64_ (E_PASTE_RESULT_NOT_TOKEN)
“/home/fred/install/berkeley_upc-2.10.0/upcr_atomic.h”, line 876: warning: result of paste undefined and not portable: 64_cswap (E_PASTE_RESULT_NOT_TOKEN)

To get rid of it apply patch not_token.patch00 (find link below). It’s described in the same 2696.

Links to patches

Unfortunately WordPress doesn’t allow to upload .txt files due to security reasons. Other formats, such as .doc or .pdf will break the lines. So I decided to give direct links when possible and provide contents of patch in text converted to .jpg format in case direct link will break in future. The drawback is that you will have to type it yourself or OCR it.

• dbg_inst.patch direct link to Berkeley UPC site
• dbg_inst.patch in .jpg format
• Phi_To_Idx_Map.patch link to Berkeley UPC Bugzilla
• Phi_To_Idx_Map.patch in .jpg format
• sun_const_field.patch00 link to Berkeley UPC Bugzilla
• sun_const_field.patch00 in .jpg format
• not_token.patch00 link to Berkeley UPC Bugzilla
• not_token.patch00 in .jpg format
• mallocreport.patch00 in .jpg format