BoroBrain

Alternative League Table

MarkE February 19, 2017 Comments Off

A few seasons ago someone clever (though I forget who – apologies) started to post a league table on The Conference Forum (and then reproduced on BoroChat) that showed teams placed alongside the numbers of points they’d earned. This reflected their relative positions by points rather than ordinal rank of position which can appear to be misleading.

I’ve had a go at doing something similar, but in this case more graphical. Each dot in the chart below shows a team’s placing according to their progress in the league. The table on the left shows their placing by pure total points and the one on the right by average points. Hover over a dot to identify the team (broadly colour coded by primary shirt colour).

This representation in Tableau stands on the shoulders of an original visualisation by @wilsooon, who produced a similar effort for the top leagues across Europe, which you can see here.

If I’ve produced this properly, it should update every 24 hours.

Preseason Survey Results – Part 2

MarkE August 4, 2016 Comments Off

If you’ve already taken in the first part of our survey results breakdown, you may or may not be keen to hear more. I’m guessing if you’re still reading this you’re not averse to the idea.

We’ve already seen that you generally expect Boro’ to finish somewhere around upper mid-table (give or take), and are cautiously optimistic about the upcoming season. However, what some supporters would be willing to compromise on in order to achieve success can vary greatly. The survey asked fans whether they would be willing to forego cup success for a top seven finish and, conversely, whether they’d sacrifice a playoff spot for a big cup run. Here are the results:

Almost two-thirds of fans polled would take a playoff place at the expense of progress in any cup competition – whilst not a complete whitewash, that’s a pretty firm preference for league over cup, although this deliberately wasn’t a strict either/or question.

Meanwhile, a similar number also stated that a place in the 4th round of the FA Cup wouldn’t be enough to sacrifice a spot among the top seven come May next year:

What’s important to underline here is that these aren’t the same fans answering “yes” to the first question but “no” to the second – in fact fewer than half of the responses formed that pattern. Some 16% of fans polled answered “no” to both, hinting that they were thirsty for progress on both fronts. Meanwhile, 18% of respondents answered “yes” to both questions, suggesting that there is a healthy portion who would take some success in whatever form it came.

The next charts almost don’t require printing, so emphatic as the results are that they could probably be explained in a few lines of text. Two questions asked fans which they felt were Boro’s strongest and weakest areas, able to choose between the three options “Goalkeeper and Defence”, “Midfield” and “Attack”.

Suffice to say there is some consistent opinion out there in regard to this, although it will be interesting in retrospect to compare this with other seasons – with most teams aching for a killer forward line, few teams’ fans will feel totally satisfied with their attacking options, so perhaps to some extent this is to be expected. What isn’t necessarily expected, though, is the strength of feeling towards the rearguard, which is certainly impressive.

Finally, to close us out, some further inspection of the sides you felt are most likely to finish above or below Boro’. We’ve covered off who you expect to top the table, but given the choice to name as many as three sides to rack up more points, here were your suggestions:

Two-thirds seem convinced that Portsmouth will be up there, and half named Plymouth. One in four went for Luton and one in five named Leyton Orient among the most likely to finish higher.

Meanwhile, those you felt were more likely to be at the bottom come the end of season appear below:

Luton sneak in at the bottom, perhaps hinting at our first possibility of fan rivalry influencing opinions, considering that they were named the side fourth most likely to finish above Boro’!

That brings us to the end of the headline results from this survey. If I find the opportunity I may dig a little deeper into one or two features at a later date, but for now thanks again to everyone who participated. I envisage running a similar exercise in future seasons.

Preseason Survey Results – Part 1

MarkE August 3, 2016 Comments Off

In late July, as Boro’ played out their final friendly matches ahead of the new season, I polled fans on how they saw the upcoming season playing out – our first preseason survey. They offered their opinions on where Boro’ would finish, who would win the league, and which other teams would finish above or below them. The survey also delved into their mood and attitudes as the season gets underway, and the initial results are released here, in the first of two parts.

In part two, later this week, we will look into attitudes and the sides most likely to finish above or below Boro’, but in this first part we look at some of the headline results.

To open up, it helps to understand who responded to the poll. What’s clear is that respondents to this survey are not representative of the Boro’ fan base as a whole – those who are more engaged are more likely to have even noticed the poll on social media, let alone participated. Three-quarters of responses came from fans who’ve followed Boro’ for at least ten years, but even still these should give us some fascinating insights whilst being conscious of the heavy weighting towards the long-haul fan.

(tip: click the graphic to view a larger version)

Respondents were also asked how many games they expected to attend in the coming season, split between home and away. This is by no means definitive – circumstances change and levels of interest will naturally go up and down according to the team’s performance. However, it does give us a proxy for how invested the fans are, and when we come around to the second part of this analysis, can see how much of a factor this is.

As the chart above describes, just over half of respondents expect to attend at least three-quarters of the games, and a quarter of them plan to be at every home league game. Away game attendance is, as one might predict, somewhat different. 14% of respondents don’t intend to attend a single away game, with the rest split between a few and around half of games (only a few committed stragglers anticipate more than that):

Okay, so with that under our belts, what did you all have to say? Well, let’s begin at the top and work our way down – who do Boro’ fans expect to win League Two this season? 75% of responses put forward the same three names, suggesting that the field contesting the top spot is not exactly wide open this year – Boro’ fans are backing either Portsmouth, Doncaster or Plymouth for the title, with one in ten rating their Bedfordshire neighbours for a tilt. The first two are perhaps not a surprise, but Plymouth are rated only 6th favourite by the bookies ahead of opening day.

You may notice a sprinkling of exceptionally confident Boro’ fans also adding their name into the mix. 5% of respondents appear confident that Boro’ will secure automatic promotion this season, and a huge 77% predict a top-half finish, seriously belying the current odds offered of 9/4 against that outcome.

Closing out the first part of our analysis, the survey asked participants to “sum up your mood in a few words as Boro’ head into the new season” – the following word cloud demonstrates the most commonly used words. Here, the larger and redder the word, the more responses that used it (note, only those with at least two occurrences appear which does indeed include “meh”).

Just a shade over half of you described your mood as “optimistic”, allowing for various spellings, with “excited” and “hopeful” checking in as the next most popular terms. There are plenty of positive words among those and very little caution applied. All in all, things would appear to be on the up for Boro’ fans’ fortunes.

Part 2 of the results will look at answers to the following questions:

Which teams do you think are most likely to finish above/below Boro’ this season?
Which is Boro’s strongest/weakest department?
Would you be happy to go out of every cup at the first round if it meant earning a playoff place?
Would you be happy to miss out on a playoff place if it meant playing in the 4th round of the FA Cup this season?

Those results should be published later this week. Thanks so much to all who took part, I hope you find the initial findings interesting.

Poll Summary Details:
Poll was conducted between 24th-31st July, 2016
Number of respondents was 80

The 2016/17 Preseason Survey

MarkE July 20, 2016 Comments Off

Regular visitors will likely have taken part in the Satisfaction Index surveys that have occasionally appeared on the site, but this season I’d like to go a step further. I’ve launched a more comprehensive survey ahead of the new season, and then intend to follow it up with a related one at the end of the season, to see how hopes and expectations meet with reality.

Until 31st July you can visit the survey and contribute your predictions, feelings and perspectives and I’m looking to see not only how Boro’s fans see the fate of their own team but also that of the competition, gauging what matters to them, and collecting further data that might help to understand what drives their motivations.

So, here’s the link to the survey. Once again, it’s hosted in Google Forms and you’ll need to have a Google account to participate. However, you will also be able to edit your answers until the poll closes.

Contributions will be anonymous, but please try to answer as honestly and openly as possible. I will look to publish some initial findings within a few days, just before the season begins, and then some more detailed insights shortly after.

Full privacy note:

The survey is hosted on Google Forms and, in an attempt to restrict abuse, it requires a Google account. If you already have one then just login and plough on – if you don’t have one they’re easy to setup. Apologies if you wish to take part but don’t wish to have a Google account – if that applies to you, get in touch and we’ll see what we can do to address that.

For those concerned about personal details, by using Google Forms we cannot see who takes part, nor any of their account details, so your privacy will remain intact.

Goal Difference Progressions

MarkE July 19, 2016 Comments Off

I’ve been getting used to a new piece of data visualisation software and in the middle of last season stumbled upon an idea I’d had, about how goals scored (and conceded) were distributed throughout the 90 minutes of a game. Graham Westley was supposedly famed for having teams that would be fit enough to last to the end of the season and perform better in the latter stages, but what about within games themselves? Do sides have a habit of scoring more goals in the first half than in the second? Are they great at defending for 80 minutes and then let the floodgates open in the dying moments?

And it’s led to something I’ve found to be quite fascinating, but requires a short explanation, so first please allow me to build up a picture. The following chart shows the number of goals scored by Boro’ (in blue) over the course of the 2010/11 season, Boro’s first in the Football League, grouped into five minute chunks. That is to say that the left-most blue bar counts the goals scored in the first five minutes (there were two), the next bar shows how many goals were scored between then and the tenth minute and so on.

The red bars hanging down represent the goals that were conceded during those same five minute intervals. So between the 5th and 10th minutes of league games that season, Boro’ scored eight goals and conceded just one. So far, so straightforward, and you can pick out various other features if you so wish:

Goals Scored in Boro’ League games during 2010/11 regular season, grouped in five minute intervals

Where this led to, however, was a thought around how those sets of bars relate, and what happens when you consolidate them over time. To the naked eye it appears that Boro’ significantly outscored their opponents in the first half but then things evened out a little in the second half. But how to depict that in the chart without requiring the reader to pick it apart? Well, instead of considering two sets of bars, why not one? To that end, I repeated the exercise, but just looked at the difference between goals scored and conceded in each of the periods.

Goal Difference in Boro’ League games during 2010/11 regular season, grouped in five minute intervals

Examining this chart crudely, it’s evident that there were more five-minute intervals where the Goal Difference was positive than when it was negative, and a couple of particularly big spikes early on. Translating that into what it meant in footballing terms, Boro’ were definitely scoring more goals than their opponents in the first half, and broadly the same number in the second half. The visualisation still falls a long way short of telling a complete story, though, as we end up having to spend some time interpreting the data and performing some manual arithmetic to make sense of it all.

What would be much clearer would be a cumulative total – we can see that Boro’ outscored their opponents by eight goals in the first ten minutes with some simple addition, but to work out the equivalent number ove, say, 60 minutes takes time and gets dull very quickly. So, here’s that same chart, but instead the bars represent the cumulative goal difference since the start of the game. The numbers at the tip (or base) of each of the bars indicate the goal difference at that point:

Progressive Goal Difference in League games, 2010/11

Remember that this consolidates all 46 games over the course of a league season, but by doing so we get a generalised picture of how Boro’s games went in that spell. The story here is that they were effective at getting ahead within the first half an hour of games, and then were even more effective at holding on to that lead. Their goal difference at 30 minutes in was +17, and was the same at full time, an hour later.

Obviously, seasons in which Boro’ have had a positive goal difference are generally more successful and see them winning more games, but there are many ways to skin a cat. The following season Boro’ ventured into League One and made another dash for the playoffs, also with a healthy positive goal difference, this time of +25. The profile, however, differed greatly to that from the previous season – an hour in, Boro’ were generally scoring as many goals as they conceded, but in the final half hour of games totally dominated, scoring 20 more goals than they conceded:

Progressive Goal Difference in League games, 2011/12

As alluded to earlier, Graham Westley’s sides were known for having the stamina to improve as the season went on, but not much was said about this side’s endurance within games. Westley departed just over halfway through the season, but had clearly imposed some superior fitness in those players.

The following season (12/13) saw significant transition as Gary Smith took his side towards the top of the table in the early months before fading rapidly as the season progressed and closing out in a lower-mid table spot. With a final goal difference of -16 the chart looks much different, as you’d expect, but is almost a mirror image of the previous one, with only a difference of -5 in the first hour and some serious fade as games reached their conclusion. The squad had featured a serious overhaul in that period, and little of that staying power transferred from one season to the next.

Progressive Goal Difference in League games, 2012/13

By 2013/14 Westley had returned again, but saw his side miserably relegated. This time the deterioration was worse – Boro’ were generally in their games until half time but then capitulated quite gradually. Were we to see a perfectly consistent drop as the minutes passed we’d be able to argue that there was little profile to the way games were played out at all, but as we can see Boro’ were certainly hanging on until the latter stages of the first half before the rot set in.

Progressive Goal Difference in League games, 2013/14

As Boro’ then returned to League Two, Westley took one final shot at success and hauled Boro’ into the playoffs in 2014/15, with a relatively featureless season in terms of goal distribution. There was still a positive surge in the final third of games, but not nearly as significant as in previous seasons. You’ll notice a few peaks and troughs, but those peaks are relatively low compared with the earlier seasons:

Progressive Goal Difference in League games, 2014/15

And, finally, the most recently completed season saw a combination of Sheringham and Sarll flirt with relegation. Once again, however, the swing towards a negative goal difference took hold relatively late in games. Incredibly, after an hour they were level with their opponents, scoring exactly as many as they conceded, yet shipping 15 more than they netted in the final half an hour. This would really represent the opposite of that first League One season, where fitness could well have told late in games and in spite of a positive start, Boro’ really faded when it mattered most.

Progressive Goal Difference in League games, 2015/16

Six Football League seasons then, and six quite different experiences, borne out by the charts presented above. To save scrolling up and down the page to compare between them, here the are once again in a single gif, this time scaled to the same y-axis:

Rolling Goal Difference progression 2010/11 – 2015/16

If you’d like to share this gif, it’s posted to Twitter here.

2015/6 In Numbers

MarkE July 15, 2016 Comments Off

Having a whole season’s worth of data to consume in one chunk is at least a more substantial way of getting back into the swing of things. 2015/6 was a severe low point from many fans’ perspectives, but does at least provide some interesting insights that might otherwise be harder to come by in a season of mid-table mediocrity.

Appearances
Some 47 players appeared in a Boro’ shirt during the League Two, eight more than the previous highest total since joining the Football League (39 in 13/14). Some playing in most games (Fraser Franks – 38), while others were peripheral (Jack Storer – 12 minutes).

Goals
In what wasn’t a particularly high-scoring season, there were still some interesting points of note:

Boro’ came away scoreless in 17 of their 46 league games, keeping clean sheets in just nine of them. Five of those clean sheets came in the final eight matches.
24 different players got their name on the scoresheet (plus a few own goals), but only two of them racking up more than three in the league – Whelpdale (8) and Gnanduillet (5) the two saving face. Whelpdale scored Boro’s third hat-trick since joining the Football League.
Boro’s overall goal difference on the season was -15, and only one player had a personal GD worse than that – Mark Hughes was on the field to see 37 goals shipped while his team only netted 21 times at the other end. However, some six players did come away with a positive goal difference during their time on the field, the most notable of whom being Ben Kennedy (+2) and Dean Wells (+1).

Records
A clutch of milestones and significant events over the course of the season:

Boro’ earned 26 points from their 23 home games, equalling the totals from their second and third seasons in League One (2012/3 and 2013/4). These are tied for the lowest hauls from home games since entering the Football League.
Ronnie Henry went 2,716 league minutes (equivalent to just over 30 complete games) without receiving a single booking.
Steven Schumacher’s six cards in just 1,313 minutes means that he has had the most indisciplined season in a Boro’ shirt, receiving a caution every 219 minutes.
Gnanduillet’s goal in the first minute at home to Accrington was the third such since Boro’ joined the Football League. In the previous season Chris Whelpdale also scored in the first minute… also against Accrington.

Managers
Both Teddy Sheringham and Darren Sarll opened their managerial accounts during the season, with mixed results, though neither with overall winning habits:

Sheringham managed 29 league games (63% of the total), Sarll 17 (37%)
Sarll’s win rate was 29% compared with Sheringham’s 21%, but his loss rate down to 31% compared with Sheringham’s 45%
35% of Boro’s points earned from Sheringham’s reign came from drawn matches (10 from 28 points), whereas Sarll only relied on 25% of those points coming from stalemates
Boro’ scored half a goal fewer per game under Sarll than under Sheringham (0.8 per game vs 1.3), but concended only half as many (0.9 per game vs 1.8)

Performance By Month
Looking at how the results break down month-to-month, each is effectively a mini-series of five or six league games. It’s apparent that Boro’ were consistently poor during this time with minor exceptions – there were only two months where Boro’ won more games than they lost, being November and April.

Also notable is the fact that, in six consecutive months from August to January, Boro’ conceded more goals each month than games played. Conversely, in the four months from February to May that happened only once (March). Of all the statistics covered in this article, this is probably the most representative of the season as a whole.