Accrington Stanley, Gautam Gambhir & David Ferrer

12th May, 2017.

A few things that I've read this week caught my eye, so I thought I'd give my thoughts on them.  The interesting aspect with all the three subjects is that despite Accrington Stanley, Gautam Gambhir and David Ferrer having a variety of different problems, they all have some relevance towards sports data.

This week, Accrington Stanley chairman Andy Holt criticised Premier League clubs over their spending, with particular reference to agents fees, which have themselves attracted publicity in recent days.  You can read more about Holt's comments here.

Essentially, Holt is complaining that Premier League clubs spend too much on themselves and don't give enough to smaller clubs, which in my opinion is a strange point of view.  Putting this into a retail context, this would be like Tesco giving money to a local corner shop, simply because they are a bigger company, which obviously is utterly absurd.

Ironically, Accrington, as well as other clubs in League Two, receive 'solidarity payments' of £430,000 per season from Premier League clubs, so the big clubs in the Premier League do actually give to these smaller clubs.

Living in Lancashire myself, it is pretty easy to pinpoint where Accrington have a major problem.  The town of Accrington lies on the M65, between Blackburn and Burnley.  In fact, the M65 starts in Preston, so in the space of a small motorway around 25 miles long, this season there has been a Premier League club, and two Championship teams, in Accrington's catchment area.  Considering this competition, it isn't a huge surprise that their average crowd this season was 1,699.

Accrington Stanley's average crowd this season was 1,699...

While there is definitely an argument that agent spending at the top levels of football is over the top, it is the clubs money to do with as they wish, and I fail to see the relevance between this and lower league clubs.  

According to the article linked above, Accrington expect to lose around £500,000 per season.  With any business, whether it is a sporting club, or in retail, there are several reasons why they lose money - either financial mismanagement (usually over-spending or over-estimating revenue/being too ambitious too quickly) or poor recruitment, or in many cases, both.

Quite simply, with attendances being low in a competitive area for football, with many clubs playing at a higher level, the only way Accrington will make a profit as a business is from player trading.  According to Transfermarkt, the average age of their signings this season was 22.3, so they are already being pretty smart at getting players at the right end of the age curve, where wages are likely to be low, with potential transfer fee upside and player gratitude should be high.

I haven't worked with football analytics, mainly due to time constraints, but I'd imagine that very few, if any, League Two clubs use analytics data to generate player acquisitions.  In their situation, perhaps this could work quite well for Accrington...

Analytics would also probably benefit the Indian batsman, Gautam Gambhir, having read this article on Cricinfo.  Gambhir is, quite amazingly, quoted as saying he doesn't even know his strike rate in this IPL, and that stats don't matter at the auction.  

Gautam Gambhir is unaware of his strike-rate this year in the IPL...

After his eight runs from 18 balls used up almost 20% of the Kolkata Knight Riders' batting resources in their match against Kings XI Punjab on Tuesday, Gambhir's strike rate this IPL - according to my database -  is 124.1 against pace, and 134.1 facing spin.  Last season in the IPL, he was striking 115.7 against pace and 128.0 versus spin, suggesting that teams would do very well to bowl pace to Gambhir, as he is striking at around a 120 strike rate against pace and lost more wickets against pace than spin in this time period.  

To add context to these numbers, if a team had an overall strike rate of 120, they would score 144 in an innings, a total which rarely is good enough to win a T20 match.

Furthermore, stats certainly do matter at an auction, particularly that of the IPL, where money is thrown around virtually like confetti.  Despite Gambhir's assertion that "Three years back, Russell's stats were absolutely shocking", stats suggest this wasn't the case at all.

The West Indian all-rounder's batting data from 2013 show that he was in fact a magnificent striker of the ball.  While his two IPL innings in 2013 didn't amount to much (11 runs from 14 balls faced), he destroyed Big Bash bowlers scoring 203 runs in four completed innings from just 107 balls faced, while his data in the English T20 competition was also superb, with 240 runs scored in seven completed innings from 151 balls.

Quite clearly, with an average between these three leagues of 34.92, and a strike rate of 166.91, stats also indicated that Russell was a superstar in the making.  Perhaps if more T20 franchises used quality data, the number of expensive mistakes at auction would dramatically decrease...

Andre Russell's 2013 data, was in fact, magnificent...

Finally, to one of my favourite Tennis players, David Ferrer.  I don't usually have favourite Tennis players, largely because I can't quite understand the rationale of anyone who does - however, what I respect about Ferrer is his work ethic and determination to succeed, treating virtually every match like it is a Grand Slam final.  Sadly, Ferrer is likely to retire without a Grand Slam title to his name, but there can be little doubt that he will do so with the knowledge that he managed to extract the maximum from the level of talent he was given.

Numerous people on social media have been slating the Spaniard of late, suggesting that time to retire is virtually immediate, while Ferrer himself was quoted as being frustrated with his physical situation.

However, the markets are likely to find themselves struggling to price Ferrer in the short-term, and again, we can use data to find out exactly where he is at right now.


Service Hold %

Service Break %

Combined Hold/Break %

Service Pts Won %

Return Pts Won %

Combined SP/RP Won %

Over/Under Performance on Serve BPs 

Over/Under Performance on Return BPs 

Net Over/Under Performance on BPs





























































The table above illustrates Ferrer's numbers across a variety of different metrics.  

We can see that his level from 2012 to 2015 was relatively consistent, with combined hold/break numbers between 111.3% and 116.8% across all surfaces in every year.  Combined service/return points won between 105.9% and 108.9% was also pretty consistent in this time period, and these are very solid top 10 numbers.

However, a large drop-off started in 2016, and has continued into 2017.  If this rate of around 7.5% decline in combined hold/break continues each year, Ferrer's ranking is likely to be so low in 2018 that he will not be able to play a number of main tour events.  

David Ferrer's data has been declining rapidly in 2016 and 2017...

Having said this, we can also see that he has won 100.3% combined serve and return points won, so his hold/break percentage should also be over 100% as well, on that basis.  The reason why it isn't is the -4.2% underperformance across break points this year.  

As can be seen above, Ferrer's underperformance on return break points certainly isn't nothing new - he's under achieved in this metric in every season in the sample - but there is no doubt that this has had some impact on his 2017 results at the time of writing.  He's 7-9 for the season, when probably an 8-8 or even 9-7 record would be fairer.  

Ferrer's level right now is probably top 40-50, but will need constant re-assessment to see if this hold/break decline continues.