Badgers Cricket Club – The Statistical Sett

Affiliated to the CCC

As time has gone on my ramblings regarding the ‘behind the scenes’ details of the data used to build this web site have been scattered across the various pages of this site, with no rhyme nor reason. When I came to record the differences from the input of the full data for 1985 I decided it was time to create a single home for such trivia, and here it is. The links below break the information we have here down into several categories – currently a handful of unusual statistical categories that are not otherwise visible on the rest of the site, the changes made to the season averages by the retrospective entry of full scorecard information for seasons prior to 1988, other changes made to results between original publication and now, a record of those matches for which we know the full information is unavailable and, for more recent times, changes made to information in the scorebooks in the light of processing the bowling analyses to recreate the whole scorecard – some of which are still held on this main page of the Sett and some of which have been split off to their own pages.

The links above will either take you to specific areas of this page or to sub-pages amd should land on some sort of mini-index to the information therein. At some point, as the amount of information recorded here increases, I may well split it into more sub-pages – probably based on the list of categories above.


The Treasure Trove – some less obvious statistical gems

There are a number of statistical milestones that are much prized in the cricket world – centuries, either individual or in partnership, five wicket hauls – and we already have information on those scattered throughout the site. However, there are some achievements that are harder to pin down and either require considerable spelunking through the data, or in some cases aren’t readily identifiable at all.

Furthermore some feats can only be confirmed where there is full individual information in the database for each game, and this is currently only true back to 1984. We have overcome that requirement for best performances and best partnerships by reviewing every scorebook entry and manually creating data to be used to generate those particular pages but other more esoteric examples are dealt with as they crop up, to be included in match reports but then left to fester without further updates unless something similar happens again.

The intention of this section therefore is to gather together such gems into a single location and to try and keep them updated over time.


Bowling hat-tricks

This is one of the most prized and rarest of cricketing feats, but often does not get officially recorded anywhere. When Mark took one to finish off the Dormansland innings in 2019 I asked for anecdotal evidence of others and the subsequent responses have inspired me to create this section of the Sett and populate it with such information as I’ve been able to track down.

Hat-tricks for the Badgers
BowlerOpponentMatch date Overall figuresDetails
Pete LeggeSt. Johns13 May 19612-1-4-3 Pete took his only wickets with the last three balls of his first over, a maiden, all of them bowled
Allan ButtHayes Green23 Jun 196213-2-45-4 The second, third and fourth balls of his 10th over, the third of a second spell – two bowled and one caught at the wicket
Brian MooreMontrose13 Jun 196412-7-10-6 The second, third, fourth and fifth balls of his 11th over, with the score on 42 for 4 – the first caught at the wicket and the three golden ducks all bowled
Brian MooreMontrose2 Jul 19665-4-2-10 The fifth and sixth balls of his second over and the first ball of his next saw numbers four, five and six all bowled for nought
Brian MooreMontrose2 Jul 19665-4-2-10 The fourth, fifth and sixth balls of his fifth over removed nine, ten and jack for the fifth, sixth and seventh ducks of the innings, ending it on 20 all out – the first two were bowled, the last man LBW
Simon Fox Old Freemans14 May 198811-3-19-5 This one is purely anecdotal, from out of Foxy’s memory and cannot be properly confirmed because of the lack of relevant detail in the scorebook. From the fall of wickets and the batsmen’s scores it looks more than possible
Alan TicknerF&CO10 Jul 20047.2-0-22-4 With the opposition needing just seven to win with four wickets in hand as Al ran in for the third ball of his sixth over, the 38th of the innings, the cause looked forlorn, but (as eloquently described in the match report of the game) two skied catches and an LBW combined to put the Badgers in position for an unlikely victory
Mark GordonDormansland19 May 20193.5-0-13-6 The first three balls of Mark’s fourth over produced a caught behind and two clean bowled to end the home side’s innings
 


Streaks – both team and personal

This probably ought to have a proper page of its own, generated from the database, especially as I first wrote the code to calculate consecutive game streaks back in July 2009. However, it has never come to the top of the To Do list and maybe never will, so herewith a quick table of the top streaks, team only for now, in club history.

Longest streaks for or by the Badgers
Consecutive gamesByLength StartedBrokenNotes
Won(team)9 14 May 200517 Jul 2005  
Lost(team)7 26 Aug 196130 Jun 1962  
Drawn/tied(team)5 2 Jul 1977 6 Aug 1977 (two tied games bookended three draws)
  5 8 Aug 201019 Sep 2010  
Not winning(team)15 30 Jun 199526 May 1996 (6 draws and 9 losses, the last five consecutive)
Not losing(team)19 30 Jun 199118 Jul 1992 (7 draws and 12 wins)
Not drawing(team)57 27 Sep 201528 Apr 2019 (includes limited overs games)
With no streak(team)12 14 Jul 2013 4 May 2014 (arguably 13 because there was a No result in the middle)
 

Three or more outfield catches in a single game

This is one of those categories that can be readily determined where the full information has been added to the database but is harder to figure out for games prior to that because of the need to trawl through the scorebooks one game at a time. Since I was prompted to complete that exercise for catches by Daniel taking five in a game against Horley in 2021, then a summary of what I have discovered is presented here.

Frequency of taking three or more outfield catches in a game
FielderTimesNotes
Mark Gordon29 first in 1995, most recent in 2019, includes four catches on four occasions
Jake Gordon7 first in 2014, most recent in 2020
Alan Tickner5 first in 1973, most recent in 2005
Graham Ward4 first in 2003, most recent in 2021
Darrell Pitts4 first in 2007, most recent in 2021
Ian Gregg3 first in 2000 which was the first ever four in a game, most recent in 2011, includes four catches on two occasions
Ben Valentine3 first in 2010, most recent in 2014
Peter Ward2 first in 1962 which was the first ever three in a game, most recent in 1967
Brent Noble2 first in 1964, most recent in 1965
John Bailey1 in 1968
Pete Legge1 in 1970
Dennis Hopkins1 in 1977
Steve Pitts1 in 1978
Dave Clark1 in 1980
Steve Card1 in 1983
Adrian Cowell1 in 1986
Steve Ducker1 in 1989
Darren Hanley1 in 1993
Allan Butt1 in 1999
Pat Redding1 in 2005
Barry Davenport1 in 2005
Matt Mann1 in 2012
Rob Knew1 in 2019
Daniel Ward1 in 2021, the first ever five in a game
NOTE that wicket keeper catches are excluded from the above but may end up being recorded separately at some point in the future

Fun facts – from the statistical archives

This is a catch-all slot for other items that have been dug up over the years, but again is currently just a place-holder until I find a little more time.


Changes made after the fact to published data

One of the disadvantages of publishing the statistics more often than once per year is that mistakes creep in that only get picked up in the end of season cross-checking process. Fortunately in recent times that has generally been mitigated by various changes to the way we handle the scorebook data but this section details changes made under those circumstances, and also other alterations to figures that have been available on the site for some time but which have been changed as the result of investigations into particular areas or the manual filling in of missing information.


June 2020: Superbase replacement project

Thanks to the COVID-19 pandemic the summer of 2020 looked like being a complete wash out as far as cricket was concerned, although in the end we managed to get some games played in the latter couple of months of the season, but it had one small positive from the Badgers web site point of view in that Darrell and I decided to spend at least some of the Sunday afternoon time that we would have spent cricketing in doing things related to the website. This started out with my explaining how the existing setup hung together, making sure that Darrell could access all of the source material, making sure that he could use and understand the tools that I used to pull everything together and generally just trying to ensure that, morbid as it might sound, he would be able to take over were anything to happen that meant I could no longer act as webmaster.

Once the initial burst of enthusiasm had waned somewhat we agreed that the best use of our time was in eliminating the ancient 16-bit Superbase for Windows from the tool chain. Superbase, in its various forms down the years, has been the bedrock on which the Badgers database has been constructed – dating all the way back to 1988 and the Atari ST on which the first thirty seasons of averages data were initially digitised and the thirtieth season report produced. However, it has become increasingly difficult to keep a 16-bit program, with no meaningful official upgrade path, running in a 64-bit world (requiring a 32-bit Windows 10 virtual machine running on my 64-bit Windows 8.1 system under Hyper-V).

The database itself had been cloned into a mySQL (now MariaDB) equivalent more than a decade ago, so that I could start working on dynamic web pages, but it was the 30,000+ lines of Superbase Basic spread across 67 different programs that presented the road block to getting rid of the program altogether. After some strategic thinking and a brief discussion we agreed that the best bet was to rewrite the essential (12 programs that accounted for roughly half the lines of code) and useful (another 25 programs of just over 6,000 LOC) routines in Python, primarily because it might be beneficial for Darrell to learn the language for his working life and as one of the top half dozen programming languages in the world in terms of various measures of popularity, and open source to boot, it represented a mainstream long-term base for the web site in the future.

Inevitably rewriting code throws up anomalies that need to be reviewed and either fixed (nothing to see here, honest) or accounted for. Sadly one of the errors that I found, whilst rewriting the code for the season record included in every season report, was rather egregious in that the average per game calculation for both total runs scored and total runs conceded was using the number of games played up to but not including the season being reported upon. This has now been fixed so that all such season report sections, like the 2019 version, now show the correct averages per game for all seasons combined.

That is the most serious issue I found and many of the problems identified were of lesser importance and had no material effect on the information currently visible on this web site. For example, the new season averages calculator threw up a few cases where the, only recently tracked, wicket keeping details did not cater for the new flags introduced when I added support for a change of keeper in mid innings, so that was fixed by awarding each player who took the gloves half a game kept instead of a whole one.

The program also processes the fall of wickets data to build the best partnerships for each wicket in each season and that too threw up some fairly recent anomalies related to player retirements, which turned out to be bugs in the existing logic. Fortunately none of the stands involved were considered significant enough to show up on either the associated partnerships pages or anyone’s personal pages, so in this case it was just the database data that was out of whack.

However, I have come across one or two things that needed to be corrected here – the most significant of which was the fall of wickets for the June 2000 game at Loose, where David Aldwinckle scored 48 unbeaten runs batting at number four, but the fall of wickets had him fifth out. Checking back with the score book I suspect that I must have perpetrated this one as it looks to me like the dismissed batsman numbers for the last five wickets to fall were added after the fact. To make matters worse, not only is the scorcard wrong but the eighth wicket stand was the highest scored in that season and has been credited to Greggy for the past twenty years, when it should have been David who shared the glory with Graham Davenport.

The Blindley Heath game in September of that same year also has an error in the number of the batsman dismissed for the fifth wicket, showing 5 when it should have been 6, but that has fewer implications since it was the last wicket to fall and the last pair only added 20 odd.

One other item that was fixed during the same time period, but doesn’t really relate to the project, was that I corrected which Phil Smith played in the Reigate Cavaliers game in 2010. The odd thing is that I had spotted at the time that there was already a Phil Smith in the database, communicated with the new Phil to get a distinguishing middle initial and created a unique entry for him but then chose the wrong entry when I entered the data {sigh} As a side effect I have also corrected a Blindley Heath game on Play-Cricket where the wrong Phil Smith had been selected when they entered the match details (looks like someone else has been busy on data related work during lockdown).

Credit for spotting that anomaly originally goes to Wardy, who apparently commented during one of the lockdown quiz teleconferences that he felt my mention of the gap between catches by Phil – in last year’s Stoke D'Abernon match report – was almost certainly spurious because the Old Sutts Phil was unlikely to have turned out for the Badgers before Wardy himself had ever done so. Darrell reported this back to me at some point but it didn’t trigger a response until I was checking up on parent sibling relationships following Nat Cibardo’s catch off his Dad in the 2020 Newdigate game and noticed that there were two different Phil Smiths in the players table but that one of them had no games played.


November 2016: corrections from best performance filling in

As a side effect of building a manual database of best performances for seasons prior to 1984 there were a significant number of best bowling figures added to the season averages – too many to detail here – plus one correction, to Phil Walters’ best bowling figures for 1977 which had been incorrectly entered when the historical data was first input back in 1988 as six for 24 rather than the six for 25 shown in both the scorebook and the 1977 end of season report.

In addition the high score details for Brian Hodges in 1959, Brian Moore in 1974 and Brent Noble in 1979 were changed to reflect that they were not out in the innings concerned.


October 2016: corrections from end of season checking

The end of season checking for 2016 involved a three-pronged approach with each Play-Cricket results page being compared by program code to the Badgers database and a similar process performed between the database and the scorecard checking program’s version of events. Once that was completed the scorecards were printed out from the database and Darrell and I had the lovely chore of checking that version of events with the paper scorebook. Unfortunately, but not unexpectedly, that process found 75 discrepancies which needed to be investigated, most of which were errors in the scorebook that had been corrected by me whilst checking the photographed version of the book without physical acccess to it to make corrections.

All of the anomalies have been checked and double-checked before being resolved according to the best information available, normally involving a review of one or both scorebooks. For the record, the following mistakes were found and corrected:


October 2015: corrections from end of season checking

The end of season checking, for the first time since I’ve been doing the averages, did not involve trawling through the scorebook(s) – which was a good thing as I had no willing volunteer(s) to help with the process this year. Instead, because I’ve processed every game played through my scorecard checking program on a week-by-week basis, I decided that it made more sense to check the data in the database against the output from that program. Since the weekly process involved manually entering and/or cross-checking the output from that program into Play-Cricket, that was where errors were going to be introduced. This involved writing seven hundred or so additional lines of code into the program to read the database version of events and perform the comparisons.

The resultant validation found four discrepancies, which was slightly more than I had anticipated. All of the anomalies have been checked and double-checked before being resolved according to the best information available, normally involving a review of one or both scorebooks. For the record, the following mistakes were found and corrected:


28th Nov 2014: corrections from Play-Cricket scorecards

All seasons with complete results entered into Play-Cricket (1986-7, 2004, 2008-14) now have full Extras details, where they are available from the scorebooks, and some information held by Play-Cricket but not historically part of the Badgers’ database (eg. boundaries hit, balls faced, wides and no balls bowled) have also been filled in for those seasons (although not yet exposed anywhere on this site).

There were also two minor corrections made on the back of re-visiting those scorecards, with Graham Davenport being awarded a maiden that had been missed during the 2004 Woldingham Village game and Daniel Ward allowing one less run during his bowling at Reigate in 2012 where a bye had been incorrectly recorded against his analysis.


30th Oct 2014: corrections from end of season checking

The end of season checking found around twenty discrepancies, most of which were a case of the paper scorebook not being available to update when I made decisions about discrepancies or logical inconsistencies during the course of the season. All of the anomalies have been checked and double-checked before being resolved according to the best information available, with a couple more games being fed through my scorecard checking program to validate the numbers. For the record, we found and corrected the following mistakes:


27th Oct 2013: corrections from end of season checking

Our checking did find the better part of fifty discrepancies, many of which arose because, absent any evidence to the contrary, I generally chose to go with the CricHQ version of a card when there were differences with the paper scorecard but often without changing the latter to bring it in line, whereas the checking was done from the scorebooks. All of the anomalies have been checked and double-checked before being resolved according to the best information available, which sometimes involved feeding the bowling from a whole innings through a program that I created which then builds the batting and fall of wicket details. For the record, we found and corrected the following:

For completeness, the following differences between CricHQ and the paper scorebooks were resolved by accepting the data entered on the tablet, which matched the data already entered on Play-Cricket and reported here, because of lack of proof that one or other version of events was correct:


Adjustments made to statistics after entry of full season data

When this site was first put together I had ‘full’ data (ie. details of all Badgers batting, bowling, catching and fall of wickets) for games going back to the 1988 season and in recent times I have slowly started to back-fill the preceding seasons. For those wondering why I would be doing such a thing, when I first created the Badgers database back in 1988 (using Superbase Personal on an Atari ST with a massive 1MB of RAM and 1.44MB floppy disks for permanent storage!!), in order to produce lifetime averages for the thirty year report, I minimised the chore by only entering full data for that season (since I also needed to create a season report for 1988) and using the existing season average reports to provide the bulk of the information for the preceding 29 seasons. This had the effect that when I created the historical areas of this website the seasons were split into two groups, with the older ones not appearing for other than the batting and bowling averages at that time (summary results for 1959 to 1987 were added in September/October 2002 and other gaps have been filled in since) and the newer ones having more complete information either from the beginning or as features were added that required full data.

The areas affected by that missing data are fairly minor these days (because I’ve found ways, often involving a lot of scorebook spelunking and manual data entry into the database, to mitigate the other elements that used to be limited, including partnerships, individual batting and bowling performances, fielding dismissals and games played) but there are facts that cannot be identified without full data (eg. games bowled, average batting position, plus games played and best bowling for those not added by hand). It has always been my intention to add those missing details over time, although progress has been slower than I’d have liked it to be and enthusiasm for the task is often tempered by the fact that the data will never be complete due to the missing 1981 scorebook.

The idea is that as each historical season is completed (by entry into the ECB-supplied Play-Cricket system) I will update the Badgers database (now held in MariaDB rather than Superbase) with the full details and then compare the resultant season averages with those that were manually entered thirty-two and a half years ago. Once I’m satisfied with the accuracy of the data then this site will be updated with the statistics, scorecards and revised results details. No promises on when 1983 will appear, as there is considerable effort required simply to enter the information from the scorebooks into Play-Cricket, let alone the cross-checking and other peripheral activities that are necessary before the information lands here.

Almost inevitably the entry and processing of the old scorecards throws up a number of discrepancies in the end of season averages for that year. Each of those has to be investigated but we are generally left with a number of changes to the previously published figures, which are now documented here for the official record:


Adjustments made to statistics after entry of 1984 season data

The full 1984 season details took a long and tortuous route to get here, with the first game being entered into Play-Cricket some time after the end of the 2016 season, that process stalling because of discussions with the API developers around player naming in early 2017 (they didn’t like me using year numbers in the player names), not picking up again until after the 2018 season when I completed the data entry but then fell foul of the fall of wickets issue at Hook Norton, expanded on below. The complexity and wide-ranging nature of the fixes for that meant that it wasn’t until August 2020 that I have finally put myself in the position to finish this off.

The following changes were made to the 1984 statistics following the entry of the scorecards for that season into Play-Cricket (the alterations described are also generally marked on the image of the scorebook page using an obviously computer generated text overlay in red):

This has the following knock-on effects that show up as differences between the manually entered season averages and those generated from the newly added scorebook details:


Adjustments made to statistics after entry of 1985 season data

Generally the problems found during the entry of full data are either tallying errors in one of the original score sheet or the season summary, or the incorrect transposition of data between the two. However, with 1985 there were a couple of unusual issues that cropped up multiple times. Firstly, whoever did the summarising chose to ignore partial overs, so we have a number of bowlers with extra balls bowled. Secondly, this was one season that had a number of long time Badgers amongst the also batted or bowled lists and thus I entered a number of additional season statistics records manually. Unfortunately in several cases I mistakenly included performances from the President’s eleven game, which obviously should not count towards the averages.

The following changes were made to the 1985 statistics following the entry of the scorecards for that season:


Adjustments made to statistics after entry of 1986 season data

The following changes were made to the 1986 statistics following the entry of the scorecards for that season:


Adjustments made to statistics after entry of 1987 season data

The following changes were made to the 1987 statistics following the entry of the scorecards for that season:


Matches with missing details

Unsurprisingly not all scorecards are filled out with equal zeal and that has meant that down the years we have some games with incomplete information, and others that might have. Some of those situations have been called out in the section above, but others are more recent and were never flagged anywhere. This section aims to rectify that and will be added to whenever I find anomalies.

OpponentMatch dateProblem(s)
Newdigate4 Aug 1984 just five players recorded in our batting card and only five others can be derived from the bowling card
Wallingon Cottagers25 Aug 1984 no modes of dismissal or fall of wickets recorded for our innings
Englefield Green27 Aug 1984 just three players recorded in our batting card and ‘only’ seven others can be derived from the bowling card
Sun Allon15 Sep 1984 just eight players recorded in our batting card and only two others can be derived from the bowling card
Blindley Heath14 Sep 1985 just four players entered into our batting card and only four others can be derived from the bowling card
Old Xaverians20 Sep 1986 just seven players in the batting card, only three can be derived from the bowling card
Nat West Bank19 Sep 1987 just ten names in the batting card, with the number ten slot left blank
Nat West Bank14 Jul 1990 no catchers were recorded in our scorebook, although somehow they have been assigned in the database (possibly erroneously, but too late to tell)
Southall Muslims25 Aug 1990 no catchers were recorded in our scorebook, although somehow they have been assigned in the database
Dormansland18 Jul 1992 our eleventh man (who I seem to remember was a youngster and not associated with the Badgers) has only been recorded as Ben and remains immortalised as the only entry in the Players table in the database with a wholly alphabetic code
Crondall 9 Jun 1996 a member of our team has been recorded ever since as AN Other, but the scorebook shows just James. I suspect we borrowed him from the opposition but why we ended up without a full name is a mystery. He has been given his own identity in the database now though