A summary of the analysis on this blog:

Quantifying the link between Test and First Class averagesCC D1 avg of 40 is worth 31 in Tests
Explaining the Underperformance of Overseas batsmen in County CricketEnglish conditions mean SA / NZ / Aus batsmen find it easiest to slot into county cricket
Using CricViz False Shot % as an alternative to AveragesProportion False shots times wickets per False Shot times Strike Rate equals average
Batting: 35 is the new 4035 is a decent Test average. Most players averaging 35-40 should keep their place
England’s current Test batting, in the context of the last 20 yearsTechnique for putting a team in context – take sum of top seven players’ averages, adjusted for age
Kohli’s ODI run ranges are as expected for a phenomenal batsmanIgnore a player’s distribution of scores. Focus on their average.
Should Jennings’ expected average be reduced after the Bridgetown Test?Take a big sample size over a smaller one. There’s good opening bowlers in county cricket, don’t rush to judge someone based on a few Tests.
England Lions – red ball selection analysis“A” teams should pick players close to Test standard, not just exciting young guns
Modelling tail end battingExpected value of the tenth wicket: Weaker batsman average + 20% of the difference between both batsmen’s averages. Link and Link
Home AdvantageHome advantage gets bigger as a series goes on Home advantage is worth 14% on average. Link and Link
University matches – statistical analysisThe gap between counties and universities is about 73%. Too significant for these matches to be useful practice.
ODI trendsHigher averages allow more risk taking which drives up strike rates
ODI boundary hittingSix hitting is rare (<2% of balls) until after the 35th over
ODI dropped catch impactImpact of dropped catch is biggest at the start of the innings
Matchups in ODIsOpening with a spinner reduces the rest each bowler gets in the innings. Also features a nice visualisation of bowling spells.
Pool together associate nations to make a competitive teamLink
Analysis of England ODI players at Lord’sThe regional county cricket structure means northerners may have very little experience of southern grounds
Should Buttler bat up the order in ODIs?Move Buttler up the order if second wicket falls after 30th over. Generally, be more fluid with batting orders
Comparing Jofra Archer and Ryan HigginsRyan Higgins is massively under-rated. It’s easy to miss players who chip in with 2-45 every time.
Could Woakes open the batting in Tests?I concluded “maybe”. This is interesting, because it would now be easy to answer with my ball-by-ball database sourced from Cricsheet.
Lyon vs Moeen Ali by inningsLyon amazing in the first innings, not great in the 4th. Ali the opposite. I missed the point that fourth innings spin gets more rewards in England that Australia.
Do hundreds matter?No.
Sibley or Roy?Openers on debut score three runs more per match than expected: feel free to throw players in if they’re good enough.
Mythbusting: Vaughan and Trescothick selected for England despite modest First Class RecordsIt’s a myth
England’s bowling optionsBowling options. England have gone on to pick most of these players – showing they are (mostly) using county data
Debutants in Away Tests have shorter careersStart away from home, have a weaker average, have a shorter career (avg six Tests vs nine for home debutants)
Recovery time in One Day CricketFewer than 2 rest days between 50 over games costs you 10 runs in the field
Northamptonshire’s Promotion DriversNice template for reviewing why Northants had such a good season.
Duck analysisNot sure what my point was.
On the decline of Test Batting being driven by T20Top order Test batting declined after white ball scoring speeds increased
Batting: All County Cricketers RatedDan Lawrence was 46th on the list, now in the Test team
County grounds ranked by ease of battingFC averages should be adjusted for where someone’s home ground is. Surry easy. Canterbury hard.
Opening batsmen: the divergence of ODI and Test playersODI averages now a poor predictor of Test averages
Boost to average from batting down the orderBoost to average from batting down the order
Getting your eye in across formatsHas a chart for the three formats
The case against Zak CrawleyI savage the selectors. Then wonder if they might have played a blinder when Crawley hit 267. I end up proposing to use judgement to rate players within their range of plausible averages (see error bars analysis below). Link And Link
 Rating Wicket KeepersUsing byes per 100 overs, % of dismissals, batting average to judge keepers
Test batting: do some players gain an extra boost at home?Unable to find extra home advantage for openers when playing at home.
Some flawed analysis of leg spin and off spinI misunderstood some Cricinfo data showing what scores batsmen were on when dismissed (I thought it was their average). Link and Link
Best bowlers of the last 50 yearsMethodology to appraise bowlers based on who they got it. This is a way of going beyond averages (which are a bit blunt). This approach calculates what the impact a bowler had on a batsman’s performance.
How many innings before an average is accurate?Averages after 30 innings are surprisingly weak indicators of averages over the rest of a player’s career. Even 30 innings are too few to judge a T20 player’s career SR Volatility by year in average can be explained by low sample sizes. xW can smooth that out
How much recent data should you use when assessing a bowler?Last four years looks like the best balance between sample size and stale data
Reasons to play in “A” toursFour reasons to pick players: Stay sharp in off season, suited to those particular conditions, gain tour experience, or just people already in the country after white ball tours.
How to win a super overStrategies for Super Over success, based on analysis of all Super Overs to date
Test cricket’s evolution and professionalismI have a theory that as the talent pool increases there are fewer extreme performances. My analysis showed Test cricket improved markedly from 1930 to 1960 on this measure.
Adding error bars to averagesFormula: 1 Standard Deviation = Average x (Wickets ^ -0.5)
Batting ability in Test cricket is not normally distributed (it just looks like it is).A poorly explained argument of something I think is true.
IPL analysisT20 death overs scoring rates are driven by boundary %. But – a handful of batsmen can strike 150+ with low boundary percentages. Big hitters should refuse singles.
Do right-left pairings score faster in ODIs?Right-Left combinations don’t impact strike rates
Test partnerships – does it matter who bats with whom?No.
Pace bowlers struggle in back to back Tests7% increase in bowling average if playing in back to back Tests
Test bowling averages by month in EnglandSpinners average measurably less as the English summer goes on
Top innings bowler in TestsOne for the gamblers – trends in the top bowler market
Conferences: what we can learn from the 2020 Bob Willis TrophyPutting 18 counties at the same level would leave too big a gap between FC and Test standard in England.
Rating the Blast teams … Lancashire’s batting is better than it looksA strong bowling attack means small chases, which can trash batsmen’s stats by giving them lower Strike Rates than they would otherwise have.
The ODI “who is winning?” formulaStole a formula from @cric_analytics and stuck it in my model as a check
The perils of batsmen switching countiesMoving counties late in career negatively impacts a batsman’s stats, possibly through loss of home advantage
T20 batting: running out of steamBalls per wicket data can be used to work out in running how aggressive batsmen can be in T20 without having to resort to letting the tail bat
DRS: The story so farMyths and truths from analysing a database of DRS data
Country vs Country matchups in Test CricketHome advantage for specific pairs of countries

Note that this list excludes series/match previews, which you can find through the “Preview” tag if you’re that way inclined.