Posts

CricViz Analysis: How Sri Lanka Can Beat Australia

Ben Jones analyses the key areas where the tourists could steal a march on Tim Paine’s side.

Sri Lanka arrive in Australia as afterthoughts.

That is not a slight against their hosts, who have had more than enough to think about in the last nine months. It’s not a slight against the media, who are justified in focusing on this Indian side, the world’s best team achieving something truly historic. And it’s certainly not a slight against Sri Lanka.

Yet Dinesh Chandimal’s side have won all of their last three Tests against Australia. Of course, those matches took place in home conditions – the pitches in Colombo and Galle a far cry from Canberra and the Gabba – but it’s remarkable how quickly this has been forgotten.

Equally, Australian cricket hasn’t been at a lower ebb in the last 30 years. Whilst the bowling line-up still has something of an aura, the gulf in ability between the two batting line-ups is not that great. A combined top seven may include more Sri Lankans than Australians. The tourists may struggle, they may thrive, but right now, if it’s a match involving Aussie cricketers, it’s an unpredictable affair.

So let’s put aside the idea that Sri Lanka are no-hopers. They can win this series, and would be doing themselves a disservice if they didn’t try to eek out every fine margin, to try and sneak a Test victory through the back door. So here are are the five ways – according to CricViz data – that Sri Lanka can compete in these two matches.

BOWLING PLANS

Usman Khawaja is Australia’s best batsman. He’s their most attractive stroke-maker, their most substantial run-scorer, their most likely candidate to step into the void left by his former teammates. Yet asked to step up this summer, he’s not been up to the task, and Sri Lanka will be eager to keep his subdued run going.

For the spinners, it’s tough, because Khawaja is famously strong against slow bowling in Australia. However, there is a chance to get one over on him – but the key is in the detail. If you are going to bowl spin to Khawaja you have to have the ball turning away from him, i.e. an off-spinner or a left-arm wrist spinner.

If you’re going to bowl pace at Khawaja, it’s better to err on the side of being too full rather than too short. He is an extremely competent puller (averaging 161) and cutter (averaging 122), and dropping your length back will be punished. Turn it away from him – and don’t drop short.

In the last decade, the average opener is dismissed every 72 balls against seamers with the new ball in Australia. So far in his career, Marcus Harris has recorded a figure of 93.3, and that was against an extremely good Indian bowling attack. He is a young man of clear talent when it comes to seeing off the first hour. Against seamers, Harris has only been dismissed by balls above 140kph – touching the upper end of the speed gun could be important in removing the Australian openers, Lahiru Kumara the most likely candidate. Raw pace is needed.

Tim Paine is an effective lower-order batsman. Australia’s tail is in no way a weakness, but the ability of Paine to bind together the two halves of his batting line-up is significant – but he struggles against pace. Indeed, he’s dismissed almost twice as often when facing pace rather than spin.

Interestingly, Paine is extremely good against balls on his stumps. He averages 42 against balls from seamers that would have hit – all other top-order players in the Tests Paine has played in average 14.35 against those balls. No, the big difference-maker when bowling to Paine is not line or length, but seam movement. The average seam movement in Australia is around 0.75° – when bowlers exceed this, Paine’s average plummets. Chandimal should turn to whichever bowler he feels is getting the most off the pitch when Paine comes to the crease. Seam is key.

By contrast, Matthew Renshaw is weak against the deliveries Paine dominates. In his first stint in Tests, Renshaw struggled against balls targeting his stumps, averaging just 9.75 against deliveries from seamers that would have gone on to hit. The average for the other openers in the games he’s played is 20.30. Target the stumps.

To Travis Head, a patient approach is advisable. Repeatedly in the India series, Head was dismissed to deliveries outside his off-stump from the seamers. Indeed, to good length balls in the channel, he averaged just 10.50. As such, Sri Lanka should try to keep him away from spin, and continue to punish his weakness outside off. Hang the ball on a wide line, and wait.

To date, Joe Burns holds an even record against both pace and spin, but on closer inspection he’s really struggled with full and good length balls from the seamers. Against short balls he’s averaged 57, but anything fuller and he’s averaged 24.09. It’s an occupational hazard for an opener (five of the seven dismissals were in the first 10 overs), but it’s still an area Sri Lanka may be minded to try and exploit while the ball is still fresh. Pitch it up, and the Sri Lankan seamers may get some success.

FINDING THE RIGHT LENGTH

Anecdotally, when overseas seamers arrive in Australia they can get carried away with the pace and bounce on offer and struggle to find a threatening length. In reality, the story is different depending on the pace you actually bowl at.

In the last year, Suranga Lakmal’s average speed has been 130kph. You can still have success bowling at that speed in Australia – James Anderson and Vernon Philander have shown this recently – but you need to be precise with your length. In last year’s Ashes, Anderson hit a good length with 47% of his deliveries, the highest figure of any seamer on show. Don’t get tricked into thinking you’re fast and nasty because of the carry through to the keeper. Lakmal cannot forget that as a default rule a ‘good’ length – that is, 6-8m from the batsman’s stumps – is a good length. Keep hitting that in-between zone.

For Lahiru Kumara, it’s a different story, because in the last 12 months, Kumara is the sixth fastest bowler in the world.

What’s more, he puts that pace into bowling wicket-taking deliveries; of the six seamers to appear for Sri Lanka in Tests over the last 12 months, Kumara has the lowest Expected Dismissal Rate (according to CricViz’s Wicket Probability Model). Lakmal may be a more rounded bowler, but Kumara has a very high ceiling.

This relates to his bowling length because his extra pace gives him more options. It’s still statistically better to hit a good, in-between length, but the second best option is to bowl shorter, rather than fuller – the opposite of bowlers of Lakmal’s pace. This gives Kumara more tactical variation.

SPINNERS SPEEDS

As an off-spinner visiting Australia, the key is your speed, because you have a tiny margin for error. It’s the Goldilocks conundrum – too hot, or too cold, and you’re in trouble.

Nathan Lyon understands this. For the last five years 58% of his bowling in Australia has been in that middle bracket. He has adapted his bowling to suit the conditions he plays in most – but it’s taken time, and Sri Lanka’s spinners need to go through this process inside a month.

Dilruwan Perera is the most suited to bowling this pace – across his career 45% of his deliveries have fallen into that speed bracket. However, whereas Lyon uses the slower ball as a rare variation (just 11% of his bowling at home is slower than 83kph), Dilruwan drops below 83kph with 48% of his deliveries. Clearly that is a product of his home conditions, but it’s a big mechanical switch to suddenly remove those deliveries from your game. It could be even more of an issue for Dhananjaya de Silva, who has bowled 68% of his career deliveries at that speed.

However, if that speed issue can’t be resolved through coaching, there is a tactical way out. If you’re going to bowl that slowly, then you cannot get drawn into thinking you can toss it right up and beat the batsman above the eyeline. It’s counter-intuitive, but it’s the case. For spinners, it’s not a binary issue of being allowed to pitch it up more the slower you bowl – the slower you are, the more accurate you have to be.

ATTACKING THE NEW BALL

One of the few areas in which Australia were better than India was in their new ball bowling. Despite India’s superior opening batsmen, Mitchell Starc and Josh Hazlewood consistently took wickets and outperformed their opposite numbers.

Thus, the absence of Hazlewood for this series is significant. The New South Wales seamer has taken the new ball in 72 of the 82 innings in which he’s bowled for Australia, and he leaves a hole in the attack that needs to be filled. There has been talk this summer that Pat Cummins deserves a shot with the new ball (he’s taken it just nine times in 33 Test innings), but it was largely suggested that he replace Starc, not Hazlewood. You’d imagine that Paine and Justin Langer will be reluctant to open up with the relative inexperience of both Cummins and Jhye Richardson, so it’s likely Starc will retain his spot with the new ball. An out of form bowler and one new to the job will begin proceedings for the hosts.

So, Sri Lanka need to go hard at Starc, because he’s vulnerable. Since the start of 2018, attacking shots against the left-armer’s new ball deliveries have averaged 69.50, leading to a dismissal every 45.5 shots. As shown below, that is a worse record than most established bowlers.

Yet as you can see, Cummins has a great record in this regard – an attacking shot against him is very likely to bring a dismissal. As such, Sri Lanka’s openers need to be willing to go after Starc, given it will make it easier to soak up the pressure from Cummins at the other end. If you lose wickets to Starc early on when trying to punish the bad balls, you shouldn’t worry – it’s the prudent approach.

This should inform Sri Lanka’s selection because to execute this strategy, Karunaratne needs a strokemaker alongside him. In the practice game against the Cricket Australia XI, the visitors opened with Lahiru Thirimanne, a man who has played 29 Tests and averaged 23.56 – hardly world-beating. What’s more, he’s rarely been fluent at Test level, playing just 18% attacking shots, not exactly a man well-suited to an aggressive strategy.

A far better option would be to promote Dhananjaya de Silva. Despite never opening in Tests, Dhananjaya has opened on 57 occasions in FC cricket, with an average of over 40; more significantly, he is perfectly suited to the role of taking on the new ball. He has has the highest average with attacking strokes of anyone in the Sri Lankan squad, as well as the best dismissal rate. Given that other more traditional opening candidates have clear issues at this stage (Sadeera Samarawickrama averages just 8.62 against pace bowling in Tests), a bold decision is required.

USING YOUR FEET AGAINST LYON

Disrupting Nathan Lyon is hugely important for any side facing Australia. Within Paine’s bowling quartet, Lyon bowling lots of economical overs is fundamental; if Lyon can’t bowl, the knock-on effects are severe, forcing the fragile Starc and Cummins into longer and longer spells.

The key to playing Lyon in Australia is to come down the track. Across his career, batsmen coming down to Lyon average 32.29. His overall average in Australia is 33.31, so it’s not more effective in a purely numerical manner, but the scoring rate for these shots is 7.06rpo. This puts Paine in a difficult situation. With the likely absence of an all-rounder other than Travis Head, Lyon is Paine’s only relief option. He has to make a call – wait for the breakthrough, or take Lyon off.

What’s more, this tactic naturally suits Sri Lanka. Their batsman come down the track more often than most – as shown below. What’s more, the average in Sri Lanka is to come down to 8.5% of deliveries, so this isn’t a measurement informed only by home conditions. This is a group that have the natural inclination to play in this manner.

Conclusion

Sri Lanka have a chance – but that’s all it is. If they have good plans and execute them well, Australia may not be strong enough right now to resist. However, the hosts are still likely to come out on top; the absence of Angelo Mathews may come to be crucial both in terms of batting heft and bowling depth, and Langer’s revamped squad may just be fresh enough to allow them to move on from the India defeat. Regardless, Sri Lanka should go into this series believing that the smash ‘n’ grab is on – because right now, everything in Australia is up for grabs.

Ben Jones is an analyst at CricViz.

CricViz Analysis: Who Should Australia Pick?

Ben Jones considers who the Australian selectors should be 

calling ahead of the Sri Lanka series.

In the coming days, Australia are expected to announce their Test squad for the home series against Sri Lanka. As a series in isolation, it shouldn’t represent the toughest challenge for Tim Paine’s side, but the Tests will be as much about finalising Ashes preparation as they will be about victory in Brisbane and Canberra. Because Australian batting is in a bit of a state.

Two of their last three series have seen no centuries from Australian batsmen, and the collective batting average they recorded in 2018 is the lowest for any calendar year since 1978. India were better than them in pretty much all departments throughout the four match series, but ultimately the series was lost in the final two Tests with the Aussies unable to respond to India posting substantial first innings totals. Broadly, the batting was to blame.

And it’s because of this that the team Australia choose for the Sri Lanka series is a big deal. With the potential addition of Steve Smith and David Warner, this side will form the guts of the touring party for next summer’s Ashes. True, the conditions at the Gabba will be rather different to the greentops England will prepare at Edgbaston and Old Trafford, but this is as close to a rehearsal as Australia will get. They need to get it right.

Some players will be retained. Marcus Harris has the highest average of  any player to appear more than once, and has generally shown a level of control at the crease that suggests he can maintain his moderate success, and improve on it. Tim Paine has had a decent summer – though perhaps his rise has been overstated given what preceded him. There is little case for him remaining in the side from a cricketing perspective, but he seemingly has the backing of the players, and in a disrupted period for the side that has significant value. Usman Khawaja is one of very few senior players in that side, and whilst his form dipped briefly in this series, he is clearly a good player.

But the others’ places should be up for discussion. Travis Head may have top scored but an average of 33.85 is not substantial. Shaun Marsh failed to step up when younger, less experienced teammates needed him to. Peter Handscomb’s technique is still questionable. They are not out of contention, but their place should not be assumed.

The question of who to replace them has been a hot topic in the past weeks, and one phrase has come up time and time again. Many have complained that ‘nobody is banging the door down’, the selectors frustrated that they can’t simply call upon batsmen averaging 75 over three Shield seasons.

But if a player is performing in this manner then, frankly, a selector’s job is all but unnecessary. Picking a player in that sort of form doesn’t require any skill or judgement. Weighing up relative strengths when a player has weaknesses as well is quite literally the job. To suggest that a player must present a perfect case before they are allowed into the team is perverse. Particularly when that team has won just one of their last nine Test matches.

Sure, no player is banging down the down, but plenty are knocking pretty damned hard. You’ve just got to know what you’re listening out for – and that’s where data can help.

So many names have been bandied about in the past few weeks that it pays to keep an open mind, so let’s start with a big pool of candidates. Our shortlist of 24 Sheffield Shield players possesses a wide variety of skills and experience; there are classical red ball performers in there, T20 specialists, youngsters and veterans. We’ve then looked at their data for the last three Shield seasons, including this current season, to try and work out the merits of their case for selection. The data we have included extends beyond the raw numbers of runs and averages (though they are included), offering a more in-depth assessment of each player’s relative quality.

Welcome to the CricViz selection meeting.

Runs and Balls Faced

Firstly, let’s look at the most basic statistic available – how many runs these chaps have scored since the start of the 2016/17 Shield season. It’s a measure which rewards those to have been consistent across a prolonged period of time, accumulating runs not in clumps or purple patches but just as standard. It penalises players to play international cricket of any form, given those matches eat into the Shield season.

Of course, it’s not all about runs – as Australia have seen in this series when bowling to Cheteshwar Pujara, occupying the crease is key. The ability to face lots of deliveries is just as important as being able to score off them.

If we rank each player on our shortlist by their position in these two charts, then average out their ranking, then we can make an overall ranking. Kurtis Patterson was ranked 2 for runs scored, then 1 for balls faced, so his average ranking is 1.5. Thus, our table looks like this.

Batting Averages and Dismissal Rates

Of course, runs and balls faced only measure some things. They measure quantity, but not always quality, and focusing just on that area unfairly penalises international players, whose commitments with Australia mean they miss out on matches. So of course we consider batting averages.

We can also look at the average length of innings for all of our players. The average dismissal rate takes the number of balls faced and divides it by dismissals – what you’re left with is a good measure of a batsman’s ability to occupy the crease over time.

When we factor in the rankings for Batting Average and Dismissal Rate, our list changes.

False Shot Percentage

The way that a batsman makes his runs is also important to consider. If a player is edging or missing the ball a lot, they could be riding their luck with an average of 40, whilst a player with an average of 30 could be suffering a run of poor fortune.

We can try to understand a bit more of what’s going on by looking at false shot percentages. This measures what percentage of a batsman’s deliveries does he miss or edge – essentially, a measure of control. The average false shot percentage for top-order players in Shield cricket is 13%, so anything below that is a sign of quality.

When we add in these rankings, our selection table looks like this:

Average Comparison

Sometimes, averages aren’t always indicative of a player’s quality. If you play exclusively on friendly wickets, or in high-scoring games, there is a reduced value to your runs. As a result, it’s helpful to compare how a player has done relative to other batsmen who’ve played in the same game.

For instance, George Bailey has averaged 38.88 in Shield cricket over the past three seasons. Moises Henriques has averaged 42.18. But the average for all top seven batsmen in games where Bailey’s played was 32.53; for Henriques, that figure is 37.25. This shows that generally the pitches Henriques have played on have been more batting friendly, whilst Bailey has overperformed compared to others in the games he’s played. Bailey has averaged 6.35 more runs per dismissal than everyone else in his matches, whilst Henriques just 4.93 more. Henriques has averaged more, but Bailey’s been worth more.

We can find this figure for all of our contenders, and get a better picture of when they have made their runs.

When we add this to our table, the list looks like this:

Consistency

It’s important to be able to play a wide range of bowling if you are to succeed in Test cricket. Often players can have exorbitant FC batting averages against either pace or spin, hiding a weakness against the other, and while players will always have a preference for a certain style of bowling, consistency across the game is a valuable asset.

When we add this to our table, it looks like this:

Conclusion

The CricViz selected five batsmen for the series against Sri Lanka would be: Joe Burns, Kurtis Patterson, Will Pucovski, and Matthew Wade, with Peter Handscomb as the spare.

It’s a mixture of a list. Pucovski and Wade are at different ends of their careers, representing both youth and experience. Handscomb has already been in and out of the Test side, whilst Patterson has never played an international match of any sort. This isn’t about “putting faith in youth”, or “trusting experience”. It’s about trying to take a more objective look at every player, and ignore irrelevant factors. This is a middle order picked on weight of runs, control, ability to perform in tough conditions and against all kinds of bowling. Those are pretty solid criteria.

It also has the benefit of being a relatively exciting group. Pucovski’s youth is alluring, and the story of Wade’s return alongside the man who replaced him is an interesting one. Patterson’s status as a relative unknown to those on the international stage makes him an interesting prospect; Burns is so clearly the best batsman in domestic cricket that giving him the opportunity to prove that is filled with anticipation.

Of course, this is just a suggestion. Data offers just one perspective, and this should not be read as a foolproof means of selecting talent. It’s important, as always, that these sort of assessments interact with cricketing judgement, in order for both perspectives to have the best chance of getting the decisions correct.

Equally, you can read this article and think that certain metrics should be worth more. Perhaps you think pure runs should be worth more than false shot percentage. Perhaps you think the opposite. There is scope to make this broader, or more specific, weighting certain skills ahead of others, but for now, this will do.

Regardless, Australia really do need to get this correct. Sri Lanka have some very talented batsmen, and if Langer et al take a wrong step then they could be punished both now and then again in England. It’s a hugely significant moment for the Australian selectors; will they get it right?

Ben Jones is an analyst at CricViz.