The chief villain of the summer was supposed to be Virat Kohli, but that changed following the Boxing Day Test, with focus instead turning to the Australian men’s selection panel.
Two washed-out days in Sydney and endless hours of talkback radio further intensified the critiques.
Most of the criticism has focused on the lack of communication and inconsistent decision-making. Why was Aaron Finch asked to open the batting despite not playing there for his state? And why was the request for players to score runs at Shield level despite many fringe players, particularly Glenn Maxwell and Marcus Stoinis, only having the opportunity to play two Shield games before the Big Bash League started?
However, communication issues and inconsistent decision-making are symptoms of the ingrained philosophical approach to selecting cricket teams. Why do we still pick teams based on opinion and not analysis?
Like James Silver was in his article yesterday, I’m reminded of the baseball film Moneyball. Billy Beane (Brad Pitt), the general manager of the Oakland A’s, walks into a room of old scouts, who inform him that they have found new players for the club based on their “swing technique” and “gut instinct”. Beane promptly dismisses their advice and begins the movement to sabermetrics and the use of advanced analytics in baseball.
As my mind envisages the selection meeting room at the MCG, the image of Trevor Hohns, Greg Chappell and Justin Langer sitting around the table doesn’t look too dissimilar from the one Beane saw in Oakland. Decisions made on ‘technique’ and ‘gut instinct’.
Like baseball 15 years ago, there is a desperate need for advanced analytics in cricket. The most commonly reported cricket statistics are still batting averages and number of hundreds or bowling averages and number of wickets. These statistics do not take into account the context of the match, the quality of the opposition or the pitch and ground conditions. What is the value of scoring 40 on a green seamer? Or taking a wicket to break a significant partnership? These performances are washed over in the reporting of averages.
In the recent Test series, Cheteshwar Pujara has been lauded for his three 100s, most notably his 193 in Sydney. But I hypothesise that advanced analytics would suggest that his two innings in Adelaide (123 and 71) were more valuable to India winning a match. Pujara scored 35 per cent of India’s total runs in Adelaide, including 50 per cent of India’s first-innings runs on a pitch that was difficult to bat on early.
The sabermetrics movement in baseball led to the development of new and improved analytics. One of the most important statistics is wins above replacement (WAR), which incorporates a huge range of batting, pitching and fielding data to assess a player’s contribution to winning matches. A similar measure could be developed for cricket.
Ultimately cricket teams seek to win matches; runs and wickets only enable that to happen. Therefore, a statistic or statistics that measure the average contribution of a player to enabling a win would be more important than the actual number of runs or wickets he or she contributes. In a low-scoring match, scores of 75 and 25 might be more valuable to winning a match than 250 not-out on a flat deck.
A cricket ‘wins above replacement’ statistic would enable more evidence-based and consistent decision-making at the selection table. The data could clearly articulate the contribution of a player to winning matches and thus his or her value of being selected.
With a transformation of the Australian men’s team selection process expected, now could be the opportunity for Cricket Australia to begin leading the advanced analytics movement in cricket.
We live in a digital age where data is everywhere. The challenge is to enable data to inform our decision-making, not just opinions.
Simon Massey
Roar Rookie
Completely agree Sheek. A players contribution towards winning (not their average runs / wickets) needs to be weighed against team dynamics, willingness to learn / adapt, leadership, etc.
sheek
Roar Guru
Selection in cricket, or any other sport for that matter, is simply not just about picking the highest scoring batsman or highest wicket taking bowler in Sheffield Shield. When selectors sit down to discuss the next test team, it already has a history, a narrative. There are already key players in the team, & those chosen to join them must add to that value of the team already there, not detract from it. It's not simply a matter of dropping A, B & C, & replacing them with D, E & F. Back in 2007, journalist/writer Mark de Rond spent many months with the Cambridge University Boat Club as they decided on their final xcrew to meet Oxford University in the annual 'head of the river' clash. There were 39 world class rowers competing for just 8 seats in the boat, of which about 6 were already imbedded. In the end, CU didn't pick the next fastest rowers to join those already assured of selection in the boat. They chose instead, rowers of less optimum skill, but whose social attributes helped bring out the better in the best rowers in the boat, plus also providing a more cohesive team spirit. Those of you who have rowed would understand that rowing is one of the most delicate team sports there is. Getting 4, or 8 guys to row to the extent they become "as one", is one of the most difficult technical feats to achieve. On race day, CU nearly lost, but they also won, breaking a several years grip on the race bu OU. So this is something to keep in mind next time you want to rage against the selectors. The process of selecting is more complicated & less sure-thing than you might appreciate.
Chris Love
Roar Guru
I really like the thought process behind this and often wondered the same myself. There are so many factors that could be weighed upon to consider how valuable certain runs are and wickets are. Things like a number 5/6/7 scoring a 40 after a batting collapse but doing so soaking up 150 balls or more and tiring out bowlers which makes it easier for 6/7/8 batsmen to score the 20/30/40 runs. Or similarly, a bowler who only gets two wickets for the match but both wickets broke 100-150run partnerships should be considered more valuable than a bowler that was ineffective against the top order but manages to swathe through a weak tail taking the last 4 wickets. The later ends up with respectable bowling figures of 4-100 at an average of 25 where as the former ends up with 2-100 at an average of 50. One looks like twice the bowler as the other but the 2-100 could be far more valuable. Similarly bowling first the opening bowlers get the best use of the new ball and probably doing a bit more off the pitch in the first session of a match (Think Starc and Hazlewood). Those wickets are easier to come by than say a Nathan Lyon in the second innings in Perth that had to produce wonderful bowling to put Australia in prime position. There are lots of factors that would need to be considered and a huge amount of tweaking along the way. Just the fact that a pitch wears, or flattens out after the first two sessions need to be considered on a test by test basis.
Simon Massey
Roar Rookie
I feel like there is enough cricket data available to calculate a 'WAR statistic' both presently and in the past. Taking a wicket to break a partnership is an underrated aspect of the game. Much more important than taking a bag full of wickets to clean up tailenders?
Simon Massey
Roar Rookie
Totally agree re the old scouting techniques selection of Oakland's pitching staff (at least that year). But it's probably fair to say that Beane's approach started a new movement in baseball? Much like Kevin Cash's introduction of the 'Opener' in 2018 is being adopted by many other teams (including Oakland)?
Simon Massey
Roar Rookie
Agree that baseball has a much larger sample size, which makes it easier to develop advanced analytics. Cricket does have discrete movements like baseball though - one pitch / one ball. This would make it easier to calculate the contribution of a player to a win than other more dynamic sports.
Tony Tea
Guest
I have always found it puzzling that "wins above replacement" was largely ignored by the wider sporting public and punditry. We glorify the big scores in cricket and the goal bags in Aussie Rules (I'm Victorian) while often overlooking the lower scores in tougher conditions. Without any statistical analysis it seemed to me that Andrew McDonald would often pinch a couple of wickets or make a tight 50 at just the right time to get Victoria a win - I wonder what his WAR would be.
JohnB
Roar Rookie
Plus the Oakland As in using sabermetrics weren't looking for the best players, as national selectors unencumbered by salary caps or budgets are, but for undervalued players who they could afford. All the hype about moneyball also glosses over the fact that a lot of Oakland's success that year was down to its pitching staff - who for the most part were picked and developed the old fashioned way.
peeeko
Roar Guru
agree to an extent but stats are more usable in baseball due to the large number of batting opportunities. in one season a batter will have 500 plus chances to score or get out and he will be out in 300 of these. Cricket has 90% less innings then baseball and one out and thats it it is amazing how many people on the roar have opinions about shield cricket yet never watch it