Australia have bossed the first three Tests of the Ashes series, and one of the key reasons for that has been the difference between their bowling attacks. The Australian attack, with their extra pace and an incisive spinner, have induced more false shots from England's batsmen than England's batsmen have managed from the Australians. Before the series, it was clear that both teams had weaknesses in their batting line-ups, but Australia's bowlers have clearly exploited them better.

ESPNcricinfo's control stats from this series underlines the fact. The control stats record on a ball-by-ball basis if the batsman was in control of the delivery bowled to him, where "in control" means a ball left alone or middled, with the ball going where it was intended to go by the batsman. Any play-and-miss, edge or mistimed stroke is a not-in-control delivery.

In the first three Tests of this series, Australia's batsmen have been in control of 85.5% of the deliveries bowled to them, while the corresponding percentage for England's batsmen is only 80.9%. A difference of 4.6 percentage points looks small, but is a fairly difference given that each team has faced more than 3200 deliveries in the series. In fact, this difference so far is the highest in the last six Ashes series; in 2013-14, when Australia blanked England 5-0, the difference was only 3.3. The previous-highest difference during this period was in the 2013 series in England which the hosts won 3-0, when the difference was 4 percentage points.

Control stats in the last six Ashes series*
Series Aus control Eng control Diff Series result
 2017-18  85.5  80.9  4.6  Aus 3-0
 2015  80.0  79.0  1.0  Eng 3-2
 2013-14  89.0  85.7  3.3  Aus 5-0
 2013  83.2  87.2  -4.0  Eng 3-0
 2010-11  80.8  84.3  -3.6  Eng 3-1
 2009  80.9  77.0  3.9  Eng 2-1
*Control stats are in percentages, for the batting team. The difference is calculated as Aus control minus Eng control

Also, while it is easy to assume that the batsmen from the winning team have a higher in-control factor than those of the losing team, that isn't always the case. In two of these six series, the batsmen from the losing team - Australia, in both cases - had a higher control percentage than the winners. Both series were in the England, in 2009 and 2015, and the home team emerged winners by a one-Test margin. In both series, Australia's bowlers drew a higher percentage of errors from England's batsmen, but couldn't quite convert that into a positive series result. (The 2009 series was especially unusual, as Australia had a much higher batting average and took more wickets, but failed to clinch the crucial moments.)

The batsmen
In the current series, though, Australia have done all the running, and have clearly been the better team with bat and ball, a fact which is reflected in their control stats as well. Steven Smith has been their star - not only has he scored mountains of runs, he has also looked almost completely untroubled while doing so. That explains his incredibly high control percentage of nearly 91. Among those who have batted more than once in this series, the next highest for Australia is Shaun Marsh's 86.9%. The only batsman who has looked completely out of sorts is Peter Handscomb, and that is reflected in his control percentage of 77.2: not only did he not score runs, he didn't look like he would get any. It is hardly surprising that he was dropped for the Perth Test. (His replacement, Mitchell Marsh, had a control percentage of 89.4 in his 181 at the WACA.)

For England, while Dawid Malan has stood out in terms of runs scored, in terms of control percentages James Vince tops everyone else. His control factor is higher than those of all the other England batsmen, and in fact, among batsmen from either team who have batted more than once, is next only to that of Smith. His returns obviously haven't matched the assurance with which he has batted at the crease - in six innings he has only managed 182 at 30.33. However, as his superb 83 in Brisbane showed - he had a control factor of 88.8% in that innings - he is capable of far better.

Malan's control factor is a relatively low 82.7%, which isn't very different from those for Joe Root (83) or Alastair Cook (82.2). However, Malan, whose control factor during his 140 in Perth was higher at 86.3%, has been able to get away more false shots that Cook: he has played 112 false shots in the series and been dismissed just six times, compared to 38 by Cook for six dismissals. Moeen Ali in second place is a surprise, though.

Control percentages for England's batsmen in the series
Player Inns Runs Ave Control %
 James Vince  6  182  30.33  87.7
 Moeen Ali  6  116  19.33  84.0
 Joe Root  6  176  29.33  83.0
 Jonny Bairstow  6  241  40.16  83.0
 Dawid Malan  6  302  50.33  82.7
 Alastair Cook  6  83  13.83  82.2
 Mark Stoneman  6  193  32.16  79.1

It needs to be mentioned here that not all false strokes are equal, even though they all add to the not-in-control total: leaving a straight ball on middle stump and heading towards the wickets is obviously worse than missing a leg-side ball and being struck on the pads, for example. However, the numbers do suggest that Malan has had his share of good fortune in this series: he has been beaten 51 times in the series, the most among all batsmen. Mark Stoneman is next on 46. Stoneman has had a mixed time too: he has scored two fifties, but his control percentage during his 56 in Perth was only 72.7, as he was given a thorough working over by the Australian fast bowlers, and especially Josh Hazlewood.

Root, on the other hand, hasn't had those lucky breaks in this series so far. In 2015, when he scored 460 runs at 57.50, his control factor was a relatively low 81.5, but the rub of the green went his way. In the 2015 Ashes, he was only dismissed once every 16 false strokes; in this series, that ratio has fallen to one in 9.3 false shots.

The bowlers
Among the Australian bowlers, Pat Cummins has the fewer wickets than Starc, Hazlewood and Lyon, but he is the bowlers against whom England's batsmen have the lowest in-control percentage. Against him, the control factor is only 78.4, which gives him a not-in-control (NIC) percentage of 21.6. The corresponding numbers for Starc and Hazlewood are 21.0 and 19.3, while it is 16.6 for Lyon.

Two of England's top five batsmen have control percentages of less than 80 against Cummins: Root (75%) and Stoneman (79.4). However, Cummins has only taken a wicket every 14.9 balls when he has induced a false shot, compared to one every 8.3 balls for Starc, and one every 9.3 balls for Hazlewood. That suggests Cummins has been quite unlucky in this series, though sometimes beating the bat and not managing a nick can also be the result of bowling a wrong length. In Cummins' case, though, it does seem that 11 wickets is scant reward for the way he has bowled in the series.

The England bowlers all have NIC percentages of less than 20. It is 17.8% for Anderson and 17.4% Broad. For Chris Woakes and Moeen Ali it is less than 12%, which indicates they haven't really troubled the Australian batsmen at all in this series. Anderson and Broad have similar NIC percentages but vastly different averages, thanks to their conversions on those NIC deliveries: Anderson has taken a wicket every 11.6 times he has induced a false shot, compared to 23.8 for Broad.

Control percentages for England's bowlers in the series
Bowler Wkts Ave NIC %
 James Anderson  12  25.83  17.8
 Stuart Broad  5  61.80  17.4
 Craig Overton  6  37.66  15.9
 Jake Ball  1  115.00  14.7
 Chris Woakes  7  51.57  11.7
 Moeen Ali  3  105.33  10.2

Three of Australia's top-order batsmen have control percentages of less than 80 against Anderson: Cameron Bancroft (79.1%), Usman Khawaja (78.3) and Handscomb (76.2). However, Smith (88.3) and Shaun Marsh (90.2) have done very well against him. Broad has enjoyed bowling to Handscomb, who has had a control percentage of only 66 against him, but not so much Smith (92) or Bancroft (88.1). Sadly for England, Handscomb is unlikely to take any further part in the series, while Smith still has some unfinished business to complete.

S Rajesh is stats editor of ESPNcricinfo. @rajeshstats.