Unfortunately I didn’t see the comment, but this is something which can be analysed.

Double headers of the WBBL and BBL have been regularly scheduled throughout the last two seasons. Twenty-four out of seventy BBL games across 2015/16 and 2016/17 were scheduled immediately after WBBL matches at the same ground.

I assume that the same pitch was used for both matches in all of these double headers. So for each Big Bash League match held as part of a double header, the pitch has already been used for up to forty overs of women’s cricket that same day. At the one double header I attended, the steam roller was run over the pitch between matches, but it isn’t inconceivable that these pitches could be different.

A ‘slow pitch’ suggests to me that scores would be lower, and we can test this premise by comparing scores in matches preceded by a WBBL match, and those not preceded by such a match.

Firstly, let’s examine the total figures. My dataset doesn’t include the semi-finals and finals played in the last week, so just under 1/3 of the sample were double headers.

Double header? | Matches | Median runs | Median balls | Median run rate |
---|---|---|---|---|

Yes | 21 | 317 | 235 | 8.18 |

No | 45 | 320 | 234 | 8.35 |

The number of runs scored and the number of balls faced are roughly the same at the median point. The difference in the median run rate translated to 6.8 runs per match, which is slightly more than the run gap, but not much more.

Obviously this isn’t enough to answer this question. We don’t know the distribution of matches, and whether one is consistently less than another. More importantly, we don’t have a breakdown by venue. If pitches are having an impact, it seems far more plausible that there is greater difference between venues, rather than difference depending on the presence of a women’s team earlier in the day.

This first chart shows the difference between the median run rate in double headers and single matches at each of the eight grounds that hosted BBL matches in 2015/16/17.

A larger bar indicates that run rates were higher in double headers than in other matches. What we find is that there is no difference in run rate at the Adelaide Oval and the Sydney Showground (aka Spotless Stadium) while substantially more runs were scored at double headers at the SCG and MCG. Double headers produced lower scoring rates only at Docklands, Bellerive and the WACA.

This next chart goes into a bit more detail. It shows the run rate for every match played in BBL05 and BBL06, broken down by venue. Red dots represent double headers, blue dots represent standalone matches.

I see no clear evidence of a trend. While the highest score was at a standalone game (445 runs scored at Renegades vs Hurricanes at Docklands), there are plenty of grounds where double headers produced more runs than single matches.

The lowest performance was a standalone match at Bellerive, the Gabba, the MCG, the Sydney Showground and the WACA. Only at the SCG, Docklands and Adelaide did this honour go to a double header match.

This isn’t definitive evidence of anything, but the available data suggests to me no proof that playing a women’s match on a pitch has any impact on the scoring in the following men’s match.

]]>I’ve updated my analysis and also looked into how often teams win when batting first or second, when you break down the matches based on how close they are. In short, close matches are overwhelmingly won by the team batting second.

**I’ve written up this analysis in my first cricket piece at the Guardian Australia.**

And here is the most interesting chart, showing break down of results based on margin of victory in BBL06 and WBBL02:

And as a bonus for blog readers, here is the same chart for the last six seasons of the BBL and the WBBL and its predecessor:

]]>So my question: is this a broader trend, with more runs being scored across Twenty20 cricket? In this post I’ll attempt to find out, and also look at how different ways of scoring are becoming more or less prominent.

Firstly, let’s just look at the median number of runs scored in each innings. I wanted to simplify the results, but keep together whole seasons, so each year covers the northern summer in that year and the southern summer at the end of that year, so each year covers April-March.

The number of matches in 2003 and 2004 was particularly low, but once Twenty20 cricket became more stable we saw the median score stabilise around 150. It dropped from 2010 to 2013, but since 2014 the median score has shot up, to close to 160 in the last ten months.

I wanted to isolate where this growth is happening: is it because matches are being played in different places, or is there a general increase in scoring rates?

This next chart shows the run-scoring rates for the top-level men’s domestic T20 league in five big cricketing countries, as well as the scoring rates for T20 Internationals.

Scoring rates are higher in Australia, India and England compared to the West Indies, South Africa and in international matches. While the data is noisy, it does appear that there has been a general spike across the last three seasons.

Secondly, I wanted to get a sense of how run-scoring has changed in terms of what sort of runs are scored. Runs can be split into four categories: fours, sixes, extras and ordinary runs scored by running between the wickets.

Over the last decade, we’ve seen an increase in the proportion of runs which are scored as sixes, while there has been a decline in the proportion of extras and the proportion of fours. Ordinary runs remain the primary way of scoring, and this has remained roughly steady just over 40% of all runs scored.

(It’s worth noting that the drop in fours and spike in sixes is off a very small sample – 8 BBL matches, 6 matches in the NZ domestic competition, and 2 internationals between New Zealand and Bangladesh).

Finally, this chart shows the same men’s data as the above chart alongside the same data for women:

Ordinary runs make up a much larger share of runs in women’s cricket. Fours and extras make up similar shares as in the men’s game (with four-scoring increasing over the last three years), while six-hitting is a much less important part of the women’s game.

]]>This is the table as it stands now:

Team |
Played |
Won |
Lost |
NRR |

Melbourne Stars | 6 | 4 | 2 | +0.786 |

Brisbane Heat | 6 | 4 | 2 | +0.557 |

Perth Scorchers | 7 | 4 | 3 | +0.331 |

Sydney Sixers | 7 | 4 | 3 | -1.062 |

Sydney Thunder | 7 | 3 | 4 | -0.097 |

Hobart Hurricanes | 7 | 3 | 4 | -0.179 |

Melbourne Renegades | 6 | 2 | 4 | -0.003 |

Adelaide Strikers | 6 | 2 | 4 | -0.157 |

In the case of teams on the same number of points, the net run rate will be used to break ties.

The teams have been split into four tiers. The Stars and Heat stand in a clear lead having won four matches and with two to come. The Renegades and Strikers are clearly at the bottom of the table, with only two wins, but with two matches to come. The remaining four teams have only one match left to come, with the Sixers and Scorchers on four wins and the Thunder and Hurricanes on three wins.

In this post, I’ll show the proportion of scenarios where a team would clearly make the finals, clearly miss out, or find themselves in a tie to be broken by their net run rate. It should be noted that I’m not attempting to judge the chance that each team will win any of the matches – just the possible outcomes if they win or lose.

Firstly, this shows how often, out of the 64 different scenarios, how often a team would make the finals, not make the finals, or depend on their net run rate.

Team |
Finals |
No finals |
NRR decides |

Melbourne Stars | 49 | 0 | 15 |

Brisbane Heat | 48 | 0 | 16 |

Perth Scorchers | 32 | 0 | 32 |

Sydney Sixers | 34 | 0 | 30 |

Sydney Thunder | 0 | 34 | 30 |

Hobart Hurricanes | 0 | 32 | 32 |

Melbourne Renegades | 0 | 48 | 16 |

Adelaide Strikers | 0 | 51 | 13 |

The Stars and Heat have approximately 75% chance of cleanly making the finals. Each team needs to win one of their two remaining matches to make the finals without any doubt. The Scorchers and Sixers will make the finals if they win their remaining match, and if they lose they will depend on their net run rate. There is a narrow scenario where the Sixers lose their remaining match and still cleanly make the finals – if the Renegades beat the Strikers, the Strikers beat the Thunder, the Scorchers beat the Hurricanes and the Heat beat the Renegades, there will be no need for any tiebreakers. In that case, the Sixers-Stars match on the final day will only matter in determining exactly where the finals are played, and which teams play each other.

The Renegades and Strikers need to win both of their matches to have a chance of making the finals (so tomorrow night’s head-to-head between those teams will eliminate one of them). The Thunder and Hurricanes also need to each win their remaining match to have a chance, but will also rely on a high net run rate to make the finals.

We don’t know exactly where the net run rate will end up, but we do have the information about how teams have gone so far. The Stars, Heat and Scorchers all have a healthy net run rate, well above zero. The Sixers, despite winning a majority of their matches, has a terrible net run rate thanks to a massive defeat to the Thunder last night. The Thunder had a very bad net run rate, but that has been largely neutralised thanks to their defeat of the Sixers in ten overs.

Amongst the teams in the bottom four, the Renegades have the best net run rate, while all four teams have a better net run rate.

For my final table, I ranked teams in each tie according to their **current net run rate**, although this may change in the remaining six matches.

Team |
Finals chance |

Melbourne Stars | 100% |

Brisbane Heat | 100% |

Perth Scorchers | 100% |

Sydney Sixers | 53.1% |

Sydney Thunder | 18.8% |

Melbourne Renegades | 12.5% |

Hobart Hurricanes | 9.4% |

Adelaide Strikers | 6.3% |

The Sixers’ terrible net run rate means that they will not make the finals if they are forced into a tie, unless they can manage a massive defeat of the Stars in the final match of the season. The Stars, Heat and Scorchers will make it through any tie, so they will all make the finals unless their net run rate significantly worsens in the remaining matches.

If the Sixers are stuck in a tie, that fourth finals position could go to any of the four remaining teams, with the Thunder and Renegades having the best chance of pulling through.

Tomorrow night’s Strikers-Renegades match will definitely eliminate one team from the tournament. If the Renegades win, their chance of making the finals will double from 12.5% to 25%. If the Strikers win, the Sixers will have a 50% chance, the Thunder will have a 25% chance, and the Hurricanes and Strikers will each have a 12.5% chance.

]]>Today I’m going to show which players have performed the best in the Big Bash League and the Women’s Big Bash League, when their score is adjusted to reflect their opportunities.

This metric is based on this graph: how many runs does each batsman score, broken down by the over in which they entered the innings:

And here’s the same chart for women’s Twenty20 cricket:

The WBBL data includes the last four seasons of the state-based competition which preceded the WBBL, and it does appear that players in domestic Australian women’s Twenty20 cricket entering in the middle of the innings have performed better over the last five years.

This metric is simple. How does each batsman’s score compare to the median score for batsmen coming in to the innings in the same over for the period since July 2011? These measures (as a percentage) are then averaged to produce a figure for the entire 2016/17 season.

(Two notes: firstly, runs per innings is not the same thing as a batting average, as it doesn’t care whether the batsman was dismissed or survived without losing their wicket. Secondly, the data does not cover the matches played on Sunday 8 January 2017.)

This first table shows the top ten batsmen amongst those who usually bat in the top six for their team. Chris Lynn and Rob Quiney have both scored over 60 runs per innings. While Quiney has scored slightly more runs on average, Lynn normally enters around over 3, so his runs are slightly more impressive.

Sunil Narine and Mitch Marsh usually come in later in the innings, and Narine in particular is valued more highly thanks to late runs.

batsman | team | times batted | total runs | median entry point | runs per innings | scaled runs (%) |
---|---|---|---|---|---|---|

Chris Lynn | Heat | 5 | 309 | 3 | 61.80 | 346.82 |

Rob Quiney | Stars | 2 | 126 | 0 | 63.00 | 331.58 |

Sunil Narine | Renegades | 4 | 45 | 9.5 | 11.25 | 326.75 |

Mitchell Marsh | Scorchers | 5 | 210 | 11 | 42.00 | 275.79 |

Daniel Hughes | Sixers | 3 | 155 | 0 | 51.67 | 271.93 |

Ben Dunk | Strikers | 5 | 232 | 0 | 46.40 | 240.50 |

Kevin Pietersen | Stars | 3 | 148 | 2 | 49.33 | 238.89 |

Brad Hodge | Strikers | 5 | 204 | 5 | 40.80 | 231.47 |

Tim Paine | Hurricanes | 5 | 218 | 0 | 43.60 | 229.47 |

Brendon McCullum | Heat | 5 | 198 | 0 | 39.60 | 208.42 |

Here are the top batsmen who bat lower down the order. Ashton Agar has only scored 45 runs in four appearances, but for someone who usually only appears at the end of the innings this ranks him very highly. Interestingly Strikers bowler Billy Stanlake has only scored four runs in his first BBL season, but this is enough to put him above average thanks to his late entry.

batsman | team | times batted | total runs | median entry point | runs per innings | scaled runs (%) |
---|---|---|---|---|---|---|

Ashton Agar | Scorchers | 4 | 45 | 17.5 | 11.25 | 388.81 |

Kane Richardson | Strikers | 1 | 45 | 13 | 45.00 | 375.00 |

Beau Webster | Hurricanes | 1 | 67 | 6 | 67.00 | 304.55 |

Chris Green | Thunder | 4 | 39 | 16 | 9.75 | 230.56 |

Pat Cummins | Thunder | 5 | 133 | 13 | 26.60 | 207.19 |

Clive Rose | Hurricanes | 2 | 18 | 18.5 | 9.00 | 171.43 |

Jack Wildermuth | Heat | 2 | 37 | 14 | 18.50 | 168.68 |

Cameron Boyce | Hurricanes | 3 | 26 | 17 | 8.67 | 135.78 |

Billy Stanlake | Strikers | 3 | 4 | 20 | 1.33 | 133.33 |

Johan Botha | Sixers | 4 | 62 | 11 | 15.50 | 127.18 |

Before I move onto the WBBL, here are the worst ten rankings amongst batsmen who bat in the top six:

I have excluded anyone who has only batted once in the season, so Dom Michael and Jono Dean top the list thanks to each of them scoring two ducks in their two innings.

Particularly notable appearances on this list are two Sydney Thunder mainstays. Shane Watson had managed only six runs in his first three appearances as captain of the Thunder, and he did this despite usually coming in during the third over. Andre Russell was also performing poorly before his injury earlier this week, scoring only 25 runs in four appearances. (Shane Watson went on to score over 50 in last night’s match, but this data wasn’t available at the time of writing).

Shaun Marsh has scored 51 runs in four appearances, but this is particularly penalised because he has played as an opener for the Scorchers. This is a disappointment for them after he **scored 29% of their runs in BBL05**. Finally, Kumar Sangakkara has also performed poorly for someone who usually enters the match in the second over, managing only 68 runs in five appearances, before he was dropped by the Hurricanes for last night’s match against the Thunder.

batsman | team | times batted | total runs | median entry point | runs per innings | scaled runs (%) |
---|---|---|---|---|---|---|

Dom Michael | Hurricanes | 2 | 0 | 9.5 | 0.00 | 0.00 |

Jono Dean | Strikers | 2 | 0 | 9 | 0.00 | 0.00 |

Jake Lehmann | Strikers | 3 | 2 | 12 | 0.67 | 3.03 |

Shane Watson | Thunder | 3 | 6 | 3 | 2.00 | 11.01 |

Trent Lawford | Renegades | 2 | 6 | 13 | 3.00 | 34.45 |

Andre Russell | Thunder | 4 | 25 | 10.5 | 6.25 | 43.37 |

Shaun Marsh | Scorchers | 4 | 51 | 0 | 12.75 | 67.11 |

David Willey | Scorchers | 3 | 11 | 13 | 3.67 | 68.95 |

Ben Cutting | Heat | 3 | 27 | 12 | 9.00 | 72.49 |

Kumar Sangakkara | Hurricanes | 5 | 68 | 2 | 13.60 | 77.74 |

Now here are the similar stats for the women, starting with those who bat in the top six.

While Meg Lanning has been the top run-scorer, she ranks behind Sophie Devine and Amy Satterthwaite, who have each played fewer innings while also entering their innings at a later point.

Most batsmen playing in the WBBL score above the average for their role – which would match with other evidence that WBBL players score more runs overall than in other women’s Twenty20 included in the database.

batsman | team | times batted | total runs | median entry point | runs per innings | scaled runs (%) |
---|---|---|---|---|---|---|

Sophie Devine | Strikers | 5 | 230.00 | 4.00 | 46.00 | 375.63 |

Amy Satterthwaite | Hurricanes | 6 | 160.00 | 9.50 | 26.67 | 310.47 |

Meg Lanning | Stars | 9 | 384.00 | 0.00 | 42.67 | 304.76 |

Harmanpreet Kaur | Thunder | 8 | 194.00 | 9.50 | 24.25 | 303.15 |

Lauren Ebsary | Scorchers | 6 | 96.00 | 11.50 | 16.00 | 280.85 |

Katie Mack | Stars | 9 | 167.00 | 14.00 | 18.56 | 253.21 |

Ashleigh Gardner | Sixers | 8 | 225.00 | 5.00 | 28.12 | 250.11 |

Katherine Brunt | Scorchers | 6 | 105.00 | 13.00 | 17.50 | 242.46 |

Ellyse Perry | Sixers | 8 | 259.00 | 0.00 | 32.38 | 231.25 |

Alex Blackwell | Thunder | 9 | 253.00 | 5.00 | 28.11 | 217.73 |

And finally, the top ten batsmen who play in the lower order:

batsman | team | times batted | total runs | median entry point | runs per innings | scaled runs (%) |
---|---|---|---|---|---|---|

Sammy-Jo Johnson | Heat | 3 | 25.00 | 19.00 | 8.33 | 333.33 |

Veronica Pyke | Hurricanes | 4 | 60.00 | 17.00 | 15.00 | 281.25 |

Emma King | Scorchers | 2 | 8.00 | 19.50 | 4.00 | 250.00 |

Morna Nielsen | Stars | 2 | 6.00 | 19.50 | 3.00 | 175.00 |

Megan Schutt | Strikers | 4 | 33.00 | 16.50 | 8.25 | 161.67 |

Emma Kearney | Stars | 5 | 39.00 | 18.00 | 7.80 | 156.67 |

Megan Banting | Scorchers | 2 | 12.00 | 16.00 | 6.00 | 155.56 |

Kara Sutherland | Heat | 2 | 9.00 | 19.00 | 4.50 | 150.00 |

Jemma Barsby | Heat | 5 | 55.00 | 16.00 | 11.00 | 143.94 |

Annabel Sutherland | Renegades | 2 | 13.00 | 17.50 | 6.50 | 140.00 |

So what do you think? Is this a useful metric, and does it tell us anything more than a simple batting average?

As a final demonstration, these graphs identify players scoring over 300% in both the BBL and WBBL, on a chart comparing the runs scored per innings to the point where the batsman entered the innings.

]]>

There’s a lot of focus in Twenty20 cricket on a player’s strike rate, and that is undoubtedly important. But it’s also important that a batsman is able to last a reasonable amount of time. Players who hit a six and then get out off their second ball on a regular basis will have an extremely high strike rate, but won’t be of much value to their team. So the ability to stay also has value.

When you consider this point, the old-fashioned batting average (the number of runs scored over the number of times your wicket has been taken) has value. Ideally we’d come up with another metric which can mix together these two simple measures to give a sense of the ability of a batsman to score fast but also stick around long enough to make an impact.

In this post, I’m going to focus on a particular datapoint which I think has value when making assessments of players: when they come into the match.

Not all balls are the same, and not all overs are the same. Generally matches follow a pattern where the number of runs scored speeds up as you head towards the end of the match (barring the loss of significant numbers of wickets). You can see that in this graph:

The powerplay covers the first six overs in Twenty20 matches. During this time teams may only place two fielders outside of the circle which marks out 30 yards from the pitch. Following the powerplay, teams may have up to five fielders outside the circle. This clearly has an impact on the game. While the first over is the lowest-scoring over of the match, runs are scored quite quickly in overs 3-6, before collapsing in over 7. It takes until around over 15 before the batting team usually surpasses the scoring rate of the powerplay, but the average number of runs scored per over exceeds nine runs for the final overs.

Obviously you would expect different performances from batsmen depending on when they enter the match. If a batsman enters earlier in the match, they are not expected to score as quickly, but they have more time to play, while those coming in later are expected to score at a faster rate but may not have time to score as many runs.

The next few charts show the various scoring statistics broken down by the over in which the batsman entered the match. Opening batsmen are judged to have entered in over 0, while those who have come in later have been categorised by matching up their position in the batting order to the fall of wickets of preceding batsmen. I have included the metric for all men’s Twenty20 since mid-2011, as well as just for matches played in the Big Bash League since the first season in 2011/12.

Firstly, median runs scored per batsmen:

The curves are similar, although in the BBL we see more erratic data (likely due to a much smaller sample) and higher scoring for batsmen entering earlier.

Next, the number of balls faced:

Again, the curves are similar although BBL batsmen entering the match between over 5 and over 9 face more balls than the global average.

Next, this shows how often these batsmen get out:

The trend is very clear – batsmen entering in the first fifteen overs have a very high chance of losing their wicket, while those entering in the last five have a better show of surviving.

Finally, let’s take a look at the strike rate. This one surprised me. While, as expected, the average strike rate gradually increases for games played globally, you see a much weaker effect in the BBL.

So what does this all tell us?

Firstly, it is clear that the point in the match that a batsman comes in matters – it’s not reasonable to judge a batsman by the same standards if they open or if they come in late in the innings. Those coming in earlier have longer to bat and score more runs, but those coming in late score faster. Those coming in early are also more likely to lose their wickets. Indeed, you’d expect them to lose their wicket, as not losing their wicket may suggest they have wasted too much of their time at the crease with defensive play. Because of this, it may not be ideal to use a traditional batting average, which rewards batsmen for holding on to their wicket.

My first idea for a simple metric is one which simply takes the median number of runs scored by batsmen entering in that over, and measures the batsmen’s performance as a percentage of that score.

At the moment, I’m thinking I would measure that performance against all cricket played in the previous five years, back to 2011, but that will need to be finessed.

I’m going to come back tomorrow with some tables of the best and worst performing batsmen in the BBL this season according to this measure.

What other metrics would you like to see which adjust based on when a batsman arrives at the crease? I’m open to all suggestions.

]]>In an ideal world we would have ball-by-ball data which would include data on what sort of shots players attempt, and I would expect you would see a trend where batsmen attempting a lot of aggressive batting (whether or not those shots turn into runs) would be at a higher risk of losing their wicket. Unfortunately we don’t have that data, so we can only rely on the total number of runs scored, and the number of wickets lost, in each over.

I have to make a few assumptions here:

- Teams don’t dramatically change their batting behaviour from over to over, particularly if a wicket hasn’t been lost. So if a team is batting in a risky manner in one over and this results in the team losing a wicket, they would be more likely to have been batting in a similar way in the previous over.
- There is a correlation between batting in a more risky way, and scoring more runs. So while we don’t have enough data to actually categorise play according to risk, we can use the run rate immediately before the loss of the wicket as a proxy.

I have over-by-over data for just over 4000 men’s Twenty20 matches.

Firstly, let’s look at every over in the data set (excluding the first over of each innings), and the number of runs scored in the previous over, to identify how many of those overs resulted in the loss of a wicket.

There is a clear relationship between runs scored in the previous over and the loss of wickets. If a team scored no runs in the previous over, they lost a wicket in the next over about 25% of the time, whereas teams scoring seven or more runs in an over have a 30% chance of losing a wicket in the next over.

It’s important to not over-emphasise this effect. The chance of losing a wicket is below 40%, even for teams scoring twenty runs in an over, while the risk is about 25% in the over following a maiden. But there is a relationship.

For a slightly different perspective, this next chart shows the risk of losing a wicket broken up by the run rate in the five preceding overs. Again, you see a clear relationship, with teams losing a wicket in 25% of cases following a very slow runrate, and losing a wicket in 33% of cases following a runrate of 13 runs per over.

]]>

I’ve noticed teams winning the toss and bowling in BBL. Is that a regular tactic?

— Josh Pinngle Bells (@JoshCPinn) December 23, 2016

This is a good question, and one that is easy to answer with my dataset. In this post I’ll look at how the fashion in men’s T20 cricket has shifted towards teams preferring to bowl first, and also whether there is evidence that this works.

Firstly, let’s take a look at how often teams choose to bowl when they win the toss in the Big Bash League.

It’s a dramatic shift. Up until 2011/12, it was rare for teams to choose to bowl. Around 20% took the option in 2007/8 and 2008/9, but in 2009/10 a team chose to bowl first in only one out of 16 matches. We came closer to parity in 2012/13, but only a quarter chose to do so in 2013/14. In 2014/15 the team who won the toss chose to bowl in exactly half of the matches. Last season a small majority of matches saw teams choosing to bowl.

**In the first five matches of this season, every team has chosen to bowl.**

(It’s worth noting that I’ve excluded matches affected by rain for some of the later analysis, which means these numbers don’t cover every single match).

So let’s look at the same analysis but for all men’s Twenty20 cricket globally, broken down by year:

The trend is less dramatic – we don’t see a strong preference for batting first before 2012, but it is a clear favourite. In 2013 bowling first became the slight preference, and in 2016 so far almost 64% of teams chose to bowl first.

It’s also worth noting that we don’t have data on who won the toss for many games. This is why I can’t do a similar analysis for women’s cricket, since we don’t have toss information for domestic cricket in Australia prior to the first season of the WBBL. For what it’s worth, only three out of 55 matches in WBBL01 saw the team choose to bowl, but so far this season the team winning the toss has chosen to bowl in 6 out of 14 matches.

So why have teams been opting to bowl first? I won’t try to cover every possible explanation, but let’s start with something simple.

Firstly, does winning the toss give a team an advantage in terms of winning, ignoring all other factors?

It’s very marginal, but it exists in most years. 51.8% of teams winning the toss in 2016 have gone on to win the match. Teams winning the toss have won a majority of matches in all years except 2008 (narrow majority of losses), 2012 (solid majority of teams lost) and 2014 (both losses and wins made up a minority, with the rest being ties). So winning the toss appears to give a slight edge.

So what do you do with it? This next chart is the same as the chart above, but it breaks down the results based on the decision the team made after winning the toss:

In both 2015 and 2016, teams choosing to bowl won a majority of the time, and teams choosing to bat lost a majority of the time. This trend hasn’t always existed – from 2009 to 2011, it appears that teams did better when they chose to bat.

Finally, I wanted to look at whether there is a general tendency for a team batting first or second to win, taking out the question of whether there is extra information that the team winning the toss might be using to judge whether batting first or second is in their best interest. This graph shows the number of matches won by the teams batting first or second:

One shouldn’t overestimate the gap, but teams batting second have tended to win in a majority of matches in most recent years.

]]>In this post I’m going to run through some random stats about these matches (no deep analysis here) and also touch on a remarkable match played yesterday in New Zealand.

There were two impressive debutants who have played in the first two days of the Big Bash League. Ryan Gibson opened for the Sydney Thunder on Tuesday, scoring 53 off 43 balls. Gibson was the backbone to the Thunder line-up, holding on while the remaining batsmen crumbled before building up the score with batsman Pat Cummins. Jake Weatherald opened the batting for the Adelaide Strikers, putting on a 133-run opening partnership before going out for 52. Gibson has the tenth-highest ever score for a debutant in Australian domestic Twenty20 (including the Big Bash League and its predecessors. Weatherald is eleventh.

Pat Cummins is known as a bowler but showed his ability to bat on Tuesday, scoring 30 runs to help the Thunder put on a more defensible target. Cummins has played 49 T20 matches, and has batted in 21 of those matches. Thanks to injury, he hasn’t batted since he played for Australia in August 2015. He also played for the Kolkata Knight Riders in the Indian Premier League, and his last experience batting in the BBL was a score of 1* for the Thunder in January 2015. Out of those 21 innings, Cummins has only cracked ten runs twice, with a previous high score of 14. He had only ever hit two sixes. On Tuesday, he hit three sixes and a total of 30.

Sixers captain Moises Henriques scored a very fast 76 runs off 41 balls to lead the Sixers to victory. This is his second-highest score, but much faster than his highest – 77 runs off 57 balls for the Sixers in January 2015. This is primarily due to a much higher boundary rate. He scored 32 runs on the boundary in that previous performance, but 52 runs off the boundary on Tuesday night.

The Sydney Sixers only lost one wicket as they chased down the Sydney Thunder’s 159 runs. There have been 298 matches in the Big Bash League and its predecessors that produced a regular result. A team has managed to only lose one wicket in sixteen cases, and in one case the Scorchers lasted without losing any wickets in December 2015. Tuesday night’s Sixers score was the sixth-highest score reached while losing no more than one wicket.

The 133-run opening partnership by Weatherald and Dunk for the Adelaide Strikers wasn’t close to a record. The highest ever such score in the BBL was 172 runs by Quiney and Wright for the Melbourne Stars in 2012. The world record was set at 215 by Virat Kohli and AB de Villiers in 2015, and broken by the same two batsmen at 229 in 2016, both times batting one wicket down for the Royal Challengers Bangalore in the IPL.

The most remarkable stat from the last two days came from New Zealand, where a **match involving Central Districts and Otago** put on a total of 497 runs – a world record total score for a T20 match. Otago scored 249 runs, and Central Districts managed 248 in response – losing by one run.

The previous world record was 489 runs scored in a T20 International played between the West Indies and India in Lauderhill, Florida, earlier this year. The West Indies opened with 245, and like this week’s game, India came within one run of beating that target, scoring 244.

The previous record had stood much longer. An Indian Premier League match in 2010 saw 469 runs scored, and remains the third-highest score.

With this amazing score, it seems like only a matter of time before a match results in over 500 runs scored.

]]>In today’s post, I want to explore the role of losing early wickets in determining how many runs a team scores, and their chance of victory.

As a taste, this graph shows the likelihood of victory for teams based on how many wickets they have lost after five overs:

Teams that have maintained all of their wickets have roughly two-thirds chance of winning. Teams that are down one wicket are still favoured to win, but the chance of victory drops gradually further. A team which has lost four wickets has less than 20% chance of winning. There have only ever been six innings where a team had lost six or seven wickets after five overs, and in all six cases that team went on to lose.

Obviously this is a simplification: I haven’t separated out first and second innings matches, and it doesn’t take into consideration how many runs the other team will score – if both teams lose a lot of wickets they may cancel out these effects. But when you zoom out to the entirety of men’s Twenty20 cricket over the last sixteen years, the trend is clear.

Here’s another version of the same graph, but showing the state of play after ten overs:

Any team that has lost no more than two wickets after ten overs, but any time down five wickets or more has a slim chance of winning.

Finally, the same metric at fifteen overs: by this point you expect teams to have lost wickets. There aren’t many cases of a team lasting fifteen overs without losing a wicket, but when they do they are almost guaranteed of a win. Out of 53 cases, that team won in 50 cases, along with two losses and one tie.

Finally here is the same information for the fifth, tenth and fifteenth over, split between the two innings for both men and women:

Now obviously this gives you a general idea of what is a “normal” loss of wickets at a particular point in the game, but it doesn’t tell you how many runs to expect to score when you’re trying to calculate your chance of reaching a particular target.

It’s time for some box plots. If you’re not familiar with box plots, the box represents the middle 50% of cases in the sample – the upper end of the box is the 3rd quartile, and the lower end is the 1st quartile, so 25% of samples are above the box and 25% below. The thick line in the middle represents the median point – if you line up every score in a row, the median is the one in the middle. The dots represent outlier cases.

I measured this in two ways. Firstly: at a particular point in the innings and having lost a certain number of wickets, how many runs (as a raw number) are left to be scored in the remainder of the match. Secondly: at that some point, how many runs are left to be scored as a proportion of the runs scored so far. If a team has scored 50 runs and will end up on 185, then the first metric will produce a score of 135, and the second metric will produce a score of 2.7. I’m not really sure which metric is the best at this point in time.

The chart above shows how many more runs are scored in the last fifteen overs of the match relative to the runs scored in the first five. What you see is that the ratio of runs is roughly the same for teams who have lost between zero and five wickets, with slight evidence that teams who have lost a handful of wickets actually do better than those who have avoided losing any.

This chart also shows the five-over mark, but shows the number of runs to come in the remainder of the match as a raw number. When you look at scores this way, teams steadily score less runs if they have lost more wickets. If you assume that teams who have lost a handful of wickets will have scored less runs in the same period than those who haven’t lost any (a reasonable assumption), then it makes sense that the ratio of runs to come over runs scored is roughly the same but the raw numbers of runs is higher for those teams who haven’t lost a wicket.

For the sake of completeness I’ll end with two charts, showing how many runs are scored in the remainder of the innings (as in the above chart) broken down by wicket at the 5-over, 10-over and 15-over point in the innings, both for men and women.

So what can we conclude from this? Firstly, there is a clear trend that teams which lose more wickets in the early and middle parts of their innings have less chance of winning the match.

Secondly, teams that have lost up to four-five wickets usually have the ability to score a similar ratio of runs in the remainder of the innings as teams that have lost no wickets. Roughly, this means a team who has lost these number of wickets will score four times their score at the 5-over mark, double their score at the 10-over mark, and 40% more than their score so far at the 15-over mark. Teams that have lost more than 5 wickets at the 5-over mark, 6 wickets at the 10-over mark or 7 wickets at the 15-over mark are capable of scoring less runs in the remainder of their innings.

When you look at raw numbers of runs, every subsequent lost wicket reduces your power to score more runs, with the drop-off becoming particularly dramatic at the points outlined in the previous paragraph.

Overall none of this is probably very surprising, but it makes it very clear that the loss of early wickets in a Twenty20 match can have an impact. A team that comes into the end of their innings with plenty of wickets in hand is capable of scoring more runs, and thus having a greater chance of winning.

]]>