This article is part of a series on cricket analysis at club, age group performance and school level. To see part one, click here.
As you have noticed from previous articles, cricket stats analysis can get complicated quickly. That’s fine for the cricket badger with time and spreadsheets, but what about the cricketer who just wants to know how well they are doing?
The answer is a measure called Impact.
Impact uses the power and improved accuracy of expected scores but wraps them up in a simple number. Averages are the old way of doing this, but as we know, averages are short of context, leaving us to guess. For example, did a player who averaged 15.37 with the bat make a positive difference to the result or not?
Impact instantly tells them if they are meeting, exceeding or falling below expectations.
Here are the Impact scores for the West of Scotland team in 2019 (six innings or more) compared to their average:
You can see the professional Kleinveldt had the biggest impact overall, as you would expect. You can also see positive impacts from players with averages in the 20s, and even one at 16. You can also see one player who averaged 19 but had an overall negative impact on game outcomes (although not by much). The take-away is much clearer here - “I need to build innings at three to give the team a better start” - than just trying to improve average.
In fact, even if average is the main driver for a player, you can point out that a “good” average is much clearer. For example Farndale batted six times at number three and averaged 19. Had he scored four more runs per innings (average 23) his Impact would have been +4. In comparison, So whichever way you look at it, Impact is helpful.
This measure of Batting Impact is calculated in the following way per innings:
Impact = xR - Runs
Where xR is the expected runs for each batsman. You calculate this by dividing the Par (or Target) by 11. If you want to be fairer, you can give a weighting to higher order batsmen. For example, make top order batsman score 11% of the Par each, while lower order are expected to contribute almost nothing.
To get an Impact score for the season, you can average by dividing by each innings.
For a player brought up on averages, this number is simple to understand. That average of 15.37 had an Impact of +3 so you are doing better than expected. Well done. Your team mate with an average of 18.71 had an Impact of -2 so they need to do more work if the team is to win games while they are batting. This is despite the better average.
For bowling Impact the calculation is slightly different:
Impact = (Balls x xRpB) - Runs
Where xRpB is the expected number of runs conceded per ball to win the match (Par / 300 in a 50 over game).
The obvious flaws with this calculation are:
No account of when the bowler is bowling (at the death when the opposition need 12 an over is different from opening when the opposition need three an over)
No account for wickets.
If you don’t have access to over-by-over information, that will have to do. Although there is a possibility we can factor in the result of the match to make up the difference because by definition, a winning team ha had a positive Impact and a losing team a negative one. More on that another time.
However, both these issues can be solved by using a DLS sheet to calculate xRpB at any stage of the innings (look up the DLS Par at that over, minus the actual score, and divide by ball remaining). This will give you a more accurate Impact by removing outliers, but is not essential (and is impossible without over-by-over scores).