February 28, 2017

# We need a better measure than economy rate in T20

The traditional metric has served well, but it doesn't take context into account
25

In T20 cricket, economy rate is the most widely used metric by which to assess and compare bowlers. However, despite its simple utility, it is a flawed and misleading measure. In only taking into account balls bowled and runs conceded, economy rate disregards the context of the match in which those balls were bowled and runs were conceded. The match venue, conditions, strength of the opposition batting and the nature of the match itself are critical determining factors of economy rate that are not considered by the measure.

The great strength of economy rate is that its normalisation across six-ball overs allows for straightforward comparison between bowlers. However, using a new measure, economy-rate differential (ERx), it is possible to maintain comparison while placing economy rate within the context of the innings.

ERx is calculated by subtracting the innings run rate from the bowler's economy rate to give a more accurate representation of that bowler's performance in that innings. For example, in the final of the 2016 IPL between Royal Challengers Bangalore and Sunrisers Hyderabad, Yuzvendra Chahal returned figures of 1 for 35 from his four overs, at a typically mid-to-high range economy rate of 8.75. However, placed within the context of the innings, Chahal's economy rate was 1.65 runs per over fewer than the innings run rate of 10.40; in other words, a fairly good showing.

Although such a conclusion could be reached from an analytical observation of the scorecard, the presentation of traditional economy rate over a season or competition would fail to illustrate this nuance the way ERx can: a bowler with a negative ERx is a bowler who performs above average in the innings in which they bowl; a bowler with a positive ERx is a bowler who performs below average in the innings in which they bowl.

It is possible to apply ERx to a career by subtracting the overall run rate of all the bowling innings a player is involved in from their overall career economy rate; similarly, it can be applied to a competition by doing the same with the competition economy rate. I have analysed the bowlers in the 2016 IPL using this method.

 Bowler Overs ER (rank) Innings RR ERx Adam Zampa (RPS) 17.0 6.76 (1) 7.89 -1.42 Sandeep Sharma (KXIP) 50.0 7.32 (7) 8.56 -1.24 Sunil Narine (KKR) 42.4 7.12 (4) 8.25 -1.13 Sreenath Aravind (RCB) 29.1 7.40 (9) 8.51 -1.11 Chris Morris (DD) 44.0 7.00 (3) 8.10 -1.10 Dhawal Kulkarni (GL) 49.0 7.42 (10) 8.52 -1.10 Mustafizur Rahman (SRH) 61.0 6.90 (2) 7.95 -1.05 Rajat Bhatia (RPS) 34.0 7.17 (5) 8.12 -0.95 Ravichandran Ashwin (RPS) 44.0 7.25 (6) 8.10 -0.85 Ravindra Jadeja (GL) 40.1 7.74 (20) 8.56 -0.82 Praveen Kumar (GL) 51.3 7.78 (21) 8.46 -0.68 Amit Mishra (DD) 46.0 7.47 (12) 8.13 -0.66 Ashok Dinda (RPS) 30.3 7.57 (14) 8.15 -0.58 Tim Southee (MI) 43.0 7.65 (16) 8.23 -0.58 Yuzvendra Chahal (RCB) 49.1 8.15 (29) 8.72 -0.57

Bowlers who traditionally have good economy rates will generally continue to be rewarded under ERx (unless they play in a really strong bowling team) so it is unsurprising to see the list of top ERx bowlers dominated by those with top economy rates. However the presence of Ravindra Jadeja, Praveen Kumar and Chahal in the list illustrates the value of ERx in that these are four bowlers outside the top 20 economy rates who make an appearance on this list of top 15 ERx bowlers because their economy rate was better than the average run rate of innings in which they bowled.

 Bowler Overs ER (rank) Innings RR ERx Kyle Abbott (KXIP) 16.0 11.06 (59) 9.14 +1.92 Ishant Sharma (RPS) 15.0 9.86 (57) 7.96 +1.90 Thisara Perera (RPS) 33.1 9.82 (56) 8.09 +1.73 James Faulkner (GL) 17.0 9.82 (56) 8.22 +1.49 Karn Sharma (SRH) 16.2 10.46 (58) 8.99 +1.47 Mohammad Shami (DD) 25.1 9.69 (54) 8.32 +1.37 Shadab Jakati (GL) 20.0 8.90 (46) 7.95 +0.95 Umesh Yadav (KKR) 26.0 9.00 (48) 8.11 +0.89 Chris Jordan (RCB) 28.0 9.21 (51) 8.33 +0.88 Varun Aaron (RCB) 23.4 9.59 (53) 8.73 +0.86 KC Cariappa (KXIP) 17.0 9.35 (52) 8.56 +0.79 Barinder Sran (SRH) 49.3 8.34 (34) 7.62 +0.72 Hardik Pandya (MI) 16.4 9.18 (50) 8.54 +0.64 Brad Hogg (KKR) 16.0 8.87 (45) 8.36 +0.51 Dwayne Bravo (GL) 56.0 8.82 (44) 8.45 +0.37

Similarly, there is a strong correlation between the worst ERx bowlers and the ones with the worst economy rates, but ERx gives greater context to the economy rates. For example, Ishant Sharma's economy rate of 9.86 is made to look considerably worse by the fact that the innings run rate in matches in which he bowled was almost two runs per over lower.

By comparing a bowler's ranking under economy rate with their ranking under ERx it is possible to ascertain those whose rankings change most significantly as a result of the new measure. In the case of positive movers - those whose ERx ranking is higher than their economy rate ranking - the size of their move illustrates how significantly economy rate undervalues their performance.

 Bowler Overs ER ERx Move Tabraiz Shamsi (RCB) 16.0 9.18 (49) -0.19 (32) +17 Shivil Kaushik (GL) 23.0 8.34 (32) -0.54 (16) +16 Shabhaz Nadeem (DD) 15.0 8.46 (39) -0.39 (24) +15 Iqbal Abdulla (RCB) 25.0 8.52 (41) -0.27 (26) +15 Yuzvendra Chahal (RCB) 49.1 8.15 (29) -0.57 (15) +14 Shane Watson (RCB) 56.3 8.58 (42) -0.24 (28) +14 Imran Tahir (DD) 16.0 8.62 (43) -0.20 (31) +12 Ravindra Jadeja (GL) 40.1 7.74 (20) -0.82 (10) +10 Praveen Kumar (GL) 51.3 7.78 (21) -0.68 (11) +10 Axar Patel (KXIP) 47.5 8.11 (28) -0.45 (20) +8

It is unsurprising to see Royal Challengers Bangalore bowlers feature prominently on the above list because RCB's home ground, the M Chinnaswamy Stadium, was the best venue for batting last season, which meant RCB bowlers therefore had relatively high economy rates. However, when the economy rates of Tabraiz Shamsi, Iqbal Abdulla, Chahal and Shane Watson are compared with the innings run rates of the innings they bowled in, it shows the four of them to have bowled considerably better than the innings average.

 Bowler Overs ER ERx Move Jayant Yadav (DD) 17.0 7.35 (8) -0.05 (37) -29 Nathan Coulter-Nile (DD) 15.0 7.60 (15) 0.25 (42) -27 Barinder Sran (SRH) 49.3 8.34 (34) 0.72 (48) -14 Moises Henriques (SRH) 52.0 7.98 (26) 0.02 (39) -13 Shakib Al Hasan (KKR) 31.0 7.83 (23) -0.16 (35) -12 Carlos Brathwaite (DD) 23.1 8.15 (30) 0.02 (38) -8 Morne Morkel (KKR) 35.3 8.36 (35) 0.26 (43) -8 Marcus Stoinis (KXIP) 23.0 8.43 (37) 0.36 (44) -7 Bhuvneshwar Kumar (SRH) 66.0 7.42 (11) -0.54 (17) -6 Krunal Pandya (MI) 31.1 7.57 (13) -0.52 (19) -6

The biggest negative movers are bowlers whose performances have been flattered by their economy rates. For example, Nathan Coulter-Nile's economy rate of 7.60 is, at first glance, a good one; however, ERx shows that figure to have been 0.25 runs per over worse than the innings average--not so impressive.

It is long overdue in T20 coverage for career and competition economy rates to be presented by match phase, which are typically considered to be: Powerplay (over one to six), middle overs (seven to 15) and death overs (16 to 20). ERx by phase would be a further improvement on this.

The value of doing this is illustrated by Dwayne Bravo in the 2016 IPL, whose economy rate of 8.82 is typically high. However, analysing Bravo's bowling by phase shows him to have bowled 49% of his deliveries in the death overs - the most expensive phase of the innings - which will have significantly affected his overall economy rate. Closer analysis of his phase breakdown shows him to have recorded negative ERx (better than the innings average economy rate for the phase) in the middle overs (economy rate 6.29, ERx -1.45) and the death overs (economy rate: 10.52, ERx -0.68). He only bowled two overs in the Powerplay phase. Calculating ERx in this manner adds another layer of depth to the measure and helps provide an even more accurate reflection of a bowler's performance.

Economy rate is a measure that has served cricket well for decades. However, in T20, where the margins are so fine, it is a measure that lacks the nuance to give an accurate enough reflection of a bowler. ERx provides for a more complete picture by factoring in the opposition batting strength, conditions and the match scenario, to offer a fairer representation of a bowler that remains simple to understand.

Freddie Wilde is a freelance T20 journalist. @fwildecricket

• Hamish on March 3, 2017, 1:36 GMT

Spinners bowl a lot less of the closing overs which are still the most expensive overs in most innings.

Also bowlers in a strong bowling team will come off worse than the same bowler in a weaker team.

• Hannes on March 2, 2017, 20:57 GMT

I've been making similar arguments for the past couple years. The other factor I believe is critical in rating a bowler's impact in T20s is strike rate. Wickets change the context of the match by slowing down runs and forcing the batting team to adjust its plans. Obviously, top order wickets, as well as those of key middle order power hitters, are more valuable than, say, those of tailenders, but this would be difficult to standardize easily. However, overall, I believe a measure that couples context-relevant differential economy rate (ERx) with SR goes a long way to determining a bowler's true value relative to other bowlers. It has the benefit of being completely quantitative and objective, and I think it's universal enough to easily calculate on a running basis. I'd suggest a simple division: ERx / SR; the further this index moves below zero, the better the bowler's career performance, suggesting a greater overall value relative to other bowlers.

• Adam on March 2, 2017, 14:48 GMT

"Startling thing is that the list of best ERx is dominated by spinners, and the worst ERx list is dominated by quicks. "

That's entirely due to the periods of the game where they bowl.

• Chris on March 1, 2017, 23:07 GMT

Startling thing is that the list of best ERx is dominated by spinners, and the worst ERx list is dominated by quicks.

This flys in the face of the early - and reasonable - prediction that T20 would see spinners belted out of the game.

Maybe it is time to limit bats in T20 to give quicks more of a chance.

At the moment, those stats show it's the quicks who are going to get belted out of the game.

• niaz on March 1, 2017, 21:43 GMT

As some have pointed out how flawed. SRH had the best bowling lineup with economy rate of 7.95. Only one of their bowlers showed up in the list 20 or so bowlers. This is how it should measure

NER (Economy rate): (Summation of Runsgiven*SR (of batsmanwho scored it))/Balls bowled by the bowler NAvr (Average for bowlers): (Summation Runsgiven *Average batting of the batsman) / (Summation of average of the batsmen whose wickets were won by the bowler).

Similarly batting SR should be weighted against the Economy rate of the bowler and batting averages should be weighted against the wicket taking ability of the bowler.

These should be used along with normal statistics like economy(or Strike) rate and averages.

• Anurag on March 1, 2017, 16:07 GMT

A huge basic flaw in this system. A good bowler in a good bowling team will not get well-ranked, while a bad bowler in bowler in the worst bowling team looks good. This is the reason that 3 SRH bowlers make the list of biggest negative movers, even though they were central in forming the strong bowling lineup that won SRH their IPL title. Also, a good bowler in a bad bowling line-up gets too much glory, case in point being Adam Zampa and Ravi Ashwin in RPS. This will especially be the case, because opposition teams will be cautious against the best bowlers, and go after the not so good bowlers, thus magnifying the difference between the good and the not-so-good bowler.

• Ashhar on March 1, 2017, 4:29 GMT

If you are doing phase wise then strike rate of bowlers should be recalculated also because there is high chance of getting wickets in powerplays and in death overs. Same goes with batsmen, a strike of 140 in middle overs is better than same strike in powerplays and death overs. This will open a pandora box. I think its better to keep it simple.

• David on March 1, 2017, 1:37 GMT

ERx is full of problems. It fails to capture the acceleration at the end of the innings. It fails to capture the batting quality. It measures bowling quality relative to other bowlers in the side which probably isn't useful. (It's more of a problem in international T20s than club T20s; Richard Hadlee's ERx would have been amazing.) It does not reward dismissals. There are too many problems here for the measure to be worth saving.

Some of these issues could be addressed by using Duckworth-Lewis estimates of final score before and after the over. Good bowlers tend to make the estimate fall, bad bowlers make it rise.

• mayank on March 1, 2017, 1:14 GMT

Thanks Freddie - I couldn't have guessed that. Everyone knows this. By this logic, economy rates should be counted differently during last 10 overs. Random article just for the heck of it.

• David on March 1, 2017, 0:26 GMT

Narrowing the differential down to just look at their performance in blocks of the game sounds problematic. If Dwayne Bravo bowls 49% of his overs in the death period with an ERx of -0.68 that just means that he's slightly better than his colleague at the other end. It may even be that Bravo usually bowls overs 17 and 19, and 18 and 20 are generally more expensive.

For example in the T20 World Cup final Chris Jordan had an ERx of -12.5 for the death overs. That looks like Jordan bowled spectacularly, whereas the reality of course is that Stokes had the misfortune of bowling the 20th over and had a complete nightmare.

I'm all in favour of developing new and better stats for us to pore over and distract ourselves from reality, but I'm not sure that ERx is the statistic that we need.

• No featured comments at the moment.