ESPNcricinfo Smart Stats - a new way to understand T20 cricket

Traditional metrics don't do justice to T20 performances. We need a new language to understand them Getty Images

Cricket is blessed to have three different formats, which, while retaining the core aspect of the game - bat v ball - are distinctly different in character. But while the game has evolved continuously, and more rapidly than ever before in the last decade, the tools for measuring performances have lagged behind, and the absence of relevant data metrics is felt most acutely with Twenty20, a game defined and decided by shorter units than any other. The traditional measurements - averages, strike rates, and economy rates - don't always make sense.

Twenty20 needs a more evolved and a more nuanced system which places numbers in a more situational and contextual spaces without sacrificing cold objectivity. It needs a new stats language.

Enter ESPNcricinfo's Smart Stats. Starting IPL 2018, you will see a new set of metrics introduced on the site which will quantify batting and bowling performances. Many of these are contextual numbers, which will take into account the match stats - whether it was a high-scoring or low-scoring match - and the phase of the game in which the batsman/bowler performed. These will therefore allow for the fact that bowlers concede more runs in the slog overs, and batsmen are expected to strike harder then too.

The metrics you see below are only the start of this new stats journey for ESPNcricinfo. Over the next several months, there will be more such measures introduced to evaluate batting and bowling performances across various parameters which will be contextual and relevant to the T20 format.

To start with, the new metrics below look at the strike rate and runs scored for batting, and the economy rate and dot-ball factor for bowling. Here is a look at these metrics, plus a top-five and bottom-five list in each from the last three IPL seasons.

Smart Strike Rate
Smart strike rate is a measure to calculate the scoring rate of a batsman, but it goes beyond dividing runs scored by balls faced. Smart strike rate takes into account the overall scoring patterns in the game, and during the period when the batsman was at the crease.

The two factors considered are:

a) The match strike rate, excluding the batsman's numbers in the game
b) The strike rate at the other end while the batsman was at the crease

The first factor takes into account the overall conditions during the game. If conditions were batting-friendly and both teams scored at ten an over, then the batsmen who played the game are measured against that yardstick. The second factor takes into account when the runs were scored during a game, and what was happening at the other end during the time that the batsman was at the crease. By combining these two factors, it is possible to find out the "smart runs" scored by the batsman, which, when divided by balls faced and multiplied by 100, gives the Smart Strike Rate.

All the calculations are based on numbers from the match itself; there is no dependency on historical data.

Smart Runs Index
It calculates the number of runs by which a batsman has outperformed the average batsman who has come in to bat at a similar stage of the innings in past T20 matches. It can have a positive or a negative value, with a negative value indicating a below-par performance.

Again, there are two parts to this calculation:

a) Calculating the smart runs of a batsman in an innings/group of innings
b) Calculating the average smart runs historically made by all batsmen who have come in to bat at the same stage of an innings (in terms of overs remaining in the innings).

The smart runs for the batsman is calculated based on the strike rate of other batsmen in the entire match, and the strike rate at the other end while the batsman was at the crease. Combining these factors, it is possible to find out the number of extra/fewer runs he scored in the balls he faced. This is then compared with the historical (typically, over a three-year period) average for smart runs scored from the same point of entry, in terms of overs remaining. The difference (which can be positive or negative) is the value by which the batsman has outperformed the mean. This difference is the Smart Runs Index.

The point-of-entry comparison ensures that top-order batsmen don't get the advantage of more overs to play as they are compared with other top-order batsmen, while middle-order batsmen are compared with those who bat at similar positions. Thus, the top-five list has a fair mix of top-order and lower/lower-middle-order batsmen.

Smart Economy Rate
This is a bowler's economy rate after taking into account the rate of other bowlers in the match, and also the phase of the game when he bowled (Powerplays, middle overs, death overs). Thus, due consideration is given to bowlers who bowl the tough overs - the last five/Powerplays.

There are three factors considered when calculating the smart economy rate:

a) Runs conceded in the previous over (not applicable for the 1st, 2nd, and 7th overs)
b) Runs conceded by other bowlers in the phase in which the over was bowled
c) The match economy rate

The runs conceded in the previous over gives an indication of the momentum of the innings immediately before that over was bowled. The run rate in the over phase is meant to factor in the stage of an innings when the bowler bowled. If a bowler bowled in the fourth over, for example, the runs he conceded in that over is compared with the run rate in the other five overs in the Powerplay. This measure, along with the previous-over factor, accounts for the type of overs the bowler bowled, and gives him credit for bowling the tough overs. The match economy rate is used to measure the bowler's overs in the overall match context.

These three factors combine to give the "smart runs" conceded by the bowler in a match, which, when divided by the overs bowled, gives the Smart Economy Rate.

Dot Pressure Index
In T20s, stringing together successive dot balls is vitally important as it builds pressure on the batting team, which often leads to wickets. While the economy rate reflects the overall rate at which the bowler concedes his runs, it doesn't factor in this particular ability to build pressure.

The Dot Pressure Index is a simple metric that rewards bowlers for bowling consecutive dots in an over - one point for three dots in a row, 1.33 for four, 1.66 for five, and two points for a maiden over. The total such points earned is divided by the number of matches played, for a per-match value.

Apart from these metrics, there will also be other smart numbers to enhance ESPNcricinfo's T20 stats coverage. An example of that is mentioned below:

Strike rate in first 5/10 balls
This is simply the strike rate of a batsman in the first 5/10balls when he comes in to bat. It is especially useful in determining which batsman is best suited to bat in which position: a batsman coming in during the slog overs needs to have an especially high strike rate in the first five balls, while that might not be as important for a top-order batsman.

This is just a sampler of the sort of stats analyses and numbers you will see on ESPNcricinfo starting with the 2018 IPL.