Method

Five innings is not a career: reading Associate numbers

Associate players often have short international records against opponents of very different strength. Here is how to read their numbers without over-reading them.

The short version: an Associate player's international record is often short, and it is built against opponents of very different strength. Both make the headline numbers swing far more than they would for an established Full Member player. That does not make the numbers useless. It means they need to be read with ranges, context and a little patience.

Why Associate records are short

Associate teams play fewer internationals than the big Full Members, and their squads change more. Players move countries for work or study, combine cricket with full-time jobs, and may play only a season or two of international cricket. A player can be important to his team and still have fewer than ten innings on record.

How much can one innings move an average?

Here is a deliberately simple, made-up example to show the arithmetic.

Calculation Hypothetical batter

Five innings, out every time: 45, 3, 0, 12, 60. Total 120 runs, so the average is 120 ÷ 5 = 24.0.

Take away the 60, and the average of the other four is 60 ÷ 4 = 15.0. Swap the 60 for a 5, and the five-innings average becomes 65 ÷ 5 = 13.0.

One innings changes the picture from "useful contributor" to "struggling". Nothing about the player's ability changed; only one result did.

With fifty innings, a single score of 60 moves the average by about one run. With five, it moves it by double figures. The fewer the innings, the more any single day dominates.

The same is true for bowlers. A bowler with four wickets can see his average halve or double in one spell. Strike rates and economy rates are steadier, because they are built on balls and overs, which pile up faster than dismissals, but they still need a reasonable number of balls before they settle.

The opposition problem

In a European qualifying tournament, a side might play a well-established team one day and a team in its first international season the next. Runs and wickets against both count the same in the career record.

So two players with identical numbers may have had very different jobs. One may have made his runs mostly in one-sided matches, the other against the strongest attacks in the group. Before comparing them, look at who the runs came against.

Four habits that help

1. Always show the sample size

Put the innings, balls or overs next to every number. "Strike rate 140 (96 balls)" and "strike rate 140 (900 balls)" are very different statements. If a number rests on very few balls, say so plainly or leave it out.

2. Prefer measures built on balls over measures built on dismissals

Strike rate, economy rate, dot-ball percentage and boundary percentage each use every ball faced or bowled. Averages use only the dismissals, which are much rarer. With a short record, the ball-based measures are usually the more stable starting point.

3. Pull extreme numbers towards the middle

Statisticians call this shrinkage. The idea is simple: with little evidence, assume a player is probably closer to a typical player than his raw numbers suggest, and let the evidence pull him away from "typical" as it builds up. A batter with a strike rate of 200 from 20 balls is more likely a good batter who had a few good days than a batter who will keep striking at 200.

In practice this means blending a player's own numbers with a sensible baseline, such as the average for his team or competition, and giving his own numbers more weight the more balls he has faced. We will explain how we set up such baselines when we use them in an article.

4. Split by opposition strength

Where the sample allows, show a player's numbers against stronger and weaker opponents separately, or at least list the opponents. If almost all of the runs came in one or two matches against much weaker teams, readers should know.

What we do on this site

  • Every table shows the sample size next to the measure.
  • We name the data source and the date it covers.
  • We flag samples we think are too small to support a firm conclusion, and we separate facts, calculations and our own opinions.

Associate cricket deserves serious analysis. Part of taking it seriously is being honest about how much a short record can tell you.