See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/273525724

Rating players in test match cricket

Article in Journal of the Operational Research Society · April 2015

DOI: 10.1057/jors.2014.30

CITATIONS READS

31 2,363

3 authors:

Sohail Akhtar Phil Scarf

University of Haripur Cardiff University
26 PUBLICATIONS 393 CITATIONS 148 PUBLICATIONS 4,292 CITATIONS

SEE PROFILE SEE PROFILE

Zahid Rasool
Kemin Industries
1 PUBLICATION 31 CITATIONS

SEE PROFILE

All content following this page was uploaded by Sohail Akhtar on 21 March 2015.

The user has requested enhancement of the downloaded file.

Journal of the Operational Research Society (2014), 1–12 © 2014 Operational Research Society Ltd. All rights reserved. 0160-5682/14
www.palgrave-journals.com/jors/

Rating players in test match cricket

Sohail Akhtar1*, Philip Scarf2 and Zahid Rasool2
1
University of Malakand, Malakand, Pakistan; and 2University of Salford, Salford, UK
In general, the evaluation of player performance in test cricket is based on measures such as batting and bowling
averages. These measures have a number of limitations, among which is that they fail to take into account the
context in which runs are made or conceded and wickets are taken or given away. Furthermore, batting and
bowling averages do not allow the comparison of performances in these two disciplines; this is because batting and
bowling performances are measured using different metrics. With these issues in mind, we develop a new player
rating system for test cricket. We use multinomial logistic regression to model match outcome probabilities session
by session. We then use these probabilities to measure the overall contribution of players to the match outcome
based on their individual batting, bowling and fielding contributions during each session. Our measure of con-
tribution has the potential for rating players over time and for determining the ‘best’ player in a match, a series or a
calendar year. We use results from 104 matches (2010–2012) to illustrate the method.
Journal of the Operational Research Society advance online publication, 30 April 2014;
doi:10.1057/jors.2014.30

Keywords: ranking; rating; sport

The online version of this article is available Open Access

1. Introduction determined in the presence of a team effect and due to the fact
that individuals may make different types of contributions
Rating of teams and players is an important part of most sports.
according to their specialisms (McHale et al, 2012). The rating
Ratings are used for the selection of teams and players for
of individual players in test cricket is the focus of this paper.
tournaments, and for seeding players and teams within tourna-
In test cricket, each team consists of 11 players, a combina-
ments (eg, Association of Tennis Professionals (ATP), 2012).
tion of batters and bowlers and a wicket keeper. The standard
They are of interest in their own right (eg, Castrol Football,
measures of batting and bowling averages have been used to
2012; FA Premier League (FAPL), 2012; Federation
assess player performance over many years. The Reliance ICC
International of Football Association (FIFA), 2012). A rating is
(2011) player rankings are the official ICC rankings, and therein
a numerical value allocated to each player or team based on their
are separate measures for batsmen, bowlers and all-rounders.
contributions, while a ranking is the ordinal position based on
All-rounders are players who bat and bowl to a frontline
the ratings (Stefani, 2011). Broadly speaking, the rating of teams
standard; all-rounders are rated using a combination of their
in team sports and the rating of individual players in individual
batting and bowling averages. Interest in estimating measures of
sports is relatively straightforward and well developed
player performance in cricket goes back to Johnston et al
(eg, Stefani, 1977; Elo, 1978; Harville, 1980). Recently,
(1993), who assessed player performance in 1-day cricket using
Stefani (2011) reviewed rating systems and discussed the three
dynamic programming. Beaudoin and Swartz (2003) proposed
basic types of sport rating systems, namely, the subjective,
a new statistic for assessing players’ performances in 1-day
objective and accumulative. Typically, objective rating systems
cricket using the Duckworth/Lewis (D/L) methodology. Lewis
use data on results to estimate competitors’ strengths. Estimated
(2005) calculates player performance in 1-day cricket. The
strengths then directly provide a rating, and rankings can be
author also uses the D/L methodology to evaluate players’
established from these (eg, Koning, 2009). Many proprietary
performances in relation to the runs expected for the stage of the
systems use a combination of the objective and subjective, with
innings and in relation to the resources that are consumed when
results forming the objective component, and the scaling of
batting, or contributed when bowling. Lewis (2008) extended
results according to other factors, such as the importance of a
this work to evaluate player performance over a longer term.
match or tournament, the subjective component (eg, ATP, 2012;
However, test matches are different because in 1-day matches
FIFA, 2012). The rating of individual players in team sports is
there is no notion of playing out the time remaining for a draw.
more challenging since individual contributions have to be
Recently, Borooah and Mangan (2010) proposed new ways of
computing test match batting averages to account for two
*Correspondence: Sohail Akhtar, Department of Statistics, University of
Malakand, Lower Dir 25000, KPK, Pakistan. deficiencies. First, they take account of a player’s career
E-mail: [email protected] consistency. Second, they take account of a player’s batting
2 Journal of the Operational Research Society

Table 1 Batting scorecard for first test, India against Australia in India in October 2010

Australia batting scorecard India batting scorecard

Players First innings Second innings Players First innings Second innings

SR Watson 126 56 G Gambhir 25 0

SM Katich 6 37 V Sehwag 59 17
RT Ponting ‡ 71 4 R Dravid 77 13
MJ Clarke 14 4 I Sharma 18 31
MEK Hussey 17 28 SR Tendulkar 98 38
MJ North 0 10 SK Raina 86 0
TD Paine† 92 9 MS Dhoni‡,† 14 2
MG Johnson 47 3 Harbhajan Singh 0 2
NM Hauritz 9 9 Z Khan 6 10
BW Hilfenhaus 20* 6 VVS Laxman 2 73*
DE Bollinger 0 5* PP Ojha 0* 5*
Extras 26 21 Extras 20 25
Total 428 192 405 216
Wickets 10 10 10 9
*denotes a not-out innings.
†wicket keeper.
‡team captain.

contribution relative to their team’s contribution (total). How- developed by Scarf and McHale (2005) and McHale et al
ever, fundamentally these adjustments do not consider the (2012) for rating soccer players.
influence of a contribution on match outcome. To illustrate the importance of different match contexts,
We propose a new rating system based on players’ perfor- consider the following example: the first test match between
mances (batting, bowling and fielding) during each session of a India and Australia at Chandigarh, India in 2010. The batting
test match. (Note that each day of a 5-day test match is divided scorecard for this match is shown in Table 1. The match was
into three sessions: morning, afternoon and evening, each very closely contested and India won by 1 wicket. Watson
nominally 2 h in length.) There are advantages to looking at scored 126 runs in the first innings and 56 runs in the second
players’ performances session by session. First, a session-by- innings. Tendulkar scored 98 runs and 38 runs in his two
session analysis can take account of the state of the batting and innings. These players scored the maximum number of runs for
bowling conditions; thus, scoring runs on the first day in a their respective sides during the match. However, it is difficult to
match may be quite different from scoring runs on the fourth or determine how important these batting contributions were in the
fifth day because test pitches deteriorate with the passage of match from Table 1; this is because the state of the match at the
time and batting becomes more difficult in the later stages of the time it was scored is not clear. The session-by-session scorecard
match. Second, a session-by-session analysis can take account (Table 2) shows that Watson and Tendulkar scored runs in
of the state of the match (pressure situation); thus, a player who relatively easy situations in the match. On the other hand, the
scores 50 runs when his team score is 100 runs for 5 wickets most crucial and decisive period appears to be the last four
may be making a more valuable contribution and should be sessions in which Laxman played a very important role, scoring
rated more highly for this contribution than a player who scores 73 runs and arguably playing the decisive innings for his team in
50 runs when his team score is 100 runs for 1 wicket. A similar this match. Traditional measures of performance ignore the
argument may be put forward for bowlers and wicket-taking context of the match, and thus the Reliance ICC player ranking,
contributions. The system we develop also takes account of the which is based on traditional measures, does likewise.
wicket value, that is, whether a wicket of a recognized batsman In order to take account of the match situation when runs are
or ‘tail-ender’ is taken. Furthermore, we propose a rating system scored or wickets taken, we analyse the net contribution of
that measures batting and bowling (and fielding) performances players using a model of match outcome probabilities given the
on the same scale. Then in principle the rating system can be position at the end of each session that was proposed by Akhtar
used to choose the best player in a match or in a series (of and Scarf (2012). This model was developed to predict match
matches), or to determine the new emerging player in a outcomes ‘in-play’. It is similar in principle to those developed
particular year, or the best player in a year, whether a batsman, by Brooks et al (2002), Scarf and Shi (2005), Scarf and Akhtar
a bowler, an all-rounder (a player who is a frontline batsman (2011) and Scarf et al (2011a). In particular, Brooks et al (2002)
and a frontline bowler) or a wicket keeper. Best batting and best use an ordered probit model to predict match outcome given the
bowling performances over a period can also be determined. batting and bowling strengths at the start of a match, Scarf and
The system we propose has certain similarities to that Shi (2005) use logistic regression to predict match outcome
 Sohail Akhtar et al—Rating players in test match cricket 3

Table 2 Session-by-session scorecard for match in Table 1

Day 1 Day 2 Day 3 Day 4 Day 5

Session At lunch At tea End At lunch At tea End At lunch At tea End At lunch At tea End At lunch At tea

Lead 101 179 224 295 390 318 237 148 23 123 188 160 53 −1
W1 1 3 5 6 8 10 10 10 10 13 16 20 20 20
W2 0 0 0 0 0 2 3 4 10 10 10 14 18 19
W1 is Australia total wickets down at end of session and W2 is India total wickets down at end of session.

given the end of third innings position, Scarf and Akhtar (2011) Table 3 Best fitting multinomial logistic regression model at each
extend this latter work to the end of first and second innings stage, and showing Nagelkerke R2 and number of matches used to
positions, and finally Scarf et al (2011a) consider declaration estimate the model
strategy using the third innings position model. Day Session Model Nagelkerke Number
Using the session-by-session match outcome probabilities, R2(in of matches
we measure the effect of a contribution (runs scored or wickets percentage)
taken during a session) on the match outcome probabilities; the
Day 1 At lunch G + L + W1 24.4 146
greater the effect, the greater the value of the contribution and the At tea G + L + W1 32.3 146
more highly the contribution should be rated in the rating system. End of day G + L + W1 39.6 146
This principle of rewarding contributions by players in a team Day 2 At lunch G + L + W1 + W2 44.0 146
according to the effect of the contribution on the match outcome At tea G + L + W1 + W2 53.4 146
was essentially put forward in the development of the EA Sports End of day G + L + W1 + W2 56.7 146
Day 3 At lunch G + L + W1 + W2 60.7 146
Player Performance Index for soccer (McHale et al, 2012). The At tea G + L + W1 + W2 66.5 146
novelty of our paper lies in implementing this principle in the End of day G + L + W1 + W2 72.7 143
context of a sport in which players make fundamentally different Day 4 At lunch G + L + W1 + W2 74.8 131
kinds of contributions. Early development of the work is At tea L + W1 + W2 74.2 122
described in Scarf et al (2011b) and Akhtar (2011). End of day L + W1 + W2 80.5 111
Day 5 At lunch L + W1 + W2 84.0 91
The structure of the paper is as follows. The session-by- At tea L + W1 + W2 94.9 76
session prediction model is briefly described. We then
describe how the prediction model is used to determine Covariates here are L, lead of reference team; G, ground effect; W1, total
wickets down by reference team; W2, total wickets down by opponents.
player contributions. We then compare different potential
method measures of player contribution. Furthermore, we
illustrate player rating with examples. Lastly, we introduce a
weighted average rating scheme. assumes Y has a multinomial distribution, that is, Y ~ MN
(p1, p0, p − 1; ∑pi = 1), where p1, p0 and p − 1 represent the
probability of a win, a draw and a loss, with
2. Measuring batting, bowling and fielding contributions p1 ¼ PðY ¼ 1 j XÞ
We use match outcome probabilities to measure the contribu-

exp α1 + βT1 X
tions of players. The probabilities are predicted using a nominal ¼


multinomial logistic regression fitted session by session to data 1 + exp α1 + βT1 X + exp α - 1 + βT- 1 X ;
on 146 test matches. Data were obtained from the ESPNcricinfo p0 ¼ PðY ¼ 0 j XÞ
website (ESPN Cricinfo, 2010), and the complete data set of
146 matches relates to all the test matches in the period between 1
¼


November 2005 and March 2010, excluding those matches for 1 + exp α1 + βT1 X + exp α - 1 + βT- 1 X ;
which the session-by-session data were not available and in
which more than 90 overs were lost to poor weather. Session- p - 1 ¼ PðY ¼ - 1 j XÞ
by-session information is not generally available before

November 2005. The data were processed in order to obtain exp α - 1 + βT- 1 X
¼


session-by-session information (eg, Table 2) for each match. 1 + exp α1 + βT1 X + exp α - 1 + βT- 1 X :
The nominal multinomial logistic regression model is as
follows. With match outcome Y taking values (1, 0, − 1) to A sequence of models with covariates lead, ground effect
denote a win, draw and loss, respectively, covariates denoted by and total wickets lost for each team is obtained. These
X and taking a draw (0) as a reference category, the model models are given in Table 3. The coefficients are shown in
4 Journal of the Operational Research Society

Table 4 The fitted parameter estimates for each session of the match with covariates lead (L), total wickets down by reference team (W1), total
wickets down opposing team (W2) and ground effect (G)

Win/draw Loss/draw

Constant G L W1 W2 Constant G L W1 W2

Day 1 Lunch 2.719 − 3.485 − 0.009 − 0.062 — 1.839 − 4.977 − 0.012 0.504 —
Tea 1.977 − 3.830 − 0.004 0.196 — 1.042 − 5.345 − 0.007 0.578 —
End 2.566 − 3.815 − 0.005 0.139 — 2.551 − 4.952 − 0.012 0.394 —
Day 2 Lunch 2.562 − 3.712 − 0.003 0.027 0.161 3.454 − 5.648 − 0.012 0.263 − 0.119
Tea − 0.262 − 3.766 0.001 0.183 0.236 0.447 − 5.585 − 0.011 0.496 − 0.146
End 0.126 − 4.238 0.001 0.050 0.310 0.656 − 5.414 − 0.010 0.300 0.037
Day 3 Lunch − 1.461 − 4.467 0.002 0.129 0.298 − 0.932 − 5.957 − 0.010 0.393 0.052
Tea − 2.990 − 4.661 0.004 0.067 0.485 − 3.334 − 4.620 − 0.009 0.421 0.179
End − 6.920 − 4.288 0.002 0.260 0.583 − 6.100 − 4.947 − 0.010 0.669 0.088
Day 4 Lunch − 8.926 − 3.404 0.003 0.168 0.759 − 7.823 − 4.690 − 0.012 0.695 0.124
Tea − 9.260 — 0.003 0.176 0.569 − 8.690 — − 0.015 0.899 − 0.341
End − 11.614 — − 0.001 0.323 0.554 − 16.800 — − 0.025 1.370 − 0.330
Day 5 Lunch − 20.321 — − 0.007 0.617 0.759 − 23.964 — − 0.031 1.653 − 0.223
Tea − 130.711 — − 0.025 5.896 1.154 − 404.939 — − 0.302 24.858 − 4.288
End − 130.711 — − 0.025 5.896 1.154 − 404.939 — − 0.302 24.858 − 4.288

Table 4. There are two points to note here. First, the Thus, the hypothetical position for the batting side at the end
win–draw comparison and the loss–draw comparison depend of session t is (L = lt − 1,W1, = w1,t, W2 = w2,t); if the batters score
on the covariates in different ways—this is a property of nothing in the session then their lead does not increase and
nominal as opposed to ordinal multinomial logistic regres- remains at lt − 1, which is the lead at the end of the previous
sion. Second, the wickets covariate is transformed to wicket session t − 1.
resources before fitting—this allows the value of the tenth Similarly for the bowling side, we define a hypothetical
wicket, for example, to be less than that of the first wicket in position at the end of session t in which bowlers have not taken
the wickets-lost effect. The models of Akhtar and Scarf any wickets in session t and obtain match outcome probabilities
(2012) included further covariates: pre-match team strengths given the hypothetical position
and home advantage. We do not use these here because our

t;bowl ð yÞ ¼ Pt Y ¼ y j L ¼ lt ; W1 ¼ w1;t - 1 ; W2 ¼ w2;t - 1 : (3)
PH
view is that the worth of the player should be based on the
current performance. Finally, note that the number of Thus, the hypothetical position for the bowling team in
matches available for model fitting for later sessions session t is (L = lt, W1 = w1, t − 1, W2 = w2, t − 1). In this way, if
decreases because not all matches last the full 5 days. the bowling team take no wickets in the session then the wickets
Using these models, we can predict the match outcome down for the other team remains at wt − 1. Of course, the wickets
probabilities (win, draw and loss) given the match position at down for the bowling team remains unchanged during the
the end of each session t (t = 1, 2, 3 … 15). These are given by session.

 At the end of each session, to assess the worth of players’
PAt ð yÞ ¼ Pt Y ¼ y j L ¼ lt ; W1 ¼ w1;t ; W2 ¼ w2;t (1) contributions, we use the difference between the hypothetical
match outcome probabilities and the actual match outcome
where PAt denotes the probabilities at the end of session t, and probabilities. These differences, for batters and for bowlers, are
W1, W2 and L are, respectively, the total wickets down of the our measures of total player contribution during the session.
reference team, the total wickets down of the opponent team The probability difference (total contribution) is then awarded
and the lead of the reference team. The reference team bat first to a batter according to his share of the runs scored in the
in the match. The actual match position at the end of session t is session, and to a bowler according to his share of wickets taken
(L = lt,W1, = w1, t, W2 = w2, t). in the session. For bowling contributions, however, a share is
Next, to measure performance session by session, we define also given to fielders. The details of these procedures are
hypothetical positions for the batting team and for the bowling given below.
team. For the batting team, we define a hypothetical position at However, matters are not straightforward. This is because we
(the end of) session t in which batsmen have not scored any can look at probabilities in relation to alternative outcomes:
runs in session t, and obtain match outcome probabilities given winning, drawing or at least one of these, or equivalently, not
the hypothetical position: losing. Hence, we can calculate in the manner described above

 the value of a contribution to winning or the value of a
t;bat ð yÞ ¼ Pt Y ¼ y j L ¼ lt - 1 ; W1 ¼ w1;t ; W2 ¼ w2;t : (2)
PH contribution to not losing. Different situations have been
 Sohail Akhtar et al—Rating players in test match cricket 5

a b
1.0 1.0 0

total wickets-taking points (not lose)

0.9 0.9 0

0.8 0.8 0
batting points (not lose)

-1 1
0.7 0.7 1 1
0
1 1
0.6 -1
0.6 00 0 -1 1 -1
1 -1 1 1
-1 1
0.5 -1
0 0.5 0
0
1 11
-1
1
1 00 1 1 -1 1 -1 1 1
1 0
1 1 0 1 11
0.4 1 0.4 -1 1 1 1 -1-1 1 1 -1 1
1 -1 1 1 0 -1-1 1 -1 1 -1 -1 1 1 11 1
0 1 01 1 0 0 1 1 1 -1-1 1 1
1 -1 1 -1 1
1 -1 -10 -1 111 1 10 1 -1-1
0.3 0 1-1
-1 -1 -1
010-1 -1 -1
0 0 -1
-1 1 10 1 00-1
0 0.3 -1 0 10 1 0 1 -1-1
1
0 1 -1 1
-1
1 -1
-1
1 010
1
0 -1
1111
-1
-1 1 1
1 1
1 1 0 1
10 -1 0 11 11 -11 1-10 0 0 0 -1 -1
1 11 1 11
-11-1-1 1 -1 1 1
10-1111 111-1 -10 0 0 110-1 -1-10-10
1-1 01111 11 1 -1 11 1 0 01 0 0
-1 1 -1-1-1
11
0 -1 1 -1 1 1 11 1 -1 -1 1 0
0.2 0 0 -1 -1 1
-1 1 0 -1 1
-1
111 0 0
01 001 1 1 -1
-1
1 -1 11 1 1
1 1011000 1
01
1 01 1
1
1
1-1 10 0.2 11
1 -1
0 -1
1
1 1 -1-1
0 -1 1
1
1
-1 -1
1
11-1 0
1
0
-1 0 0
-1 1
0 -10 0 0 -1 10101 0 00-1
0
1 -1
0-1101101 1
1001-1 -1
01-1
01-101-1-1 1 1
11
1-1 0 1 1-111 00 1
1 0 1 -1 -1 1 -1 1 0 -1-1 0
1
0 1 -1
1 -1 -1 1 1
00-1011-1 -1-10 -1 11
00-1 -1 0-1
11-1 11-1
10-1
-1 1-1
1-10
-1-1
-1 -11-1
1
0-1
11-1 0-1
0 -1-1 0
-11111-1-1
011
-1 1-1 -1
0 -1
10-1 0 1
0 -1 -1-1
1 -1
1 1
0
1
1 -1 0
-1
0
1 -1
-1
0 -1 1 01 0 -1
1 1 -1
1 -1 11 1 0
0.1 1 -1
00
10
-1
-1 0
-1
0
-1
11-1-1
0-1 -11
011
00 01
1-1
-1
0 -1
1
11011
00-1
1-1
1-1
11
0-11
-1
0-1 -1
1 -110
1 0
-1 00
1
-1
1-1 -1
-1-1
-111111 -1
1 0
10
0 -11 1 0 10 1
0 -1
0 1 0 0 -11 1 0 0.1 01 1 -1
1 1 0
-1
1
-1
1
-1
1
1 -1 -1
0
0 -1
1
-1 1
-1
0 -1-1
1 -1
0 -1 -1
11 1 -1
-1 1 -1 1 1 111
-1
0 1
-1 1-1
-1
-10
0-10
11111
11
-1
0-1
-1 0 11
1
0-1011
-1
1-1 11
-11 11
-1
0 0
-1
-1
0-1
111
00 1-1 1
-1 -1-1 0
-1
-1
1001-1
-110111
1-1 0
11
11111-1
0
-1 0 101 0 00
00-1 0
1 1001-1 -101-1
0-1
-1 -10 01 1 1 0 01000 0
0 1-1 0 -1-1-1-1
1
1 -1
0 1
-1
0 1 1
1
-1
1 -1
1 0
-1
-1
0
-1
1
1
0
00
1 -1
1 -1 -1
0 -1
-1
-1 1 0 -1
1 -1 0
0 -1
-1 0 0 -1 0 -1
1 1 1 011 11
0-1 -1-1
-11-1
00-11 1
-1
01
1
01-1
11
-1
1-1
00 101
-1 1
11
-1
110 0 1
0
-11
-1
11111-1
1 0
01
-100-1
-1
00-1
11
-1 1
11-111
-1 -1
101
-1
0
01 1-1
0 -1
1
-1 10-1
0
-1
0-1
1 0-1
1 -1
1
-1 1001 -1 10-1 -1-110-1111-1-1 1 11-1
1
11-1
-1 0
-1 0
-1
1 -1
1
0
-1
1 0
0 -1
-1 1 1
-1 -1
0
1 10
0 -1 0
1 -1 1 -1
1 01 1
-1 -1 0
1-1-1
1-1 -1
1
--1
-1
-1
11
-1
1
-11
1-1
1
-1
0-111-1
111
-1 1
-11
11
011 1
-1
0-1-1
-1
1
0
1 0
1
0
-1
1 1
-1
-1
1
0 1
0
-1-1
1
1
0
1-1
0-1
0
-1
-111
1
-1
-1 0
0
11
01
-1
0
0-1
1 -1
-1
-110
1
0-1
-1
0
111
-1-1
00 011
11
0
-1
1-1
-1-1
-1 -1
0-1
0-1
1
-1
1 11
-1
1
-1 11-1
0 01
0
-11-1
0 0
0-10
0-1
-11
0-1 1101
-1
-1
-1
0 0
-1
100
-1
1
-1
0
-1 0-1
00
-1 1-1
-100-1 -1
-1
1
0
00-1
-1 0-1
-100-1
1
-1 0 0-1 0 0-1 0
110-1
-1 -1 1 1 -1
10 -1
0
1
11
0
-1
1
0
-1
1
-1
-1
0
1-1
1
-1
0
1
-1 1
0
1
01-1
1
1 -1 0
-1
1
1 0
-1
0
1
-1
1
-1 1
0
-1
1
1
0
-1
1
-1
0 1
-1
1 -1
0
-1
-1
1
-1
0 -1
-1
1
-1
1
-1
0
0
0
-1
1
-1 0 1
01
1 -1
1
0
-1
1
1
1 -1
1
-1 0
01
1
0
1 1
1
0
1 00 1
0
-1 1 1 -1
1 1 -1 -1
0.0 0
-11
11
-1
0
0
-1
-1 1
-1--1
-1
-1-1
1
11
11
-000
--1-1
11
00-1
000
00
11
-1
-111
0
1
1 -1
00 1
--1
00
-1
-1-1-1
-1
11100
-1-1
00
0
-1-1
-1-1-1
1-1
0-1-1
-1-1
-1-1 0
-1
01
1
11
00
1
-1-1
0 0
00
1
0
01
-1-1
-10
-1-1
-1-1-1
1-1
0000 1-1 -1
-10
11
-11
0
1
-101
-1-1
-11 0
--1
1
0-1
11
0
-1
01
00
-1-1
-1-1
-1 1
-1
1
0-1
01
1-1 0 -1
1
--1
-1
1 1
-0
0
-1
1
01
10
-1-1
1
00
-0000
-1
-1 1
1
--1
1 0
1
-1
-1
10
-1
1
00
-11
00
0100
-1
10
-1-1
-1
-1
-1
-1
-1-1 -1
0
000
1
-1
-1
-1
-1
100-1
0-1
1
-1
-1
0
1-1
1 -1
0
10
-10
-1
1-1
--1
1-1
-1
0 1
0-1
100
-11-1
-1
1-1 -1
-1
0
0-1
0
-10-1
-1-1110
1-1
-1
-1
-1 1
-1
-1
-1 0-1
11-1
0
-11
-1-1
0
-1 0
11
00
-1
-1 100-1
-1
-1 1-1
010000-1 0
-1
-1 00
1-10 001
-101 110-1
1-1
0 --1
0-1 000
10 00 -11 00-1
0-10 0-10 01 0 0 0.0 1
--11
10
0
-10
00
-1-1
-111
-1
00
1-1
111
0
1
1-1
-10
11
00
0
-1
-1
1
1-1
111
1
-1
0
11
00
1
11 1
0
-1
1
0
1
-1
0
0-1
1
0
1
-1
0 0
1
0
-1-1 1
0
1
1
-1
-1
1
0
-1
00
1
0
0
1
-1
0
1
0 1
1
0 0
0
1
1
0 0
-1
1 0 -1
1
0 -1
-1
1 1 1
0
11
1 0
1 0
1 -1
0
-1
1
0 1 1
0
1 0 0 0
1 1
0 0 0
-1 111 1 1
1 1
00 -100
-1-1-1
00 -1 0
-1-1
0010
-1 -1-1
0
-1 0
-1
-1 00
00 0
-1-1 -1
-1 -1 -1
-1
-1-1
0 010-1
-1
0
0
00
-1-1
1
-1
-10 0
00
-1
0 -1
1 -1
-
-1
0 -1
1 0
0
1-1
1
01
-1
0
-1
0
-1
0
1 -1
1 1
-1
0 -1
0 -1
-1
0
0 -1
0 -1
0 00
0 0
0 -1
-1
0 0 -1
-1
0
-1 0 0-0
-11 0
0 -1
0
-1 -1 1
0 -1
0 1 -1 -1 0 -1
1
0 -1
-1
1 010 0 0
-1 -1
1 0 0 -1 1 00 -1
0 1 00 0 0
-0.1 -0.1 -1 00 0 0 -1
0 0000 0 0
0 -1 1 1
0 1 -1 1 0
-0.2 -0.2
0 50 100 150 200 250 300 350 0 2 4 6 8 10 12
total score wickets share

c d
1.0 1.0
0 1
0.9 0.9 -1
1 1
1

wickets-taking points (win)

0.8 0.8 0
0 1 1 1
0 0
0 1 11 0
0
0.7 0.7
batting points (win)

0 1 -1 1 1 1
0 0 0
0 0 -1 1
0.6 01 0.6 1
1 -1
1
0
1 1 1 -1 1 1 1
1 1 0 -1 -1 -1 0
0.5 1
0
0 0 0.5 0 00 0
1 0 1 -1
1 1
-1
1 0 -1 0 1
-1 -1 -1
1 1 1 1 0 1
1 -1
1 -1 -1
-1 0
-1 -1-1
1
0
1 11 1 0 0 -1
0.4 1
1
1 1 0.4 0
0 1 0 1 1 1 0 1 -1
0 -1 -1
0 0 0 -1
0
0 -1
-1-1
01 11 1 -1
-1
-1 1 1 -1
0 0 1
-1 1 1 1 1 1
1
0 1 -1 1
1 -1 -1 1
0 -1
1
0 -1 1
1 -1 11
0 1 1 1 -1 1
0 0 -1 1
0.3 0
1 1 -1
-1 1 1 1
1 0.3 1 1 0 1 -1 -1
-1 0 0
1 11
1 -1
1 1 1 -1
-1 1 01-1
1 -1
-1-1-1
0 01
1 0 1 -1 1 -1
0 1 -1-1-1
1 -1
1 1 0
0 01 -1 1101 11 0 -1 -1 0 1 -1
0
1 -1 0
-1
1 -1 111 1 1 1
0 0 0-1
0-1-1 1-10 0 1 1 0 0 -1 11
0
1 -1
01 0 0 -1
1 0 -1
0 0 -1
0 1
1 -1
0 0
1 0 1 -1 1 -1 1 1 1 1 -1-1 1 1
0.2 0 0 1-1 -1 00-1 -1 -1
01-10 -1 -1
0 -101
0 0.2 1 -1
0
0 1 -11
0
1
-1
-1
-1 -1
1 0 -1
0 -1
0
0 -1-1
1 -1 0
11
1
1 1
1 0 -1-1
0 -1
1 -1 1
0
1 01
0 1
-1
1 1 1
1 0 1 1
0-1 01 1 01 0 1 -110
1 0
1 -1
0
1 1 -1
-1
1
0
0 -1 0 -1
0 -1
0
-1 00
1 1 -1-10 1 -1
1 1 -1 11 1 1
-10
-1 1 -1 -1
11
0 1 1 110-1 1
-1
10 1-1 -11 1 0 0
01
1 1
0
-1 1-1
1
0 0 -1
0
1 -1
1
-1
1
0
-1
1
-1
-1
0
1
-1
0 0 -1
1
1
-1
1 1
0
-1
-1
1
1 -1
0
-1
1
0
1
-1
1 -1
0 1 -1
-1 -1
11 0 1 0 -1 1
1 -1
0
1
1
1
0.1 -1 1 -1 -10-1 -1-1-1
0
1
-1 -1
-1 0-1
1 -1 -1
-1
01-1
-1
1 11 111-1 -1 -1-1-11-1 -1 -1 -1 -11 1 0.1 10 0 -1
1
01-1
1
1-1 1-1
0-1
1 1
0 0
-1
1 1
1 -1
-1 0 -1 0 0 -1
-1
1 -1
1 0
1 1
0
-1 11 1 1 -1 1 0 1 -1 1 1
1-1 1 1
0 -1
0-1 -11-1 1 0-1 -1
-1
-1-1
-1
1 1
1 01
11111-1-1
-1 1 1
-1
-1 1 1 1-1 -1 01-1
1
1 1
0
-1
1 1
0
10-1
0
0 -1
-1
1 -1
1
0 0
1
-1
0
-1 0 -1
1
1 0
0 -1 0
-1 1 -1
0 -1 1 -1
0 11 11 0 1 1
0 -1 -10 -1
10
-1 0-1
11-1
0-1
-1
-1
-1 -1
-1
-1
1-1
-1
-1 -1 11
-1
-1
111-1
-1
-1 1
-1
1-100-1
0 -1-1-1
111-1 0-1 -1
-10 -1
-1-11110 -11001-1 -1 01-1 01 1 1 0 0-1 1 -1
0
01-1
1
0
-11 1
-1
1
0
-1 1 -1
1
0 -1
0
-1 0
-1
1
0 1
0
1
-1 1
-1 0
1 -1
-1 1 0 -1
1
0 1
0 -1
0 01
-1
1 0
0 -1 -1-1-1
-1 0 1
1 1
0 -1 1
-1
1-1
-1
0
0-1
-1-1
-1 0-1
0-1
1 -1
-1
-11 1
0-1
1-1
-1
1-1
-111-1
-1
1-1
-1
1-1
1-1
1 -1 -1
-1
0-110 1-1
-1
-1 1-1
0 0-1
-1 -1-1
0
1-1 -1
001-1
1 -1 -1 11
111
1
-1
-1 11-1 1101 1101-1-1-10 11 1-1
0-11-11-1
0 1 1 1 -1
1
0
-1 -1
0
1
0
0-1
-1 1
0 1
-1
1
-1
0 1
-1
0
-1
-1-1
0 1
1
0
-1 0
0 -1
1
-1 0
1
1 -1 1 -1
1
1 -1 -1
0 0 -1
0
0 -1
1 1
0 -1
1 -1 01 1 01 1
0.0 -1-1
-100
00
1
11
-1
-1
11
111
00000
-1
-1
1-1
01
00
0
-1--1
--1-1-1-1
-1-1
-1-1
11
1
000
-1-1-1
0
0-1-1
1-1-1
1
--1-1
00
-1 11
1-1
000
0000
11
11
0-1
1
0
0-1
001
-1-1
11
1
-1
-1
--1
1
-1-1-1
1
-1-1
-1
-1
-0
1
-1
-1
1
0-1
0 1
1
-1-1
-1-1
11
-1
-1-1
1
011
000
1
-1
00
00
-1
00
11
0
1
0
--1
111
0
-1
-1
1
-1
-1
-1-1
0
11
1 10
-1
-1
-1
1-1
1
-1-1
111
0 1
-11
1
-1
0-1
-1-1
0 101
-11
01
11
-1
-1
1
01
0
11
-1
-10
-1
1
000
-1
-10
0
1-1
1
-1
-1
0
1
--1
-1
01
0
0000
0101-1
-1-1
11 -1
-1
0
1 01
-1-1
11
-1
1-1
1
-11
00
1
0
1
111
-1
-1-1
-1 1111
-1
0
-1
-1
-1
-10 0-1
1-11
-1
1-1
1
01
-1
0
0
1-1
-1
1
-1
1
1
0 01
-1
-1 0
0
01
1
11
-1
000 00
-1
1
-1
-1
-1
0-1
0
1
10
-1
1
11
01
1
1
-1
1-1
1-1
-1-1
-1
1-1
1-1-1
0111
-10
-1
-1
0
-1
1
-1-1
-1
0 1-1
-1
1
1
0-1
11
11
-1
-1
-1
0
-1
1
1-1
0
1
-1
10
00
0
-1
-1-1
1-1
111
0-1
111
1 11
00-1
1
-1
1
-10-1
-1
11
1-1
-1 -1
1-1
10 10
11
1
0100
-1 1
0-1
-1 0
-1
-1
0
1
-1
0 -1011
-1 -1
01-1
11
1 -10 1
11-1
0 001 -0 0-1
11 0- 1011
0-1
-1 0 1 1-1 0111-111 1
0 1 00 1
0 0 0 0.0 1-1
000
-1-1
11
0
11
1-1
1
00
0
1
-1
-1-1-1
-1
0
00
-10
1 00
0 0
1
-1-1
00-1
1
0
-1
1
-1
0
1
0
1 -1
-1
0
1 1
-1
1 -1
1
0 -1
11
-1
0
1
-1
1
0
1
1 -1
1
0
0 -1
1 0 1
1 -1
0
-1 0
1 0
1 0
-1 110
1 1 0
-1
1
-1-11
0-1
-10
-1
1
1-1
-1-1 1 1
-1
-1-1
-1
00
1
11-1
0
1 0
-11
-11
00
11
-1
-1
-111
1-1
-11
-1
-1
-1
0 1
0
11
-1
00
1
-1
-11
01
0
-1
0
001
-1
-1
11
-1
-1
-1
001
-1
-1
-1
1
-1
10-1
11
0
1-1
0
-1
1
111
-1
-1
0
1
-1-1
0
10
-1 11
-1
-1-1
-1
1
1-1
-11
-1
1
-110
-1
0
01
-1-1
-1
0
-1
001
1
0
0
-1
0 -1
0-1
-1
1
10
00
-1
-1
0-1
0
-10
010-1
1 -1
-1 1
0
1
1
0 -1
-1
-1
111-1
0 01
0
1 1
10-1
-1
-1
10
1-111-1
-1 1-1
111
1
0
0010
-1 1 0
1-1 1 0 0-1 -1
1-111-1 1 -1
010111-1
-1
-1 0
1 001110 -1
01
-1 -1 0 010-10 1 -1
0 1 -1 00 1 01-10-1 1-1
-10 01100 1
0 -1-1 1 -1 1 1 1
-1 0
1 1 0 1
-0.1 1 0-1 0 -1 10-1 0-1 1-1100 -1 -1 0 1011 -1
1 1 01 1 0 -1 1
1
-0.1
-10-1 -10 10 -1 -1 1
1 1
-1 00 0
1 -1 0 0
-0.2 -0.2
0 50 100 150 200 250 300 350 0 2 4 6 8 10 12
total score wickets share

Figure 1 For individual players: (a) total batting points (not lose criterion) versus runs in the match; (b) total wicket-taking points (not lose
criterion) versus wickets share; (c) total batting points (win criterion) versus runs in the match; (d) total wicket-taking points (win criterion)
versus wickets share; 1 (win), 0 (draw), − 1 (loss) from point of view batting and bowling team; respectively (104 matches).

investigated, and we propose one criterion for batters and another the batting and bowling contributions of all players in session
for bowlers. For batters, we look at probability differences with t are
respect to ‘not losing’ and at bowlers with respect to winning.
Ct;bat ¼ PAt ðfnot losegÞ - PH
t;bat ðfnot losegÞ
We believe that looking at not losing is a reasonable approach to
rate the batters in test cricket. This is because batting teams act Ct;bowl ¼ PAt ðfwingÞ - PH
t;bowl ðfwingÞ:
conservatively and early in a game will aim to bat towards a
position from which they cannot lose. Even later in a game, a The value of an individual player’s batting contribution is
team will bat for a win only if a target or a lead is moderate;  
ri;t
otherwise they play for a draw. On the other hand, bowlers and Ci; t; bat ¼ Ct; bat ´ :
rt
fielders predominantly play for a win—the notion of bowlers and
fielders playing for a draw does not really exist, despite the fact where ri,t is the runs scored by player i in session t and rt is the
that fields might be set to be defensive. The numerical results total runs scored by his team in session t.
considered later in Section 4 (Figure 1) provide a further For wickets taken, we proceed in the same way in principle to
justification of these choices of criteria. Alternatively, one might runs scored, sharing the total points among those who take
use an expected ‘points’ gained criterion for both batters and wickets. However, we make allowances for fielding contribu-
wicket takers that assumes 1 point for a win and 2 points for tions (catches, run-outs and stumpings), and award a share of
a draw. the points for a wicket to the fielder if indeed a fielder
The value of a player’s contribution is computed at the end contributed. The choice of the number to express such a share
of each session in the following way. Defining the event for catches and stumpings is arbitrary. We suppose there are
{win} that the match outcome is a win for the reference j = 1, …, n types of wicket-taking contribution, each of which
side, and likewise for {lose}, and {not lose} = {win ∪ draw}, has share αj of the points. Thus, j = 1 corresponds to a wicket
6 Journal of the Operational Research Society

taken by the bowler (bowled, lbw, α1 = 1), j = 2 corresponds to (Australia, New Zealand, Pakistan, India, South Africa, West
a catch taken by a fielder (α2 = 0.25) and j = 3 corresponds to Indies, Sri Lanka and England). Figure 1(a) clearly shows a
the bowler contribution in this case (α3 = 0.75). Run-out positive relationship between batting points under the {not
contributions are shared between the fielders involved. Stump- lose} criterion and total score. On the other hand, using the
ings are treated similarly to catches. For player i in session t, the {win} criterion for batters, some of the players earn negative
number of contributions of type j, zitj, can be obtained from the points (Figure 1(c)). This is because batters do not always play
scorecard. If the total number of wickets taken in session t is zt for a win; sometimes batters bat slowly for a draw when the
then the wicket-taking points awarded to player i in session t is opponent’s lead is large. When a batter bats slowly the win and
Pn loss probabilities decrease and the draw probability increases.
j¼1 zitj αj
Ci;t;bowl ¼ Ct;bowl ´ : Therefore, those batters who play for a draw can receive
zt negative points under the {win} criterion. By using the {not
The net contribution of player i in the match is then the sum of lose} criterion for wicket taking (Figure 1(b)), some players
contributions from all sessions: received negative points. On the other hand, Figure 1(d)({win}
criterion for wicket-taking points) gives a stronger positive
X
15

Ci;t;bowl + Ci;t;bat : relationship. Thus, these plots would appear to support the idea
t¼1 of using the {win} criterion for wicket-takers and the {not lose}
criterion for batting points.
We assume in these expressions that there is no change of
innings during a session. When there is, we calculate the actual
and hypothetical probabilities in a slightly different way. We 4. Examples
suppose in effect that the within-session change of innings
occurs at a virtual session-end. The outcome probabilities for 4.1. Example 1
the virtual session-end are calculated by interpolating between Consider now the third test between Australia and England,
the probabilities that can be calculated for the actual session- Perth, 2010. Table 5 provides the summary of players’ batting,
ends immediately before and after the virtual session-end. We bowling and fielding contributions in the match. Batting
carry out the interpolation linearly with weights determined by contributions use {not lose} and bowling and fielding contribu-
the number of overs into the session at which the change of tions use {win}. Australia batted first, scoring 268 runs in their
innings occurs. We omit the details of the calculation here. A first innings and 309 runs in their second innings. In reply,
player will only score batting and wicket-taking points in the England scored 187 runs in their first and 123 runs in the second
same session if there is a change of innings in the session. The innings. Australia won this match by 267 runs. Hussey scored
following should also be noted. First, in modelling match maximum runs in the match and received maximum batting
outcome, lead is calculated session by session throughout the points (0.53). On the other hand, Johnson scored 63 runs and
match with respect to the reference team, the team batting first took nine wickets in the match. He received the maximum
in the match. Therefore, all the above formulae are strictly points (0.66) with this all-round performance and was declared
applicable to the reference team. For the opponents, the team man of the match. Note that wicket-taking points are adjusted
bowling first, the bowling criterion {win} of the reference team (rescaled) so that total batting points and total wicket-taking
becomes {lose} of the reference team and vice versa. Second, points are equal; we do this to standardize the contributions.
the change in the probabilities at the end of a match (last Such standardization might be carried out over a series of
session) is calculated by comparison of the hypothetical posi- matches, or even over all matches played in a specified period.
tion with the actual match result. Third, we could use different The rationale for this is that ultimately, if enough matches are
criteria, for example, {win} for batters and {not lose} for considered, run-scoring and wicket-taking contributions ought
bowlers. This is explored in the example below. Finally, our to be balanced.
intention is to produce a rating system where total batting and
wicket-taking points over all time should be the same. How-
ever, we are not yet in a position to do this as we have not 4.2. Example 2
collected data on sufficient matches. Consider now the 2010/2011 Ashes, a five-match test series
between England and Australia played in Australia. The results
of adopting the contribution system are shown in Table 6. This
3. Comparison of potential measures of contribution
series was an interesting one for all stakeholders: fans, teams,
We now compare the approaches described above. Batting and bookmakers and the communication-media organizations.
wicket-taking points for individual players are calculated by the Squads for each team contained up to 17 players so that more
above stated approaches, and plotted in Figure 1, for 104 test than 22 players took part in at least one match. For brevity, only
matches played in 2010, 2011 and 2012. To calculate the the top 20 players are presented. Note also that wicket-taking
ratings of not too large a number of players for a reasonably points are rescaled so that total batting points and total wicket-
large number of matches, we only consider the top eight teams taking points over the series are equal in order to standardize the
 Sohail Akhtar et al—Rating players in test match cricket 7

Table 5 Player contributions in the third Ashes test, Australia versus England, Perth, 2010

Runs Wickets Run outs Catches Fielding Batting Wicket-taking Fielding Total Rank
share (fielding) (fielding) share points points points points

Australia
Watson 108 0 0 2 0.5 0.42 0.00 0.04 0.46 3
Hughes 14 0 0 0 0 0.06 0.00 0.00 0.06 20
Ponting(*) 13 0 0 2 0.5 0.01 0.00 0.04 0.05 21
Hussey 177 0 0 2 0.5 0.53 0.00 0.02 0.55 2
Clarke 24 0 0 0 0 0.10 0.00 0.00 0.10 18
Smith 43 0 0 2 0.5 0.15 0.00 0.03 0.18 12
Haddin(+ ) 60 0 0 3 0.75 0.11 0.00 0.06 0.16 14
Johnson 63 8 0 0 0 0.15 0.51 0.00 0.66 1
Harris 4 7.5 0 0 0 0.01 0.40 0.00 0.41 4
Siddle 43 1 0 0 0 0.13 0.09 0.00 0.21 9
Hilfenhaus 13 0.75 0 0 0 0.04 0.07 0.00 0.10 16

England
Strauss(*) 67 0 0 1 0.25 0.26 0.00 0.03 0.28 6
Cook 45 0 0 1 0.25 0.15 0.00 0.02 0.17 13
Trott 35 0 0 0 0 0.06 0.00 0.00 0.06 19
Pietersen 3 0 0 0 0 0.00 0.00 0.00 0.00 22
Collingwood 16 0.75 0 3 0.75 0.04 0.02 0.04 0.10 17
Bell 69 0 0 2 0.5 0.19 0.00 0.01 0.20 10
Prior( + ) 22 0 0 1 1 0.04 0.00 0.07 0.11 15
Swann 20 1.5 0 2 0.5 0.03 0.13 0.02 0.18 11
Tremlett 3 7 0 0 0 0.01 0.40 0.00 0.41 5
Anderson 3 3.25 0 1 0.25 0.00 0.23 0.02 0.24 7
Finn 3 4 0 0 0 0.00 0.24 0.00 0.24 8
substitute 0 0 0 0 0 0.00 0.00

contributions. Using the combined contribution measure to match and received the maximum batting points (0.187). On
evaluate the player of the series or tournament award, we see the other hand, Steyn received the maximum wicket-taking
that Alastair Cook, the England opening batsman, is placed in points (0.878) in the match. Note that Sehwag and Ten-
top position. Behind him in second is his countryman Ander- dulkar scored almost the same number of runs but received
son, a bowler whose contribution with the ball was also quite different numbers of points. This is because Tendulkar
outstanding. It is difficult to argue against these players being made 100 runs in the second innings when the defeat of his
the best batsman and bowler, respectively. Cook in fact was team was almost certain. Table 8 summarizes the net batting
awarded the man of the series accolade. England players and wicket-taking contributions in the second test. India
occupy four of the top five positions; Hussey is the exception won the match by an innings and 57 runs. Amla once again
at third. England won the series 4–0. scored maximum runs in the match but received fewer
batting points than Sehwag and Tendulkar. The Sehwag
and Tendulkar partnership was the crucial and decisive one
4.3. Example 3
from the point of view of the match result. Laxman and
Consider the two-match test series between South Africa Dhoni also made hundreds in the match but received fewer
and India in India, February 2010. The series was tied at points than Sehwag and Tendulkar. This is because India
1–1. Tables 7 and 8 summarize the net wicket-taking were already in a strong position in the match and faced the
(bowling and fielding) and batting contributions of the less pressured situation.
players. To calculate the batting points, we tried both the Table 9 shows the total scores over the 2-match series for
earlier stated criteria ({win} and {not lose}). However, the top 10 players in rank order, with wicket-taking con-
under the {win} criterion, a significant number of batsmen tributions rescaled so that the total batting and wicket-taking
received negative batting points. Therefore, batting contri- contributions over the series are equal. Harbhajan Singh is
butions are calculated using the {not lose} criterion. placed in top position with maximum total points on the
Wicket-taking contributions are calculated using the {win} basis of his 7.5 wickets (bowling plus fielding) in the second
criterion. Raw unadjusted contributions are shown. test; arguably, all these wickets were decisive and played a
Table 7 summarizes the net batting and net wicket-taking vital role in the win. Amla scored a massive 490 runs in the
contributions in the first test match. South Africa won this two-match series but is some way behind Singh in the
match by an innings and 6 runs. Amla scored 253 runs in the adjusted contribution score. Sehwag, who scored 290 runs
8 Journal of the Operational Research Society

Table 6 Ashes series 2010/2011: Players in rank order of total contribution, batting and wicket taking (bowling and fielding)

Players Country Batting Wicket-taking Fielding Adjusted wicket-taking Adjusted fielding Total net Rank
contribution contribution contribution contribution contribution contribution

Cook England 0.885 0.000 0.106 0.000 0.059 0.944 1

Anderson England 0.031 1.400 0.063 0.774 0.035 0.840 2
Hussey Australia 0.791 0.000 0.048 0.000 0.027 0.818 3
Swann England 0.039 0.932 0.110 0.515 0.061 0.615 4
Strauss England 0.528 0.000 0.149 0.000 0.082 0.610 5
Tremlett England 0.005 0.950 0.000 0.525 0.000 0.530 6
Johnson Australia 0.200 0.602 0.000 0.333 0.000 0.533 7
Watson Australia 0.434 0.107 0.077 0.059 0.043 0.536 8
Siddle Australia 0.089 0.491 0.002 0.271 0.001 0.361 9
Pietersen England 0.232 0.133 0.164 0.074 0.091 0.396 10
Finn England 0.001 0.650 0.032 0.359 0.018 0.378 11
Prior England 0.061 0.000 0.528 0.000 0.292 0.353 12
Haddin Australia 0.322 0.000 0.069 0.000 0.038 0.360 13
Trott England 0.257 0.000 0.124 0.000 0.069 0.326 14
Bresnan England 0.005 0.515 0.000 0.285 0.000 0.290 15
Bell England 0.287 0.000 0.014 0.000 0.008 0.295 16
Harris Australia 0.010 0.426 0.000 0.235 0.000 0.245 17
Hilfenhaus Australia 0.095 0.192 0.007 0.106 0.004 0.205 18
Collingwood England 0.059 0.075 0.156 0.041 0.086 0.187 19
Clarke Australia 0.163 0.000 0.010 0.000 0.006 0.169 20

Table 7 Raw player contributions for 2010 test series between India and South Africa in India, first test match

Batting contribution Wicket-taking contribution Batting contribution Wicket-taking contribution

South Africa Runs Points Wickets Points India Runs Points Wickets Points
(bowling + fielding) (bowling + fielding)

Smith 6 0.005 0.25 0.026 Gambhir 13 0.004 0 0.000

Boucher ( + ) 39 0.007 0.75 0.048 Vijay 36 0.008 0.25 0.011
Amla 253 0.187 0 0.000 Sehwag 125 0.081 0.75 0.010
Kallis 173 0.152 1 0.097 Tendulkar 107 0.022 0 0.000
Prince 0 0.000 0.25 0.026 Bedrinith 62 0.048 0.25 0.003
Duminey 9 0.003 0.25 0.030 Saha 36 −0.023 0 0.000
de villiers 53 0.030 0.25 0.023 Dhoni (+ ) 31 0.013 0.25 0.021
Harris 0 0.000 3.5 0.299 Mishra 0 0.000 0.25 0.002
Pernell 0 0.000 2.5 0.224 H Singh 47 −0.016 1.75 0.048
Steyn 0 0.000 9.25 0.878 Z Khan 35 −0.024 2.5 0.153
Morkel 0 0.000 2 0.168 Sharma 0 0.000 0 0.000
Substitute 0 0.000 0 0.000 Substitute 0 0.000 0 0.000
Extras 25 Extras 60
Match total 558 0.384 20 1.819 Match total 552 0.114 6 0.249
South Africa won by an innings and 6 runs.

in the series, achieved nearly as many adjusted contribution position strong. This suggests that bowlers are more likely
points as Amla. This is because Sehwag’s runs came at to receive the maximum points, even in the adjusted
a more decisive time than Amla’s, particularly in the second contribution system.
match of the series. It is quite clear that Amla was the
most consistent batsman in the series, but he received
fewer points than Harbhajan Singh. This is perhaps because
5. Rating test match players
one good bowling spell can change the situation in the
match completely and in a very short time; therefore, To rate players on the above measures of contribution may not
under the system bowlers can accrue a large number of be totally satisfactory. For instance, consider the above stated
points. On the other hand, batters take time to make the Example 3, the two-match series between India and South
 Sohail Akhtar et al—Rating players in test match cricket 9

Table 8 Raw player contributions for 2010 test series between India and South Africa in India, second test match

Batting contribution Wicket-taking contribution Batting contribution Wicket-taking contribution

South Africa Runs Points Wickets Points India Runs Points Wickets Points
(bowling + fielding) (bowling + fielding)

Smith 24 0.023 0 0.000 Gambhir 25 0.018 0 0

Petersen 121 0.079 0.5 0.019 Vijay 7 0.005 0 0
Amla 237 0.101 0 0.000 Sehwag 165 0.199 0 0
Kallis 30 0.030 0.5 0.031 Tendulkar 106 0.144 0 0
Prince 24 0.007 0.25 0.018 Bedrinith 1 0.001 0.25 0.023
Duminey 6 0.001 0.75 0.053 Laxman 143 0.048 0.25 0.018
de villiers (+ ) 15 0.026 0.75 0.029 Dhoni ( + ) 132 0.034 1 0.073
Harris 5 0.002 0.75 0.053 Mishra 28 0.023 3.75 0.340
Pernell 34 0.032 0 0.000 H Singh 0 0.000 7.5 1.753
Steyn 6 0.011 1 0.071 Z Khan 0 0.000 4.5 0.273
Morkel 23 0.025 1.5 0.070 Sharma 0 0.000 2.5 0.583
substitute 0 0.000 0 0.000 Karthik (sub) 0 0.000 0.25 0.083
Extras 61 Extras 36
Match total 586 0.338 6 0.344 Match total 643 0.473 20 3.146
India won the match by an innings and 57 runs.

Table 9 Total adjusted player contributions for the 2010 test series Such a scheme could be developed in a number of ways.
between India and South Africa in India (top 10 players) Under one such scheme, the batting, bowling and fielding rating
points might be calculated in the following way. Let the batting
Country Player Wickets Runs Adjusted Rank points of player i (i = 1, 2, … ,22) under a weighted average
(bowling + contribution
fielding) points rating system in a match be
P15 !
India H Singh 9.25 47 0.430 1
t¼1 Ci;t;bat ri 1
South Africa Amla 0 490 0.288 2 P22 P15 + ´
India Sehwag 0.75 290 0.282 3 i¼1 C
j¼1 i;t;bat
R T 2
South Africa Steyn 10.25 6 0.240 4
South Africa Kallis 1.5 203 0.209 5 where ri is total runs (raw) scored by player i in the match and
India Tendulkar 0 213 0.167 6
India Sharma 2.5 0 0.143 7 RT is the total runs scored by both teams (match total). Note that
India Mishra 4 28 0.104 8 t is indexing each of the 15 sessions in a match. Let the wicket-
India Z Khan 7 35 0.101 9 taking (bowling and fielding) points of player i under this
South Africa Pernell 2.5 34 0.089 10
weighted average rating system in the match be
P15 !
t¼1 Ci;t;bowl wi 1
P22 P15 + ´
i¼1 j¼1 Ci;t;bowl
40 2
Africa. First, Amla scored 490 runs in the series at an average of
490 runs per innings and received the maximum batting points where wi is total wickets share of player i in the match. To
(0.288), while Sehwag scored 290 runs in the series at an standardize the contribution scores under this system, we
average of 97 runs per innings and received 0.282 points, nearly multiply the net player contribution (sum of the batting,
as many. There is a large difference between their runs but they bowling and fielding points) by 1038—the average score per
received almost the same number of contribution points. This is match, based on all test matches from 1877 to 2007. In this way,
in part because the contribution points system takes little points achieved are put on a ‘runs-like’ scale. In addition, the
account of contributions after a point when the win or draw maximum batting (runs) and wickets points will be reserved up
probability of any team is close to unity. Second, in both to 1038 points per match.
matches, bowlers stand out in the contribution points table: To illustrate this rating system, Table 10 summarizes the net
Steyn in the first test (Table 7) and Harbhajan Singh (Table 8) in contributions of the top 26 players in the 104 test matches in
the second test. Furthermore, there is the fact that negative 2010, 2011 and 2012 played by the top eight teams. The top
points can be achieved, and this appears to be more prevalent two positions are secured by bowlers. DW Steyn, in the number
for batters. To overcome these problems, it may be sensible to one spot, scored 381 runs and had a wicket share of 109.25. GP
use the contributions as one component of a weighted average Swann secured second place with 593 runs and a wicket share
rating system. The other component might be the raw runs and of 102.75. MJ Clarke in third scored 3120 runs. JH Kallis
wickets in the match. secured the seventh position with his all-round performance.
10 Journal of the Operational Research Society

Table 10 Batting, wicket taking and net player contributions (points) for 104 matches in 2010, 2011 and 2012

Player Country Raw Contribution Points Rank

Runs Bowling share Fielding share Batting Bowling Fielding Batting Wicket taking Total

DW Steyn South Africa 381 109.25 3.83 0.52 6.27 0.19 396.46 3628.10 4024.57 1
GP Swann England 593 102.75 11.83 0.79 7.00 0.62 649.85 3247.09 3896.94 2
MJ Clarke Australia 3120 10.75 12.75 3.10 0.80 0.72 3200.41 651.23 3851.64 3
JM Anderson England 276 100.75 7.00 0.27 5.93 0.53 270.26 3008.72 3278.98 4
M Morkel South Africa 277 87.25 2.25 0.36 4.96 0.33 312.72 2894.34 3207.06 5
HM Amla South Africa 2678 0.00 6.33 2.71 0.00 0.39 2859.11 217.43 3076.54 6
JH Kallis South Africa 2148 20.25 11.50 1.67 1.08 0.65 2115.72 939.87 3055.58 7
SCJ Broad England 794 76.75 1.50 0.97 4.77 0.04 878.90 2132.11 3011.00 8
AN Cook England 2664 0.00 8.08 2.83 0.00 0.51 2592.61 218.97 2811.59 9
HMRKB Herath Sri Lanka 442 88.25 3.75 0.93 3.60 0.15 555.51 2189.52 2745.03 10
MEK Hussey Australia 2317 4.00 11.00 2.58 0.32 0.57 2239.66 451.16 2690.82 11
KC Sangakkara Sri Lanka 2336 0.00 2.50 2.42 0.00 0.45 2534.22 150.07 2684.30 12
AB de Villiers South Africa 2073 0.00 14.33 2.00 0.00 0.84 2220.27 461.68 2681.94 13
Saeed Ajmal Pakistan 191 75.00 2.25 0.23 4.75 0.11 241.17 2357.11 2598.29 14
RT Ponting Australia 2234 0.00 9.00 2.80 0.00 0.47 2312.36 239.48 2551.84 15
MJ Prior England 1749 0.00 25.42 2.15 0.00 1.61 1799.75 729.51 2529.26 16
GC Smith South Africa 1815 0.00 11.75 2.10 0.00 0.61 2079.49 443.99 2523.48 17
KP Pietersen England 2383 4.75 5.50 2.48 0.14 0.34 2287.78 224.46 2512.24 18
PM Siddle Australia 328 65.00 3.00 0.44 4.29 0.12 412.50 2082.17 2494.67 19
IJL Trott England 2177 2.00 4.75 1.90 0.52 0.30 2175.08 280.40 2455.48 20
PP Ojha India 80 74.25 2.25 0.06 4.70 0.12 69.32 2329.51 2398.83 21
V Sehwag India 2158 5.75 6.50 2.05 0.36 0.42 2010.56 373.50 2384.06 22
VD Philander South Africa 267 61.75 0.75 0.46 2.68 0.05 289.09 2076.79 2365.88 23
R Ashwin India 567 58.50 0.75 0.54 3.63 0.03 490.39 1717.50 2207.89 24
MS Dhoni India 1629 0.00 24.50 1.52 0.00 1.49 1478.99 695.85 2174.84 25
Harbhajan Singh India 601 48.50 2.25 0.48 4.13 0.15 508.93 1575.95 2084.89 26

The wicket-keeper batsmen MJ Prior and MS Dhoni were in Table 11 2 ×2 cross-classification table for 104 test matches in
sixteenth and twenty-fifth position, respectively. 2010–2012 played by the top eight teams (30 drawn matches are
Another potential application of our rating system that may ignored), classified by actual man of the match (MOM)
on winning or losing side and highest rated player on
be appealing relates to the selection of man of the match, as the
winning or losing side
choice is often between a batsman or a bowler and occasionally
between a player on the winning team or the losing team. Actual MOM award
Selecting the highest rated player as man of the match has an
objective quality. Counting the instances in which the man of On winning team On losing team Total
the match is on the winning or losing team as opposed to the Highest rated player
highest rated player being on the winning or losing team On winning team 57 0 57
(Table 11), we can see: (a) in only 2 matches (of the 104 On losing team 15 2 17
matches in 2010–2012) was the man of the match on the losing Total 72 2 74
team; (b) if the man of the match was awarded to the highest
rated player (by our method) then there would have been 17
occasions for which the man of the match was on the losing
side. Thus, the table suggests that the subjective choice of man in which the advantage often changes between the teams, while
of the match is biased towards winners. contribution points will be increased standardization moderates
Not only would a man-of-the-match award based on max- this effect. Thus, the unadjusted contributions method will give
imum rating points be fairer to losers, but also it brings a better greater rewards to players in tight contests—this would seem to
balance between batters, bowlers and all-rounders. Among be an advantage of this method. To overcome these drawbacks,
these 104 matches, the man of the match was a batter, bowler, different components of a weighted average scheme might be
all-rounder in 51, 44, 5% of occasions, respectively. The combined once total scores on each component are determined
corresponding figures for the highest rated player are 41, 48, for a series of matches. Thus, suppose in a five-match series r
11%, respectively. Thus, these figures suggest that bowlers play runs are scored and w wickets taken, and thus 1 wicket is
a more decisive role than they are given credit for in man-of- equivalent to r/w runs. Suppose also that the raw total (over all
the-match awards. five matches) contribution points are n for batting and m for
A drawback of our rating system is that the total number of wicket taking. Then a possible scheme that equally weights
points achieved is the same for every match. Thus, for matches batting and wicket taking, and actual runs scored and wickets
 Sohail Akhtar et al—Rating players in test match cricket 11

taken and contributions for each, would be: second (Akhtar and Scarf, 2012). Therefore, the bowlers and




  fielders of the reference team will not be awarded at this stage.
ri + wr wi + nr ni + mr mi
Ti ¼ The collection of data on more matches would be beneficial to
4
overcome this problem.
where ri, wi, ni and mi are the runs scored, wickets taken, raw In addition, bowlers are not penalized for giving away runs,
batting contribution points and raw wicket-taking contribution and fielders are not rewarded for saving runs, or penalized for
points of player i in the series, and Ti is their total weighted giving them away through miss-fields. The latter events are not
average points scored in the series on a ‘run-equivalent’ scale. recorded in match scorecards and thus their incorporation
Note again that no adjustment accrues to batsmen for wickets requires a different form of data recording. For runs conceded
lost or to bowlers for runs conceded. by bowlers, the concept of a hypothetical match position at
session-end cannot be extended because hypothetically a
bowler could concede an unlimited number of runs during a
6. Discussion
session. One might instead consider runs conceded relative to
The purpose of this paper is to evaluate player performances certain norms; this is an avenue for further investigation. For
based on batting, bowling and fielding contributions in the example, one might consider the runs conceded by a bowler
match, session by session. We propose a new rating system for during a session relative to a figure for runs conceded that is
bowlers and batters using a common scale. Nominal multi- ‘loss neutral’ over the session.
nomial logistic regression models are used to model match However, more positively, it is our view that the rating
outcome probabilities for hypothetical and actual positions. system shows potential for the more reliable assessment of
Measures of player performance are then calculated using player performances in batting and wicket taking and in
changes in such probabilities session by session. The advantage comparing performances between these two disciplines. This
of looking at the contribution at the end of each session is that potential has been shown not just for an individual match but
we can evaluate the performance of the player taking into also for a series of matches. The contributions approach might
account the stage of match in which runs and wicket are earned be used in combination with more traditional performance
and conceded. Another advantage of this rating system is that it measures, as we illustrate. If the methodology is to be
provides measures (points) for all-rounders, wicket keepers and incorporated into or to supplant the existing standard and
fielders, again on the same common scale. The ICC rating does commercial measures of performance and rankings, further
not take into account fielding contributions of players. Further- evaluation of the method is necessary, particularly regarding
more, our rating system has the potential to select man of the how the measure would perform in the longer term. To achieve
match, man of the series, best player of the year and new this, many more matches need to be analysed.
emerging player of the year. These measures could also be It is implicit in the system proposed that only those who play
helpful for team selectors and management to select a best can make a contribution, and furthermore that frontline batters
possible team based on their current performance. and bowlers have a greater opportunity to make a contribution.
A limitation of this work is that batting contributions Thus, the system will not measure who are the best players.
and wicket-taking contributions are measured using subtly Instead, it will measure who played the ‘best’ cricket. If one
different criteria: wicket takers are assumed to contribute to were to consider a rating based on contribution rate (per match
winning while run scorers are assumed to contribute to the played) then the best players might emerge, although again,
contrary, that is, to not losing. These different contribution frontline batters and bowlers would tend to be favoured. An
criteria necessitate the rescaling of contributions to a common average (eg, mean) contribution per match would also tend to
base. This can be achieved in principle by choosing a relative favour (and disfavour in equal measure) those who play few
rescaling factor such that the total run-scoring contributions matches due to the increasing variability in the mean as the
balance with the total wicket-taking contributions over the number of observations decreases. However, one could in
‘history’ of the game. This would present a challenge for principle adjust for this by considering the problem of rating as
implementation. A more fundamental challenge, however, is one of estimating player strengths and shrinking the estimated
that in assigning wicket-taking contributions to fielders (for strengths towards an overall mean strength in the manner
catches and run-outs) we must subjectively assign a value for described, for example, by McHale and Szczepański (2013) in
the share due to a fielder. If run-scoring and wicket-taking the context of football.
contributions were determined using the same criterion then we Finally, the contribution measure might be used to deter-
might have chosen the fielder’s share objectively in the same mine the top cricketers of all time. It would be fascinating
way. Unfortunately, we cannot resolve two unknowns with to determine, for example, whether the great all-rounders
a single constraint. of the past (eg, Imran Khan, Ian Botham, Garry Sobers)
Another limitation of this work is the weakness of the match head such a list, and where the best batters (eg, Don Bradman,
prediction model. For example, if both the teams lose wickets Brian Lara, Sachin Tendulkar) lie in relation to the best
on the first day of play then selected models are not able to take bowlers (eg, Muttiah Muralitharan, Richard Hadlee, Michael
into account the number of wickets down for the team batting Holding).
12 Journal of the Operational Research Society

Acknowledgements —We are grateful to two referees for their comments; Lewis AJ (2008). Extending the range of player-performance measures
these helped to improve the paper. We would like to thank Mr. Danny Bland in one-day cricket. Journal of the Operational Research Society
(University of Salford), Mr. Abdul Salam and Habibullah Khan (University
of Malakand) for helping us in data collection. Finally, thanks to the Higher 59(6): 729–742.
Education Commission of Pakistan who provided financial support. McHale I and Szczepański RL (2013). A mixed effects model for identify-
ing goal scoring ability of footballers. Journal of the Royal Statistical
Society Series A (Statistics in Society) 177(2): 397–417.
McHale I, Scarf PA and Folker DE (2012). On the development
References
of a soccer player performance rating system for the English
Premier League. Interfaces 42(4): 339–351.
Akhtar S (2011). Statistical modelling in test cricket. PhD Thesis,
Scarf PA and Akhtar S (2011). An analysis of strategy in the first three
University of Salford, UK.
innings in test cricket: Declaration and the follow-on. Journal of the
Akhtar S and Scarf PA (2012). Forecasting test cricket match outcomes
Operational Research Society 169(11): 1–10.
in play. International Journal of Forecasting 28(3): 632–643.
Scarf PA and McHale I (2005). Ranking football players. Significance
ATP (2012). South African Airways ATP Rankings, http://www
2(2): 54–57.
.atpworldtour.com/Rankings/Singles.aspx, accessed 25 May 2012.
Scarf PA and Shi X (2005). Modelling match outcomes and decision
Beaudoin D and Swartz TB (2003). The best batsmen and bowlers in
support for setting a final innings target in test cricket. IMA Journal of
one-day cricket. South African Statistics Journal 37(2): 203–222.
Management Mathematics 16(2): 161–178.
Borooah VK and Mangan JE (2010). The ‘Bradman class’: An
Scarf PA, Shi X and Akhtar S (2011a). On the distribution of runs scored
exploration of some issues in the evaluation of batsmen for test
and batting strategy in test cricket. Journal of the Royal Statistical
matches, 1877–2006. Journal of Quantitative Analysis in Sports
Society, Series A 174(2): 471–497.
6(3): 1–21.
Scarf PA, Akhtar S and Rasool Z (2011b). Rating players in test match
Brooks RD, Faff RW and Sokulsky D (2002). An ordered response
cricket. In: Proceedings of the 3rd International Conference of
model of test cricket performance. Applied Economics 34:
Mathematics in Sport, (eds. J. Reade, D. Percy and P. Scarf), Salford,
2353–2365.
UK, pp 208–214.
Castrol Football (2012). The Castrol Edge Ranking, http://www.youtube
Stefani R (1977). Football and basketball predictions using least
.com/castrol, accessed 25 May 2012.
squares. IEEE Transactions on Systems, Man and Cybernetics 7(2):
Elo A (1978). The Rating of Chess players, Past and Present. Arco:
117–121.
New York.
Stefani R (2011). The methodology of officially recognized inter-
EPSN Cricinfo (2010). Test match records, http://www.espncricinfo.
national sports rating systems. Journal of Quantitative Analysis in
com/ci/engine/series/index.html, accessed 25 May 2012.
Sports 7(4): 1–22.
FAPL (2012). FA Premier League EA Sports Player Performance Index,
http://www.premierleague.com/en-gb/players/ea-sports-player-perfor-
mance-index.html, accessed 25 May 2012.
FIFA (2012). FIFA/Coca-Cola World Ranking, www.fifa.com/world
Received 7 June 2012;
ranking/rankingtable/index.html, accessed 25 May 2012. accepted 4 March 2014 after two revisions
Harville D (1980). Predictions for National Football League games via
linear model methodology. Journal of the American Statistical Society This work is licensed under a Creative Com-
75(371): 516–524.
ICC (2011). Reliance Mobile Test Championship, http://www.reliance mons Attribution 3.0 Unported License The
iccrankings.com/, accessed 28 January 2011. images or other third party material in this article are included
Johnston MI, Clarke SR and Noble DH (1993). Assessing player in the article’s Creative Commons license, unless indicated
performance in one day cricket using dynamic programming. Asia- otherwise in the credit line; if the material is not included under
Pacific Journal of Operational Research 10(1): 45–55. the Creative Commons license, users will need to obtain
Koning RH (2009). Sport and measurement of competition. De Econo-
mist 157(2): 229–249.
permission from the license holder to reproduce the material.
Lewis AJ (2005). Towards fairer measure of player performance in one- To view a copy of this license, visit http://creativecommons
day cricket. Journal of the Operational Research Society 56(7): 804–815. .org/licenses/by/3.0/

View publication stats