Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
9
, No
.
5
,
Octo
ber
201
9
, pp.
4423
~
44
32
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
9
i
5
.
pp
4423
-
44
32
4423
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Data v
is
ualiz
ation and to
ss relat
ed
analysi
s o
f
IPL
te
ams and
batsmen
perform
ances
Vidit
Kanun
go
,
Tul
as
i B
Depa
rtment
o
f
C
om
pute
r
Scie
n
ce,
CHRIS
T
(De
e
m
ed
to
b
e
Unive
rsit
y
)
,
Ind
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Feb
10
, 201
9
Re
vised
A
pr 16
, 2
01
9
Accepte
d
Apr
25
, 201
9
Sports
play
a
ve
r
y
signif
icant
rol
e
in
th
e
dev
el
op
m
ent
of
the
hum
an
per
son
a
.
Gett
ing
invol
v
e
d
in
games
li
ke
Cric
ket
and
oth
er
var
ious
sports
hel
p
us
to
buil
d
ch
aract
e
r,
disci
pli
n
e,
conf
i
denc
e
and
ph
y
si
ca
l
fit
n
ess.
Indi
an
Prem
ier
Le
agu
e,
IPL
pr
ovide
s
the
m
ost
succ
essful
for
m
of
cri
cke
t
a
s
it
give
s
opportuni
ties
to
y
oung
and
tale
nte
d
pl
a
y
ers
to
show
ca
se
the
ir
ta
l
ent
s
on
var
ious
pit
ch
.
Dec
ision
-
m
ake
rs
a
re
the
utmos
t
cu
stom
ers
for
al
l
funda
m
ent
als
in
the
sports
an
aly
tics
fra
m
ework.
Sports
ana
l
y
t
ic
s
has
bee
n
a
sm
ash
hit
in
shaping
succ
ess
for
m
an
y
play
ers
and
te
ams
in
var
ious
sports.
Sports
ana
l
y
t
ic
s
and
d
a
ta
visualiz
at
ion
ca
n
pl
a
y
a
cru
cial
rol
e
in
sel
e
cti
ng
the
best
play
ers
for
a
team
.
Thi
s
paper
is
about
th
e
To
ss
Rel
at
ed
ana
l
y
sis
and
th
e
bre
adt
h
of
da
t
a
visua
li
z
at
ion
in
supporting
the
d
ecision
m
ake
rs
for
ide
nti
f
y
i
ng
inhe
r
ent
p
lay
ers
for th
ei
r tea
m
s.
Ke
yw
or
d
s
:
Data
vis
ualiz
at
ion
Spor
ts
a
naly
ti
cs
Play
er
pe
rfor
m
ance
“Spyde
r” t
oo
l
“R
” too
l
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Vidit Ka
nung
o,
Dep
a
rtm
ent o
f C
om
pu
te
r
Scie
nce,
CHRIST
(Dee
m
ed
to
be Un
i
ver
sit
y),
77
/
38 Piu
s
Vill
a,1
st
Ma
in
, 11t
h
Cr
os
s
, Bha
ra
thi Lay
out,
DRC p
os
t, S
G Pal
ya
, Ind
ia
.
Em
a
il
:
vid
it
work5@gm
ai
l.com
1.
INTROD
U
CTION
Spor
ts
a
naly
ti
cs
an
d
Data
Visu
al
iz
at
ion
has
prov
i
ded
a
gr
eat
e
r
plat
form
fo
r
Play
er
sel
ect
ors
,
m
anag
ers
a
nd
al
so
the
play
er
s
to
increase
on
fiel
d
pe
rfo
rm
ance.
De
ci
sion
m
aker
s
an
d
an
al
ysi
s,
the
nex
t
piece
of
t
he fram
ework, is the
proc
e
ss of a
pp
ly
in
g st
at
ist
ic
al
too
ls
an
d
al
go
rithm
s
to
data to
g
ai
n
insi
ght i
nto
wh
at
is
li
kely
to
happe
n
in
the
fu
t
ur
e
.
Spor
ts
a
naly
tics
[1
]
is
bei
ng
app
li
ed
i
n
va
r
iou
s
s
p
or
ts
li
ke
So
cce
r
,
ba
ske
tball
and
c
ricket.
Ea
ch
m
ov
em
ent
of
t
he
ball
,
t
he
play
er
strike
r
at
e,
run
rate,
e
ver
yt
hi
ng
is
ca
ptured
us
in
g
s
pecial
ca
m
era
syst
e
m
s
and
o
the
r
rec
ordin
g
m
echani
s
m
s.
This
data
is
run
thr
ough
var
i
ou
s
sta
ti
sti
cal
al
go
rithm
s
,
too
ls
and
vi
s
ualiz
at
ion
te
ch
niques
to
pro
vi
de
de
e
per
i
ns
ig
ht
a
nd
pav
e
way
for
rec
omm
end
at
ion
s
t
o
the
pla
ye
r
or
te
a
m
.
W
it
h
t
he
ease
of
obta
in
ing
a
nd
sto
rin
g
data
,
ad
va
nc
ed
an
al
yt
ic
s
and
m
achine
le
ar
ning
te
ch
nique
s
are
app
li
ed
t
o
en
gi
neer
a
pre
dicti
ve
m
od
el
fo
r
var
i
ou
s
te
am
sp
ort
s
li
ke
cr
ic
ket.
The
re
a
re
three
ve
rsions
of
cricket
–
Test
m
a
tc
hes,
O
ne
-
day
In
te
rn
at
i
on
al
s
a
nd
T
w
enty
20.
Test
Cric
ket
is
one
of
the
highes
t
-
le
vel
form
at
s
wh
ic
h
is
play
ed
b
et
tw
een
tw
o
co
untr
ie
s
ov
e
r
the d
urat
ion
o
f
five
da
ys,
ODI
is
co
ns
ide
red
a
s
a
li
m
it
ed
ov
e
r
form
at
s
of
cric
ket
an
d
T2
0
is
one
of
the
la
te
st
and
su
ccess
fu
l
for
m
s
of
cricket.
The
T
20
f
orm
at
gav
e
birth
to
I
nd
ia
n
Pr
em
ie
r
League
(I
PL
)
a
pr
ofe
ssion
al
le
ag
ue
con
te
ste
d
duri
ng
A
pri
l
and
Ma
y
of
ever
y
ye
ar
[2
]
.
It w
as i
niti
at
ed
by BC
CI (
B
oa
r
d of
C
ontr
ol for
C
ricket in
I
nd
ia
)
in
2008. T
his s
horter
ve
rsion of c
ricket
is o
ne
of
t
he
m
os
t
su
ccessf
ul
one
in
te
rm
s
of
fa
n
eng
a
gem
ent
and
busines
s.
E
ver
y
on
e
e
njo
y
s
this
shorte
r
ver
si
on
of cricket.
The
m
ai
n
obj
e
ct
ive
of
t
his
le
agu
e
is
t
o
pro
vi
de
a
plat
f
or
m
for
yo
ung
a
nd
ta
le
nted
play
er
s.
IP
L
w
orks
on
the
fr
a
nch
i
se
syst
e
m
of
hiring
play
ers
.
Ther
e
are
ei
ght
te
a
m
s
in
IP
L.
Each
te
am
is
a
gr
oup
of
el
even
play
ers
co
ns
ist
ing
of
batsm
en,
bowler
,
an
d
al
l
-
rounde
rs.
T
his
tournam
ent
is
being
play
e
d
in
dif
fer
e
nt
ci
t
ie
s,
because
of
this
,
there
is
a
huge
fan
f
ollow
i
ng
with
a
lot
of
m
edia
interest
and
business
i
nvolv
em
ent.
IP
L
is
a
m
ixtur
e
of
ta
le
nt
an
d
opport
unit
y
so
basical
ly
play
er
pe
rfo
r
m
ance
is
the
ke
y
factor
in
t
his.
Va
rio
us
oth
e
r
key
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
201
9
:
4
4
2
3
-
4
4
3
2
4424
factors
are
the
ty
pe
of
pitc
he
s
–
Flat
pitches,
pitches
that
fa
vor
fast
bo
wlin
g,
s
pin
bo
wlin
g
an
d
swi
ng
bowling
and
w
hethe
r
they
are
be
ne
fici
al
fo
r
bats
m
en,
non
-
stri
ker
batsm
en,
and
bowle
rs
for
holdi
ng
a
good
par
t
ner
s
hip.
Al
l
these
nat
ur
al
par
am
et
ers
an
d
histo
rical
data
of
play
ers
wi
ll
help
t
he
te
am
m
anag
em
ent
i
n
the
sel
ect
ion
proc
ess.
Wh
en
it
com
es
to
IP
L
or
any
ki
nd
of
s
ports,
Tea
m
streng
th,
S
pecial
(K
ey
)
Play
ers,
Ho
m
e
Crowd
play
s
an
i
m
po
rtant
ro
le
in
th
e
pr
edict
io
n
of
a
m
a
tc
h.
An
a
ly
ti
cs
is
on
e
of
the
m
os
t
i
m
p
or
ta
nt
factors
in
Cric
ket
histor
y
[
3].
Ther
e
will
al
ways
be
so
m
e
so
ught
of
un
ce
rta
inty
at
ta
ched
t
o
bowler
or
bat
sm
en
aver
a
ge
perfor
m
ance.
Last
over
’s
a
nd
po
w
er
play
s
a
re
th
e
tur
ning
po
i
nt
of
the
m
at
ches.
Sele
ct
ing
th
e
rig
ht
play
er
f
or
thes
e
cru
ci
al
over
’
s
is
no
t
easy
.
An
al
yt
ic
s
can
help
in
al
l
thes
e
tough
sit
uati
on
s
.
A
naly
ti
cs
br
i
dg
es
the
ga
p
f
or
te
a
m
sel
ect
or
s,
coach
e
s,
a
nd
m
anag
ers
.
Anal
yt
ic
s
giv
es
us
them
a
cl
ear
er
idea
a
bout
play
e
r
consi
ste
ncy,
fa
st
scor
i
ng
an
d
finis
hing
a
bili
ty
.
To
m
anag
e
the
ris
k
in
a
bette
r
way
an
d
to
get
the
pro
bab
le
winners
,
analy
ti
cs
play
a
cruci
al
ro
le
in
the
fiel
d
an
d
ou
t
of
t
he
fiel
d.
D
at
a
Visu
al
iz
at
ion
is
one
of
th
e
m
ajo
r
ou
tc
om
es
in
s
ports
a
naly
ti
cs
[
4].
T
he
visua
l
form
of
dat
a
is
m
or
e
easi
ly
unde
rstan
da
ble
ov
e
r
num
ber
s
a
nd
te
xt.
T
his
pa
pe
r
e
xp
l
or
e
s
th
e
data
visu
al
i
zat
ion
te
c
hn
i
ques
,
Toss
rela
te
d
analy
sis
li
ke
plo
tt
ing
f
or
the
data colle
ct
ed
.
2.
RE
LATE
D
W
ORK
Spor
ts
analy
sis
is
a
hu
ge
cl
ust
er
of
s
pecific
data
an
d
sta
ti
st
ic
s.
Spor
ts
ana
ly
ti
cs
are
the
pr
esent
a
nd
fu
t
ur
e
of
the
pr
of
essi
onal
spo
r
ts
era.
O
n
-
fiel
d
an
d
Off
-
fiel
d
analy
ti
cs
hav
e
gone
beyo
nd
pro
vid
i
ng
play
e
r
an
d
te
a
m
analy
sis
and
pre
dicti
ng
correct
re
su
l
ts.
The
aut
hors
in
[5
]
discu
ssed
a
fact
or
analy
sis
ap
proac
h
to
stu
dy
the
pe
rfo
rm
ance
of
c
ricket
play
ers
a
nd
fi
nd
i
ngs
of
his
stud
y
say
t
hat
batti
ng
ca
pabi
li
t
y
do
m
inate
s
over
bowlin
g.
T
he
s
tud
y
re
veals
th
at
the
pe
rfor
m
ance o
f
bowler
s
is
one
of
the
c
ru
ci
al
an
d
si
gnific
ant
facto
rs
wh
ic
h
m
ay
change
th
e
scena
rio
of
m
at
ches.
Coac
hes
a
nd
sel
ect
ors
ca
n
i
nclu
de
good
al
l
-
rou
nder
play
ers
to
i
m
pr
ove
te
a
m
resu
lt
s.
The
wor
k
in
[6
]
c
om
par
ed
crickete
r
s
ba
tt
ing
an
d
bow
li
ng
perform
a
nces
us
in
g
graph
ic
a
l
m
et
ho
ds.
Ba
tsm
an
an
d
Bo
wl
er'
s
record
of
s
easo
n
20
08
ha
s
bee
n
util
iz
ed
for
the
analy
sis
an
d
inte
rpretat
ion
of
the
grap
hs
. Twel
ve
bowle
rs
a
nd
t
welve b
at
s
m
en
wer
e
sel
e
ct
ed
w
ho
bow
l
ed
at
le
ast
100
balls
an
d
to
ok
at
le
ast
four
wic
kets
a
nd
batsm
en
fa
ced
at
le
ast
10
0
balls
ha
d
at
le
ast
four
com
plete
d
inn
i
ngs.
To
pr
e
dict
the
play
er
perform
ance in
ODI usi
ng v
a
r
iou
s
Mac
hin
e
Learn
i
ng A
l
gorithm
techn
ique
s is do
ne
in
[7
]
.
Naïve
Ba
ye
s,
Decisi
on
tree,
m
ul
ti
cl
ass
SV
M
and
Ra
nd
om
fo
rest
are
use
d
to
ge
ne
rate
the
pr
e
dicti
on
m
od
el
s
for
batsm
en
scor
e
a
nd
bowlers
wic
kets
f
or
bo
t
h
t
he
te
am
s.
Ra
ndom
Forest
gi
ves
t
he
m
os
t
accurate
resu
lt
s
f
or
bot
h
the
scena
rio
s
ou
t
of
al
l
the
fo
ur
te
ch
niques
us
e
d.
T
he
auth
or
s
in
[
8]
discusse
d
va
ri
ou
s
key
perform
ance
ind
ic
at
or
s
t
o
stu
dy
the
play
er
pe
rfor
m
ance
in
IP
L
f
r
om
d
iffer
ent
c
ountries.
Cl
us
te
r
analy
sis
ha
s
been
a
pp
li
ed
on
the
dataset
s
of
play
ers
of
I
PL
seaso
n
2010.
The
st
ud
y
r
eveals
that
play
ers
of
E
ng
la
nd
ha
d
perform
ed
well
as
a
gr
oup
a
nd
New
Zeel
a
nd
play
ers
are
the
lowest
pe
rfor
m
ers.
The
factor
a
naly
sis
us
ed
in
[9
]
with
va
ri
ou
s
sta
ti
sti
cal
t
echn
i
qu
e
s
w
hich
show
s
that
batti
ng
cap
abil
it
y
do
m
inate
s
ov
e
r
bowli
ng.
Dataset
of
85
batsm
en
and
85
bowle
rs
has
bee
n
co
ns
ide
red
f
r
om
IP
L
seas
on
20
12.
Var
i
ous
dim
ension
s
of
bowli
ng
and
batti
ng
we
re
us
e
d
–
th
ree
dim
ension
s
gr
oupe
d
i
nto
fac
tor
t
wo
(
bowli
ng),
fi
ve
dim
e
ns
io
ns
gro
upe
d
i
nto
factor
one
(
batti
ng
)
.
Var
ia
nce
ex
plained
by
factor
one
is
m
uch
higher
than
fact
or
t
wo
w
hich
cl
ea
rly
sho
w
s
that
batti
ng
ca
pab
il
it
y
do
m
inate
s
over
bowl
ing
.
T
he
a
utho
rs
in
[
10
]
m
easur
e
d
t
he
perfor
m
ance
e
val
uation
of
fast
bowlers
a
nd
s
pinners
ba
sed
on
var
i
ou
s
facto
rs
an
d
ra
nk
e
d
the
pe
rfo
rm
ance
with
t
he
help
of
A
H
P
an
d
TOP
S
IS.
Dif
fe
ren
t
crit
eria’s
and
par
am
et
ers
are
us
e
d
su
c
h
as
eco
no
m
y
rate,
bowlin
g
aver
a
ge
an
d
bowlin
g
strike
rate t
o ra
nk the
play
ers.
The
stu
dy
re
ve
al
s
that
Indian
bowlers
pe
rfo
rm
ed
well
and
the
to
p
7
bow
le
rs
are
Indian
s
in
al
l
th
e
three
seas
ons
(
2008,
20
09,
an
d
2010
).
T
he
m
achine
le
arn
i
ng
-
based
a
ppr
oach
us
e
d
in
[
11]
wh
ic
h
cl
us
te
red
the
play
ers
acco
rdi
ng
to
the
ro
le
s
and
in
orde
r
to
rank
th
e
pl
ay
er’
s
pe
rfo
rm
ance,
a
no
vel
ind
e
x,
nam
ely
Dee
p
Perfo
rm
ance
In
de
x
is
form
ulate
d.
Play
ers
from
IP
L
seaso
n
2008
ta
ke
n
up
for
the
f
or
m
ulati
on
of
perf
orm
ance
rankin
g.
201
pl
ay
ers
are
a
naly
zed
with
T2
0
an
d
IPL
as
th
ei
r
career
data.
Play
ers
got
cl
us
te
re
d
int
o
di
ff
e
rent
gro
up
s
de
pe
nding
upon
t
heir
batti
ng
a
nd
bo
wling
pe
rfo
rm
ances.
The
a
uth
ors
i
n
[12]
di
scusse
d
t
he
I
P
L
te
am
s
and
play
ers
t
o
do
t
he
e
valu
at
ion
wit
h
the
help
of
c
orre
la
ti
on
,
ass
ociat
ion
a
nd
cl
assifi
cat
ion
ru
le
s.
Naïv
e
Ba
ye
sia
n
cl
assifi
cat
ion
is
us
ed
to
predict
the
te
am
resu
lt
s
by
co
ns
i
der
i
ng
t
he
in
div
i
dual
pe
rfor
m
ances
of
play
ers.
Analy
sis of
team
p
erf
orm
ance at ho
m
e and
aw
ay
grou
nd
is also a
naly
zed.
By
suppo
rt an
d
c
onfiden
c
e
of
t
he
play
ers
,
sel
ect
or
s
get
th
e
idea
to
filt
er
ou
t
play
ers
f
or
the
ne
xt
se
aso
n.
T
he
work
i
n
[13]
disc
us
se
d
the
pr
e
dicti
on
t
oo
l
and
m
achine
le
arn
i
ng
al
gorithm
s
wh
ic
h
a
re
us
e
d
to
a
naly
ze
the
past
perform
ance
of
pl
ay
ers,
and
it
will
be
ben
e
fici
al
fo
r
te
a
m
autho
riti
es
to
sel
ect
th
e
righ
t
play
er.
HBase
an
op
en
source
,
dist
rib
uted
pre
dicti
on
to
ol
is
pr
esente
d
to
keep
the d
at
a
r
el
at
ed
to
m
at
ches
and
play
ers
of
I
PL
seaso
ns.
Past
per
f
orm
a
nces
of
play
ers
have
been
vis
ualiz
ed
by
HBase
too
l.
Stat
ist
ic
al
analy
sis
of
pla
ye
r’
s
pe
rfor
m
ed
base
d
on
dif
fer
e
nt
char
act
e
risti
cs.
Pr
edict
ion
pe
rfor
m
ed
o
n
perform
ances
of
the
te
am
dep
e
ndin
g
on
the
sta
ti
st
ic
s
of
the
ind
ivi
du
al
play
ers.
Th
e
auth
ors
in
[14]
analy
zed
the
data
of
OD
I
m
at
ches
of
I
ndia
n
cric
ke
t
te
a
m
’s
and
app
ly
associat
ion r
ul
es on
hom
e g
round o
r
a
way g
a
m
e att
ribu
te
s,
t
os
s,
b
at
ti
ng
ord
er a
nd the
final
m
a
tc
h
res
ults.
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Da
t
a
vi
s
ua
li
za
t
ion
and t
os
s
re
lated
analysis
of I
P
L tea
ms
and batsm
en
p
e
rforma
nces
(
Vi
dit Ka
nungo
)
4425
The
a
utho
rs
i
n
[
15
]
pro
-
pose
d
a
m
od
el
that
works
on
tw
o
m
et
ho
ds
w
hich
are
to
predict
the
sc
or
e
of
first
in
nings
on
the
basis
of
current
r
un
rat
e,
num
ber
of
wickets
fall
en,
ve
nu
e
of
m
at
ch
a
nd
batti
ng
te
am
.
Seco
nd
m
et
hod
pre
dicts
the
ou
tc
om
e
of
m
a
tc
h
co
ns
ide
rin
g
sam
e
at
tribu
te
s
from
the
fir
st
m
et
ho
d
al
on
g
with
the
ta
r
get
giv
e
n
to
batti
ng
te
a
m
.
A
dataset
of
O
DI
m
at
ches
f
ro
m
2002
t
o
2014
us
e
d
i
n
th
ese
tw
o
m
et
hods
.
Naïve
Ba
ye
s
and
Li
near
Re
gressi
on
Cl
assif
ie
r
hav
e
bee
n
us
e
d
to
im
ple
m
ent
the
se
two
m
et
ho
ds
.
T
he
auth
or
s
in
[16]
predict
ed
the
perform
ance
of
batsm
en
of
IPL
sea
-
so
n
4
based
on
the
perf
or
m
ances
of
play
er
in
first
three
seas
ons.
Mult
i
-
Lay
er
pe
rcep
tr
on
(ML
P)
ne
ural
netw
ork
is
use
d
t
o
pr
e
dict
the
pa
s
t
per
f
orm
ances.
This
pr
e
dicti
on
ca
n
help
the
m
anag
em
ent
and
s
el
ect
or
s
to
dec
ide
w
hich
bats
m
en
they
shoul
d
bid
for
a
nd
who
sh
oul
d
not be c
on
si
der
e
d
at
all
. Th
e aut
hors
i
n
[
17
]
predict
e
d
the r
es
ult of
a
m
at
ch
by co
m
par
ing
the str
eng
t
hs
of
two
te
am
s.
A
pe
rfor
m
ance
of
in
div
id
ual
play
ers
from
e
ac
h
te
am
is
m
e
asur
e
d
by
the
m
.
They
i
m
ple
m
ente
d
al
gorithm
s
to
pr
e
dict
the
perform
ances
of
batsm
en
and
bowle
rs
f
r
om
past
and
rece
nt
career
data.
T
he
w
ork
in
[
18
]
is
done
for
a
naly
zi
ng
the
pe
rfor
m
anc
es
of
bo
wlers.
A
m
easur
e
cal
le
d
Com
bin
ed
Bow
li
ng
Ra
te
wh
ic
h
i
s
a
co
m
bin
at
ion
of
th
ree
tra
diti
on
al
bowli
ng
par
am
et
ers:
bowling
a
ver
a
ge,
strike
rate
and
ec
onom
y
i
s
us
ed
for
the
expe
rim
ent.
The
aut
hors
in
[1
9]
f
or
m
ulate
d
a
sta
ti
sti
ca
l
m
od
el
to
est
i
m
at
e
the
value
of
play
er
by
consi
der
i
ng
diff
e
re
nt
sta
ti
sti
cs
of
batsm
en
,
bowle
rs
a
nd
al
l
-
rou
nd
e
rs.
They
t
ried
t
o
buil
d
a
syst
em
at
ic
log
ic
al
decisi
on
m
od
el
to
sel
ec
t
bette
r
play
ers
fo
r
a
uctio
n.
A
m
ulti
-
ob
j
ect
iv
e
op
ti
m
iz
a
ti
on
evo
luti
ona
ry
m
et
ho
d
[
20]
us
ed
i
n
this
pa
pe
r
to
optim
iz
e
batti
ng
an
d
bowlin
g
s
tren
gth
s
of
a
te
a
m
and
to
fin
d
t
he
te
am
m
e
m
ber
s.
Per
f
orm
ances
of
eac
h
play
er
are
al
so
e
valua
te
d
by
usi
ng
N
SGA
-
II
al
go
rithm
.
The
auth
ors
in
[21]
us
e
s
om
e
string
sim
il
arit
y
m
et
rics:
Le
-
ven
sh
te
i
nS
im
(L
EVS),
LeeSi
m
(LEES
),
Jacc
ard
Coe
ff
ic
ie
nt
(JA
CC
),
Dice
Coeff
ic
ie
nt
(
DI
CE
)
,
Jaro
-
W
i
nkl
er
Dista
nce
(J
W
D)
to
com
par
e
and
dif
fer
e
ntiat
e
the
per
f
orm
ances
of
un
know
n
perf
or
m
ers
to
that
of
e
xperts.
T
he
y
us
ed
t
he
co
ncep
ts
of
Lear
ning
A
naly
ti
cs,
Gam
e
An
al
yt
ic
s,
Product
ive
An
al
yt
ic
s
an
d
Data
Visu
al
iz
at
ion
t
o
a
naly
ze
the
S
erio
us
G
am
e
An
al
yt
ic
s
from
U
ser
Ge
ne
rated D
at
a.
T
he
w
ork
i
n
[22]
is d
on
e
o
n
arti
fici
al
and
real
-
w
orl
d
dat
aset
including
diff
e
ren
t
Vis
ualiz
at
ion
te
chn
i
qu
e
s:
un
ce
rtai
nty
visu
al
iz
at
ion,
ensem
ble
data
visu
al
iz
at
ion
and
m
ulti
di
m
ension
al
/m
ul
ti
var
ia
te
data
visu
al
iz
at
ion.
They
co
nclu
de
d
tha
t
diff
e
re
nces in
ensem
ble d
ist
ribu
ti
on a
re
m
os
t
cru
ci
al
a
nd im
po
rtant f
act
ors fo
r
the
pr
op
e
r
a
naly
sis of a
gam
e.
3.
ABOU
T
TOO
LS A
ND ME
THOD
OLOG
Y
IP
L,
one
of
th
e
biggest
le
ag
ue
s
in
T2
0
cric
ke
t
with
m
il
li
on
s
of
fans
al
l
over
the
w
orl
d.
Aroun
d
69
6
m
at
ches
hav
e
ta
ken
place
fro
m
20
08
-
2018.
Ther
e
is
a
huge
data
wh
ic
h
i
nclu
de
ball
by
ball
insigh
ts
of
each
m
at
ch
of
each
inn
in
gs
with
m
at
ch
locat
ion
and
al
l
oth
er
necessa
ry
detai
ls.
Sp
yder
,
th
e
fr
ee
integrat
ed
an
d
Scie
ntific
Pyt
hon
De
velo
pm
e
nt
En
vir
on
m
en
t
has
bee
n
us
e
d
to
do
t
he
dat
a
exp
l
or
at
io
n
a
nd
plo
tt
in
g
f
un
ct
ion
s
for visuali
zat
io
n.
Sp
yde
r offers
var
i
ou
s
popula
r
sci
entifi
c
pac
kag
e
s for
dee
p i
ns
pecti
on a
nd
exp
l
or
at
io
n of
data. P
rope
r
An
al
ysi
s
an
d
Visu
al
iz
at
ion
per
f
or
m
ed
in
Sp
y
der
wit
h
num
ero
us
pack
a
ges
s
uch
as
Nu
m
Py,
Pandas,
Ma
tplotl
ib
and
Seaborn.
Thes
e
pack
a
ges
hel
p
to
do
the
ba
s
ic
and
m
od
er
n
visu
al
iz
at
ion.
I
n
m
y
wo
rk,
Seabor
n
is use
d for T
oss R
el
at
ed
A
na
l
ysi
s A
ppr
oac
h and M
at
plo
tl
ib
is u
se
d f
or
pla
ye
r
vis
ualiz
at
ion
.
4.
DA
T
A COLL
ECTION
Data
ha
s
be
en
colle
ct
ed
f
ro
m
w
ww.iplt
20.c
om
,w
ww.crics
heet.or
g.
Data
consi
sts
of
the
ball
by
ball
detai
ls
for
a
to
ta
l
of
696
m
at
ches
f
ro
m
2008
-
2018.
Ba
ll
by
ball
data
pro
vid
es
i
n
de
pth
detai
l
of
al
l
th
e
balls
thr
own
in
that
par
ti
cula
r
ov
e
r
.
The
ball
cou
l
d
be
ei
ther
wi
de,
dea
d,
no
ba
ll
or
a
play
er
go
t
sin
gles,
do
ub
le
s
,
triple
s,
six
or
four
on
t
hat
ba
ll
.
Ther
e
a
re
two
cs
v
file
s
of
dataset
s.
Ma
t
ches
.cs
v
giv
es
the
detai
ls
of
m
at
ch
venue,
l
ocati
on,
Sea
son,
c
on
te
sti
ng
te
a
m,
a
bout
toss
winn
er
an
d
tos
s
dec
isi
on
,
m
at
ch
re
su
lt
,
wi
n
got
by
runs
or
wic
kets,
pla
ye
r
of
t
he
m
atch
,
detai
ls
of
a
ll
the
thre
e
ump
ires
a
nd
m
at
c
h
W
i
nner
et
c.
Deli
ver
ie
s
.c
sv
is
the
ball
by
ball
da
ta
and
the
co
m
bin
at
ion
of
a
ll
the
deliveries
for
al
l
the
m
at
ches
from
2008
-
18.
It
co
ns
ist
s
of
diff
e
re
nt
at
tribu
te
s
Ma
tc
h_
id
,
bowling
te
am
,
batti
ng
te
a
m
,
batsm
en,
bo
wl
er,
N
onstrike
r,
no
ball
r
uns,
pe
nalty
runs,
E
xtra
runs,
ov
e
r,
total
runs
et
c.
I
nning
s
te
ll
if
the
first
te
a
m
was
goin
g
on
fiel
d
or
sec
ond
one
.
Ov
e
r desc
ribes
the c
urren
t
ov
er
nu
m
ber
. Bal
l descri
bes
t
he c
urren
t
ball
num
ber
o
f
t
he
c
urren
t
over.
Table
1
decr
i
bes
the
total
of
te
n
at
t
ri
bu
t
es
wh
ic
h
we
r
e
us
ed
for
th
e
visu
al
iz
at
ion
of
batsm
en
perform
ances
and
to
ss
relat
ed
analy
sis.
To
ss
decisi
on,
T
os
s
W
i
nn
e
r
an
d
W
in
ne
r
are
the
key
at
tribu
t
es
use
d
for
to
ss rel
at
ed
an
al
ysi
s for 6
96 m
at
ches f
r
om
2
00
8
-
2018.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
201
9
:
4
4
2
3
-
4
4
3
2
4426
Table
1.
Desc
ription o
f
t
he
at
trib
utes used
ATT
RIB
UT
ES
DESCRIP
TI
O
N
Dis
m
iss
al
Kin
d
Bo
wled
Cau
g
h
t
Cau
g
h
t and
Bo
wle
d
Hit wicket
LBW
Ob
stru
ctin
g
the f
ield
Retired
Hurt
Ru
n
Out
Bo
wler
The Pla
y
er
who
deliv
ers the ball to
ba
ts
m
en
keep
in
g
dif
fer
en
t con
d
itio
n
s an
d
scen
arios
(in o
rder of
f
ield
)
su
ch
as W
id
e ball
Ru
n
s, No b
all Ru
n
s, E
x
tra
Ru
n
s.
Total Ru
n
s (For an
ind
iv
id
u
al Play
er)
The f
in
al Score of
th
e Play
e
r
c
alcu
lat
ed
at
th
e end
of
ea
ch
ov
er.
Extra Ru
n
s
The Ru
n
s sco
red b
y
bats
m
en
oth
er
th
an
hittin
g
the b
all.
Thes
e r
u
n
s are
no
t
credited
to
an
y
of
the b
ats
m
en
.
Bats
m
en
Ru
n
s
Ru
n
s Sco
red b
y
B
ats
m
en
on
eac
h
bal
l on
particu
lar
d
eli
v
ery
.
No
Ball Ru
n
s
Bo
wler
m
u
st
th
row th
e ball in
resp
ect of
the ar
m
;
Oth
e
rwise
U
m
p
ire
will
d
eclar
e it
as
No
ball an
d
r
u
n
s w
ill be p
rocess
ed
to
Battin
g
tea
m
.
Match De
cisio
n
Match
W
in
n
in
g
Co
n
d
itio
n
W
in
by
Ru
n
s
W
in
by
W
i
ck
ets
Tos
s Decisio
n
Decisio
n
m
ad
e b
y
Tea
m
Cap
tain
af
te
r
win
n
in
g
the Tos
s, either
to
bat f
irst
o
r
f
ield
.
Tos
s
W
in
n
er
The Tea
m
wh
o
won
the to
ss
.
W
in
n
er
The Tea
m
wh
o
won
the
m
atch
.
4
.
1.
Pre
-
proc
essing
ph
as
e
In
this phase
fi
lt
rati
on
an
d
cl
e
anin
g
of
m
at
ch
es
and
d
el
ive
ries
dataset
s
took
place.
T
his phase
m
ai
nly
deals
with
sta
nd
a
r
dizat
ion
,
t
ran
s
f
or
m
at
ion
and
c
orrecti
on
of
data.
The
re
was
no
m
ajo
r
pre
-
processi
ng
done
for
the
d
at
a
col
le
ct
ed
as m
os
t of m
uch
w
as
nor
m
al
iz
ed.
4
.
2.
D
ata
visu
aliz
at
i
on
The
m
os
t
i
m
po
rta
nt
an
d
si
gnific
ant
pa
rt
of
data
vis
ualiz
at
ion
an
d
pr
e
dicti
ve
a
naly
sis
is
to
re
present
the
data
i
n
for
m
of
char
ts
an
d
gr
a
phs
to
get
a
vis
ual
prese
ntati
on
of
data.
Th
e
c
ollec
te
d
data
is
visu
al
i
zed
t
o
get
a
bette
r
an
d
cl
ear
unde
rst
and
i
ng
ab
out
a
ll
the
param
et
e
rs
of
t
he
Seas
on,
t
he
te
am
,
All
-
rou
nd
e
rs,
bat
sm
en
and
bowle
rs
s
o
that
it
will
be
helpful
f
or
the
te
am
sel
ec
tors,
Ca
ptains
and
m
anag
ers
for
the
ne
xt
a
uction.
Diff
e
re
nt p
ack
ages ar
e use
d
t
o
get the prope
r
analy
sis and
visu
al
iz
at
ion
f
or
play
ers
an
d
te
a
m
s.
Nu
m
Py
is u
sed
as
nu
m
erical
c
om
pu
ti
ng
f
or
t
he
giv
e
n
datas
et
s.
Pandas
us
e
d
as
the
data
proces
sin
g
an
d
I
/O
for
both
csv
file
s.
Ma
tplotl
ib
us
e
d
as
the
basic
visu
al
iz
at
ion
f
or
play
ers.
Sea
bor
n
pac
ka
ge
us
e
d
as
the
m
od
er
n
visu
al
iz
at
ion
for
Toss
relat
ed
a
naly
sis
as
well
as
fo
r
te
am
a
nd
play
er
insig
hts.
Dif
fer
e
nt
new
fe
at
ur
e
s
are
introd
uced
s
uch
a
s
the
num
ber
of
total
m
at
ches
play
ed
by
the
te
a
m
fo
r
al
l
the
el
even
seas
ons,
Ma
xim
u
m
Ma
n
of
the
M
at
ches,
Ma
xim
u
m
Ce
ntu
ries
Score
d
by
Ba
ts
m
en,
M
axim
u
m
Pla
yer
of
t
he
Ma
tc
h
Aw
a
rd
s
,
Ma
xi
m
u
m
Cou
nt
of
Toss
W
i
ns
by
Diff
e
ren
t
Team
s,
D
eci
sion
ta
ken
by
each
te
am
after
winni
ng
t
oss
et
c.
Ta
ble
2
and
Ta
ble
3
li
s
ts
to
p
play
ers
ha
ving
Ma
xi
m
u
m
ce
nturie
s
sco
red
and
m
axi
m
um
Ma
n
of
the
m
at
ch
titl
es
c
onquere
d.
C
H
Gayl
e,
AB de Vil
li
ers a
nd S
K
Ra
ina
are
on the t
op for b
oth t
he
ti
tl
es.
Table
2
.
Ma
xi
m
u
m
m
an
of th
e m
a
tc
hes
Play
e
r
Co
u
n
t
CH
Ga
y
le
20
AB d
e Villie
rs
18
YK Pathan
RG Sh
ar
m
a
DA
W
arne
r
MS
Dh
o
n
i
SK Rain
a
SR W
atso
n
G Ga
m
b
h
i
r
ME
K
Hus
sey
16
16
15
14
14
13
13
12
Table
3
.
Ma
xi
m
u
m
centu
ries sco
red by
bats
m
en
Play
e
r
Co
u
n
t
CH Ga
y
le
SR W
atso
n
AB d
e Villie
rs
SK Rain
a
RR
Pant
V Koh
li
V
Seh
wag
DA
W
arne
r
BB
M
cCu
llu
m
M
Vija
y
8
4
3
3
2
5
3
3
2
2
Analysis
:
By
com
par
ing
Fi
gure
1,
Ta
ble
2
a
nd
Ta
ble
3
Ch
r
is
Gayl
e,
on
e
of
the
best
T2
0
batsm
en
fr
om
West
Indies
has
w
on
Play
er
of
the
Ma
tc
h
Aw
a
rd
for
20
tim
es
w
hich
is
the
best
reco
r
d
in
IP
L
so
far.
Var
i
ous
oth
er
All
-
Ro
unde
r
Play
ers
-
Yu
s
uf
Pathan,
S
hane
W
at
s
on,
Jacq
ues
kalli
s,
Kie
rron
P
ollard
,
Aj
i
nk
ya
Ra
ha
ne
hav
e
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Da
t
a
vi
s
ua
li
za
t
ion
and t
os
s
re
lated
analysis
of I
P
L tea
ms
and batsm
en
p
e
rforma
nces
(
Vi
dit Ka
nungo
)
4427
al
so
w
on
Play
er
of
th
e
m
a
tch
showi
ng
sov
ereig
nty
of
bat
sm
an
on
the
gam
e.
On
ly
Ami
t
Mi
sh
ra
an
d
Sunil
Nar
i
ne
(
S
pinn
er)
is
full
tim
e b
ow
le
r
in
t
op 20 list
.
Figure
1.
Ma
xi
m
u
m
p
la
ye
r
of
the
m
at
ch
awards
Analysis
:
By
Com
par
ing
Ta
ble
2,
Table
3,
Figu
re
1,
Fig
ur
e
2
a
nd
Fi
gure
3,
Cl
ear
picture
of
al
l
su
ch
sta
r
play
er
(Bat
sm
en)
go
t
vis
ualiz
ed
w
ho
will
be
the
first
pr
e
f
eren
ce
f
or
Tea
m
sel
ect
or
s
and
m
anag
e
m
ent
to
bid
on
them
and
ta
ke
them
in
th
ei
r
co
ur
t.
Ta
ble
4
sho
ws
the
li
st
of
play
ers
(Bat
sm
en)
ha
ving
best
stri
ke
rat
e
durin
g
t
he
s
pa
n from
2
008
-
2018 a
nd Ta
ble
5
s
hows
the
To
p 10 play
ers
w
i
th Maxim
u
m
r
un
s
.
F
igure
2. To
p
b
at
sm
en
(2008
-
18)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
201
9
:
4
4
2
3
-
4
4
3
2
4428
Figure
3. Ba
tsm
en
with
t
op s
trike r
at
e
(
2008
-
18)
Table
4
.
Be
st s
trike r
at
e
Play
e
r
(Bats
m
en
)
Sp
an
An
d
re
Ru
ss
ell (
D
D,
K
KR)
Ris
h
ab
h
Pant (DD)
Glen
n
M
ax
well
(D
D,
K
XIP,
M
I)
JC
Bu
ttler
(
MI
,
R
R)
AB d
e Villie
rs (
D
D,
RC
B)
CH Ga
y
le
(KX
IP,
KKR, RCB
)
2012
–
2
0
1
8
2016
-
2
0
1
8
2012
-
2
0
1
8
2016
-
2
0
1
8
2008
-
2
0
1
8
2009
-
2
0
1
8
Table
5
.
T
op
10
play
ers wit
h m
axi
m
u
m
r
un
s
Play
e
r
(Bats
m
en
)
Total Ru
n
s
SK Rain
a
V Koh
li
RG Sh
ar
m
a
G Ga
m
b
h
i
r
RV Uth
ap
p
a
S Dhawan
MS
Dh
o
n
i
CH Ga
y
le
DA
W
arne
r
AB d
e Villie
rs
5014
4962
4504
4223
4144
4090
4041
4037
4014
3974
Table
6
giv
e
c
le
ar
idea
of
t
oss
relat
ed
anal
ysi
s
for
al
l
te
am
s,
Mum
bai
I
nd
ia
ns
an
d
Kolkat
a
Kn
i
gh
t
Ri
der
s a
re
on the top li
st
hav
i
ng m
axi
m
u
m
c
ount
of toss
w
i
ns
.
Table
6
.
C
ount
of
t
os
s
w
i
ns
Tea
m
s
Co
u
n
t
Ch
en
n
ai Sup
er
Kin
g
s
Gu
jarat
Lion
s
Ko
lk
ata Knig
h
t Rid
ers
Rajas
th
an
Ro
y
als
Ro
y
al Ch
allen
g
ers
Ban
g
alo
re
Deccan Ch
argers
Kin
g
s XI
Pu
n
jab
Mu
m
b
ai
Ind
ian
s
Ris
in
g
Pun
e Sup
ergian
t
Su
n
risers H
y
d
erab
ad
Delh
i Dar
ed
ev
ils
Ko
ch
i T
u
sk
ers Ke
r
ala
Pu
n
e W
ar
riors
Ris
in
g
Pun
e Sup
ergian
ts
77
15
87
69
77
43
75
90
06
42
80
08
20
07
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Da
t
a
vi
s
ua
li
za
t
ion
and t
os
s
re
lated
analysis
of I
P
L tea
ms
and batsm
en
p
e
rforma
nces
(
Vi
dit Ka
nungo
)
4429
Analysis
:
By
Com
par
ing
Ta
ble
6
an
d
Fig
ure
4,
M
um
bai
In
i
dan
s
captai
n
an
d
Ko
l
kata
Kn
i
gh
t
Ri
der
s
Ca
ptain
hav
e
a
good
hold
of
run
wit
h
the
c
oin
.
In
T2
0
gam
es,
toss
play
s
a
cruc
ia
l
ro
le
so
m
et
i
m
es
dew
fact
or
on
the
gro
und,
or
the
m
os
it
ur
e
co
ntent
in
first
10
hour
s
ca
n
ch
ang
e
t
he
gam
e
.
Ca
ptain
an
d
oth
e
r
te
a
m
m
e
m
ber
analy
ze
the
sc
or
e
be
fore
sta
rting
of
the
m
at
c
h.
Dif
fer
e
nt
ty
pes
of
pitche
s
play
dif
fer
e
nt
r
oles
f
or
ba
tsm
en
a
nd
bowlers
.
By
w
inn
in
g
t
he
toss
Ca
ptain
ca
n
a
naly
ze
that
on
this
pa
rtic
ular
pitch,
bat
fi
rst
or
fiel
d
first,
wh
ic
h
on
e
can
g
i
ve
t
hem
an
ad
va
nta
ge.
Figure
4
.
Maxi
m
u
m
co
un
t
of
t
os
s
wins
by
dif
fer
e
nt team
s
Table
7 dep
ic
ts
the
decisi
on ta
ken b
y ea
c
h
te
a
m
after
wi
nn
i
ng the t
os
s.
Ch
enn
ai
super
k
i
ngs
decide
d
m
axi
m
u
m
tim
e
s to bat fi
rst r
at
her tha
n
fiel
ding
because
of
ke
y Pl
ay
ers
li
ke M
S
D
honi a
nd
Su
r
esh
Rai
na who
analy
zes the
pi
tc
hes very
well
.
Table
7
.
Decisi
on
ta
ke
n
by ea
ch
te
am
after win
ning the
to
ss
Tea
m
s
Bat
Field
Ch
en
n
ai Sup
er
Kin
g
s
Gu
jarat
Lion
s
Ko
lk
ata Knig
h
t Rid
ers
Rajas
th
an
Ro
y
als
Ro
y
al Ch
allen
g
ers
Ban
g
alo
re
Deccan Ch
argers
Kin
g
s XI
Pu
n
jab
Mu
m
b
ai
Ind
ian
s
Ris
in
g
Pun
e Sup
ergian
t
Su
n
risers H
y
d
erab
ad
Delh
i Dar
ed
ev
ils
Ko
ch
i T
u
sk
ers Ke
r
ala
Pu
n
e W
ar
riors
Ris
in
g
Pun
e Sup
ergian
ts
45
01
30
30
20
24
26
41
00
20
29
03
11
03
32
14
57
39
57
19
49
49
06
22
51
05
09
04
Analysis
:
Fig
ur
e
5
a
nd
Ta
ble
7
il
lustrat
es
the
tru
e
m
ental
ly
of
the
IP
L
as
well
as
the
T
20
gam
e.
Af
te
r
winnin
g
the
toss
,
te
am
s
are
pr
efe
rri
ng
to
fiel
d
fi
r
st
so
that
they
can
plan
thei
r
inn
i
ng
s
well
wh
il
e
chasin
g.
T
her
e
are
dif
fer
e
nt
ver
si
ons
of
pit
ches
a
re
avail
able
Pit
ches
that
favor
s
pin
bo
wli
ng
w
hic
h
are
m
os
tly
fo
un
d
in
the
I
nd
ia
n
S
ub
c
onti
nen
t
,
F
la
t
pitches
w
hi
ch
are
batsm
an
f
rien
dly,
Pit
ches
that
fa
vor
swing
bowlin
g,
Pit
ch
es
that
favor
f
ast
bowling.
S
o
basical
ly
,
fie
lding
first
over
batti
ng
ca
n
be
com
e
the
adv
a
ntage.
In
the
la
st
thre
e
seasons
(
2016,
20
17,
20
18)
te
a
m
strat
egie
s
are
quit
e
si
m
il
ar.
They
anal
yz
e
the
pitch
and
th
e
venue
ve
ry w
el
l.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
201
9
:
4
4
2
3
-
4
4
3
2
4430
Figure
5
.
To
ss
decisi
on seas
on
wise
Analysis:
T
his
gr
a
ph
c
on
ti
nu
es
the
sa
m
e
tre
nd
as
of
the
pr
evio
us
vis
ualiz
at
ion
.
F
ro
m
Fi
gure
4,
Table
7
an
d
Figure
6,
o
ne
point
w
hich
s
hould
be
note
d
th
at
on
ly
Chen
na
i
Su
pe
r
Kings
i
s
the
te
a
m
wh
o
pr
efe
r
s
to
bat
after
winnin
g
the t
oss. Out o
f 147
m
at
ches,
77 ti
m
es Chennai won th
e
toss
a
nd
45 ti
m
es th
ey
d
eci
de
to
ba
t first a
nd
32
ti
m
es
to
bowl.
This
can
be
becau
se
of
the
captai
ncy
of
MS
Dhon
i
w
ho
rely
on
his
bowlin
g
an
d
fiel
di
ng
un
it
.
Win
ning
count
of
90
m
at
ches
an
d
l
oss
co
un
t
of
57
m
at
ches
are
th
e
sta
ts
for
the
Chen
nai
te
am
.
MS
Dho
ni
and
S
K
Ra
ina
wh
o
w
on
14
ti
m
es
Ma
xim
u
m
m
a
n
of
the
m
at
c
hes
an
d
they
bo
t
h
are
in
th
e
li
st
of
Ma
xim
u
m
centur
ie
s
as
well
.
So
,
batti
ng
firs
t
fo
r
C
hennai
is
al
ways
the
righ
t
decisi
on
.
T
hr
ee
ti
m
es
Ch
enn
a
i
su
pe
r
Ki
ng
s
is
the
winn
e
r
of
IP
L
(20
10,
2011
an
d
20
18).
All
oth
e
r
te
a
m
s
especial
ly
Roya
l
Chall
e
ng
e
rs
Ba
ng
al
or
e
w
ho
play
ed
a
total
of
16
6
m
a
tc
hes
and
they
al
so
won
the
toss
77
tim
es
ou
t
of
wh
ic
h
57
tim
e
s
they
decide
to
fiel
d
first
an
d
20
ti
m
es
to
bat.
CH
Gayl
e
and
V
Kohli
of
Ba
ng
al
or
e
sc
or
e
d
th
e
m
axi
m
u
m
ce
nturie
s
by
a
c
ount
of
8
a
nd
5
an
d
Ba
ng
al
or
e
ha
ve
a
winnin
g
c
ou
nt
of
79
a
nd
loss
c
ount
a
s
87.
So,
thei
r
de
ci
sion
com
par
ed
t
o
C
hennai
is
no
t
a
t
a
perfect
le
ve
l.
As
t
hey
al
so
hav
e
sta
r
pla
ye
rs
li
ke
V
K
oh
li
a
nd
CH
Gayl
e
,
they
can
m
ov
e
their
decisi
on s
ta
ts t
o
bat
first
rather t
ha
n
a
fi
el
d
in t
he u
pcom
ing
m
at
ches o
r
Seas
on
s
.
Figure
6
.
To
ss
decisi
on team
wise
5.
CONCL
US
I
O
N
In
t
his
pa
per,
the
pe
rfor
m
ance
of
c
ricket
play
ers(batsm
e
n)
a
nd
toss
relat
ed
analy
sis
in
IPL
f
ro
m
seaso
n
2008
-
2018
has
bee
n
vi
su
al
iz
ed.
Fin
di
ng
out
the
hidden
par
am
et
ers,
patte
rn
s
a
nd
a
tt
ribu
te
s
that
le
ad
to
the
outc
om
e
of
a
cricket
m
at
c
h
he
lps
t
he
te
am
ow
n
ers
a
nd
sel
ect
or
s
to
rec
ognize
bette
r
pl
ay
ers.
A
sal
ar
y
of
Evaluation Warning : The document was created with Spire.PDF for Python.
Int
J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Da
t
a
vi
s
ua
li
za
t
ion
and t
os
s
re
lated
analysis
of I
P
L tea
ms
and batsm
en
p
e
rforma
nces
(
Vi
dit Ka
nungo
)
4431
IP
L
c
ricket
pl
ay
ers
is
decid
ed
th
r
ough
the
aucti
on
proce
ss.
T
hus,
it
is
a
par
t
of
f
ra
nch
ise
an
d
m
at
te
r
of
decisi
on
m
aking
ab
out
w
hic
h
play
er
to
be
bid
e
d
for
a
nd
at
wh
at
c
os
t
by
the
pa
st
pe
rfor
m
ance
of
pl
ay
ers
in
IP
L.
Eve
ry
Sel
ect
or
nee
ds
yo
ung
a
nd
dyna
m
ic
play
ers
w
ho
can
ha
ndle
the
press
ur
e
ca
l
m
l
y,
and
go
t
ow
a
r
ds
the w
i
nn
i
ng li
ne
.
This
pap
e
r
highli
gh
ts
t
he
play
er
pe
rfor
m
ance
especial
ly
ba
ts
m
en
an
d
ad
dr
ess
es
the
a
na
ly
sis
that
is
done
f
or
Ma
xi
m
u
m
Ma
n
of
t
he
Ma
tc
he
s,
Ma
xim
u
m
Ce
ntu
ries
Sc
ored
by
Ba
tsm
en,
Top
Ba
tsm
en,
Ba
ts
m
en
with
To
p
Strik
e
Ra
te
,
Top
10
Play
ers
with
Ma
xim
u
m
Run
s.
Stat
ist
ic
s
of
696
m
at
ches
hav
e
bee
n
us
e
d
in
this
exp
e
rim
ent
and
e
ven
f
or
toss
relat
ed
a
naly
sis
su
c
h
as
Co
un
t
of
To
ss
wins,
De
ci
sio
n
ta
ke
n
by
each
te
am
after
winnin
g
the
toss,
T
os
s
Dec
isi
on
Seas
on
W
i
se,
T
os
s
D
eci
sion
Team
W
i
se
.
Ba
sed
on
the
ab
ove
analy
sis,
the
Indian
bats
m
en
are
ver
y
good
an
d
are
on
top
c
ho
ic
e
by
the
sel
ect
or
s.
SK
Ra
ina
consi
der
e
d
as
the
fines
t
batsm
en
wh
o
i
s
second
in
th
e
top
li
st
of
ba
ts
m
en
hav
in
g
m
axi
m
u
m
runs,
m
axi
m
u
m
m
an
of
the
m
at
ches,
m
axi
m
u
m
centur
ie
s
sco
re
d,
V
Kohli
at
t
he
first
posi
ti
on
of
m
axi
m
u
m
runs
a
nd
e
ve
n
he
is
in
the
li
st
for
m
axi
m
u
m
centur
ie
s.
All
ot
her
Indian
Star
ba
ts
m
en
MS
Dhon
i
(Best
Ca
pt
ai
n,
Ma
xim
u
m
runs
a
nd
Ma
xim
u
m
m
an
of
the
m
atch
es)
,
Ri
sh
a
bh
Pant
(sec
ond
best
stri
ke
rate
and
m
axi
m
u
m
centu
ries),
RG
Sh
a
rm
a,
S
D
ha
wan,
G
Gam
bh
ir,
Y
K
Patha
n
a
nd
M
Vij
ay
perf
or
m
ed
ver
y
w
el
l
at
the
end
of
la
st
five
overs
.
Sele
ct
ors
hav
e
the
cl
ear
ch
oic
e
to
giv
e
pref
eren
ce
t
o
I
ndia
n
Play
ers
at
first
as
they
pe
r
f
or
m
ed
ver
y
well
in
seaso
n
f
ro
m
2008
-
2018. We
al
so
pr
ese
nte
d
toss
relat
ed
a
naly
sis,
i
n
w
hich
MS
D
honi
is
the
best
capta
in
for
CSK w
ho
w
on
the
toss
m
axi
m
um
tim
es
hav
i
ng
co
unt
of
77
an
d
el
ect
ed
to
bat
first.
Their
ch
oice
of
bat
f
irst
m
os
tl
y
resu
lt
s
in
win.
M
os
t
of
the
tim
es
file
d
first
is
el
ect
ed
by
the
captai
ns
so
that
they
ca
n
pla
n
an
d
pe
r
f
or
m
well
by
chasing.
RC
B,
KK
R,
MI
and
K
XI
P
el
ect
ed
fiel
d
fi
rst
m
os
t
of
the
tim
es
hav
ing
cou
nt
of
57
an
d
49.
Sele
ct
ors
hav
e
the
cl
ear
cho
ic
e
to
sel
ect
bat
sm
en
fr
om
M
um
bai
In
dians
and
Ki
ngs
XI
P
unj
a
b
as
this
two
te
am
s
han
dl
e
d
the
press
ur
e
ve
ry
well
durin
g
a
ll
the
seaso
ns
from
20
08
-
20
18.
By
co
ns
ide
rin
g
al
l
this
vi
su
al
iz
at
ion
a
nd
toss
relat
ed
a
naly
sis,
Team
Ma
nag
em
ent
can
se
le
ct
the
rig
ht
play
ers
a
nd
ri
gh
t
s
te
am
s
at
the
tim
e
of
a
uc
ti
on
.
A
go
od
a
nd
str
ong
cric
ket
te
am
can
be
f
or
m
ed
withi
n
a
gi
ve
n
bu
dg
et
,
w
hich
will
ha
ve
th
e
highest
cha
nc
e
of
winnin
g.
REFERE
NCE
S
[1]
A
Micha
e
l
J.
Fr
y
,
et
al
.
,
“
In
troduc
ti
on
to
th
e
Spec
ia
l
Iss
ue
on
Anal
y
t
ic
s
i
n
Sports,
Part
I:
Gene
r
al
Sports
Applic
a
ti
ons,”
I
nstit
ute
for
Operations
Re
sear
ch
and
the
Ma
nageme
nt
Sc
ie
n
ce
s
(
INFORMS)
,
vol.
42
,
n
o.
2,
pp.
105
-
108
,
Ma
rch
-
April
2012
.
[2]
Prince
Kansal
,
et
al
.
,
“
Pla
y
er
Valua
t
ion
in
In
dia
n
Prem
ie
r
L
ea
gue
Auc
ti
on
using
Data
Mi
ning
Techni
que
,
”
Inte
rnational
Co
nfe
renc
e
on
Con
te
mpor
ary
Computing
and
Informatic
s (
IC3I)
,
2012.
[3]
Loh,
C
.
S.
and
Sheng,
“
Mea
su
rin
g
expe
rt
-
p
erf
or
m
anc
e
for
Serio
us
Gam
es
A
naly
ti
cs:
From
data
t
o
insight
s,”
Seri
ous
Gam
es
Anal
ytic
s:
Me
thodol
og
ies
for
Pe
rform
an
ce
M
easurement
,
Assess
ment,
an
d
Impr
ove
ment
,
[Chapt
er
5]
,
Ne
w
York,
NY
:
Sprin
ger
,
2015
.
[4]
David
H.S.
Chu
ng,
et
a
l
.
,
“
Know
le
dge
-
As
sisted
Ranki
ng:
A
Vi
sual
Anal
y
t
ic
A
ppli
c
at
ion
for
S
ports
Eve
nt
Da
t
a,
”
IEE
E
Computer
Gr
aphic
s and
A
ppli
cations
,
vo
l.
36,
no
.
3
,
2016
.
[5]
Sricha
ran
Shah,
et
al.,
“
A
St
ud
y
on
Perfor
m
anc
e
of
Cricket
Pl
a
y
ers
usi
ng
Fact
or
Ana
l
y
sis
Appro
ac
h,
”
Inte
rnational
Jo
urnal
of Adv
an
c
ed
R
e
-
search in
Computer
Scien
ce
,
vol
.
8
,
n
o
.
3
,
2017.
[6]
Van
Stade
n
,
P.
J.,
“
Com
par
is
on
of
cr
ic
k
et
er
s’
bowling
and
batting
p
erf
or
m
anc
es
using
g
rap
hical
d
isplays
,
”
Curr
ent
Scienc
e
,
vol. 96, n
o.
6,
2
009.
[7]
Kala
pdrum
Pass
i
and
Nir
avkuma
r
Pande
y
,
“
Predic
ti
ng
p
lay
e
rs
per
form
anc
e
in
one
da
y
in
te
rn
at
ion
a
l
cricke
t
m
at
ch
e
s
using Ma
chi
n
e l
ea
rning
,
”
Computer
Sc
ie
nc
e
&
I
nformation
Tech
nology
(
CS
&
I
T
)
,
2017.
[8]
Chira
g
Go
y
al
,
“
IPL
-
2017
Cross
Countr
y
Clust
er
Anal
y
sis
,
”
Int
er
nati
onal
Journal
of
Computer
Sci
ence
Tr
ends
an
d
Technol
ogy
(
IJCST)
, v
ol.
5
,
no.
4
,
2017
.
[9]
Sujee
t
Kum
ar
S
har
m
a,
“
A
fac
tor
ana
l
y
s
is
appr
o
ac
h
in
per
form
a
nce
anal
y
s
is
of
T
-
20
cri
ck
-
et,”
J
ournal
of
Re
li
ab
il
ity
and
Stat
ist
ic
a
l
St
udie
s
,
n
o
l. 6,
no.
1,
pp.
69
-
76
,
20
13
.
[10]
Pabit
ra
Kum
ar
De
y
,
Dipendr
a
Nath
Ghos
h
and
Abho
y
Chand
M
ondal
,
“
A
MCD
M
Approac
h
for
Eva
lu
at
ing
Bowl
ers
Perform
anc
e in
I
PL,
”
Journal
o
f Emerging
Tr
end
s in
Computing
and
Information Sci
en
ce
s
,
v
ol. 2
,
n
o.
11
,
Nov 201
1.
[11]
C
Dee
p
Praka
sh,
et
al.,
“
A
new
Ca
te
gor
y
base
d
Dee
p
Perform
anc
e
Inde
x
using
Mac
hine
L
ea
rn
i
ng
for
ran
king
IPL
Cric
ke
te
rs,
”
Int
e
rnational
Journal
of
Elec
tronic
s,
El
e
ct
rica
l
and
C
omputati
onal
Sy
stem
, v
ol
.
5
,
no
.
2,
Feb
2016.
[12]
Ja
y
an
t
Praka
sh,
et
al.,
“
Eva
lu
ation
of
IPL
Te
a
m
s
and
Pla
y
ers
using
A
ss
oci
at
i
on,
Corre
l
at
ion
and
Cla
ss
ifi
c
ati
on
Rule
s,”
IE
EE
Int
ernati
onal
Confer
enc
e
on
Compu
te
r,
Comm
unic
a
t
ion
and
Con
trol
,
IC4
-
2015.
[13]
Shubhra
Singh
a
nd
Parm
ee
t
Kau
r,
“
IPL
Visuali
z
at
ion
and
Prediction
Us
ing
HBase
,
”
In
-
formatio
n
Technol
ogy
a
nd
Quantit
ative
Ma
na
geme
nt,
ITQ
M
2017.
[14]
K.
Raj
and
P.
Padm
a.
,
“
Applic
a
ti
on
of
associa
t
ion
r
ule
m
ini
ng:
A
ca
se
stud
y
on
te
am
In
dia
,
”
Inte
rnational
Co
nfe
renc
e
on
Co
mputer
Comm
unic
ati
on
and
In
formatic
s,
ICCCI
,
2
013.
[15]
Ti
je
nd
er
Singh,
et
al.,
“
Score
and
W
inni
ng
Predic
ti
on
in
Cric
k
e
t
t
hrough
Data
Mining,
”
Int
ernati
o
nal
Confe
renc
e
on
Soft
Computing
Techni
ques
and
Imple
mentations
-
(
ICSCT
I
)
,
Oct 2015.
[16]
Hem
ant
a
Saik
ia
and
Dib
y
oj
y
o
ti
Bhat
tacha
r
jee,
“
An
appl
i
ca
t
ion
of
m
ult
ila
y
er
p
e
rce
ptron
neur
al
net
work
to
pre
d
ic
t
the
per
form
ance
of
bat
sm
e
n
in
India
n
pre
m
ie
r
le
agu
e,”
Int
ernati
onal
Jour
nal
of
Re
searc
h
in
Sci
ence
and
Technol
ogy
,
vo
l. 1, no. 1,
2014.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
201
9
:
4
4
2
3
-
4
4
3
2
4432
[17]
M.
G.
Jhanwar
and
V.
Pudi,
"P
red
ic
ti
ng
the
Outcome
of
O
DI
Cric
ke
t
Mat
che
s:
A
T
ea
m
Com
positi
on
Ba
sed
Approac
h,
"
Eur
opean
Confe
ren
ce
on
Ma
chi
n
e
Lea
rning
and
P
rinci
ple
s
and
Pr
act
i
ce
of
Knowl
edge
Disco
ve
ry
in
Databases
(
ECMLPKDD
)
,
2016.
[18]
H.
H.
Le
m
m
er,
"
The
combined
b
owling
rat
e
as
a
m
ea
sure
of
bowling
per
form
anc
e
in
cri
cke
t
,
"
Sout
h
Af
rican
Journal
for R
ese
arch
in
Sport,
Ph
ysic
al
Educ
ati
on
and
R
ec
reat
ion
,
vol. 2
4,
n
o
.
2
,
pp
37
-
4
4,
Jan
2002.
[19]
R.
R.
Mara
the,
B.
Bansal,
B.
B.
Bansal,
and
T.
Ina
n
i,
“
Bu
y
i
ng
ipl
play
e
rs
i
n
auc
ti
on
:
Cri
c
ket
ing
gambl
e
o
r
s
y
stematic
logi
c
al
d
ec
ision
?”
Int
ernati
onal
Journ
al
of
Busine
ss
In
-
sights
&
Tr
an
sf
orm
ati
on
,
vol
.
6
,
no.
1,
2012
.
[20]
F.
Ahm
ed,
A.
Jindal
,
and
K.
Deb,
“
Cric
ket
t
ea
m
sele
c
ti
on
u
sing
evol
uti
on
ar
y
m
ult
i
-
ob
je
c
ti
v
e
opti
m
izati
on
,
”
Inte
rnational
Co
nfe
renc
e
on
Swa
rm
,
Ev
olu
ti
onar
y,
and
Me
m
et
i
c Com
puti
ng
,
Spri
nger
,
2011
,
pp.
7
1
–
78.
[21]
Loh,
C
.
S.
and
Sheng,
“
Mea
suring
the
(dis
-
)si
m
il
a
rity
b
et
we
e
n
expe
rt
and
n
ovic
e
b
eha
v
iors
as
serious
ga
m
es
ana
l
y
t
ic
s
,
”
Y
.
Ed
uc
Inf
Techno
l,
Educ
ati
on
and
I
nformation
Tech
nologi
es
,
Spring
er
US
,
2015
.
[22]
Haidong
Chen,
Song
Zha
ng,
and
W
ei
Chen,
“Unc
ert
a
inty
-
Aw
are
Multi
dimens
iona
l
Ensemble
Data
Visualiza
t
i
on
an
d
Exp
lora
t
ion,”
IE
EE Tr
ansacti
ons on
V
isuali
z
ati
on
and
Computer
Gr
aphic
s
,
v
ol.
21
,
no
.
9
,
Sep
2015.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
Vidit
Kanungo
,
MCA
,
Student
,
Christ
(Dee
m
ed
to
be
Univer
si
t
y
)
,
vid
it
work5@g
m
ai
l.
com
Tul
asi
B
,
MCA
,
MP
hil
,
As
soci
at
e
Profess
or
o
f
Com
pute
r
Scie
nce,
Christ
(D
ee
m
ed
to
be
Univer
sit
y
),
tula
si.b@c
hristuni
v
e
rsit
y
.
in
Evaluation Warning : The document was created with Spire.PDF for Python.