Resource Selection

Group 10

Why is Resource Selection

important?
A first round selection is far more valuable than any other
selection
a player taken in that range has a slightly better than
60% chance of turning out to be a decent or better
player; nearly three times the rate of the second round.
Second and third round picks are a cut above the rest of the
draft, although the chances of landing a decent or better
player are quite low
a little better than 1 in 5 for the second round, and
slightly better than 1 in 7 for the third round.
Theres little difference between picks made from the fourth
to ninth rounds of the draft
These rounds players are roughly half as likely to develop
into a decent or better player as a third round pick.

Why is Resource Selection

important?

Why is Resource Selection

important?
BUST(Fewer than 100 games)
SHORT LIVED(Out of league and
less than 300 Games played)
DECENTPLAYER
DOMINANT PLAYER(considered
an elite player by fans/media
outside their city)

Why is Resource Selection

important?

FIFTH ROUND (325 picks)

FIRST ROUND (286 picks)
1became a dominant player. (0.03
25became dominant players. (8.8%)
22turned into decent players.(6.8
148turned into decent players. (51.7%)
302were short lived or busts.(92.9
113were short lived or busts (39.5%)

SECOND ROUND (326 picks) SIXTH ROUND (295 picks)

4became dominant players. (1.2%)
1became a dominant player. (0.03
70turned into decent players. (21.5%)
26turned into decent players.(8.8%
252were short lived or busts. (77.3%)
268were short lived or busts.(90.9

THIRD ROUND (307 picks)

SEVENTH ROUND (308 picks)
2became dominant players. (0.6%)
2became a dominant player. (0.06
47turned into decent NHLers.(15.2)
28turned into decent players.(9.1%
259wereshort livedor busts. (84.2)
268were short lived or busts.(90.3

EIGHTH ROUND (267 picks)

FOURTH ROUND (315 picks) 1became a dominant player. (0.04
27 became decent players. (8.6%)
19turned into decent players.(7.1%
288 were short lived (91.4%)
247were short lived or busts.(92.5

Resource selection
How do you select the best resources in your
team?

onduct a process and from the good, take the necessary resource

Process
Good
Necessary resources

Good Resources

Baseball
Wins Above Replacement
If this player got injured and their team had to
replace them with a freely available minor leaguer or
a AAAA player from their bench, how much value
would the team be losing
WAR = (Batting Runs + Base Running Runs + Fielding
Runs + Positional Adjustment + League Adjustment
+Replacement Runs) / (Runs Per Win)

WAR: Batting Runs

Player
wOBA, PA, Home Park Factor
League
Average wOBA (lgwOBA), the wOBA
Scale, MLB R/PA (lgR/PA), and the
specific league (AL or NL) wRC and PA
for non-pitchers

WAR: Batting Runs

Weighted On Base Average

Not all hits are created equal

Weighted On-Base Average combines all the different
aspects of hitting into one metric, weighting each of
them in proportion to their actual run value

 Base Running Runs

Ultimate Base Running + Weighted
Stolen Base Runs + w. Grounded Double
play Runs
Fielding Runs
Zone rating, Stolen Base Runs above
average

 Positional Adjustment

League Adjustment

 Runs per Win

the value that allows us to convert runs
to wins
The process reflects the average
number of runs a team needs to score in
order to add one win to their total
RPW = 9*(MLB Runs Scored / MLB
Innings Pitched)*1.5 + 3

WAR: Notes
WAR is park and league adjusted
WAR is context neutral, meaning players
do not get additional credit for hitting with
men on base or in high leverage situations
WAR is an estimate of value and should be
treated as the center of a range of
possibilities rather than a precise measure.

Optimization

 https://www.youtube.com/watch?v=-4QPVo
0UIzc

Player position Assignment model

Max
Subject to
, for all =1,, (1)
, for all =1,,
()/

,for all =1,,

0,for all =1,,,=1,,
V Contribution of player j to position I
X the fraction of innings player j is assigned to position I
C No. of innings player j can play per season
D No. of innings for which a player should be assigned to position I

Value of flexibility in absence of

injuries
Platoon advantage Left handed player facing right handed
pitcher

Value of flexibility in presence of

simulated injuries

Position assignment comparison

Green against left handed pitchers
Red against right handed pitchers

Protection Level
5% PL - the smallest value of that results
in a 5% loss in RAR

The search for tomorrows

champions

 https://www.youtube.com/watch?v=5gaFa
YAlSaw

Late bloomers struggle to catch up

Teenagers mature at different rates, and children of the same age can vary
greatly in biological maturity
With strength and speed on their side, early maturers have a major
advantage as clubs decide which young players will make the cut
Late developers aren't just at a physical disadvantage, but a psychological
one too
The focus on physical ability can also be a disadvantage for early-maturing
footballers as it creates unrealistic expectations for a young player's potential
By age 18, physical differences between players have usually evened out.
Late developers can eventually overtake their peers in physical tests and
technical skills
But by this point, many talented players have already missed their shot at
the stadium. And England may have missed out on its own Messi

Understanding Player Potential

For all players to have the best chance to
reach their potential, understanding
growth is key.
Clubs must consider maturity when
making decisions about training,
competition and team selection.
Clubs have to make decisions based on
attributes which young footballers havent
yet developed.
Being able to predict each young player's

Bio-banding creates a more

balanced competition
Player Management Administration (PMA) database - 5,000 young
footballers across the academy system. The system assesses
players aged nine to 17 on growth, fitness and performance
throughout their early careers
PMA also allows academies to group players for training and
competition by biological age, not chronological age.
By matching training and competition to the athletes level,
practitioners can optimise development and reduce injuries
Once a club better understands a player's maturation, they can
tailor their training
Late-maturers get the chance to use their technical skills, play
creatively and take on leadership roles.
Early-maturers find the competition more physically challenging,
so they have to work harder and rely more on their technical
abilities

Coattail Effect
In 2011, University of Connecticut won
its first Big East Conference regular
season championship and reached
super regional round of NCAA
tournament.
Few months later time 10 members
were selected in the Major League
Baseball amatrur draft- earlier average
was 1.7

Teams are said to rely on certain heuristics,

including familiarity to draft players for later
rounds.
- teams prefer to spend most of their time and
resources evaluating very top amateurs to
optimize their earliest draft picks. Most lateround picks will not reach the major leagues.

So what happened in
Connecticut?
Presence of George Springer and Matt Barnes
They were selected with 2 of the top 20 overall
picks of the draft
Extra exposure from playing with them helped
more teammates be drafted by pro teams
The baseball team was seen by more scouts
than usual in 2011
A scout is more likely to recommend a player he
has seen more frequently than a player of
comparable talent with whom he is less familiar

How to measure?
Top prospect effect We consider that teammates of a top prospect are more likely to be
drafted in later round than players who dont have such teammates
However, top prospects are more likely to play for good baseball
teams only
Hence, we consider the players drafted in the top 5 rounds

Number of non-top prospects(after

5th round) chosen when a star
Regression
teammate
was equation
drafted of

Number of non-top prospects

drafted in other years

D=
D is the number of players drafted from i school in year t in rounds 6+
Coattail =1 if payer was drafted in first 5 rounds else 0

What else?
Possibility of extension of coattail effect
into other years
https://www.youtube.com/watch?v=ekMx4I
SIJ20

CRICKET

Cricketis abat and ball gameplayed between two teams of eleven players
on acricket field, at the centre of which is a rectangular 22-yardlongpitchwith awicket(a set of three woodenstumps) sited at each end.
One team, designated the batting team, attempts to score as many runs as
possible, whilst their opponents field.
Each phase of play is called an innings.
After either ten batsmen have been dismissed or a set number of overs have
been completed, the innings ends and the two teams then swap roles.
The most common forms of dismissal are bowled, when the bowler hits the
stumps directly with the ball, leg before wicket, when the batsman prevents
the ball from hitting the stumps with his body instead of his bat, and caught,
when the batsman hits the ball into the air and it is intercepted by a fielder
before touching the ground.
Runs are scored by two main methods: either by hitting the ball hard enough
for it to cross the boundary, or by the two batsmen swapping ends by each
simultaneously running the length of the pitch in opposite directions whilst the
fielders are retrieving the ball. If a fielder retrieves the ball quickly enough to
put down the wicket with a batsman not having reached the crease at that

Player roles and performances

indices
Batsman
Total runs
scored
Average
Strike Rate
Highest Score
Numbers of
Centuries / half
centuries

Bowler

Others

Catches
Wickets
Run Outs
Strike Rate
Stumping
Economy
5/10
wicket haul
Average

TYPICAL PLAYER STATISTICS

A Generic Approach for Team Selection

in Multi player Games using Genetic
Algorithm
The traditional team selection process is really lengthy, awkward and
unclear due to manual process and personal judgments; which may
lead to a disaster.
The approach is the combination of the existing quantitative approaches
with some new extensions such as attributes regarding personal
performances, team performance and the combination of players.
Researchers here propose an adaptation of island genetic algorithm to
optimize the selection of multiplayer sports team having multiple
conflicting constraints with mixed crossover where the fitness of
common solution is used to drive the selection.
The manual team selection process may have many drawbacks such as
personal liking and disliking, personal grudges between team selection
committee and players and external pressures (social, political etc.)
Team selection is also an optimization problem as the selection

GA FOR PLAYER
SELECTION
Lets consider a pool of 100 players and the task of the selection

committee to select 15 players for the competition. There may be

many combinations of best team players, provided the realities and
constraints. Selecting the best team (optimal) combination is the main
objective and we also call this the fittest combination
The procedure of finding the fittest combination is a time consuming
task and the resultant or the survivor among the fit candidates.
GA works by initializing the population (Chromosomes) as elaborated
below (see Figure 1)

GA FOR PLAYER
SELECTION
Selection is made randomly (Rolette wheel selection algorithm may be used to
avoid the influence of initial population), using the fitness function.
Now the process of Crossover is to be made and at last Mutation (Optional). This
will be repeated in a cycle till the user intervene or fittest combination found
Resolving Tie between two
players
If a tie occurs at some point during
experiment, i.e., two or more players
with same fitness value gather in the
mating pool, it will be resolved using
fitness values and experience values of
players in tie

 generic approach for

multi player team
selection has been
developed using
genetic algorithm
This automated
process will select
team on the basis of
the previous
outcomes i.e. either
win or lose.

Other Research papers

http://
www.sloansportsconference.com/wp-content/uploads/2015/06/A-Self-Adapting-Intelligent
-Optimized-Analytical-Model.pdf
http://www.iitk.ac.in/kangal/papers/k2011016.pdf
http://www.cioandleader.com/articles/40155/how-sap-helped-kkr-win-pepsi-ipl-2014
http://www.fuzzylogix.com/blog/the-thrill-of-analytics-in-ipl-part-i-what-is-a-par-score-inan-ipl-match
/
http://
www.simafore.com/blog/bid/56405/Can-IPL-cricket-statistics-support-a-predictive-analyti
cs-exercise
http://
stats.espncricinfo.com/ci/engine/stats/index.html?class=1;home_or_away=1;home_or_a
way=2;result=1;result=2;result=3;result=4;team=6;template=results;type=bowling
http://research.ijcaonline.org/volume40/number17/pxc3877440.pdf
http://
www.cricketwa.com/forum/5318/how-pepsi-ipl-atom-mvp-player's-award-is-calculated.a
spx
http://www.espncricinfo.com/magazine/content/story/352877.html

Thank you