info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Role of luck in sport
16 views
Skip to first unread message
Peter Webb
Jun 13, 2012, 4:53:02 AM6/13/12
to
I am considering starting a small research project on the the subject of
"luck in sport".
It seems clear to me that in some sports, the better team almost always wins
whereas in others a significant amount of luck is involved - particularly
where balls can bounce randomly, and scores are low and subject to
probability rather than statistics.
This can potentially be measured with an arguable amount of accuracy on a
sport-by-sport basis, and sports then compared to see which involve the
most/least luck in who wins the game.
The basic strategy is to use the half time result to break each game into
two half games. Each half is played by the same people, at almost the same
time, and both sides have the same basic strategy in both halves to maximise
their points and minimise the opponent's points. Games with very little luck
should have the side which wins the game winning both halves; conversely
games with lots of luck will see results where a side wins only a single
half but still wins the match. The winners of half games and winners of full
games can be correlated and a numeric index of luck produced on a sport by
sport basis.
Most people believe in the role of luck in sports, but as far as I can see
nobody has tried to quantify this, and there does seem to be this mechanism
to do so. It will involve looking at and data entering (in some manner) lots
of historical sporting results. It will also require some knowledge of the
sports to collect the information in a manner which eliminates systemic
differences. For example in cricket, games where one side "declares"
(voluntarily stops batting to save time) would need to be eliminated.
I don't know where this would go, maybe a web page, could be of interest to
people who like to argue about sports, or people who bet on them.
Does anybody have:
(a) Any thoughts
(b) Any ideas on where to get the statistics (remember we need half time
scores) hopefully as data files
(c) Any wish to become involved ?
Peter Webb
"luck in sport".
It seems clear to me that in some sports, the better team almost always wins
whereas in others a significant amount of luck is involved - particularly
where balls can bounce randomly, and scores are low and subject to
probability rather than statistics.
This can potentially be measured with an arguable amount of accuracy on a
sport-by-sport basis, and sports then compared to see which involve the
most/least luck in who wins the game.
The basic strategy is to use the half time result to break each game into
two half games. Each half is played by the same people, at almost the same
time, and both sides have the same basic strategy in both halves to maximise
their points and minimise the opponent's points. Games with very little luck
should have the side which wins the game winning both halves; conversely
games with lots of luck will see results where a side wins only a single
half but still wins the match. The winners of half games and winners of full
games can be correlated and a numeric index of luck produced on a sport by
sport basis.
Most people believe in the role of luck in sports, but as far as I can see
nobody has tried to quantify this, and there does seem to be this mechanism
to do so. It will involve looking at and data entering (in some manner) lots
of historical sporting results. It will also require some knowledge of the
sports to collect the information in a manner which eliminates systemic
differences. For example in cricket, games where one side "declares"
(voluntarily stops batting to save time) would need to be eliminated.
I don't know where this would go, maybe a web page, could be of interest to
people who like to argue about sports, or people who bet on them.
Does anybody have:
(a) Any thoughts
(b) Any ideas on where to get the statistics (remember we need half time
scores) hopefully as data files
(c) Any wish to become involved ?
Peter Webb
quasi
Jun 13, 2012, 8:15:11 AM6/13/12
to
Peter Webb wrote:
>It seems clear to me that in some sports, the better team
>almost always wins whereas in others a significant amount of
>luck is involved - particularly where balls can bounce
>randomly, and scores are low and subject to probability rather
>than statistics.
>
>This can potentially be measured with an arguable amount of
>accuracy on a sport-by-sport basis, and sports then compared
>to see which involve the most/least luck in who wins the game.
>
>The basic strategy is to use the half time result to break
>each game into two half games. Each half is played by the same
>people, at almost the same time, and both sides have the same
>basic strategy in both halves to maximise their points and
>minimise the opponent's points. Games with very little luck
>should have the side which wins the game winning both halves;
>conversely games with lots of luck will see results where a
>side wins only a single half but still wins the match. The
>winners of half games and winners of full games can be
>correlated and a numeric index of luck produced on a sport by
>sport basis.
I think you need 3 components, not 2 ...
>It seems clear to me that in some sports, the better team
>almost always wins whereas in others a significant amount of
>luck is involved - particularly where balls can bounce
>randomly, and scores are low and subject to probability rather
>than statistics.
>
>This can potentially be measured with an arguable amount of
>accuracy on a sport-by-sport basis, and sports then compared
>to see which involve the most/least luck in who wins the game.
>
>The basic strategy is to use the half time result to break
>each game into two half games. Each half is played by the same
>people, at almost the same time, and both sides have the same
>basic strategy in both halves to maximise their points and
>minimise the opponent's points. Games with very little luck
>should have the side which wins the game winning both halves;
>conversely games with lots of luck will see results where a
>side wins only a single half but still wins the match. The
>winners of half games and winners of full games can be
>correlated and a numeric index of luck produced on a sport by
>sport basis.
(1) skill
(2) luck
(3) the human element
For teams with approximately equal skill, the human element, not
luck, will often be the key secondary factor.
I'll outline two categories of situations for which the human
element may be the key factor ...
(1) Strategic Adjustment
Team A's chosen strategy may take team B by surprise, but
by the second half, the surprise has worn off and team B has
made appropriate adjustments.
Or, for multi-player team games, perhaps some player on team B
is not playing well in the first half, so that player is
replaced in the second half, and then all is well for team B.
(2) The Emotion Factor
Suppose Team A wins the first half by a large margin. In the
second half, it's hard for team A to psyche themselves up for
maximum output. Thus, team A will tend to "coast", riding their
huge lead. At the same time, provided team A's lead is not too
much to overcome, team B may be psyched to play at absolute
maximum skill and power, thus potentially recovering some or
all of the initial score deficit.
On the other hand, suppose team A gets a huge, essentially
unrecoverable lead in the first half. Then, in the second
half, team B may have lost the heart to fight, so gets crushed
even more.
Alternatively, for the huge, unrecoverable lead scenario, the
coach may make some substitutions to try some untested players.
It's a good time to experiment since the game is lost anyway.
Thus, with team B using their "B team", the second half may be
even worse.
Bottom line -- I think the human element dominates the luck
element, and would be hard, perhaps impossible, to filter out.
quasi
Graham Cooper
Jun 13, 2012, 7:27:59 AM6/13/12
to
SKILL: win 1st half - win 2nd half
STRAT: lost 1st half - win 2nd half
LUCK: win 1st half - win 2nd half by greater margin!
Taking into account both Teams lose a little STAMINA by the 2nd half.
Certainly if they widened the soccer goals by 1 meter it would score
more like basketball, perhaps reduce riots a little!
Herc
--
http://tinyURL.com/BLUEPRINTS-TURING
http://tinyURL.com/BLUEPRINTS-CANTOR
http://tinyURL.com/BLUEPRINTS-GODEL
http://tinyURL.com/BLUEPRINTS-PROOF
http://tinyURL.com/BLUEPRINTS-LOGIC
http://tinyURL.com/BLUEPRINTS-MATHS
http://tinyURL.com/BLUEPRINTS-HALT
http://tinyURL.com/BLUEPRINTS-P-NP
http://tinyURL.com/BLUEPRINTS-GUT
Peter Webb
Jun 13, 2012, 8:12:35 AM6/13/12
to
"quasi" <qu...@null.set> wrote in message
news:99tgt75j1ncpb4ck9...@4ax.com...
it makes a good substitution, or unlucky if it gets an injury. We are trying
to measure luck, the random component; you have mostly just listed where it
shows up.
Frederick Williams
Jun 13, 2012, 10:45:53 AM6/13/12
to
Peter Webb wrote:
>
> [...]
seems possible to me that:
i) a team doing badly at half time may be motivated to do better in the
second half (and may therefore actually do better); but also
ii) a team doing badly at half time may become disillusioned and
therefore do even worse in the second half.
Sports scientists surely include psychologists among their number who
have looked in to this.
--
The animated figures stand
Adorning every public street
And seem to breathe in stone, or
Move their marble feet.
>
> [...]
>
> The basic strategy is to use the half time result to break each game into
> two half games. Each half is played by the same people, at almost the same
> time, and both sides have the same basic strategy in both halves to maximise
> their points and minimise the opponent's points. Games with very little luck
> should have the side which wins the game winning both halves; conversely
> games with lots of luck will see results where a side wins only a single
> half but still wins the match.
I should begin by remarking that I know nothing about sport, but it
> The basic strategy is to use the half time result to break each game into
> two half games. Each half is played by the same people, at almost the same
> time, and both sides have the same basic strategy in both halves to maximise
> their points and minimise the opponent's points. Games with very little luck
> should have the side which wins the game winning both halves; conversely
> games with lots of luck will see results where a side wins only a single
> half but still wins the match.
seems possible to me that:
i) a team doing badly at half time may be motivated to do better in the
second half (and may therefore actually do better); but also
ii) a team doing badly at half time may become disillusioned and
therefore do even worse in the second half.
Sports scientists surely include psychologists among their number who
have looked in to this.
--
The animated figures stand
Adorning every public street
And seem to breathe in stone, or
Move their marble feet.
quasi
Jun 13, 2012, 2:36:58 AM6/13/12
to
Peter Webb wrote:
Strategic adjustment is not luck.
>A team is lucky if it makes a good substitution,
It's only common sense to replace a pitcher who, on a given
day, can't throw a strike, or to replace a basketball player who,
for some reason (pain, hangover, tiredness, whatever) can't make
a shot.
I wouldn't call the score improvement resulting from such a
substitution luck.
>We are trying to measure luck, the random component; you
>have mostly just listed where it shows up.
The human element allows a team to _adapt_ to the situation.
It's a _function_ of the situation and, unlike luck, is both
controllable and predictable.
quasi
>Most of those variables you list are in fact just luck.
No, not really.
Strategic adjustment is not luck.
>A team is lucky if it makes a good substitution,
day, can't throw a strike, or to replace a basketball player who,
for some reason (pain, hangover, tiredness, whatever) can't make
a shot.
I wouldn't call the score improvement resulting from such a
substitution luck.
>We are trying to measure luck, the random component; you
>have mostly just listed where it shows up.
It's a _function_ of the situation and, unlike luck, is both
controllable and predictable.
quasi
David Bernier
Jun 13, 2012, 5:27:23 PM6/13/12
to
On 06/13/2012 10:45 AM, Frederick Williams wrote:
> Peter Webb wrote:
>>
>> [...]
>>
>> The basic strategy is to use the half time result to break each game into
>> two half games. Each half is played by the same people, at almost the same
>> time, and both sides have the same basic strategy in both halves to maximise
>> their points and minimise the opponent's points. Games with very little luck
>> should have the side which wins the game winning both halves; conversely
>> games with lots of luck will see results where a side wins only a single
>> half but still wins the match.
>
> I should begin by remarking that I know nothing about sport, but it
> seems possible to me that:
> i) a team doing badly at half time may be motivated to do better in the
> second half (and may therefore actually do better); but also
> ii) a team doing badly at half time may become disillusioned and
> therefore do even worse in the second half.
> Sports scientists surely include psychologists among their number who
> have looked in to this.
>
There's a simple model for some sports where for each half-game,
> Peter Webb wrote:
>>
>> [...]
>>
>> The basic strategy is to use the half time result to break each game into
>> two half games. Each half is played by the same people, at almost the same
>> time, and both sides have the same basic strategy in both halves to maximise
>> their points and minimise the opponent's points. Games with very little luck
>> should have the side which wins the game winning both halves; conversely
>> games with lots of luck will see results where a side wins only a single
>> half but still wins the match.
>
> I should begin by remarking that I know nothing about sport, but it
> seems possible to me that:
> i) a team doing badly at half time may be motivated to do better in the
> second half (and may therefore actually do better); but also
> ii) a team doing badly at half time may become disillusioned and
> therefore do even worse in the second half.
> Sports scientists surely include psychologists among their number who
> have looked in to this.
>
team A is assumed to score a number of points n_A which
follows a Poisson distribution of parameter mu_A (the mean
number of points scored), and team B analogously scores
a number of points n_B ~ Poisson(mu_B) .
Then n_A - n_B , the "point spread", follows a so-called
Skellam distribution.
Assuming mu_A > 0, mu_B > 0 , then for any integer m,
Prob[n_A - n_B = m] > 0 (any point spread is possible
in this very simple model).
I must add that it's assumed that the random variables
n_A and n_B, both Poisson, are independent r.v.s .
The Wikipedia page on the Skellam distribution is:
< http://en.wikipedia.org/wiki/Skellam_distribution > .
In the References section of that Wikipedia page,
they refer to the article below on sports and
Poisson distributions:
Karlis, D. and Ntzoufras, I. (2003) "Analysis of sports data using
bivariate Poisson models". Journal of the Royal Statistical Society:
Series D (The Statistician), 52 (3), 381�393. doi:10.1111/1467-9884.00366
In hockey, some players get injured. But is it just luck?
What if a player gets into fights a lot or skates too fast?
Dave
Graham Cooper
Jun 13, 2012, 5:44:12 PM6/13/12
to
>
> In hockey, some players get injured. But is it just luck?
> What if a player gets into fights a lot or skates too fast?
>
> Dave
Luck might be better described as inconsistent over several matches
> In hockey, some players get injured. But is it just luck?
> What if a player gets into fights a lot or skates too fast?
>
> Dave
between the same opponents.
e.g. in BOXING
TYSON beats ROCKY
ROCKY beats ALI
would have a high incidence of consistent results in rematches.
TYSON beats ROCKY
ROCKY beats ALI
TYSON beats ROCKY
ROCKY beats ALI
Maybe there are less variables in a 1 on 1 sport for luck to generate
from.
Whereas in Aussie Rules Football the results flip back and forth more
like tossing a coin.
CARLTON beats SWANS
SWANS beats CARLTON
SWANS beats CARLTON
CARLTON beats SWANS <<< Luck
Incidentally I went to the fitness shop last week to get some boxing
mitts, but came out with a pair of MMA Gloves.. Put your face in
front of one of those "mitts" and you won't be having much luck for a
long time!
Herc
David Bernier
Jun 13, 2012, 6:43:46 PM6/13/12
to
How about:
"Modelling Association Football Scores and Inefficiencies in the
Football Betting Market", by Mark J. Dixon and Stuart G. Coles.
The abstract can be read here:
http://onlinelibrary.wiley.com/doi/10.1111/1467-9876.00065/abstract
Dave
Michael Press
Jun 14, 2012, 1:38:10 PM6/14/12
to
In article <jra04d$jv5$1...@news.albasani.net>,
Not having a good quarterback is bad luck.
--Don Shula
--
Michael Press
"Peter Webb" <r.peter...@gmail.com> wrote:
> Most of those variables you list are in fact just luck. A team is lucky if
> it makes a good substitution, or unlucky if it gets an injury. We are trying
> to measure luck, the random component; you have mostly just listed where it
> shows up.
Sure, luck means a lot in football.
> Most of those variables you list are in fact just luck. A team is lucky if
> it makes a good substitution, or unlucky if it gets an injury. We are trying
> to measure luck, the random component; you have mostly just listed where it
> shows up.
Not having a good quarterback is bad luck.
--Don Shula
--
Michael Press
dilettante
Jun 15, 2012, 2:47:02 PM6/15/12
to
"Peter Webb" <r.peter...@gmail.com> wrote in message
news:jr9ke5$s70$1...@news.albasani.net...
>I am considering starting a small research project on the the subject of
>"luck in sport".
>
> It seems clear to me that in some sports, the better team almost always
> wins whereas in others a significant amount of luck is involved -
> particularly where balls can bounce randomly, and scores are low and
> subject to probability rather than statistics.
>
> This can potentially be measured with an arguable amount of accuracy on a
> sport-by-sport basis, and sports then compared to see which involve the
> most/least luck in who wins the game.
>
> The basic strategy is to use the half time result to break each game into
> two half games. Each half is played by the same people, at almost the same
> time, and both sides have the same basic strategy in both halves to
> maximise their points and minimise the opponent's points. Games with very
> little luck should have the side which wins the game winning both halves;
> conversely games with lots of luck will see results where a side wins only
> a single half but still wins the match. The winners of half games and
> winners of full games can be correlated and a numeric index of luck
> produced on a sport by sport basis.
I think there is a huge problem with this analysis. Teams are not trying to
>"luck in sport".
>
> It seems clear to me that in some sports, the better team almost always
> wins whereas in others a significant amount of luck is involved -
> particularly where balls can bounce randomly, and scores are low and
> subject to probability rather than statistics.
>
> This can potentially be measured with an arguable amount of accuracy on a
> sport-by-sport basis, and sports then compared to see which involve the
> most/least luck in who wins the game.
>
> The basic strategy is to use the half time result to break each game into
> two half games. Each half is played by the same people, at almost the same
> time, and both sides have the same basic strategy in both halves to
> maximise their points and minimise the opponent's points. Games with very
> little luck should have the side which wins the game winning both halves;
> conversely games with lots of luck will see results where a side wins only
> a single half but still wins the match. The winners of half games and
> winners of full games can be correlated and a numeric index of luck
> produced on a sport by sport basis.
win half-games, but whole games. If a team has a 30 point lead at the end of
the first half, they will not play a strategy that necessarily maximizes
their chance of winning the second half. They will (try to) play a strategy
that minimizes their chance of losing the second half by more than 29
points. Similarly, their opponent will be playing not to try to win the
second half, but to try to win the second half by at least 30 points.
Differences in correlation of halves and games won between two sports may
reflect the effects of this kind of strategizing being more pronounced in
one than in the other.
Graham Cooper
Jun 16, 2012, 2:26:00 AM6/16/12
to
On Jun 16, 4:47 am, "dilettante" <n...@nonono.no> wrote:
> "Peter Webb" <r.peter.webb...@gmail.com> wrote in message
if there's a high margin then a strategy towards a low scoring game
for each side might be used by the leading team.
But you might still be able to determine factors like:
does the bounce of an oval shaped ball compared to a round ball
randomize the chances of winning?
does a zero strategy game, eg. 10 pin bowling where a player's game
environment is determined only by his own play increase consistent
winning streaks?
Herc
--
http://tinyURL.com/BLUEPRINTS-TURING
http://tinyURL.com/BLUEPRINTS-CANTOR
http://tinyURL.com/BLUEPRINTS-GODEL
http://tinyURL.com/BLUEPRINTS-PROOF
http://tinyURL.com/BLUEPRINTS-LOGIC
http://tinyURL.com/BLUEPRINTS-MATHS
http://tinyURL.com/BLUEPRINTS-SETS
> "Peter Webb" <r.peter.webb...@gmail.com> wrote in message
for each side might be used by the leading team.
But you might still be able to determine factors like:
does the bounce of an oval shaped ball compared to a round ball
randomize the chances of winning?
does a zero strategy game, eg. 10 pin bowling where a player's game
environment is determined only by his own play increase consistent
winning streaks?
Herc
--
http://tinyURL.com/BLUEPRINTS-TURING
http://tinyURL.com/BLUEPRINTS-CANTOR
http://tinyURL.com/BLUEPRINTS-GODEL
http://tinyURL.com/BLUEPRINTS-PROOF
http://tinyURL.com/BLUEPRINTS-LOGIC
http://tinyURL.com/BLUEPRINTS-MATHS
Graham Cooper
Jun 16, 2012, 2:29:06 AM6/16/12
to
On Jun 16, 4:26 pm, Graham Cooper <grahamcoop...@gmail.com> wrote:
>
> does a zero strategy game, eg. 10 pin bowling where a player's game
> environment is determined only by his own play increase consistent
> winning streaks?
>
technically you might opt out of going for a difficult split with the
>
> does a zero strategy game, eg. 10 pin bowling where a player's game
> environment is determined only by his own play increase consistent
> winning streaks?
>
pins if you only needed 1 pin to win a match!
still some strategy in 10 pin bowling!
Herc
MajorOz
Jun 26, 2012, 1:24:14 PM6/26/12
to
Even if all he variables are accounted for (unlikely), unless they are
quantifiable, the exercise becomes one of mysticism.
0 new messages