Analysis on "Number of Serves per
Serve Game"
rev E, by Hans, 2005-06-16
_________________
Links related to this issue:
- Top Player Opinion Poll
- Definition of Single Toss
EXECUTIVE SUMMARY:
Below is an effort at listing all aspects that have been brought forward relevant to the proposal to change the number of serves in a racketlon serve game from the present 5 serves. Specifically, since the so called "Single Toss Rule" is very related to the subject it is included in this analysis.
The most relevant arguments have been marked with bold font and a (subjective!) effort has been made at estimating the weight of each such argument (assuming that the main battle is between 2 and 5 serves). For ex. (2+++) indicates a very strong argument for 2 serves whereas (5+) indicates a weak argument for 5 serves.
The conclusion of this analysis is that we should stay with 5 serves and introduce the Single Toss Rule. The main argument for this is probably that of "Rhythm". Perhaps Sarah McFadyen, the 2004 Racketlon World Champion expresses it best:
"I prefer the 5 serves, as I think the 2 serves idea is too bitty - in a sport like racketlon where you have to adapt to so many different things, I think it's important to still be able to get some form of rhythm going."
Other major arguments for 5 serves include "Games within the match" and "Tradition".
The main argument for introducing 2 serves is that it would solve the "Unfair lucky tosser advantage" problem. But this argument becomes irrelevant if the Single Toss Rule is introduced since it takes care of the problem - in a more elegant and more efficient way!
The only other major argument for 2 serves is that of "Side confusion". But this is, at least, partly off-set by the "Tactical choice" argument and therefore it, arguably, looks weak compared to the sum of the three arguments for 5 serves listed above.
See below for the details and the explanations of all these buzzwords.
But these conclusions are admittely based on subjective evaluations of the arguments listed. Every reader of this analysis will probably evaluate the arguments differently and might therefore come to different conclusions. It's a matter of taste. For example, to some readers the "Side confusion" argument might appear strong enough to outbalance all arguments on the other side in which case the conclusion will be 2 serves and not 5. Other readers might not like the single toss rule in which case the "unfair lucky tosser advantage" might sail up as decisive, which in turn will make him favour 2 serves also. Others, still, might chose to emphasize any of the other aspects listed below. The main objective with this analysis is not to prescribe a hard conclusion but to list all relevant aspects that have been brought forward during IRF discussions. It is open for everyone to agree or disagree with the conclusion. But before you do, it is a good idea to be aware of the aspects.
ASPECTS:
1) Fairness/Asymmetry. It cannot be denied that there are
elements of asymmetry in the present 5 serves to 21 format. One
of them (the "one-sided" servers advantage) will go
away with an even number of serves. The other one (the lucky
tosser advantage) will be alleviated with shorter serve games;
1.1) "One-sided" servers (2+). Some
players will have one serve-side that is better than the other.
These players are advantaged by an odd number of serves per serve
game since they can choose to start serving from this side in
every serve game. With the traditional 5 serves 3 out of 5 (=60%)
of their serves will come from the "best" side. (This
aspect should be most relevant to tennis and squash. It is, of
course, not relevant at all in table tennis.)
Let's compare this assymetry with similar assymetries found in
the individual sports in order to get a feel for how important it
is:
1.1.1) Squash asymmetry. It can be noted that the same asymmetry
exists in squash since it allows the players to choose from which
side to start serving every time the serve is conquered. This
should in fact be a much bigger asymmetry than in racketlon given
how many times the serve can go back and forth in squash (in
traditional counting, that is). If a player with a better side
decides to systematically use this advantage he could quite
realistically end up serving over 80% of his serves from this
side.
1.1.2) Even tennis has this asymmetry - although to a lesser
extent than in racketlon; Since all serve games in tennis are
started from the right all serve games that are won at 40-15
contain 3 serves from the right and only two from the left. In
all other cases but 40-15 (which only occurs in say 20% of the
games) the situation is admittedly symmetric, however... so in
this sense the asymmetry is weaker than in racketlon. But in
another way it is stronger; This tennis asymmetry is a clear
advantage for the right-servers (i.e. the servers that are better
from the right side) over the left-servers. In racketlon there is
no asymmetry between right and left-servers. The racketlon
advantage is an advantage for players with a better side (which
should be a less significant thing... )
1.2) Lucky Tosser Advantage (2+++). Several
players feel that the present tossing rule with four individual
tosses in each of the four sets can give an "unfair"
advantage to a person with luck. In fact, this seems to be the
major argument for shorter serve games. Tossing is, admittedly,
fair to its nature since both players have the same chance to win
it - so what this is really about is the extent at which _luck_
should influence the result of a game. Most will agree that
_skill_ should decide the winner and _not_ luck. Serious sports
traditionally strive to adopt rules that minimise the influence
of luck. Racketlon should not be different. We should aim at a fair
deal between the lucky tosser and his opponent.
Let's take a closer look at the advantages a lucky tosser gets
in order to understand the importance of the problem. It is about
serves and sides;
1.2.1) Serves.
1.2.1.1) Start-serve advantage. Often, at least in tennis and
table tennis, the server has a distinct advantage over the
receiver. When this is the case it means a general advantage to
the player who starts serving because of two reasons: a) the
"starter" will have a better opportunity to take a lead
than he would if he received, which might lead to a psychological
advantage; b) the starter will also, probably, get to serve more
times in total in the set. With 5 serves he will get to serve
between 0 and 5 times more than his opponent. On average 2,5
times more. 2,5 serves is not much (in a tight 21-19 set with 40
points it is about 6% of the total number of points but it could
be enough to decide a match) -- This start-serve advantage will
obviously decrease with the number of serves in each serve game.
1.2.1.2) Start-receive advantage. Sometimes it might be an
advantage to receive (not unusual in squash and badminton). In
this case the same arguments as for the start-serve advantage
apply. See above.
1.2.2) Sides. There are potential advantages related to the sides
as well. These will obviously not go away with shorter serve
games but stay the same.
1.2.2.1) Mathematical side advantage. The present format of one
set to 21 in each sport and switch of sides at 11 contains an
asymmetry that gives a "mathematical" advantage to the
player that _finishes_ on the better side - as seen from the
following example: Assume two players of equal standard and a
court where one side is so poor that 2 out of 3 points are lost
on that side. If player A starts on the good side he will lead
11-6 when the sides are switched while player B will win the
"second half" by 15-8 meaning 21-19 to player B who had
the good side in the end. This advantage will, of course,
increase with the difference between the sides; If the side is
really horrible A will lead by 11-0 and B will win 21-11.
1.2.2.2) Psychological side advantage. But It could also be
argued that it might be a psychological advantage to _start_ on
the best side since this will give you a good start.
1.2.3) Single Toss. There is another - and arguably better(!) -
way to deal with this lucky tosser advantage, which has been
discussed before. It's the idea (as practised in Finland for
years) about a single toss at the beginning of the match instead
of 4 individual tosses at the beginning of each set. This
"Single Toss Rule" and its advantages are described in
a racketlon.com news-item.
So, why is the Single Toss arrangement a better way to deal with
the lucky tosser advantage - compared to the shortening of the
serve games? Here are some of the aspects:
1.2.3.1) Equal Deals. The single toss rule effectively eliminates
the lucky tosser advantage. With single toss it is often not even
clear whether the winner of the toss will get the best deal since
smart thinking might well be more important. The following
quantitative analysis shows that the single toss arrangement
might well give both players equal deals; Assume, for simplicity,
that it is worth 4 points to get first choice in one sport and 3,
2 and 1 points in the other three sports respectively. Then,
player A will choose the "4 point" sport and player B
the 3 point and 2 point sports leaving first choice in the 1
point sport to player A. This means 5 points to both players and
it is therefore unclear which of the players that gets the best
deal. (This is, of course, an over-simplified reasoning - but it
should make the point.)
1.2.3.2) Short Serve Game Short-comings. Compared to Single Toss
shorter serve games fail to offer "equal deals" in both
of the two (sides and serves) relevant ways and is therefore not
as effective at eliminating the lucky tosser advantage; a)
Shorter serve games fail to deal with the side advantages. The
Single Toss addresses this. b) Although shorter serve games do
alleviate the start-serve/start-receive advantages it does not
solve the problem entirely. Even serve games with one single
serve would mean a slight asymmetry. The Single Toss effectively
takes care of this too.
1.2.3.3) Complex? The single-toss, as described above, is perhaps
a little hard to explain in a simple way, but recent tests have
shown that it is a very easy arrangement in practice. (Fears
about the complexity involved with having to memorise 4 choices
have turned out to be exaggerated. In fact, the player who gets
to choose in table tennis will only need to remember one single
thing over some time - namely the only other sport in which he
gets to choose. Furthermore, since the choices will include
tactical considerations it will make them easy to memorise.)
Nevertheless, Single Tossing might meet some resistance at
introduction (originating from players forgetting to apply it
rather than from difficulty to memorise). A way to deal with this
would be to make it optional; Only if one of the players (or the
referee) insists should the single toss be used. (Compare this to
the situation at 11 in each set when sides should be switched.
This often does not happen since none of the players are
interested in switching and absolute fairness.)
There is admittedly one instance at which the single toss rule
does make for some pretty difficult memorizing and that
is when several matches are played in parallell as is done at the
Finnish Open group play (everyone in the group plays everyone in
table tennis before moving on to the next sport etc...). Since
the Finnish system means that up to three racketlon matches go on
in parallell all three corresponding sets of choice will have to
be memorized which has turned out to be difficult in practice.
But Finnish Open is an exception and the system with parallell
matches does not exist anywhere else - and is therefore very
unlikely to make it into the official rules. Besides, even in
this case, the problem can easily be overcome by writing down the
result of the allocation procedure on the match protocol before
the match starts.
1.2.3.4) Interesting. It might be argued that the Single Toss
arrangement would not just solve the problem but also contribute
to the game in its own right through being an interesting
tactical mind game. There are a multitude of tactical aspects on
choosing side or serve... (but the normal case would probably be
that one of the players gets to serve in table tennis and the
other gets to serve in tennis).
1.2.3.5) Racketlon Spirit. The Single Toss arrangement offers an
"integrated" way to attack the lucky tosser advantage
problem - i.e. rather than to try and achieve perfect fairness in
each of the four sports it aims at achieving fairness in the
over-all racketlon match. The Single Toss forces players to
compare advantages across sports - e.g. a serving advantage in
tennis might be balanced against a side advantage in badminton
etc. This is in line with Racketlon spirit. (Racketlon is not
just combining four different sports. Racketlon is a sport in its
own right.)
1.2.3.6) Alignment. All four individual sports make use of a
toss that appears only once, i.e. a single toss,
followed (in all cases but squash) by an arrangement (i.e. the
right to exercise the remaining choice) that seems to aim at a
reasonably fair deal between the winner of the toss and
the looser and therefore a minimization of the influence of the
toss. From those perspectives the Racketlon Single Toss Rule
clearly offers an alignment with the tossing rules of
the individual sports. For an account of the tossing rules in
each of the four sports, see the Single Toss
news-item, as mentioned earlier.
2) Simplicity
2.1) Side confusion (2++). With 5 serves there
is sometimes confusion about what side to serve from (especially
in squash with very long rallies). With an even number of serves
this confusion disappears; even numbers always on the right,
uneven on the left.
2.2) Switch confusion (2/5?). There is also a
potential for confusion about when the serve should switch hands.
With 5 serves it happens at familiar combinations like 7-3, 9-6
or 13-7. By adding up the numbers it is also fairly easy to
deduce whose serve it is given who started to serve. (13+7=20
meaning that the player who started to serve is about to serve
again.) With 4 serves the serve will be switched at combinations
like 6-2, 13-7 or 16-8, which would also quickly become familiar
to us racketletes (no big problem in Vienna 2003). With 2 serves
it will be a little simpler with a switch at every second serve.
But since the games are so short and since the serve switches
hand so many times over a set maybe there is an increased risk
for confusion about whose serve it is(?)
2.3) Similarity with other sports (2/5?). It
might be argued that it is an advantage of simplicity if the
format is familiar from other sports. Games of 5 serves in sets
to 21 has a long tradition within table tennis although it has
now been abandoned for 2 serves to 11. 4 serves to 21 has no
precedence. Neither has 2 serves to 21. --- On the other hand it
could also be argued that it is an advantage with a counting that
is unique to racketlon.
2.4) Tradition (5++). The presently used format
with 5 serves has a long tradition. This is how it started in
Finland in 1986 and it is what has been used in Sweden and
Austria since the beginning of the Nineties. It could be argued
that we should not change without really good arguments. Why
confuse people that are getting used to the traditional format -
if it is not really necessary? If it ain't broke, don't fix it!
3) Image. Aspects like simplicity and fairness are of course all
indirectly related to the image of our sport. But there are
aspects for which the connection is more direct;
3.1) Interesting. Some aspects indicate that 5-serves makes for a
more _interesting_ game than 4 or 2 (both from the perspective of
the audience and that of the players themselves);
3.1.1) Balance (5+). Racketlon history contains
examples of shorter sets to 15 and 11. It seems as if when the
sets have gone to 15 there have been 3 serves while sets to 11
have had 2 serves. Why is this? It can be argued that these
formats have been chosen in accordance with a sense of
"balance" - between the size of the set and the size of
the serve games. In a tight "21-19" set 5 serves make
up 12,5% of the total of 40. In a "16-14" set 3 serves
make up 10% of the total of 30. In a "11-9" set 2
serves, likewise, make up 10% of the total of 20. It can be
argued that if this balance is disregarded the format becomes
boring and uninteresting. It should be noted that 4 serves make
up 10% of 40 so passing the test.
3.1.2) Games within the match (5++). Five serves
is more interesting for both the audience and the players
themselves in the sense that the serve games make up "small
matches within the match". You can follow, with interest,
who will _win_ each serve game (e.g. 4-1, 3-2). With two serves
it is more of a monotonous and boring sequence of identical
points. (Tennis is really exploiting the idea of "small
matches within the match" with its games that make up sets
that make up matches. The consequence is that there are few
points in tennis that are uninteresting.
3.1.3) Tactical choice (5++). The 5-serves
format brings an element of tactical choice into play, since the
server can choose from which side to start serving.
3.2) Old-fashioned (2+). It can be argued
that the 5-serves format is old-fashioned. It doesn't even exist
anymore in table tennis, which will cause confusion in the future
when new young players start taking up racketlon. In fact, the
sport will be more and more affiliated with 'old people'.
4) Tournament Management Issues
4.1) Time wasted (5+). In tennis, if you changed
serve every two serves it has the potential to increase game time
dramatically hitting the balls over to each other; time which is
always tight in tennis when people are waiting for courts.
5) Rhythm (5+++). Five serves gives the players
a fair chance to find rhythm in serve and receiving, which works
for better quality of play, which is good for both the audience
and the players themselves. Two serves is too quick and
fragmented for this. This aspect relates to that of Balance
above. It can be argued that a harmonious rhythm of a set to 21
is serve games to 5 while a set to 11 works better with 2 serves.
6) Every point focus (2+). The argument has
been brought forward that 2 serves makes it easier to focus on
every point, since there is very little emphasis left on the
serve game dimension. Furthermore, it can be argued that this is
more in line with one of the fundamental principles of racketlon;
"every point counts".
CONCLUSION
So, what are the most important aspects? Everyone should pick his
own favourites from the above - but here is also an effort at
making some sense of it all and weighing them all together to
form a conclusion (starting from the assumption that the main
battle is between 2 or 5 serves);
In the survey above I count to 5 aspects in favour of downsizing
the serve games to 2 serves while there are 6 aspects in favour
of keeping the traditional 5 serves. On the "5 serves"
side the strongest are those of "rythm", "games
within", "tactical choice" and
"tradition" while the heaviest aspects on the "2
serves side" seem to be that of the "lucky tosser
advantage" and "side confusion". But given that
there is a better way to deal with the "lucky tosser
advantage" (the single toss) this makes the 2-serves side
look weak. The "side confusion" aspect alone is simply
not heavy enough to compete (especially given the fact that
squash is dealing with this side confusion without much
problems).
My conclusion is that we should stick to the 5 serves and
introduce a single toss mechanism that would take care of the big
influence that some people feel that the tossing has on the
result of a tight game.
_________________
Revision History:
rev PA1, by Hans, 2004-11-29:
Based on Gert's analysis of 4 serves. Adapted to include 2
serves. Some new aspects added (especially within the area of
"image").
rev A, by Hans, 2004-12-05:
Added some thoughts received from Ray on the chat (the _time_
aspect, the tactical aspect of left hander vs. right hander)
Added the "rhythm" aspects (also from the chat).
Added a conclusion.
Several other minor adjustments.
rev B, by Hans, 2005-01-06:
Modified the structure to include the title "Lucky Tosser
Advantage" and extending this chapter (mainly by adding a
more complete analysis of the Single Toss).
rev C, by Hans 2005-04-01:
Removed the definition of the Single Toss rule and replaced it
with a link to the Single Toss newsitem, containing the
definition and a discussion of the single toss rule.
Added the "alignment" aspect of the single toss rule.
rev D, by Hans 2005-06-10:
Introduced an Executive Summary, covering the main points of this
analysis.
Added indications of the weight and "direction" of each
of the aspects.
Added the "Every point focus" aspect.
Several smaller modifications to improve the readability of this
analysis.
rev E, by Hans 2005-06-16:
Added a quote by Sarah McFadyen in the "Executive
Summary"
_________________________________