Something that I’ve
examined in great detail is the effect of various situational factors on an
individual match.
It frustrates me greatly
that the media, both the print and especially television media, make bold
claims that an irrelevant factor has a huge bearing on a match. An example of this is that various presenters
and commentators take a marginal head to head lead, such as 1-0 or 2-0 and
almost present it as a given that the player with a head to head lead will win
the match. I’ve mentioned this before on
twitter and actually it’s not the case at all, with back-testing showing a
small head to head lead actually generates a negative return on investment on
the men’s tour, and on the women’s tour, I found a 1 match lead giving a similar figure.
Of course, this false
assertion from the media is good for knowledgeable gamblers because less
informed punters will be influenced by this and actually think that these media
presenters know what they’re talking about (the vast majority don’t but get
paid obscene sums simply because they were fortunate enough to play the game –
often at a mediocre level).
You also get ridiculous
similar statements from gamblers on forums and social media, such as twitter
and of course that’s also a good thing because these people are the ones that
give us long-term value in the market because they’re opposing my point of
view.
I looked at around 30
different situational factors, back-testing every single one of them over a
large sample. All prices used were
Pinnacle Sports closing prices.
I try and be as open as I
can on here without giving edges completely away – I’m well aware that I’m
getting website visits from areas of the world that bookmakers are based and I
need to be careful to some extent. From
Monday 8th July, 2013, I will have a new feature on the website with
a ‘Player Watchlist’ updated each week showing the current players which fall
into the below 5 categories, which we can then use to make profitable betting
and trading decisions.
So what I thought I’d do
is, without stating the back-tested edges in a percentage term, write about
which situational factors have the biggest impact on the likelihood of a player
winning (in no particular order).
1. Travel and scheduling
have a huge factor on a player’s performance.
There are many occasions throughout the calendar where a player can play
in the latter stages of a tournament at the weekend and then have to fly
halfway across the world to play a first round match in another tournament.
A recent example is the
final of the Houston in April where John Isner defeated Nicolas Almagro. Both players then had about 36 hours to get
from Houston to Monaco to play a first round match. Clearly their flight won’t depart immediately
after the final finishes so it’s fair to assume that both players had under 12
hours to acclimatise to the time zone changes, as well as being fatigued from
playing long tournament and a long flight.
Isner lost his first round match to Ernests Gulbis, whilst Almagro
(starting at a price of 1.18) dropped the first set to David Goffin before
eventually winning, but having a very poor defeat at the hands of Jurgen Melzer
(priced at 1.17) in the next round. I
have absolutely no doubt that scheduling caused this defeat, and even though
Almagro did beat Goffin in the first round, opposing him pre-match at 1.18 and
trading out at some point in the first set would have made excellent profit as
well.
There are a number of
tournaments where players often have to travel large distances in a short amount
of time, and then get ready to play a first round match in the next
tournament. I’d look at players getting
to at least the quarter finals of a tournament (Friday play) and then having to
play either on Monday or Tuesday the subsequent week. Opposing these players shows a good return on
investment for both the men’s and women’s tours.
2. Leading on from that
situation is a similar one – generally opposing a tournament finalist the week
after shows a positive expectation.
Players, on the whole, don’t fight through fatigue. It’s logical to assume that’s the case too –
almost always when a player is tired, they will have managed to earn a nice
cheque from getting to a final, and accumulated some nice ranking points in the
process. Some might even have had a big
night out celebrating their success.
Therefore it’s pretty rare that they fight until the bitter end the week
after.
A recent example (there
are many) of this is Daniela Hantuchova, who won Birmingham in June, with a
number of tight, and arduous victories.
She then immediately played in Hertogenbosch, where (priced at 1.41) she
retired at 5-0 down in the first set against the lucky loser, Lesia Tsurenko.
Top 10 players, who are
used to defending ranking points, and tend to be fitter, cope with this better
but they also tend to plan their schedules better so rarely play week after
week. I wouldn’t include them in this
bracket.
This success related
fatigue particularly affects players outside the top 50 in the WTA, but can be
applied to the ATP on the whole.
3. Another fatigue factor
that affects a player is playing too many long matches in a short space of
time. My work on the influence of a 5
set match on men’s players is well known, and can be viewed here http://sbbcolumns.co.uk/an-insight-into-tennis-implications-of-playing-a-5-set-match-in-tennis/
To summarise, players that
have played a 5 set match in the previous round have a horrific record,
particularly the players ranked outside the top 50.
This theory can also be
applied to the 3 set format in both the ATP and WTA tours. Players that play 2 or more matches which
have over 30 games in them are definitely worth opposing against players that
are much fresher.
4. Injury, or potential
injury, tends to be something that influences a pre-match price greatly. Due to the market often giving much higher
odds on a potentially injured player than would normally be the case, a player
that the market thinks has an injury doubt can be great value if they quickly
show that they have recovered from it.
However it’s very difficult to judge whether this is the case
pre-match.
I decided to assess
several injury factors, basically splitting situations into several segments –
injuries where the player is out for over two weeks, and players playing within
2 weeks of giving a walkover or retiring.
Interestingly, the women
players playing within 2 weeks of walkover/retiring were affected much worse
than the men, showing that they play before they are fully fit and the market
doesn’t take that into account as much as it should. Could we also assume that they retire due to
more genuine reasons than men? I’m sure
those who saw Philipp Kohlschreiber’s retirement at Wimbledon against Ivan
Dodig would have an opinion on that…
However the flip side is
apparent when considering more longer-term absences. This affects the men much more than the women
and men who have been out injured for more than a fortnight can generally be
profitably opposed, particularly when the player is ranked outside the top
50. This is logical because these
players don’t have the financial freedom of the top players, who can take their
time recovering because they are very comfortable financially. The lower ranked players appear to rush
themselves back because they need to earn money.
5. Qualifiers need to be
assessed with great care. It’s logical
to assume that qualifiers have reasonable chances in a first round match as
they’ve become acclimatised to the conditions having won two or three
qualifying matches in the previous few days at the venue.
This theory was confirmed
by my WTA back-testing, showing that backing women’s qualifiers ranked outside
the top 100 in first round matches should indeed be profitable in the
long-term. It was, however, only
slightly above break-even for the men (which considering the 3% or so edge that
Pinnacle have, does show it has some positive effect).
For players ranked inside
the top 100 qualifying the results are less stellar. Men’s back-testing showed it’s profitable to
oppose these players and the women’s profits dropped (although were still reasonable). This is probably because these players are
qualifying for Masters or WTA Premier events (they wouldn’t usually need to
qualify for lower level events) and are playing a better level of opponent in
their first round matches.
However the situation changes
in the subsequent rounds. In the WTA,
you can still profitably back qualifiers after the first round, and it’s not at
all unheard of for a qualifier to win a tournament – Jamie Hampton winning Eastbourne is a recent example. In the ATP though, things change. Qualifiers really struggle in their second
round matches onwards, and it’s fairly rare for a qualifier to win past a third
round match.
It’s hard to give a
specific reason as to why WTA qualifiers do better than ATP ones, but one
possible explanation is that WTA qualifiers tend to be younger players with
potential as opposed to ATP qualifiers who are often journeyman players that
find it much harder to raise their level against better opponents. There are quite a few Challenger players with
a win-rate below 25% in the ATP tour and it’s pretty rare to see a WTA player
with this (I can only think of Grace Min currently in that bracket). I also believe the WTA is a slightly more
‘level playing field’ with a large chunk of opponents being able to beat each
other on any given day.
Those are the main areas
we can look at profitably, but there are also quite a few ‘urban myths’ that I
discount from my analysis.
1. The main one, as I alluded to at the start of
the article, was a head to head record.
Let me tell you now, if a player is leading by one or two matches I
couldn’t care less! It means absolutely
nothing – in fact those leads have a slight negative return on investment
(marginally more than the Pinnacle profit margin) when back-tested. It makes me burst out laughing when tennis
journalists, presenters, commentators talk about a 3-1 head to head lead and
then refer to matches in 2008. It’s just
lazy research. Last time I checked, it’s
2013, and in 2008 it’s almost certain that both players’ careers were in very
different places than currently. I’m
only really looking at the last 18 months and if a head to head record is
dominant (e.g. a lead by 4 or more matches).
Even that doesn’t have a fantastic record in the ATP, although it does
fare better in the WTA.
All sport tends to run in
a cyclical nature, with spells of dominance and then relative mediocrity. Fine margins often determine a result of a
match. A head to head lead is often an
exact reflection of that.
2. People in the tennis world seem to think it
is a given that a player will do well in their ‘home tournament’, the
tournament in their own country. This
isn’t the case at all and in the ATP, my sample showed a pretty much break even
result. In the WTA, players ranked
inside the top 100 had a negative return on investment. So the market definitely has WAY too much
faith in this.
3. People also seem to think that form is a huge
factor to consider when assessing a match in advance. I am more of the opinion that it is just
positive variance for the player – perhaps they had a win over a top 10 player
who didn’t play well, or had a few decent tournament draws, and suddenly they
are the best thing since sliced bread.
The L’Equipe article (a review of it is shown here http://www.couriermail.com.au/sport/lequipe-predicts-bernard-tomic-at-no7-in-2018/story-fnddkbrz-1226655205013)
showing Benoit Paire predicted to be world
number 2 in 2018 shortly after his win over the hugely below-par Juan Martin
Del Potro (and subsequent semi-final) at the Rome Masters was a prime
illustration of that. Since then, he
hasn’t got past the third round of his four tournaments since, with two defeats
priced at 1.4 or below.
In my tests, I looked at
how successful backing a player after four defeats in a row would have been –
effectively seeing if bad previous form severely affected a player’s chances of future success. For top 100 players in the ATP and WTA, the
results will make surprising reading to most.
I imagine most tennis followers would expect this poor run to continue,
but in fact there was only a very slight loss when considering the built in
profit margin that Pinnacle have in their prices. My sample showed it was very rare for a
player to continue the four defeats past eight matches, leading me to believe
the runs of form that the likes of Donald Young and Arantxa Rus have recently
exhibited are statistically very unlikely.
A player generally appears to have a very reasonable chance of pulling
off a win as a reasonably priced underdog when having lost between 4 or 8
matches in a row.
I would urge a small
amount of caution here though, for players of both ends of the age
spectrum. A young player on a good run
of form would have more chance of keeping this up than a journeyman player, and
an older player having a bad run of form would have more chance of keeping this
up than a younger player (I’m thinking Paul-Henri Mathieu and Venus Williams,
for example, currently).
4. The final myth I would like to dispel is the
‘after the Lord Mayor’s show effect’.
This is the theory that a player, after a big win as a high-priced
underdog, goes on to lose their next match. I’m as guilty as anyone for
mentioning this, as I genuinely felt that it had a big effect. However I finished my research on this
yesterday and I was wrong. It showed a
slight loss in the ATP back-testing, and actually a slight profit in the WTA
back-testing.
Hopefully you have enjoyed
reading this article and it’s given you some angles to consider in your future
betting. If it did interest you, perhaps
you might also be interested in my Tennis Trading Handbook – written in a
similar style (with much more detail), giving statistical analysis of various
different in-play situations.
