Odpalam nowy projekt z typami WTA - #favi

Postanowiłem odpalić nowy projekt dotyczący typów WTA, który ma ożywić mojego bloga i twittera i dostarczy trochę value:) potencjalnym czytelnikom. Projekt ma roboczą nazwę #favi i w jego ramach będę podawał typy na faworytów… 

 

Sezon 2020 jaki był każdy wie… kompletnie wytrąciło mnie to z rytmu turniejowego. Szczerze mówiąc początek tego roku wcale nie był lepszy i normalniejszy - pierwotne turnieje w Australii były odwoływane i/lub przesuwne, eliminacje do AO rozgrywane w Dubaju… To wszystko sprawiło, że przez długi czas nie miałem ochoty analizować całych turniejów, które były często wyrwane z kontekstu i roiło się w nich od różnych niewiadomych.

 

Metoda HoN jest OK i myślę, że z czasem jak wszystko znormalnieje, to do niej wrócę. Ale ma też one swoje wady, głównie organizacyjne - wymaga dużo pracy przed turniejem, na przełomie weekendu i poniedziałku a nie zawsze mam wtedy czas. Poza tym najlepiej gdyby drabinka była już ustalona, pierwsza runda nie nakładała się z ciągle trwającymi eliminacjami itp. - a to nie zawsze jest normą. Podawanie pojedynczych typów bez narzuconego grafika jest jednak bardziej elastyczne i wygodniejsze.

 

Typy #favi będę podawał na pewno na twitterze (@voy_tennis) i prawdopodobnie jako darmówki na blogabet (https://cellarpicks.blogabet.com/). Na koniec miesiąca zrobię na blogu podsumowanie, skomentuje wyniki i zapewne niektóre mecze.

 

Celem pobocznym tego projektu jest też zmierzenie się z powszechnym przekonaniem, że granie na faworyta, szczególnie w WTA jest do d**y i nie może być efektywne. Niskie kursy dają mniejsze zwroty (yield) to fakt, ale mają też swoje zalety i mam nadzieję że ten projekt to pokaże. 

 

No to jedziemy ...

Cellarpicks - podsumowanie kwartału q4/20 (quarterly recap)

Patrząc z nadzieją w 2021 rok, podsumujmy w kilku słowach ostatni kwartał mijającego roku.

 

Wrażenia - … pierwsze odczucia są fatalne, zarówno co do przebiegu ostatnich trzech miesięcy jak i wyników. Ale z drugiej strony, jeśli kwartał na zero (+1j) określam jako fatalny, to jak podsumowałbym taki na -15j czy -20j? Póki co nigdy się taki nie zdarzył, ale teoretycznie jest to możliwe. Najwidoczniej moje odczucia i krytyczne oceny wynikają z ambitnych oczekiwań oscylujących wokół 15 jednostek na kwartał.

Liczby - od października podałem 77 typów, w sumie 88 jednostek; osiągnięty zysk to +1,1j :)) przy yield ok 1,3%. Wynik oczywiście jest słaby a na wykresie czwarty kwartał jest prawie niewidoczny i zdecydowanie odbiega od pozostałych w 2020r. 

Analiza - poszukiwanie przyczyn i skutków w zakładach, w tym przypadku przyczyn słabych wyników, jest zawsze obarczone pewnym błędem. Taka jednak natura ludzka, że szuka przyczyn, znaczenia i sensu, nawet tam gdzie go nie ma… podążając jej tropem widzę kilka powodów słabość tego kwartału: 1) sezonowość - nie jest to mój pierwszy sezon TT i słabsza jesień nie jest dla mnie specjalnym zaskoczeniem; 2) kolejne przerwy wynikająca z COVID - TTstar i LigaPro są czeskimi turniejami, a od połowy październikach w Czechach wprowadzono twardy lockdown. Zresztą w kalendarzu rozgrywanego w Niemczech ChallangerSeries też widzę dziurę prawie od połowy października do połowy listopada. A grudzień z przerwami świątecznymi zawsze był wolniejszy i nie porywał wynikami. 2a) Sama przerwa i mniej meczy to tylko część problemu, to co przeszkadzało mi bardziej, to większa niż zwykle niepewność co do aktualnej formy, postawy czy motywacji zawodników w tak ciągle przerywanym sezonie. 

 

Free picks - zgodnie z zapowiedziami, w q4/20 nie podawałem już tzw. darmowych typów. Miało być prościej i przejrzyściej i tak też to odbieram. Nie zawracam sobie obecnie głowy co podać jako paid, co jako free, jakie są wyniki każdej z grup itd. 

 

Plany - nie zamierzam wiele zmieniać, chciałbym tylko podwyższyć stawkę bazową do 2/10. Chciałem to zrobić już w q4/20, ale czułem, że lockdown, podziurawiony grafik itd. wprowadzają dodatkową niepewność i ciągle odsuwałem tą zmianę na później.  Oby nadchodzący rok był normalniejszy i ciut stabilniejszy, żebym spokojnej mógł robić swoją robotę.

 

 

 

 
 

Cellarpicks - podsumowanie kwartału q3/20 (quarterly recap)

 

(Key thoughts in english below the main post)

Z lekkim poślizgiem, czas na podsumowanie aktywności na serwisie blogabet w trzecim kwartale. Zdecydowałem, że podsumowania będą właśnie kwartalne i po polsku. Ostatni wpis z marca (!!) obejmował większość kwartału pierwszego, wiec po tym wpisie zostałaby "tylko" dziura na kwartał drugi. W ogóle ten nieszczęsny rok, za sprawą COVID jest jedną wielką dziurą… współczuję wszystkim z branż około-rozrywkowych z udziałem większych skupisk ludzkich… w tym sportowi wyczynowemu i całej jego otoczce.

KONKRETY - 89 typów, zysk 18,3j, stawki łącznie 102j, yield 18%. WTA -1,9/11=-17% z fatalną serią ostatnich pięciu typów we wrześniu, TT* +20,2/91=22%. Średnio 6j/miesiąc wyglądałoby nie najgorzej, ale rozkład tego wyniku w miesiącach był niekorzystny, bo prawie cały został wypracowany w sierpniu. Niestety tak to już jest w zakładach, że wyniki nie rozkładają się równomiernie i praktycznie nie da się nad tym zapanować. Mimo 18% yieldu odbieram ten kwartał jako poniżej oczekiwań, może już nie tak jak na wiosnę gdy naliczyłem około 10 typów przegranych z prowadzenia 2-0..., ale ciągle za dużo meczy o włos rozstrzygało się na niekorzyść pozostawiając odczucie męczarni. Z perspektywy hipotetycznego klienta, który trzymał się zalecenia by stawkę 1/10 grać przynajmniej za 30 euro (koszt subskrypcji), serwis w trzecim kwartale okazałby się dochodowy - wygrane 6-krotnie przewyższyłby koszt subskrypcji. *TT=table tennis
 

UJĘCIE KWARTALNE - jeszcze kilka słów o perspektywie kwartalnej w patrzeniu na wyniki w zakładach bukmacherskich. Mówiąc krótko, miesiąc jest za krótkim okresem przy ocenie wyników/typera. Na przeróżnych portalach, forach, serwisach panuje perspektywa miesięczna, ale moim zdaniem jest to błąd, wynikający z chęci szybkich rezultatów, co jest z do pewnego stopnia zrozumiałe i głęboko zakorzenione w ludzkiej naturze. Moim zdaniem kwartały są metodologicznie lepsze i już na wstępie wymuszają na graczu bardziej cierpliwe podejście. Zmienność wyników w kwartałach jest o wiele mniejsza, co zmniejsza emocjonalny roller coaster i gra staje się spokojniejsza. W dłuższym terminie jest to bardziej opłacalne podejście. Dowody na to obserwuję w swoich danych od lat, dla przykładu spójrzmy na wyniki cellarpicks/blogabet w tym nieszczęsnym 2020 roku - miesięcznie nieprzyjemna huśtawka a kwartalnie wyniki łączne typów płatnych+darmowych wyglądały jak na wykresie wyżej: q1=20j; q2=15; q3=18j. Gołym okiem widać, że kwartalnie zmienność jest minimalna, co wskazywałoby na spokojną grę, choć w poszczególnych miesiącach momentami jazda była nieprzyjemna. Mało kto jest w stanie wyłączyć emocje, przez trzy miesiące grać wszystko jak leci, a na koniec stwierdzić, ze był fajny kwartał na 18 jednostek. Ja też tak nie umiem, ale jest to jakiś cel.
 
FREE PICKS - W tym kwartale, bodajże w połowie września przestałem podawać FREE picks. Może i pełnią one jakąś rolę promocyjną, ale rozkład wyników między paid i free picks mocno mnie zirytował w tym roku, a postawa kilku gapowiczów przelała czarę goryczy…
 
CZESKA LIGA-PRO - nowy rynek, który zacząłem śledzić i modelować około czerwca, pierwsze typy z tego rynku trafiły na blogabet w lipcu. Rozgrywki bardzo specyficzne, z małymi wyjątkami właściwie amatorskie. Z perspektywy typerskiej powiedziałbym, że jest to młody rynek, trochę prostszy niż Challenger Series/TT Star, ale to nie znaczy, że taki stan utrzyma się długo - w porównaniu z czerwcem/lipcem w październiku, gdy to piszę, widzę już oznaki, że oddsmakerzy lepiej kalibrują kursy. Bez wchodzenia w szczegóły techniczne powiem tylko, że zebrałem już wystarczająco dużo danych by odpowiedzialnie móc podawać typy z tego rynku, wiec w czwartym kwartale można się ich spodziewać więcej…

cellarPicks - last three months RECAP [+21u/+19%]

2020 BNP Paribas Open (Indian Wells) was postponed or even canceled, so having some spare time, I will use it to wrap up the last 3 months on my blogabet paid service.

Since 2019 December, so from switching to paid mode, I posted 110 picks (92 table tennis and 18 tennis) which brought 21.3 units of profit at ROI of 19%. Unfortunately, my official record and chart seen on blogabet site show much less profit, because 31 picks with a profit of 13u.(+42%ROI) were posted as free and are not included in main blogabet statistics :(( But of course all are documented and you can check them if dig deeper in ALL picks stats. 

Why do I sum up paid and free picks? Well, from the start I treated free picks as an incentive and also a promotion for my service. I have been doing the same analytical job in case of paid and free ones, and have no control over which picks will hit. So as I said from scratch to my paid subscribers that they should follow free picks as well.


21.3u/ROI=19% looks very good but I have really mixed feelings about this past quarter. The distribution of paid and free picks was just ridiculous. I should also be bolder and more frequently use stakes larger than 1/10.

In the last or two weeks of February, you might have noticed free tennis picks tagged as "HoN" - why does it mean? For those who read my private blog it's obvious, but I will also publish this recap on blogabet blog section, so a few words of comment - these picks come from one of my projects called Hot or Not, and I decided to document picks related to it both on twitter (@voy_tennis) and my blogabet account. It's up to you if you trust these picks, I personally would bet on it:), of course WTA is more efficient market and expected results should be lower but still in profit. I am using 1/10 stake for these picks. 

I haven't decided yet if I will wrap up monthly or quarterly in the future. Being in this business for a long time, I believe that quarters are more valid and meaningful but from yours/followers perspective it's maybe expected to hear from me more often than once in 3 months. Plan for the future is pretty much the same, although I think there will be fewer free picks and I will use stakes 2/10 and 3/10 more often. 


Hot or Not - week 10/20 (Lyon, Monterrey)

It's rough Monday so it's time for Hot or Not list. This week there are 3 tournaments in Lyon, Monterrey and 125k one in Indian Wells. The last two I have to still look at, but I made some conclusions on Lyon so these are my initial lists for this week:
HOT list: Dodin, Uytvanck, Kasatkina, Mladenovic
NOT list: Garcia, Paquet,

Dodin - she played some good tennis in Moscow another indoor event this year, also her draw looks OK, and she will have home crowd behind her...
Uytvanck - indoor specialist I would say, in decent form, lately was very close to winning with stellar Rybakina (but not only)
Garica - looks weak at least for the last 3-4 months, but the draw opens up for her quite easy so I think she may reach QF even covers some HDP along the way.
Kasatkina - she had a bad year in 2019 but in a few matches this year I see some signs of hope, so maybe one of my favs is coming back. Like most Russians, she is used to playing indoor.

Later on I will add some names referring to Monterrey and Indian Wells (maybe)

edit: Tuesday 15:30
I went through Monterrey draw, check my database😃 plus my personal cache... and here are the additional Monterrey players that draw my attention:
HOT list: Potapova, Watson
NOT list: Stephens, Konta



HOT or NOT - week 8/20 Recap and how to read this recaps?

I write this post a) to recap week 8/20 (Dubai) and b) to exemplify how HoN is to run. Maybe one picture bellow will be worth more than many words I wrote before. On a regular basis, recaps will be posted in HoN posts as week and tournaments end but this post I treat as a kind of instruction on how to read my tables and how to understand the whole idea of Hot or Not lists. In the future when answering queries it will be easier to refer to this post than to explain the same all over again. 

For start a reminder who was on the lists in week 8/20:
Hot list: Halep; Mertens; Brady; Mugu; Kontaveit; Kenin; Rybakina; Pliskova
Not list: Svitolina, Sevastova,
Not list, if QF: Strycova, Martic
Basically in the recap table below you will see all matches that picked player played in. Some exceptions are possible, eg. in week 8/20 Strycova and Martic were put on sub-list "not, if QF", so in the table below you won't see their 1-2 round's matches, only from quarter-final stage. 
So in the table you can find information such as date, round, players, "side" means just which player I played on in a given order, then "exclusions" and info if the prediction is a win or miss (1 or 0). 
In "side" column you may notice a sign "eo", which usually means the two players from hot list are to play each other and this match will be naturally excluded.
Please don't worry about the "exclusions" section marked by yellow frame - this is my own mumbo-jambo and in short it represents different types of collisions with my other models or concepts. I am showing it here to a) be transparent and b) it clarifies a little bit the transitions from the "raw subset" to final picks posted on twitter/blogabet.
So what are the main information for you in this table? I think maybe two - 1) how performed the raw subset and 2) how the posted picks went. What do I understand by "raw subset" - these are all matches that players from the lists are involved in. In week 8/20 there were 15 such matches, remember that matches marked "eo" are not counted. And the most important info concerns the performance of posted picks which you can see marked in the green oval. This is easy and should be clear for everyone - I posted 3 picks on twitter/blogabet coming from week 8/20 schedule, two of them were wins and one miss. You can check it on my twitter: https://twitter.com/voy_tennis
HoN #001 Brady to win @ 1,84
HoN #002 Kontaveit-3 Martic @ 1,9
HoN #003 Rybakina -1,5 Martic @ 1,877
So the result of the raw subset was 4 units, ROI 27% and the result of posted picks was 0.8 units and also ROI 27%. These two metrics I will put in the title of the post when the week is over and added up. (This week you will see: raw:+4; posted:+0.8) Looks like week 8/20 was pretty good mostly due to Rybakina.

I know that this method has some flaws. I am aware that I can be accused of manipulations, especially at the beginning of the week when I work on a tight schedule and have to analyze many matches and I am often late with publishing the lists. But as I said before if I would cheat then mostly myself...For you guys posted picks are all it counts and they are/will be on time and verified by blogabet (probably there are some ways to manipulate twitter posts, but just don't know twitter that much and I am not aware of it) 



cellarPICKS - no more free picks?

Easy... it's just a question. I have a dilemma - how to organize paid and free picks? Probably the time will show me the way, but for now it looks awkward. For example, in this February I have made about 8 units of profit, 4 units from paid picks and 4 form free picks. I just feel it's somehow unfair for paid subscribers and some of them may think why to buy picks if I can get a similar quality for free?

This is a very frustrating business in which buyers are very inpatient, make quick judgments and are easily fooled by the law of small numbers. All this leads to misunderstandings and simply bad decisions. For example, for the last two months my free picks have performed much better than paid pick - what does it mean? Well, for me not too much ..for me it's just variance. It's kind of luck which can be misinterpreted in many, many ways by YOU guys. So one of the options I am thinking of is to cut free picks to zero and curb all the speculations. I don't know... I would like to get to a point where free picks account for max 20% of profits, but right now I don't know how to do it. 

If there is anybody there let me know what you think...


HOT or NOT - week 9/20 (Doha, Acapulco) [RAW:-1; POSTED:+2,6]

Last week, the first after HoN comeback was quite OK. Later on I will add some recap to the previous post. But time flies and here we are already with some matches in Doha and Acapulco behind us... I don't like this rush on Monday (and Sunday evening) when I have to analyze both full draws of WTA and TT tournaments - it's a lot of matches and Sunday evening is not a convenient time.

Enough these complaints:) let's move on to my WTA hot or not lists for this week, precisely for Doha cos I give myself some more hours to analyze Acapulco's draw.

Hot list: Barty, Kenin, Martic, Świątek, Vekic, Mertens, Pavlyuchenkova, Bertens, Sabalenka,
*added (Acapulco): Potapova, Hibino, Rus
Not list: Hercog, Garcia, Strycova, Muchova, Sevastova
*added (Acapulco): Stephens (if QF)

On the hot list, there are some obvious names like Barty or Kenin - players who went deep on AO a few weeks ago, both in good form and have no problems playing on hard courts. Bertens is not hard court specialist maybe, although looking only at the last few years I would say she is a solid player on all surfaces, additionally she has a rather easy draw imo.
On the not list you can notice some players that are already out of the tournament and I can be accused of manipulating results... belive me if I would like to cheat then I will be cheating myself cos I use this method for making my own bets. Half of the first's round matches were played yesterday and I made a decision of putting Hercog and Garcia on "not list" back then. In any case, just look at the pick published yesterday on my twitter and blogabet. I run this HoN project mostly for myself, since I have very few followers (or any), I am having fun testing some concepts of mine but simultaneously I have some proofs from previous tests that this method can be effective. In the full form my lists will be hard to follow anyway, so for potential followers picks posted on twitter/blogabet are all that counts. And in case of these picks, there is no shadow of doubts that they are not on time etc.

OK, that is it for now. I hope in next weeks I will write more reasons why these players and no others ... GL. In the afternoon I will add some names regarding Acapulco. 

RECAP
I have some mixetd feelings about this week. Overall it was quite weak for my hot list - there were some early exits that I wouldn't expect. Especially three matches: Mertens-Strycova, Bertens-Zheng and Mertens-Putintseva in which my favs won the first set and then fell apart. Thus the raw metrics is not satisfying - both lists produced 24 matches with a hit rate of 50%. On the other hand the posted picks (blogabet/twitter) went exceptionally good: 4-1 and 2,6 units of profit at 52% ROI.






 

HOT or NOT - week 8/20 (Dubai) [raw:+4; posted:+0.8]

It’s high time for great come back of “Hot or Not” project. A few first weeks of 2019 were brutal in many respects for me and among other projects I couldn’t find time for HoN. Additionally my WTA predictive model (in fact system of models) got complicated and adding
a new element produce many new relations and it takes more and more time to put all the pieces together in one cohesive whole.

Let’s quickly move on to the lists, because in a moment starts the second round in Dubai. As usually I am late with these lists, but I hope I will improve at this point this year... Hot or Not project is based on more intuitive approach comparing to the rest of my modeling work, I would say it’s more loose especially at the first stage where I start just with the impressions the players left me with in the last few weeks. Then I mix it up with other relevant information (injuries, shifts in coaching staff, private stuff, travelling schedule, etc) and in the end I apply some statistical layers on it to come up with my predictions.

So who should be hot (or not) this week. My list are as follow:
Hot list: Halep; Mertens ; Brady; Mugu; Kontaveit; Kenin; Rybakina; Pliskova
Not list: Svitolina, Sevastova,
Not list, if QF: Strycova, Martic

Comparing to the last year there is one novum this week (and I don’t exclude it will be other in forthcoming weeks) – a variation of “not” list, which should be read as: Strycova and Martic are problematic this week but not necessarily in the first rounds. I am willing to bet against them but only IF they reach quarterfinals. 

How to use my lists ? For me HoN is an implementation of the concept of bets series (package of bets), I have always had an intuition that betting on series of matches can be more effective then betting on one particular match in which variance is higher. The last year’s try with GDGD project confirmed my guessing. For example if I pick Halep to my “hot list” I will bet on her a few consecutive bets, hopefully she will reach the final and will cover most of HDP along the way😊. Some may ask why not outright bets – well, I don’t like this kind of bets, handicaps on games give me more flexibility (eg. I can omit some matches) and lastly there are countless variations in which player can be “ATS-positive” (positive record against the spread) but never reach the final. And don’t forget the “not list” where I basically bet that given players will not cover HDP and probably will exit the tournament at early rounds. So personally I will record all the matches that picked players are involved in and I will check how they went against the spread. This should give me even more assurance that this method and the process behind works.

But I am aware that for potential followers this all-encompassing approach may be hard. At some points it can also be unclear, eg. when players from the lists play each other. In the upcoming weekly wrap-ups you will also see that some matches were excluded – it will come from avoiding some match-ups or avoiding collisions with my main predictive model. So for potential followers I came up with simple solutions – selected HoN predictions will be posted on my twitter account (@voy_tennis) and on blogabet. These picks will also be marked in the final table, in which you will be able to see that these picks arise form the lists and are a subset of all picks coming form HoN method.

HOT or NOT - week 39 (Wuhan, Tashkent) [+0,56]



This is the first post from a new series of "HOT or NOT" posts, which will be the continuation of GDGD series. In a separate post I will write why GDGD came out difficult in day-to-day reality (or rather week-to-week) and why it has evolved to the form of HoN.

This week as a part of Asian swing, there are two tournaments on hard courts - in Wuhan and Tashkent. The first one in China is a Premier category with quite an impressive line-up. I suspect some top players still search for some points to get to finals in Shenzen.

edit: 23.9.2019 13:45
Most of the 1st round's matches were already played, but this is just one of the advantages that the project HoN has over GDGD, that I don't need to rush in Sunday evening/night or Monday morning and I can post my list even on Tuesday, however the sooner the better (I will try but not always it's possible for me). HOT and NOT lists are one part of the concept which will give you a broader perspective but more important (I think) are particular picks that you will find on my blogabet account. And in terms of picks, there will be no excuses in the sense that only published pick counts and of course they must be published before the match (no other option on blogabet:). I think it will motivate me to post the list and publish the picks on time before the first round, but as I know this "Sunday/Monday situation" it's very likely that often there will be some picks on Sunday/Monday morning and then a few hours later I will publish the lists on the blog.
This week the lists are as follows: 
HOT list:
Peterson
Martic
Kuznetsova
Pliskova
Kenin
Babos - I like how her game changed after signing a new coach. She played well in US Open and in Korea last week, ...significant progress comparing to the first half of the year. She should be one of the favorites in smaller tournaments like this one in Tashkent

NOT (really:) list:
Bertens - in general, I don't like lack of consistency she presents since Madrid. There is something wrong with her game. Some others "girls" say that they like to practice with Bertens because she can keep the ball in the court forever, meanwhile she hits 2-meters outs in some last matches.
Zidansek
Krunic
Voegele

This is it for now. Although it doesn't seem so, believe me, that there is some thought (and analysis) behind these picks. I hope there will be more descriptions in the following weeks but to be frank it's not a priority to me. Basically, I use a similar set of factors while analyzing players and matchups and I feel I would repeat myself by adding often similar descriptions, that a player is in better form, healthy or not, has better results on a particular surface and so on, so on ... 

Week 39 Summary: This week was not bad for a start:) The lists were published September 23rd afternoon, so practically after the first round and first picks covered only the second round's matches. If someone would bet on all the matches of the players from the lists, there would be 15 picks overall which would give 3,85 units of profit. This result refers to handicaps only. If someone would also try to back the MoneyLine (I recommend odds less than @3) than it would account for an additional 14 bets which resulted in 0,56 units of profit. This is of course just a hypothetical situation - on this stage nobody would bet on that many unchecked picks, including me:) Coming down to earth, in reality during this week I posted five "HOTorNOT(HoN)" picks on my blogabet. With a little effort, if you dig through a bunch of my table tennis picks, you will find five WTA handicaps tagged as "Hon". They were 3-2 ATS (against the spread) and produced +0,56 units of profit. it's OK and in generall I will be happy with a profit of around 1 unit per week - in fact, it's quite a challenging goal.
 





2019 Mallorca - go deep/go down list & daily recaps [unfinished, sorry...]

As usual, I am behind (with time), but there is a lot of stuff on my mind on Mondays. Additionally, some qualifiers are not known yet, so you wait to get some more info...you wait ...but you can't wait to the end of qualifications because the first round already starts... it's irritating but it is what it is.
Last time I covered a tournament with my GDGD method was Lugano in Aprill... since then I have distinguished some interesting factor and it took me time to integrate it to my predicting models. I must say that for some time I have been a little bit overwhelmed by the crisscross data patterns and my thinking wasn't as clear as I wish, thus I stopped publishing GDGD post until I felt that I have straightened up my decision process to the degree that I am comfortable with it. So here I go again... my picks for 2019 Mallorca (on grass) are:

GO DEEP list: Riske, Bencic, Zheng S.

GO DOWN list: Badosa, Cornet

Recap after 2R: What a comeback! What a disaster....0-5 ATS and little hope for next rounds. This comeback was really unfortunate or maybe partially I have made some mistakes which I could have omitted. But it is easier to say afterward since the probability assessment changes after the event... I refer mainly to Riske's performance - what she did was probably tanking. She won a tournament in the Netherlands just before Mallorca and probably I should have predicted that she may be up to sth to save energy for Wimbledon since the grass courts are her favorite... if she has signed up for Eastbourne then I am even more sure that she underperformed in this first-round match with Badosa...The match was really strange to me, Riske was losing an advantage on several occasion and was not very engaged, for sure not so fresh like in the Netherlands a week before. So this way GDGD method scored two misses straight away for Badosa was also on the go-down list. *I haven't seen Sorribes with Cornet, but 06 16 on home soil where she played 2 decent matches in qualification and with Cornet who is not in best shape and grass is her least favorite surface..it's a really weird result for me. *Zheng played a decent match and I guess this was a fair miss :), but in Bencic's case lack of rhythm was very visible - much more than I expected and maybe I should have tried a lower handicap which is a good tactic when a player has not much game rhythm.
The first round was strange and that was completely confirmed by the second round matches because results came back to normal I would say. Cornet and Badosa, from the go-down list, lost their matches easily although Garcia lost her focus lost the 2set being 5-2 in the front (!!). There was still sth wrong with Bencic cos she lost 1set to Rogers also being 5-2 in front (5-2 >> 5-7 !!), but the American retired eventually. So there is only Benic left in the game and I have really mixed feelings about her. But since Anisimova doesn't look very confident on the grass I guess I will give Bencic a try in hope that she's gained some rhythm which was not there in first rounds.

Recap after QF: Benic finally got some rhythm and defeated (uncomfortable on grass I think) Anisimova 62 62. Next, she faces Kerber and I think it will be a too strong opponent for her in these circumstances so I will quit that game. If she wins with Kerber, which is doubtful, then maybe I will give her a chance in the final.

Recap after SF: Well, well, well... So it happened what was less probable in my opinion. Bencic returned from the dead after trailing 26 in the first set and somehow she defeated (I would say) unlucky Kerber. Why unlucky - cos she had like 5 break points and couldn't convert any of them, meanwhile Bencic had maybe two chances to break and took one of them. The match was very even, and as for me, Kerber had the advantage for most of the time but she lost. In the final Bencic meets Kenin and as I said in the previous recap I will give her a chance. Maybe although this was a quite unfortunate contest for GDGD method, I will be able to pick a winner again:) like in Budapest, Acapulco and Miami...


WTA - May wrap-up [+2,1] and overdue April wrap-up [+1,2]

Hi guys, after some weeks not being active I need to catch up a little bit. Due to some personal issues, I was not so active in April/May, overall I posted only 35 pick on Omnibet. Ohh.. by the way, since June I have been promoted to "Experts" again. So if you wish, you can purchase my picks for "only" 53 euro on the Omnibet site ;):).
Both these moths were "in green". In May I started to post not only handicaps picks but also predictions for straight wins, so now it's more similar to the way I bet myself. Taking this into account I guess there is no use to presents ATS results (against the spread) anymore. I will refer mainly to profits/losses in units now. I have made 4,7u. and 8,5u. in April and May respectively and since ma regular stake is 4u. this translate to 1.2 and 2.1 units FLAT. I think these are pretty decent results for a market with high limits. These days when it's harder and harder to find a reliable tipster I would buy my picks for such a price;)

One more thing - as you can check in the previous posts, in the end of 2018 Summer I introduced some significant changes to my prediction model and since October I have been using this refined model to generate pick. Since then my results have improved and have come to the yield of 9% after around 190 picks. So the 3%/538 which you can see on Omnibet comes form two eras I would say. For me it is obvious that the latter part of the data is more corresponding to the way I make prediction right know and (I hope) more correlated with the future results.

2019 Lugano - go deep/go down list & daily recaps [-0,7]

As always five minutes before midnight... but I had a tough weekend, so just the lists and then picks on blogabet, and maybe later I will write some preview...

GO DEEP list: Bencic, Świątek, Alexandorva
GO DOWN list: Kuznetsova, Cirstea, Minella


Recap after 1R: Terrible start and I guess this first round will be deciding for GDGD method during this tournament... Alexandorva lost to Kuza after tight three-setter and I am not going to risk to bet against Kuza in next rounds - she looked too sharp and she is to face Bacs in the next round, with whom Russian has comfortable 4-0 h2h, so I will quit that match.
This is not the first and last time when I will abandon a player from "go-down" list if she is through to the next round. I hope it is understandable because often that will be an underlying logic behind the "go-down" list - a player should go down in the first/second round, and if not than sth went wrong eg. she is in much better form than expected. Sometimes the player is planned to go down in the first round due to a particular match-up (negative record, others). In this particular case, Alexandrova was in good shape, had momentum and seemed to be a solid clay player, and Kuza was back from a long break and her form was very questionable, plus she is not 19 teen anymore:). But if she beat Alexandorva then all assumptions must be reconsidered - I hope it's obvious for you. A similar case is with Cirstea, who was to go down with in-form Barthel (who won 1set but then disintegrated...), but she won and in the R16 will face Hercog, with whom she had a nice record, and alike in Kuza case, I will quit that match. Bencic with Lottner... I must say it was a stunner of the month so far for me but must give her a credit to Lottner for that - a well-deserved win.
So after the first round, there is 2-4 ATS and only Świątek stays in play for me as I am not going to touch Cirstea and Kuznetsova in this tournament anymore.









and almost 2 moths later.... sorry for that again.
So as mentioned abova I quit Kuznetsova and Cirstea after the unfortunate first round, so in practice only Świątek was left to follow, and must say she did well, almost perfect until the final. She reached the final in which lost to Hercog in three-setter. On the route to the final she covered 4 spreads (including first round) so I can be satisfied with this pick.
Overall the GDGD method was 5-5 ATS and loss of -0,7 units in Lugano which is OK, considering a terrible first round.