Парсинг коэффициентов

It is necessary to parse the coefficients of several bookmakers (mainly the coefficients from Pinnacle from the site oddsportal.com with substitution, in case there are no coefficients for some Asian handicaps, from the additional line of Pinnacle itself (Pinnacle.com).

  Championships for parsing - England (EPL and Championship); Germany (Bundesliga and 2. Bundesliga); Spain (La Liga and La Liga 2); Italy (Serie A and Serie 😎; France (Ligue 1 and Ligue 2); Netherlands (Eredivisie and Eerste Divisie); Portugal (Primeira Liga and Segunda Liga).

  A)  England (EPL); Germany (Bundesliga); Spain (top); Italy (top); France (top); Netherlands (top) - a collection of 6 championships, as shown in the screenshot.

   B) Portugal (top); Italy (B); Germany (2); Netherlands (2); France (2); Portugal (2) - another collection of 6 teams.

    C) Not yet available but for the future, it needs to be -  England (Championship); Spain (2); Belgium/ if it would be possible to combine with the leagues playing in the second half of the season in Brazil A and B, and Japan (top). The absence of matches on a certain date for some championships, while other championships from my collections are playing at that time, should be recorded as a blank (empty).

    Such tables, as shown in the screenshot above, should be recorded on one level, from left to right, collection A 6 championships, twenty cells of indentation - collection B, another twenty, collection C. If in any collection there is even one postponed match on a certain date, all other collections should also be recorded in this range of cells on the right, but recorded with a blank in coefficients, dates, rounds, and results (to make it convenient to select with formulas later).

    When entering a match and reviewing the coefficients according to the line of the required bookmaker - the opening line is indicated at the very bottom of the opening window (popup window, push window, I don't know how to correctly name it, it opens without clicking), it does not change, mainly, it is better to fix it as final together with the fixation of the late line, that is, the opening line and the late line are fixed simultaneously (mainly 2-3 days before the match, sometimes a day before the matches start). Then the line is fixed 50 minutes before the match starts, and 2 hours after the match ends, the closing line is fixed.

The program should access the site before the game days, mainly in the middle of the week, if a certain championship is not held twice a week (if twice, conditionally the game will take place Tuesday-Wednesday and Friday-Saturday-Sunday, then the coefficients of the late line should be fixed on Monday and Thursday, the program should track changes in the schedule of upcoming nearest meetings of the championship to understand when it is necessary to fix the coefficients of a certain period of the line. There should be 4 lines for fixation.   Access each meeting of this championship and fix the coefficients of different periods.

1) Opening coefficient (conditionally, since the parser from oddsportal is not very fast, and essentially fixes the early line, not the opening coefficient, but let it be.   When hovering (not clicking) on the coefficient on oddsportal, it indicates the opening coefficient.

2) Late line coefficient, conditionally the parser should take the coefficient that is displayed at that moment (conditionally on Wednesday, if 1 round is played in a week, if two rounds - the first coefficient is fixed on Monday, for the other round on Thursday or Friday, depending on when the first round ends and the next one is to start.

  Coefficients should not change, that is, after fixation, they must be stable numbers (that is, something that cannot be done with formulas in Excel)

3) Coefficient 40-50 minutes before the match. Conditionally counted as the coefficient for an hour (for time). The coefficients changed after the coach announced the main squad of the team.  (if for various reasons it is not possible to fix the coefficients of this period, it is allowed to fix the average value between the late line coefficient and the closing coefficient).

4) Closing coefficient. Fixed as well as the match result 1-2 hours after the end of each match. Such a term is necessary to eliminate inaccuracies on the site with the match result.

   In the worst case, in the absence of coefficients on oddsportal for a certain type of bet and difficulty accessing the Pinnacle site, it is allowed to substitute the coefficients for these events with the coefficients from Bet 365, which are in the list of this event.

    All English team names and dates should be replaced with Ukrainian ones, the sample for replacing names is the site livescore.in/ua/ or flashscore.ua. However, there is one caveat. Team names on these sites are duplicated when copied, it is necessary to neutralize the duplicate to avoid parsing errors. Also, when copying, the number of red cards is fixed, more than 1, which also needs to be neutralized individually. The schedule of upcoming matches, changes, and additions of postponed meetings should also be fixed from these sites to timely determine the date of fixing the coefficients of the period.

  The pairs of teams that meet should be connected by a space and a dash. Conditionally Arsenal - Leicester. Teams are substituted without abbreviations, as they are on Ukrainian sites. 

  In case of complete absence of coefficients from Pinnacle and Bet 365 for a certain type of bet, an empty cell is fixed. In case of a postponed match - an empty cell is fixed instead of the result, and in the next cell on the right, the event is fixed as "postponed" ("stopped").

  The parsing program (conditionally Python)  should output the parsing results in an Excel table, adapted for reading and error-free transfer to version 2010 (or better directly to this version). Since transferring some formulas and signs from this version to Excel 365 and Google Sheets has previously produced many erroneous results. I don't want to throw money away.

 1X - this is the Asian handicap F1(+0.5) only the left digit;  (the right fixes F2(-0.5), that is something similar to P2, conditionally the victory of the second team) - ignore

2X - this is the Asian handicap F1(-0.5) only the right digit;  (the left fixes F1(-0.5), that is something similar to P1, conditionally the victory of the first team) - ignore

F1(-1) - only the left digit;     the right digit should be parsed to the value F2(+1) - we fix;

to F2(-1) it is necessary to parse from the value on the site F1(+1) - only the right digit.

  In short, to avoid mistakes, a person is needed who understands not only the basics of parsing but also has some experience and understanding of betting terms.

   Also immediately for ease or removal of manual labor, there is a desire to also fix the complete list of results of the matches of the round and the pairs of the next round. Necessary for inclusion in the main table for team selection, to obtain a forecast (this is already done manually).  Example:

Red color indicates the postponed match itself and the number of such matches at the moment. Blue color indicates the match(es) played in advance (prematurely). Purple color indicates the match or the number of matches (additionally) canceled during the meeting, with a later continuation.

   The program should verify the match results across several sites and in case of discrepancies either have a link in the code to some unfilled cell opposite, to enter the correct result manually, or come up with an alternative solution.

 The letters and the orange color of the cell at the bottom of the round simply indicate that the data of the meeting have been reviewed, there are no errors, and the data have been entered into the main tables (this I do manually).

  When a match is postponed, it is necessary to manually indicate next to the match time that the match is "postponed".

  But everything in order. The main thing is parsing the coefficients, the rest later, depending on the quality of the first work.

  Parsing from sites like Pinnacle is quite problematic, as the company's security service regularly works on security against free parsing. It may make sense to parse directly from its clone, with a free API, from the PS3838 site instead of Pinnacle's coefficients,  this site is a bit difficult for me to understand.