Good morning, hltv enjoyers! I've made some changes in my sheets so predictions to be more accurate. Calculated top-20 based on pre-blast fall stats and results occured very surprising for me. docs.google.com/spreadsheets/d/1XV0vGN9S.. What's new? 1. Added playoff rating As was mentioned by AdreN_1_major_adreN_0_major and shoom123 at previous thread, playoff rating is one of the most important stat in top-20 rating, so I calculated playoff rating for the most relevant teams. One page per team. Included: ENCE BIG NIP Moviestar Riders (for sunpayus) Cloud9 aka Players aka Gambit Furia NaVi Liquid Vitality MOUZ Astralis Heroic Outsiders aka VP FaZe G2 How I did math: I made some experiment to understand how HLTV calculates rating. Summed all maps ratings and divided by maps amount and the result occured incorrect. Then I tryed to multiply every map rating to amount of rounds in map and after it divided by total amount of rounds in series and result occured to be very close to original hltv rating. I've understood that I dont want to do this much calculations for each match so final decision was to multiply match rating with maps amount and sum up for all matches, after that just divide by total map amount. This seemed for me kinda logical because that way I got mean rating of player for entire year. But had to admit: it may be inaccurate. 2. Added points system instead of raw rating comparison. Thanks to S1mple2002 who gave me an idea to calculate points for each parameter. 3. Added points correction based on map amount Correlation between points amd maps played because more maps played = more stable perfomance or more mean rating. For example frozen got 1,24 rating against top-5 teams at big events, but he played only 4 maps what is extremely low. Thats why final points should be affected by played maps amount. How I did math: Final formula looks like this: Points = (Rating/max rating in table)^2,5 * (maps/max maps in table)^0,3 I've made exponential correlation for rating to make difference between extraordinary and mediocre rating more significant. 0,3 power in maps was added to make map correlation not this much impactfull. For example if set power to 1, faze players immidiately pop off to the top cuz they played the most maps in 2022. 4. Calculated MVP/EVP rating All calculations could be find at "Team rating" page Set different amount of points for mvp and evp at A, S and major-tier. Points could be adjusted automatically be changing points foe some category. 5. Added weights for each parameter What's next? Wait for global final results, update stats and play with weights Feel free to share your thoughts on my work! What should I add or adjust?