What are the odds?

TyDollarSignTyDollarSign Member Posts: 239 ★★


I lost the match but funny to have the same exact score but get the win.

Comments

  • TyDollarSignTyDollarSign Member Posts: 239 ★★
    @ahmynuts ha, I understand - I've just never had a score that close before.
  • MasquiradeMasquirade Member Posts: 163
    I mean you say decimal numbers but it's kind of hard to account for with the parameters we have? It might be that, in the event of a draw, it prioritises speed of the kill rather than health score, very interesting case though, crazy to get the exact same score haha.

    @RichTheMan
  • DNA3000DNA3000 Member, Guardian Posts: 19,658 Guardian
    This is actually a useful one to save. Someone recently posited the theory that when there's an apparent tie, the game uses remaining attacker health as a tie breaker. I'm pretty sure that's not true, but quite a lot of the "visual" ties happen when both sides end up at full health, and coincidentally a lot of the screengrabs of wins due to round off happen to have the winner have higher health. This proves that when the score appears to be equal, the game doesn't revert to handing the win to the player with the higher attacker health.

    This demonstrates the current best theory, which is that when it appears to be a tie but someone actually wins, it is just a round off discrepancy.
  • PikoluPikolu Member, Guardian Posts: 7,732 Guardian
    DNA3000 said:

    This is actually a useful one to save. Someone recently posited the theory that when there's an apparent tie, the game uses remaining attacker health as a tie breaker. I'm pretty sure that's not true, but quite a lot of the "visual" ties happen when both sides end up at full health, and coincidentally a lot of the screengrabs of wins due to round off happen to have the winner have higher health. This proves that when the score appears to be equal, the game doesn't revert to handing the win to the player with the higher attacker health.

    This demonstrates the current best theory, which is that when it appears to be a tie but someone actually wins, it is just a round off discrepancy.

    Would be nice if the health scoring would round up instead of having the decimal places. For example: 7* R2 Bishop with 67451 HP was lowered to 54,462 health (which is about 84.501%) and then it takes that number and multiplies it by 15,000 to give 12,675.21 points. As we can see here, Bishop got more points than that because he actually had closer to 84.5267% of his health, but the game truncates (or maybe rounds, would need more data to test) the % of health remaining. The game does the same with the points given for HP as well.

    Back to my original point, odds are that the Bishop had close to 57,015/67451 health (84.528%) then they would get 12679.204 points and the game will just truncate it visually to 12,679 points while including the decimal internally. Would be nice if the game would take the 12679.2 and round it up to 12,680 points just to avoid confusion. The marginal difference in HP would be literally 3 hp in this case. Only 57,015-57018 have a point value in 12,679. Wouldn't be game breaking, but could potentially give more ties.
  • PikoluPikolu Member, Guardian Posts: 7,732 Guardian
    To add to my previous post, here is an example of the decimals adding the total score to 1 higher than what was shown thanks to unseen decimals.


  • SummonerNRSummonerNR Member, Guardian Posts: 12,770 Guardian
    Pikolu said:

    To add to my previous post, here is an example of the decimals adding the total score to 1 higher than what was shown thanks to unseen decimals.


    Ah, yes. The original pic at top had both sides each adding up exactly to what looks like a tie.
    Whereas in the one you just provided, if you added up each “rounded/displayed” values on the left side player, they DO NOT add up to their total at the bottom.

    So can assume the un-rounded values (for Attack and Defense HP) each are internally (fully calculated) a number with ending parts of for example 49.3 and 80.3 (each one rounded DOWN for their respective categories of points).

    But when the fractional parts of 49.3 + 80.3 are added up for the Total points, the number thus ends in 129.6 which gets rounded UPWARDS to 130

    Evidence that it maintains the full fractional values all the way through to final addition.
Sign In or Register to comment.