JavaScript is required to use Bungie.net

Servicio de alertas
Mañana, Destiny 2 estará fuera de línea temporalmente para recibir mantenimiento programado. Haz clic aquí para ver el Reporte de actualizaciones y servidor de Destiny.

Destiny

Hablemos de Destiny.
2/7/2017 9:49:35 PM
9

Proper trials k/d to elo ratio

One of the most common stat checking sites in the Trials of Osiris is [url]destinytrialsreport.com[/url]. This site provides two important statistics that people factor in when analyzing a trials opponent, deciding whether or not to let someone on their trials team, etc. These two statistics are ELO and kill/death ratio (for those who don't know what ELO is, here is a link to a website that explains ELO and the math behind the system: [url]https://metinmediamath.wordpress.com/2013/11/27/how-to-calculate-the-elo-rating-including-example/[/url]). Now, I'm a halfway decent PvPer, so I like to do carries from time to time. However, I am more frequently encountering players with low k/d ratios who have high ELO ratings. When I play with them, I am more likely to lose games, and when I lose games, my ELO drop is greater. In order to address this problem, I took a sample of about 1,600 guardians from [url]guardian.gg[/url], and created a quadratic regression equation that represents the relationship between ELO and k/d. This is the equation that I came up with where k/d (represented by the variable y) is a function of ELO (represented by the variable x): y=0.0000002388x^2+0.00057748x-0.24369 For those of you looking for a target ELO based on your k/d ratio, this equation may be helpful to you. There are a couple things that I took into consideration while making this equation. First of all, since ELO is a very current statistics (not very greatly influenced by past performance) I decided to take a more recent sample of k/d ratio from each player. Hence, instead of using their overall k/ds in this equation, I used their year 3 trials k/ds. Also, I tried to take a sample from various points on the ELO spectrum from about 700 to 2200 ELO (any higher than 2200 ELO, and k/ds will become to various to create a statistically accurate equation). If you have any more questions concerning the creation of this equation, please message me on Xbox live or on bungie.net @Alpha8118.

Publicando en idioma:

 

Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

  • My k/d is poop but I would say I do alright. https://youtu.be/X0AfI6D0txo But going tryhard is only so fun :/ Maybe drinking and playing PvP/trials or doing telesto only or touch of malice only etc challenges in PvP/trials isn't a great idea for k/d or elo but man it's fun... isn't that what games are designed to be?

    Publicando en idioma:

     

    Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

  • If someone's doing a regression analysis on my stats for a trials card I'll just find another team :)

    Publicando en idioma:

     

    Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

    2 Respuestas
    • I feel like there has been a massive increase in players farming Elo and KD recently. Ive been using lfg and I've had guys with 1800 Elo who have been complete shit. The week before last I must of played with 20 different players over the weekends and the majority of them were terrible in comparison to their stats. I played with 2 guys with 1850 elos. My weekly was 1.55 after 18 games and both theirs were 0.92. I obviously left and found a new team. I've also noticed that they tend to have very few flawless runs, like 10 or 15 flawless runs but spread across 3 years.

      Publicando en idioma:

       

      Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

    • Kinda how raiding parties wouldn't let people into their fireteams because they didn't have a ghally? Your low Elo or your regression analysis means you can't join my fireteam and participate in Trials with me. 😌 Good point, better accuracy on skill.

      Publicando en idioma:

       

      Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

      1 Respuesta
      • Dude, I think you need a woman to stroke your elo for you...

        Publicando en idioma:

         

        Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

      • ELO and K/D dont matter.

        Publicando en idioma:

         

        Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

      • The issue is that people cheat for Elo. One of the highest ranked Elo players in the world plays with two players who recover the accounts of really low Elo players. This means that even though he has a 3200+ both his team mates have about 600 each so according to Elo he has very little chance of winning. Meaning he gains a lot from wins and losses little from losses. Also if he comes up against really good players with high elos, he ddoses out the lowest Elo player on the opposing team. Also some classes are just better than others. In trials My hunter has a 1.57 kd. My Titan has a 1.76 and my warlock has 1.86. This is because just shield and amplitude are insanely good. Yet I sit around 1800 Elo.

        Publicando en idioma:

         

        Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

      • Editado por championkiller: 2/8/2017 3:00:40 AM
        Quite a bit off from mine, but I literally only play when I'm trying to go flawless or help my friends go, never farmed a single point of ELO...1750ish ELO with 1.8kd... your formula says with my ELO I should be around a 1.45-1.5

        Publicando en idioma:

         

        Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

      • Just to be clear, this is my alternate account, so I likely won't receive messages or comments you may have. For the sake of clarity and repetition, reach me with any questions or comments @Alpha8118.

        Publicando en idioma:

         

        Pórtate bien. Echa un vistazo a nuestro Código de conducta antes de publicar tu mensaje. Cancelar Editar Crear escuadra Publicar

      No se te permite acceder a este contenido.
      ;
      preload icon
      preload icon
      preload icon