Greedy Goblin

Monday, July 31, 2017

I'm doing OK in PlayerUnknown's Battleground

The site pubgtracker.com has statistics about the playerbase, here is the solo statistics:

As the cumulative statistics shows, 2195 rating is pretty high 6% players are better than that. Why does that rating matters?

Not bad from someone who started last weekend, has the reaction time of a 40+ years old and haven't played FPS competitively since college and casually in 10 years. I must be Rambo reborn. Actually, not, 0.26 K:D is pretty lame, the same tracking site says that 90%!!! of the players are better than me. Yet, my official kill rating isn't that bad:
If #31K is top 6%, then #105K is top 18%. Not so bad, but raises two questions: how can someone with abysmal K:D (the average player has around 0.90, due to 10% death are environmental) be in the top 18% and how can someone with top 18% kill in top 6% composite rating? To the second question, the answer is the other part of the composite: win rating.

My win rate is very good. While I don't have a single ultimate win, being last survivor, but it's not needed, look how a guy got to top 10 (not 10%, top 10 players) with only 2 wins. This rating is calculated by your average position when you die. And I get into top 10% in 20% of the games and in the top half practically every time.

The answer for the previous question is that rating depends on how you trade with high rated players. So slaughtering a dozen morons gives you less rating than killing a single 2000+ player. While I have few kills, all of them are high-rated players, unless you assume that morons linger around in the top 10 to kill.

I have multiple goals:
  1. Getting to the top 10 in winrate. Yep, into the very best players of this game and yes, I find it possible. Also, you might guessed from my abysmal K:D that I'm not planning to "skillz" myself there, but using my brain.
  2. Getting to top 1% composite rating would be nice, but much harder. I don't expect my K:D to increase significantly, so my kill rating will probably not rise much.
  3. Getting a last survivor "win". I also find this harder than getting into the top 10 wins.
  4. Writing a comprehensive guide how can anyone rise to the top 5%. Which of course is only possible if "anyone" is a small subset of the players. But something tells me that the "for fun" players won't adapt my strategy, so it will keep working.
Tomorrow comes a surprising analysis on the player distribution, something similar to what I've found in BDO.

8 comments:

Malthan said...

Have you seen the recent PUBG controversy? https://www.reddit.com/r/PUBATTLEGROUNDS/comments/6q0otl/duo_partner_banned_for_7_days_for_stream_sniping/

While I disagree with your "gamedevs are corrupt if they agree to work in abysmal conditions" theory, there seems to be more mingling nowadays between devs and so called "content creators" like streamers, which let the socially connected players abuse the system by getting their opponents banned. I don't think this is done for any financial gain, it's just a byproduct of social interactions.

dobablo said...

Another one to add to the suspicions pile. Overly sensitive internet-famous streamers getting favourable treatment from corrupt developer or engaged community team going to the highest levels to stamp out cheating based on balance of probabilities?
http://www.pcgamer.com/pubg-studio-promises-changes-to-ban-system-after-stream-sniping-uproar/

dobablo said...

Is it a game? You are "winning" by declining to take part until the game has already completed your win conditions for you. When the win condition is "Do not lose" then the optimal solution is to avoid playing.

Have you ever played amateur poker tournaments, especially online ones? You can achieve similar "success" and delay elimination by playing tight and keeping your head down, but while you do that, everyone else is growing their chip pile and becoming stronger. A successful poker player needs to eliminate their weaker opponents to gain resources and to deny those resources to the opposition. This turns the optimal solution from "avoid contact" to "take controlled risks and judge when to escalate."
I think the similarities to PUBG might break down at this point because combat is not zero-sum so surviving killers may not necessarily be stronger than a hider.

Gevlon said...

@Malthan, @Dobablo: post tomorrow

@second Dobablo: I'm winning according to the win conditions of the game. These are mutually exclusive (if I occupy the #31201 position then I pushed everyone back one point by climbing there from the starting #99999999 position), so it's winning. If you think my strategy is "cheap", bring it up to Sirlin. Yes, I played poker and everyone hated me because while I never won big, no one ever won big against me. I folded when my cards were weak and when I didn't fold, I usually had good cards.

Anonymous said...

" look how a guy got to top 10 (not 10%, top 10 players) with only 2 kills."

In the image above that quote you've circled the number 2 in a column, but that column is for wins. Did you mean to say wins instead of kills there?

Gevlon said...

@Anon: yes, sorry, fixed.

Anonymous said...

Your logical fallacy is assuming you go into each game on the same playing field as everyone else.

Do you think your 20 years of gaming have given you an advantage over a teenager who wasn't born when you started gaming? If you truly have gained no knowledge in 2 decades of playing games, then, yes, you are actually starting at the bottom.

However, I would expect Method members to rise to the top quite fast in other games they play.
Outside of extreme examples, you say you havent played competitively since college...how about non-competitive?

Gevlon said...

FPS-es are typically considered twitch games where young age with good hand-eye coordination is the king and old farts can't keep up.

The last time I had pixel assault rifle in my hands were trying to play America's Army together with my GF. After a few weeks of trying she stopped and so did I. That was 8-9 years ago.