27 Replies Latest reply: Feb 4, 2013 7:40 PM by Harv72 RSS

Actual statistics, not speculation...

My friend and I tried an experiment.  For five nights we played a private party of hardcore Free-For-All.  Game rules were default Free-For-All with Hardcore enabled.  He is located in Maryland, I am in Nevada.  He always hosted.  Bots were enabled and set to Regular, with 8 bots on Nuketown and Hijacked and 10 bots elsewhere.  Each night we played all 15 maps.

 

Night 1:  Me- 8 wins, Friend- 7 wins

Night 2:  Me- 8 wins, Friend- 7 wins

Night 3:  Me- 6 wins, Friend- 9 wins

Night 4:  Me- 7 wins, Friend- 8 wins

Night 5:  Me- 7 wins, Friend- 8 wins

 

We used the same Class and the same perks for every game played.  We never altered our playing style, we didn't try to skew the stats.  We just played and recorded the results. 

 

I am not a mathmatician but I would say that it would be very rare statistically for us to play 75 games and end up with this even of a split, without a "player catchup" type deal in the programming.  These stats simply reinforce my opinion that it is not a lag issue in online play, but rather a calculated readjustment of hit detection to keep most players on an even keel, which is a big pile of steaming crap. 

 

They can keep their new map packs and they can keep any future versions of Call of Duty.  Any game of this nature which is not based purely on player skill, but is instead "rigged for comfort" should be sold as such.  No where on the lable does it say that this game is, to use a golf term, handicapped.  Again, a big steaming pile of crap.

 

I encourage each of you to try the same experiment for yourselves.

  • Re: Actual statistics, not speculation...
    ghamorra

    I've posted a thread stating such claims. Nothing like a conspiracy or anything but a thread that addressed some interesting coincidences that occur. Like a guy who has the lowest K/D in the lobby over the course of several matches and every time killcam comes up I'm left saying "that's impossible" or "that's nowhere near what happened"

  • Re: Actual statistics, not speculation...
    BukLau101

    Hmmm suspicous things always seem to happen when I'm on a roll. Here is a good example, just yesterday I was playing KC on Raid and was one kill from a K9 unit when some noob who I'd repeated killed managed to kill me. It was a point blank range engagement and I was using the Skorpion EVO (The SMG with the fastest TTK at close range) and he was using the M27 (one of the slowest killing weapons in the game) I shot first but didn't get any hitmarkers on him, this was at point blank range so there was no way I could be missing my shots and I was on the good side of connection that game so it couldn' t have been lag. Weird things happen not all the time but just enough to make me wonder...

  • Re: Actual statistics, not speculation...
    josh1073

    I mostly have the issue i shoot people get hitmarkers i just drop dead but the kill cam says i dont shoot.Cods main audiences is under 16 we all know this so a system that gives them a boost now and again would not surprise me. Kinda explains why my k/d has dropped alot in bo2.

  • Re: Actual statistics, not speculation...

    Further information regarding this experiment, now that I've had some time to run more numbers on our games and discussing the results.

     

    Games were played to 30 kills, no bonus points.  One of us would be on a tear, pulling ahead by a number of kills and suddenly would not be able to kill anyone and being killed with ease.  That person would also start spawning directly in front of enemy players 2-3 times in a row.  As soon as the kill counts, regardless of the number of deaths, were very close equal it was back to normal play again. It seems that Bots do not figure into this, only live players.  We will not be attempting this with 3 friends and bots to test because I couldn't give a sh#! at this point.

     

    The telling stats: Overall my KDR for the 75 games was a 2.93 vs my friends KDR of 1.84 (both rounded to the nearest hundredth) To calculate, we recorded the number of kills and deaths from each game.  An astounding 76% of the time (57/75) the final score was 30-29 between the two live players.  Only 4 times did a bot finish 2nd, and a bot never had a win.

     

    Again, this is not a lag compensation issue, it is a handicapping system built into the mechanics of the game to keep all players on a level playing field regardless of their abilities.

     

    Again, I encourage you to TRY THIS EXPERIMENT FOR YOURSELVES. 

  • Re: Actual statistics, not speculation...
    monkeylord

      Or playing with bots could be the issue as in the bots might be programmed to stop the match being to one sided. Sometimes i'll play on the local option KC with the mr's and if one of us gets to far in the lead the leaders bots will just start ignoring tags and will stand out in the open waiting to be shot.

      Also to see if there is a balancer code at work you'd have to play against someone whos "skill" level is way different to yours and see if suddenly the better player is losing more to the worse player than they should.

      It's best to accept some crazy crap happens in COD from C4 detonating in someones face without effecting the guy but killing a person behind him to spawns dropping you infront of people already firing and just get on with the game.

  • Re: Actual statistics, not speculation...

    Wow...so over an enormous statistcal sample of 5 instances, you've got your proof.  Move over, Dr. Hawking.

    • Re: Actual statistics, not speculation...

      Re-read the first post.  It was over five nights, playing all 15 maps each night, so 75 games.  Not Stephen Hawking level analysis, but good enough for me. 

      • Re: Actual statistics, not speculation...
        Eye_Gore

        Though knowing one is in an experiment, is not good for the experiment. Results will not be accurate.

        • Re: Actual statistics, not speculation...

          doesnt matter, multiplayer does have handicap stuff in it for the lil kids.....if you think they are THAT good, just challenge them toa game of grief and see just how much they fail against something that CANT SHOOT BACK

        • Re: Actual statistics, not speculation...

          In this, you'll just have to trust me and say that we played each game as well as we were capable.  Also note, we did not alter our playing styles.

          • Re: Actual statistics, not speculation...

            But you're not breaking down your evidence by 75 games, you're doing it by 5 nights.  You are using only two players as your sample, and the results are flawed both by the small sample size and by the fact that you are obviously familiar with each other's style of play. 

             

            If your premise were true, individual statistics in the game would trend toward a k/d of 1 (ignoring the rather obvious point that the emphasis should be based on winning matchs, not on improving one's own stats).  Instead, you have a number of players maintaining k/ds well above 2, and plenty who cannot even approach 1.  You have plenty of complaints that the game allegedly stacks on side in team-based games, which would not be the case where your grand conspiracy theory correct.  You have numerous instances, both team-based and individual, where a single player utterly dominates the lobby for as long as they remain there.  And, finally, you fail to acknowledge the possibility that you and your friend are simply evenly matched players (despite your "telling" k/d ratios from the sample) or that the issue could potentially be affected by the inclusion of bots in the games.

             

            My own experiences with the game in no way suggest that it is "handicapped", beyond the skill-based portion of the matchmaking system.  The overall stats, built off of millions of samples, agree with my experiences.

            • Re: Actual statistics, not speculation...

              But each of the five nights we were in the same "lobby" the entire time.  Additionally, when playing hardcore Free-For-All, being familiar with the others style of play isn't a factor.  Perhaps it would be in Team Deathmatch or one of the other game modes.

               

              My premise is that the game works to keep the players even during the life of the lobby, not necessarily the game.  One game you're a god, the next a target, sometimes this happens within a game.  KDR has nothing to do with it, aside from showing a disparity between the players skill; it's the number of kills alone, which in this case also dictated the win/loss ratio as there were only two live players involved.  As I suggested, at this point a third live player should be added, a bot removed and the test repeated. 

               

              I would also venture to say that a vast majority of the players you see with a high KDR, or no complaints about the game, are not staying in a lobby long enough to suffer the effects.  I found when playing in public forums that if I joined a lobby in progress, played the game until the end, left the lobby, joined another game in progress, played until the end, left the lobby, (repeat ad nauseum) I rarely suffered the effects of what I'm referring to as a handicapping system.  Sometimes you can get away with it two games in a row in the same lobby, but rarely three.  Join a lobby for Team Deathmatch as a clan, stay in the same lobby and watch your numbers fluxuate within the clan.  Again, god followed by target.

               

              I would argue that it is particularly evident when only two live players are represented and these two players remain consistant.  Players leaving and entering a lobby forcing a host change should have an obvious affect on the calculations as it may calculate as a new lobby, but in a consistant lobby you would see ups and downs in your individual game play even though the connection and host remain constant, i.e. hit markers and no kill.  Reference the hit detection/lagging topics.

               

              I'll admit, maybe I'm wrong but you'll have a hard time convincing me after our little experiment.  I'd say if you're happy with the game as it is, forge onward and enjoy yourself.  For me, I wanted to post the results and I'm done with the series.  I don't like having my play style handicapped.  I'm all for putting me in a lobby based on player ranking, but let me play without a choke. 

               

              Crysis 3 comes out on the 19th, Tomb Raider on March 5 and I have Skyrim until then.  Enjoy yourselves.

              • Re: Actual statistics, not speculation...

                The game is far from perfect, but I do feel like I've gotten my $60 worth (and then some) out of it.

                 

                Again, my experience is that I rarely fluctuate that wildly when in the same lobby & against (mostly) the same opponents.  I've seen nothing at all to indicate to me that there's some grand conspiracy going on to level the playing field; that would actually be directly contrary to Activision & Treyarch's best interests, as they benefit greatly from You Tube CoD videos & competitive gaming...if the game itself was trying to even things out, it would be that much more difficult for the big You Tubers to get their gameplays, and the MLG guys would notice.

                 

                It really boils down to my central point: your statistical sample size is meaningless when trying to draw such a broad conclusion.  You are sampling only two players out of the well over 1 million who purchased this game, and your total of 75 matches is outdone many times over every single minute of the day in public lobbies.  You would need thousands, if not tens of thousands, of matches & a similar number of players before you could draw any sort of statistical analysis from the results.

                 

                Incidentally, knowing your opponent is more important in HC FFA than any other mode.  You know which routes he tends to favor, which locations he wants to hold down, and how he likes to approach his engagements...in a game mode where it's everyone for themselves and you're both basically competing to kill the most bots, that's huge.  Even more so when you can often kill your opponent with 1 or maybe 2 shots.

  • Re: Actual statistics, not speculation...

    This doesnt prove anything.

     

    It may only prove that you two are evenly matched.

  • Re: Actual statistics, not speculation...

    I've blown C4 under someones feet, and not only did it not show flak jacket, I didn't even get a hitmarker. My friend did the same thing in the final killcam the other day, where the guy ran right over the C4 blast and my friend had to shoot him instead. The game is a total scam.

  • Re: Actual statistics, not speculation...
    BukLau101

    Interesting post but I don't think your experiment really proves that much. Anway I think any kind of handicap system would have been found in the game's code by now.

  • Re: Actual statistics, not speculation...

    Wow. I started thinking this last night and then I check the forums this morning and it's in the top of the list!

     

    I had a couple of games where I could tell I was on the good side of the lag comp in the first couple of engagements. I went 8-0 on my first life, then all of a sudden I'm getting hitmarkers with my Remmington from 2-3ft away. Then I was pegged back to 8-5 and again and I get the good side of lag comp and will start winning one-on-ones to end 19-10.

     

    Thats a good game for me, but still, the fact I had such a great start to then have all the lag comp switch the other way is so annoying!

     

    You just can't trust your gun - or even your eyes sometimes!

  • Re: Actual statistics, not speculation...

    This is why I always wear my tinfoil hat when I play.