Welcome to symthic forums! We would love if you'd register!
You don't have to be expert in bit baking, everyone is more than welcome to join our community.

You are not logged in.

Hey! If this is your first visit on symthic.com, also check out our weapon damage charts.
Currently we have charts for Battlefield 3, Call of Duty: Black Ops 2, Medal of Honor: Warfighter and Call of Duty: Modern Warfare 3

  • "3VerstsNorth" started this thread

Posts: 1,614

Date of registration
: Apr 12th 2013

Platform: PC

Location: Guilin Peaks, Finland

Battlelog:

Reputation modifier: 14

  • Send private message

1

Tuesday, September 17th 2013, 3:54pm

The RDS Experiment

Background
Background – Game Science @ Symthic
I, among others, feel that Symthic could represent “game science” as a broader concept than just gun statistics and mechanisms even though the gun stats are most lenient to rigorous analyses. As one attempt to this end here, I some time ago argued that player stats could be useful for addressing whether mouse sensitivity is relevant for the player’s accuracy. Although that analysis was left somewhat unfinished (because of work IRL), there was indeed a significant trend for high-accuracy players to use low sensitivity even though good accuracies were also achievable with high sensitivity.
The overall aim of this post is to test whether it is possible to acquire statistical evidence to support a commonly very feeling- or opinion-based choice, the choice of sights. What would be the stats that could be useful and what is the amount of data one needs to draw conclusions? Basing such choices on evidence is important because feelings often don’t account for all data and often get stuck on supporting the superiority of whatever one has become accustomed to. Evidence is also important because of two reasons. 1. Changing sights, sensitivities, etc. is expensive time-wise since it takes hours or tens of hours to learn into ‘muscle memory’, e.g., the recoil compensation matching a given combination of optic and sensitivity. Should one invest those hours into learning something new, which may or may not be beneficial, or stick to choices that already have been found to give adequate (even if not optimal!) performance? 2. As we see below, the game-to-game and even day-to-day variability in game performance is so large that without statistics, one simply cannot infer that one choice is better than another.

Background – RDS vs. 3.4x on AN-94
I have, for a while, used 3.4x instead of Holo on AN-94. In two relevant threads, Gameplay for feedback and How to master the AN-94, any tips / guidelines?, 10+ Symthicians voiced their opinion that RDS is better than 3.4x while only a single vote of support was given to 3.4x and mine doesn’t count. One opinion against another might not warrant actions but a consensus opinion with so much experience to back it up is much more difficult to bypass. Relating to the question 1. above, I’m thinking now “I’m doing fine with 3.4x but what if I could be doing even better with the RDS?”. How does one ‘know’ (as opposed to ‘feel’) that he’s doing better with one sight than the other?

Background – tl;dr
Is it possible to measure which sights or other human-interface-related options are better for a given gun than some others?

Methods
Design
I have acquired statistics covering ~1600+ minutes (across 16 days) of gameplay with AN-94 and the 3.4x scope. I also plotted the complete preceding history.

In the next two weeks, I will try to get the same amount of data for AN-94 and RDS. To minimize confounding variables, the games will be Noshahr TDM with 22-36 players. In TDM, the lack of objectives, need for strategic intellect, and commonly even squad/team-play gives a focus on individual gunplay, which should minimize confounders and highlight the differences between optics, if any are to be found. Ideally, after the RDS period, I should go back and get more data with 3.4x to exclude underlying learning-related trends. If BF4 beta is released before a conclusion is reached, idgaf anymore.

As measures of performance, I’m going to use bf3stats for recording daily KDR and KPM. Bf3battlereports or even Battlelog would be better but too much work and AFAIK they do not yield access to the far history. I think KDR and KPM are the good output measures for “killing effectivity” because KDR essentially measures the fraction of shoot-outs/encounters that the player survives, which often is dependent on how rapidly and reliably the player hits others while ADS. KPM measures the efficiency and controls that a ‘good’ KDR is not achieved by camping or staying away from action, which would decrease KPM. Deaths per minute (DPM) gives a complementary measure of survivability.

Statistics
For now, I’m just using linear regression and confidence interval estimates for visualization, Spearman’s Rank Correlation coefficient (rho) for assessing trends, and 1-way ANOVA for assessing time-window vs. time-window differences. Results with p < 0.05 will be considered significant.

On statistical tests - a long story short
Suppose one has two collections of numbers, how does one tell whether they are different? Statistical tests like ANOVA answer this question. The idea there is to inspect the distributions of these numbers in order to assess the probability that they originate from the same distribution of 'random' numbers. This is the 'null hypothesis' and here could be formulated: "KDR values are not different between 3.4x and RDS conditions". The test gives a probability value (p-value) for this null hypothesis to be acceptable. If the p value is small enough (say < 0.05), one can reject the null hypothesis and accept that KDR values "(statistically) significantly" are different between these conditions.

Results
Trends in ‘baseline’ data and implications of playing with G9x vs. EC1
Let’s first have a look at my complete history in BF3. The plots of the daily KDR, KPM & DPM as a function of cumulative gametime show highly significant increasing trends (White symbols: Data, Green line: best linear fit, Red lines: 95% confidence intervals for the linear fit). I’m dividing the track record here to three pieces:
“Past”: various gamemodes and guns (top-4: AN-94, M416, AS-VAL, SCAR-H), sights mostly Holo, sensitivities increasing from 4 to 23 cm/360 deg, mouse: Logitech G9x.
“G9x”: (almost) only TDM, AN-94, 3.4x scope, learning from comments in Gameplay for feedback thread, sensitivity 23 cm/360 deg, mouse: Logitech G9x.
“EC1”: (almost) only TDM, AN-94, 3.4x scope, still learning ;), sensitivity 36 cm / 360 deg, mouse Zowie EC1 EVO.



It is easy to see that the Past, G9x, and EC1 time windows are separated by clear jumps in daily KDR. For G9x, mean +/- standard error KDR = 2.25 +/- 0.10, and for EC1 KDR = 2.95 +/- 0.12. Even if learning effects cannot be completely ruled out, the difference is quite large IMHO and shows that not all mice are born equal. This G9x-vs.-EC1 effect is also statistically significant (p = 0.012, 1-way ANOVA). Obviously, this implies that the G9x data cannot be used when comparing 3.4x with RDS.

Did I just camp more to increase KDR with EC1? The plot of KPM and DPM shows that probably not:



KPM with EC1 (2.40 +/- 0.06) was greater than with G9x (2.15 +/- 0.07) although statistically this was only a trend (p = 0.07). DPM was smaller with EC1 (0.82 +/- 0.02) than with G9x (0.97 +/- 0.03) and this difference was significant at the level of p = 0.04. KDR and KPM were also positively correlated so that high KDRs were achieved in games with high KPM.

Hence, in comparison with G9x, EC1 gave a better KDR, more KPM and less DPM, which suggests that EC1 just gave an overall increase in killing effectivity.

Power estimates
The EC1 (3.4x) data can be used to estimate the how large of a change in KDR the RDS must produce in order to reach statistical significance. For significance level of p = 0.05, the change has to be > 0.40. To reach p = 0.01, KDR should improve by 0.56. If the effect of RDS on my performance is smaller, more data need to be acquired.


Gameplay with 3.4x
I uploaded two gameplay recordings to document the playstyle I'm using with AN-94 + 3.4x.
Short game
Long game

Results with RDS
Ok, I have now some gameplay with RDS. Total gametimes were 576 min for 3.4x and 731 min for RDS. The daily gametimes (82 +/- 10 and 104 +/- 19 min/day) were not significantly different. In essence, I sucked donkey balls with it.



There was a clear decrease in KDR from 2.95 +/- 0.11 to 2.43 +/- 0.11 and this effect was highly significant (p < 0.01).



There were no significant (p > 0.4) effects on KPM (3.4x: 2.40 +/- 0.06 vs. RDS: 2.32 +/- 0.06) but DPM was significantly higher with RDS (3.4x: 0.82 +/- 0.02 vs. RDS: 0.96 +/- 0.04, p < 0.01).

Conclusions

So, this small (7+7 days) sample of data clearly shows that RDS was detrimental to my performance both in terms of worse KDR and higher DPM. Period.

Can we interpret this as evidence for 3.4x overall being better for AN-94 than RDS? No, to make inferences at the population level, we would need data from many players (not feasible now). At the personal level, this should be corroborated with a retest of both conditions. It is still possible there just was a streak of subpar games with RDS or subpar opposition with 3.4x.

These data also span a relatively short time window: maybe I could learn to use RDS better than 3.4x or better adopt a more effective playstyle in a few more weeks. Introspectively, RDS did display signs of greatness in CQ encounters with multiple enemies but these benefits did not surpass the handicap at longer ranges (more time and bullets needed to kill, more assists and less kills, and, consequently more deaths.

Overall, this 'proof-of-concept' experiment (or 'soon-forgotten exercise in futility') shows that for making informed decisions, it is possible to assess and find game-option-related statistical differences in player stats with relatively small amounts of data.
"Less is more? How can that be? How could less be more, that's impossible. More is more." Yngwie Malmsten
"Many bullets help." WoopsyYaya
"most rhetorically legitimate ad hominem 2015" ToTheSun!

This post has been edited 7 times, last edit by "3VerstsNorth" (Sep 24th 2013, 12:13pm)


Posts: 1,553

Date of registration
: Sep 4th 2013

Platform: PC

Location: Germany

Battlelog:

Reputation modifier: 7

  • Send private message

2

Tuesday, September 17th 2013, 7:08pm

Background
Background – Game Science @ Symthic
I, among others, feel that Symthic could represent “game science” as a broader concept than just gun statistics and mechanisms even though the gun stats are most lenient to rigorous analyses. As one attempt to this end [here], I some time ago argued that player stats could be useful for addressing whether mouse sensitivity is relevant for the player’s accuracy. Although that analysis was left somewhat unfinished (because of work IRL), there was indeed a significant trend for high-accuracy players to use low sensitivity even though good accuracies were also achievable with high sensitivity.
The overall aim of this post is to test whether it is possible to acquire statistical evidence to support a commonly very feeling- or opinion-based choice, the choice of sights. What would be the stats that could be useful and what is the amount of data one needs to draw conclusions? Basing such choices on evidence is important because feelings often don’t account for all data and often get stuck on supporting the superiority of whatever one has become accustomed to. Evidence is also important because of two reasons. 1. Changing sights, sensitivities, etc. is expensive time-wise since it takes hours or tens of hours to learn into ‘muscle memory’, e.g., the recoil compensation matching a given combination of optic and sensitivity. Should one invest those hours into learning something new, which may or may not be beneficial, or stick to choices that already have been found to give adequate (even if not optimal!) performance? 2. As we see below, the game-to-game and even day-to-day variability in game performance is so large that without statistics, one simply cannot infer that one choice is better than another.

Background – RDS vs. 3.4x on AN-94
I have, for a while, used 3.4x instead of Holo on AN-94. In two relevant threads, Gameplay for feedback and How to master the AN-94, any tips / guidelines?, 10+ Symthicians voiced their opinion that RDS is better than 3.4x while only a single vote of support was given to 3.4x and mine doesn’t count. One opinion against another might not warrant actions but a consensus opinion with so much experience to back it up is much more difficult to bypass. Relating to the question 1. above, I’m thinking now “I’m doing fine with 3.4x but what if I could be doing even better with the RDS?”. How does one ‘know’ (as opposed to ‘feel’) that he’s doing better with one sight than the other?

Background – tl;dr
Is it possible to measure which sights or other human-interface-related options are better for a given gun than some others?

Methods
Design
I have acquired statistics covering ~1600+ minutes (across 16 days) of gameplay with AN-94 and the 3.4x scope. I also plotted the complete preceding history.

In the next two weeks, I will try to get the same amount of data for AN-94 and RDS. To minimize confounding variables, the games will be Noshahr TDM with 22-36 players. In TDM, the lack of objectives, need for strategic intellect, and commonly even squad/team-play gives a focus on individual gunplay, which should minimize confounders and highlight the differences between optics, if any are to be found. Ideally, after the RDS period, I should go back and get more data with 3.4x to exclude underlying learning-related trends. If BF4 beta is released before a conclusion is reached, idgaf anymore.

As measures of performance, I’m going to use bf3stats for recording daily KDR and KPM. Bf3battlereports or even Battlelog would be better but too much work and AFAIK they do not yield access to the far history. I think KDR and KPM are the good output measures for “killing effectivity” because KDR essentially measures the fraction of shoot-outs/encounters that the player survives, which often is dependent on how rapidly and reliably the player hits others while ADS. KPM measures the efficiency and controls that a ‘good’ KDR is not achieved by camping or staying away from action, which would decrease KPM. Deaths per minute (DPM) gives a complementary measure of survivability.

Statistics
To be determined.

For now, I’m just using linear regression and confidence interval estimates for visualization, Spearman’s Rank Correlation coefficient (rho) for assessing trends, and 1-way ANOVA for assessing time-window vs. time-window differences. Results with p < 0.05 will be considered significant.

Results
Trends in ‘baseline’ data and implications of playing with G9x vs. EC1
Let’s first have a look at my complete history in BF3. The plots of the daily KDR, KPM & DPM as a function of cumulative gametime show highly significant increasing trends (White symbols: Data, Green line: best linear fit, Red lines: 95% confidence intervals for the linear fit). I’m dividing the track record here to three pieces:
“Past”: various gamemodes and guns (top-4: AN-94, M416, AS-VAL, SCAR-H), sights mostly Holo, sensitivities increasing from 4 to 23 cm/360 deg, mouse: Logitech G9x.
“G9x”: (almost) only TDM, AN-94, 3.4x scope, learning from comments in Gameplay for feedback thread, sensitivity 23 cm/360 deg, mouse: Logitech G9x.
“EC1”: (almost) only TDM, AN-94, 3.4x scope, still learning ;), sensitivity 36 cm / 360 deg, mouse Zowie EC1 EVO.



It is easy to see that the Past, G9x, and EC1 time windows are separated by clear jumps in daily KDR. For G9x, mean +/- standard error KDR = 2.25 +/- 0.10, and for EC1 KDR = 2.95 +/- 0.12. Even if learning effects cannot be completely ruled out, the difference is quite large IMHO and shows that not all mice are born equal. This G9x-vs.-EC1 effect is also statistically significant (p = 0.012, 1-way ANOVA). Obviously, this implies that the G9x data cannot be used when comparing 3.4x with RDS.

Did I just camp more to increase KDR with EC1? The plot of KPM and DPM shows that probably not:



KPM with EC1 (2.40 +/- 0.06) was greater than with G9x (2.15 +/- 0.07) although statistically this was only a trend (p = 0.07). DPM was smaller with EC1 (0.82 +/- 0.02) than with G9x (0.97 +/- 0.03) and this difference was significant at the level of p = 0.04. KDR and KPM were also positively correlated so that high KDRs were achieved in games with high KPM.

Hence, in comparison with G9x, EC1 gave a better KDR, more KPM and less DPM, which suggests that EC1 just gave an overall increase in killing effectivity.

Power estimates
The EC1 (3.4x) data can be used to estimate the how large of a change in KDR the RDS must produce in order to reach statistical significance. For significance level of p = 0.05, the change has to be > 0.40. To reach p = 0.01, KDR should improve by 0.56. If the effect of RDS on my performance is smaller, more data need to be acquired.


Gameplay with 3.4x
I uploaded two gameplay recordings to document the playstyle I'm using with AN-94 + 3.4x.
Short game
Long game

Conclusions

So, we’ll see in a week or two what happens :D, I will play the first games with RDS tonight. Any comments, critique, etc. feedback is welcome. I will update this post both in response to comments and to show the data when I have it.



WOW seems like you put a lot of thought into it!
Unfortunately my stats is not as advanced as ....all the stuff you have in there basically.... :P
I know means and deviation and stuff, could you label the graphs maybe? Ok NO, wait I get it.....

the only thing I would say though, I agree, that after the new mouse, the old data becomes kind of irrelevant.
What you can do is collect new data by playing on the same server maybe 4 games in a row with the same number of players with both RDS and 3,4x.
You may need to discount deaths by C4 and crane snipers. Because when I tried to figure it out for me with an RDS (previously Holo), in the RDS game I kept getting killed by snipers, C4 and weapon tryhards (Assault 100SS, AEK100SS / Assault 100SS, M16A3 100SS / AUG 100SS etc...) So my K/D was like 3 where in the holo game it was 4. I'd go as far as saying stats can only give you a rough indicator, because as you change sights, you might change playstyle, go CQ with RDS, equally you may not be able to get of those mega long range kills with an RDS as you were able to with the 3,4x. Sight determine playstyle imho..

Edit: Since I recommended that you play on 64 man servers, let me clarify what I meant above, the problem wasnt even all the tryhards, it was my team. It was so bad, 16 guys stood underneath the crane, whilst there were enemy snipers on top. The rest kept getting killed. So I was constantly taken out from all sorts of angles and I only managed to go 3 - 1 but because so much of my team was bunched up on one area of the map, the enemy bipods and campers had a field day.... I dont ragequit often but I did it there, I was so pissed off with my team! :( But that happens rare enough so dont let one bad experience put you off your mission :D

Anyway just my 2 cents :)

Where did you get the data btw??
Moving ADS spread values

WoopsyYaya - YouTube(Tipps und Tricks) [Deutsch/HD]


Show Gun Master some love <3 - Vol. 3


Funny quotes


As a main hater on this forum, I say no, it's not worth it.
I have been playing Battlefield since 1942 (the original PC game, not the year).
now I'd jerk off a hobo
With a compensator and angled grip I click people to death like I was playing diablo.
AKU-12 stomps it in the nuts and posts the video to WorldStarHipHop.

This post has been edited 1 times, last edit by "Gero2013" (Sep 18th 2013, 2:27pm)


Posts: 766

Date of registration
: Sep 19th 2012

Platform: PS3

Location: Virginia, USA

Battlelog:

Reputation modifier: 9

  • Send private message

3

Tuesday, September 17th 2013, 7:26pm

Interested to see conclusions...

I'm another longtime unabashed proponent of the 3.4x scopes...They feel much more versatile to me, and since I no longer have the reflexes of a teenager, they give me the advantage of engaging enemies outside the usable range of the ever-popular red dot and kobra sights...

J35T3R

Punctual Joker

(74)

Posts: 1,213

Date of registration
: Mar 21st 2012

Platform: PC

Battlelog:

Reputation modifier: 6

  • Send private message

4

Wednesday, September 18th 2013, 4:30am

i did use 3.4 pka for about 3000 kills with an 94 and then swapped to holo as it was more forgiving of a miss
PC: Platform: PC





youtube chanel HIPFIREGAMER - YouTube

  • "3VerstsNorth" started this thread

Posts: 1,614

Date of registration
: Apr 12th 2013

Platform: PC

Location: Guilin Peaks, Finland

Battlelog:

Reputation modifier: 14

  • Send private message

5

Wednesday, September 18th 2013, 9:20am

Thanks guys!

I know means and deviation and stuff, could you label the graphs maybe? Ok NO, wait I get it.....


I'll add a little appendix on the basics of statistics and stuff that might not be commonplace to everyone here.

You may need to discount deaths by C4 and crane snipers. Because when I tried to figure it out for me with an RDS (previously Holo), in the RDS game I kept getting killed by snipers, C4 and weapon tryhards (Assault 100SS, AEK100SS / Assault 100SS, M16A3 100SS / AUG 100SS etc...)


Discounting deaths will not be possible because I'm not recording every game. I'm trying to acquire the statistics 'en masse' (hopefully up to 1000 gameplay minutes) to even out spurious events, C4 douchebags, and specific games that were very tough or very easy.

I'd go as far as saying stats can only give you a rough indicator, because as you change sights, you might change playstyle, go CQ with RDS, equally you may not be able to get of those mega long range kills with an RDS as you were able to with the 3,4x. Sight determine playstyle imho..


Yep, the playstyle change seems quite unavoidable. Already yesterday I noticed that (obviously) short-range engagements were easier with RDS and conversely I was really struggling more with reliably (first-shot) hitting M16 headglitchers at range even if long-range hitting was easier than I expected.

But then, if one choice of sights leads to a playstyle that is more effective than another in terms of the raw outcome, then I think it can be attributed to the overall "effect of sights".

Where did you get the data btw??


I'm using the 'daily' listing of bf3stats.com history that I have populated fairly regularly.

I'm another longtime unabashed proponent of the 3.4x scopes...They feel much more versatile to me, and since I no longer have the reflexes of a teenager, they give me the advantage of engaging enemies outside the usable range of the ever-popular red dot and kobra sights...


This. I'm no teenager either. Given the G3 rape I received yesterday, I'm not sure if I'm going to be able to make the RDS work...

i did use 3.4 pka for about 3000 kills with an 94 and then swapped to holo as it was more forgiving of a miss


Interesting!
"Less is more? How can that be? How could less be more, that's impossible. More is more." Yngwie Malmsten
"Many bullets help." WoopsyYaya
"most rhetorically legitimate ad hominem 2015" ToTheSun!

C0llis

Up and down. Bounce all around

(3,334)

Posts: 3,100

Date of registration
: Apr 15th 2013

Platform: PC

Location: Sweden

Battlelog:

Reputation modifier: 15

  • Send private message

6

Wednesday, September 18th 2013, 9:58am

Isn't tracking accuracy also somewhat relevant?

Things people said

And reading Youtube comments still gives me Turbo Cancer.

It really is quite frustrating when Helen Keller sets up her LMG in the only doorway in/out of an area.

What kind of question is that? Since when is cheese ever a bad idea?

Hardline is a fun and sometimes silly Cops and Robbers sorta thing and I think that's great. Or it would be if it didn't suck.

  • "3VerstsNorth" started this thread

Posts: 1,614

Date of registration
: Apr 12th 2013

Platform: PC

Location: Guilin Peaks, Finland

Battlelog:

Reputation modifier: 14

  • Send private message

7

Wednesday, September 18th 2013, 10:14am

Isn't tracking accuracy also somewhat relevant?

For mice I presume? I doubt that the hardware/sensor-related difference in G9x vs. EC1 is large enough to show in gameplay (at least at my level). With G9x I never observed non-smoothness (although, of course, that might not be prevalent enough to be perceptible and yet influence performance => shows up in statistics). One of the main reasons for the superiority of EC1 IMHO is simply that it is so lightweight and well gliding that fast movements can be made accurately and without overshooting.
"Less is more? How can that be? How could less be more, that's impossible. More is more." Yngwie Malmsten
"Many bullets help." WoopsyYaya
"most rhetorically legitimate ad hominem 2015" ToTheSun!

Posts: 230

Date of registration
: May 30th 2012

Platform: PC

Battlelog:

Reputation modifier: 5

  • Send private message

8

Wednesday, September 18th 2013, 12:29pm

Isn't tracking accuracy also somewhat relevant?

For mice I presume? I doubt that the hardware/sensor-related difference in G9x vs. EC1 is large enough to show in gameplay (at least at my level). With G9x I never observed non-smoothness (although, of course, that might not be prevalent enough to be perceptible and yet influence performance => shows up in statistics). One of the main reasons for the superiority of EC1 IMHO is simply that it is so lightweight and well gliding that fast movements can be made accurately and without overshooting.

Didn't surprise me much that you did better with the EC1 because it has a sensor without inherent acceleration, unlike the G9x. The A3090 is just better for FPS imo. Regarding the 3.4x vs RDS I can't really help, because my sens is already so low that using a higher magnification scope makes close to mid range fights terrible. I only used it on bolt action rifles so far, but prefer the Acog there now.




"I reject your reality and substitute my own"

C0llis

Up and down. Bounce all around

(3,334)

Posts: 3,100

Date of registration
: Apr 15th 2013

Platform: PC

Location: Sweden

Battlelog:

Reputation modifier: 15

  • Send private message

9

Wednesday, September 18th 2013, 12:47pm

Isn't tracking accuracy also somewhat relevant?

For mice I presume? I doubt that the hardware/sensor-related difference in G9x vs. EC1 is large enough to show in gameplay (at least at my level). With G9x I never observed non-smoothness (although, of course, that might not be prevalent enough to be perceptible and yet influence performance => shows up in statistics). One of the main reasons for the superiority of EC1 IMHO is simply that it is so lightweight and well gliding that fast movements can be made accurately and without overshooting.
No, I meant for scopes. It would be interesting to see if there was any noticeable difference in your accuracy when using scopes with different magnification.

Things people said

And reading Youtube comments still gives me Turbo Cancer.

It really is quite frustrating when Helen Keller sets up her LMG in the only doorway in/out of an area.

What kind of question is that? Since when is cheese ever a bad idea?

Hardline is a fun and sometimes silly Cops and Robbers sorta thing and I think that's great. Or it would be if it didn't suck.

  • "3VerstsNorth" started this thread

Posts: 1,614

Date of registration
: Apr 12th 2013

Platform: PC

Location: Guilin Peaks, Finland

Battlelog:

Reputation modifier: 14

  • Send private message

10

Wednesday, September 18th 2013, 1:13pm

Isn't tracking accuracy also somewhat relevant?

For mice I presume? I doubt that the hardware/sensor-related difference in G9x vs. EC1 is large enough to show in gameplay (at least at my level). With G9x I never observed non-smoothness (although, of course, that might not be prevalent enough to be perceptible and yet influence performance => shows up in statistics). One of the main reasons for the superiority of EC1 IMHO is simply that it is so lightweight and well gliding that fast movements can be made accurately and without overshooting.
No, I meant for scopes. It would be interesting to see if there was any noticeable difference in your accuracy when using scopes with different magnification.

Hmm, yes, tracking accuracy would definitely be relevant. But how to measure/quantify it? Or you mean accuracy as the game stat? I would consider it but there seems to be no access to my accuracy history! Bf3stats history doesn't show it. Battlelog shows only max 16 latest games and so does BF3 Battle Reports (that doesn't seem to work anymore anyway). Of course I can manually log the game-by-game accuracies from battlelog for the upcoming RDS data, but then I would need to do the 3.4x condition again because all but the last 7 of the the old 3.4x games are lost already.
"Less is more? How can that be? How could less be more, that's impossible. More is more." Yngwie Malmsten
"Many bullets help." WoopsyYaya
"most rhetorically legitimate ad hominem 2015" ToTheSun!