Jump to content


This could be a math question or a MM conspiracy question. ( We will see were it goes!)

session win rates statistics

  • Please log in to reply
23 replies to this topic

gpsjake #1 Posted Sep 15 2018 - 04:48

    Corporal

  • Players
  • 23803 battles
  • 82
  • [B-O-W] B-O-W
  • Member since:
    02-21-2014

 

Looking for some understanding of session win rates.  

So, overall I am right at a 50% win rate and right at 52% for my last 1000 games or so.  I have no problem with overall win rate, I feel like overall it reflects how much I contribute to wins or losses.

When I hang out in tiers six or seven my win rate responds positively compared to higher tiers, so I get the connection to overall win rates. My performance is better in the lower tiers so I win more.

What I do not understand is why my session win rates are so very random. It is constantly an all or nothing outing as far as wins or losses. My WN8 stays fairly level but I get 1-9 nights followed by 9-1 nights regularly. It is rare when I end a session around 50%. I typically try to play 10-15 games in a session if I can. 

If I am fairly consistent shouldn't my average sessions be around 50% more often than not.  I feel like I understand statistics and I understand small sample sizes, but it still seems like my mean ought to be centered more often than not. 

 

So jumping from math to the conspiracy stuff....

I have read on these forums that War Gaming does have some kind of patent on an MM that raised and lowered quality of teams/ matches based on recent win rates or performance of players. I am not saying this is happening, but is there any credence to this?

 

My question is....Am I looking at this wrong? Since it works out overall does that mean its purely due to the small sample sizes of nightly sessions? or does this seem a little wonky from a math/statistical viewpoint? ( Seems wonky to me)

 

Also since I know how some of these discussions go... I am aware that my overall stats may say that I should have a lower win rate than I do. This is due to my first 15,000 or so games being on satellite internet with 600+ping. I have substantially raised my stats/gameplay in the last 8,000 or so games. Judge me on more recent efforts than on my early stuff.

 

Thanks for your input ahead of time.

 



SnakePlisskenn #2 Posted Sep 15 2018 - 04:57

    Corporal

  • -Players-
  • 75 battles
  • 67
  • Member since:
    03-27-2016
Even if WG has a program to try and level everyone's win rate to 50% what would be the one factor to off set that attempt? The account users skill or lack of skill. No matter how much a program trys some players it cant elevate above 43% and some it cant push down below say 55%...numbers are just an example. I would not worry about it at all because if such a thing exists, it exists for all, and the effectiveness of it obviously doesn't work very well.

Bob_5000 #3 Posted Sep 15 2018 - 04:58

    Staff sergeant

  • Players
  • 33547 battles
  • 308
  • [-ZOO-] -ZOO-
  • Member since:
    11-12-2013

No this game is not rigged, dont listen to the bads. Stay positive game to game, do what you can. Any sort of match ups are not predetermined results. Trust me, I've seen many after battle results where bad players (statistically) did good in a match. And there is no rigged mm that can anticipate that. 

 

Only mind set you need is to constantly improve and that match wins are up to you. As far as skill levels, assume every enemy knows what they are doing, and assume every friendly doesnt. From there make your game a win. 



rockbutcher #4 Posted Sep 15 2018 - 05:05

    First lieutenant

  • -Players-
  • 18569 battles
  • 601
  • [69CBR] 69CBR
  • Member since:
    03-06-2015

There have been a number of very good posts in the past by people who understand statistics (the science not the in-game) better than I, where they presented thousands of recorded games and showed how it all worked out.  In most of those posts, they questioned the probability of win/losing streaks of different lengths.  I think that the main "take away" from those posts showed that if you simply haul your weight, no more no less, then your results will fall into the mean.

 

 

 

That said, flipping heads or tails 1000 times shouldn't result in numerous "10 tails" streaks (win or loss) the way it seems to happen here.  I don't know the reason why it does, but I roll with it.

Happy tanking and don't sweat the stats.


 



Nunya_000 #5 Posted Sep 15 2018 - 05:08

    Major

  • Players
  • 21033 battles
  • 12,785
  • [PACNW] PACNW
  • Member since:
    09-20-2013

View Postgpsjake, on Sep 14 2018 - 19:48, said:

 

Looking for some understanding of session win rates.  

So, overall I am right at a 50% win rate and right at 52% for my last 1000 games or so.  I have no problem with overall win rate, I feel like overall it reflects how much I contribute to wins or losses.

When I hang out in tiers six or seven my win rate responds positively compared to higher tiers, so I get the connection to overall win rates. My performance is better in the lower tiers so I win more.

What I do not understand is why my session win rates are so very random. It is constantly an all or nothing outing as far as wins or losses. My WN8 stays fairly level but I get 1-9 nights followed by 9-1 nights regularly. It is rare when I end a session around 50%. I typically try to play 10-15 games in a session if I can. 

If I am fairly consistent shouldn't my average sessions be around 50% more often than not.  I feel like I understand statistics and I understand small sample sizes, but it still seems like my mean ought to be centered more often than not. 

 

 

 

We get streaks because the outcome of battles are often reliant on how well our team plays, as well as our own play.  Some nights we get teams that work real well together and some nights we don't.....and it may not even have anything to do with the skill level of the players we are sided with.  Even our own play fluctuates from day to day.  Mind not into the battles.  Tired after a hard day at work. etc.  One mistake at the wrong time could make the difference between dying in the first few minutes of a battle or have a great game.

 

Block Quote

So jumping from math to the conspiracy stuff....

I have read on these forums that War Gaming does have some kind of patent on an MM that raised and lowered quality of teams/ matches based on recent win rates or performance of players. I am not saying this is happening, but is there any credence to this?

 

The patent does not say that.  It says that a player may be high tier more often if they have been losing and low tier more often if winning.  It says nothing about the skill quality of the teams we are placed on.  It is even questionable if that part of the patent is still in play after the 3-5-7 format was introduced....if it ever was.

 


8_Hussars #6 Posted Sep 15 2018 - 05:14

    Staff sergeant

  • Players
  • 52014 battles
  • 343
  • [SHIRE] SHIRE
  • Member since:
    09-17-2013

View Postgpsjake, on Sep 14 2018 - 20:48, said:

If I am fairly consistent shouldn't my average sessions be around 50% more often than not.  I feel like I understand statistics and I understand small sample sizes, but it still seems like my mean ought to be centered more often than not. 

 

Player consistency is a myth more often than not (and that is a further reason SBMM wont work as advertised).

Confirmation bias means players think they are consistent but they rarely are.  Humans are wired to see patterns where none exist especially over short sessions of small samples.

Edited by 8_Hussars, Sep 15 2018 - 05:16.


death_stryker #7 Posted Sep 15 2018 - 06:32

    Sergeant

  • -Players-
  • 4575 battles
  • 204
  • Member since:
    05-23-2014

I don't believe in the MM conspiracy, and so I will try to explain this mathematically. 

 

MATH WARNING

 

In theory, we can calculate the probabilities of each combination of wins and losses in a session occurring, by treating each match as a coinflip (i.e. 50% winrate). This simplifies the math substantially. Following this line of thought, the chance of a session of 10 games going 1-9, assuming 50% winrate, is (0.5^10)*C(10, 1) = ~0.00977 or ~1% (Here C(n,k) denotes the combination operator). The chance of going 9-1 is the same as this (~1%). The chance of going 5-5 is (0.5^10)*C(10,5) = ~0.246 or ~24.6%.

 

This means that, theoretically, about 1 in 4 sessions will end 5-5, but that only 1 in 100 sessions will end 1-9 or 9-1. This is obviously not the case, so there must be something going on, right?

 

HOWEVER

 

You may not have noticed, but in these calculations (that illustrate your points), I have been forced to make the same mistake you made in your post, OP. This is because of a concept in statistics known as variance, which I have totally ignored, in order to make my calculations possible. You'll see why soon.

 

Every data set, in this case the set of nightly WoT sessions, has an intrinsic property (expressed as a numerical value) called its variance. This is, in essence, how much each data point varies from the mean, or expected, value. In our case, this expected value is 50% winrate. Crucially, however, we do not know the variance of this data set. What this means is that we do not know how much each nightly session is expected, in theory, to vary from the expected 50% winrate. Here's a thought experiment to illustrate my point:

 

You've just lost 9 games in a row. You're so tilted you're falling out of your chair. In the next game, do you expect to perform as well as you usually do?

 

So, what does this mean? It means that, empirically, people tend to perform worse (and thus win less) after long losing streaks. In other words, the more you vary from 50% winrate (at least negatively), the more you tend to vary even more. This means that the variance of a nightly WoT session is higher than that of a (fair) coinflip, resulting in 5-5 sessions being rarer than expected and 1-9 sessions being more common than expected. However, without a very large sample of battles, probably so large that it would have to come straight from WG, we cannot actually calculate the numerical value of the variance, and thus cannot make any remotely accurate calculations.

 

 

Boy did this post take a while to write.

 

*Edit* Fun facts: Looking at the posting times, I estimate that this post took over an hour and a half to write. During that time, I refuted at least two of my own logical fallacies, in order to make my wall of text actually logically sound, instead of trying to make it appealing as so many do. At this point I think my reasoning is pretty tough to invalidate, though whether the effects I've discussed are statistically significant is anyone's guess.

 


Edited by death_stryker, Sep 15 2018 - 06:40.


xrays_ #8 Posted Sep 15 2018 - 06:41

    Major

  • Players
  • 48673 battles
  • 3,803
  • [FELOW] FELOW
  • Member since:
    08-02-2013

View Postgpsjake, on Sep 14 2018 - 22:48, said:

I typically try to play 10-15 games in a session if I can. If I am fairly consistent shouldn't my average sessions be around 50% more often than not.

I feel like I understand statistics and I understand small sample sizes, but it still seems like my mean ought to be centered more often than not.

 

Regardless of how you play, the matches in this game are influenced by the other 14 players on your team and the 15 players on the enemy teams. The fact that you bounce around from night to night seems to confirm randomness more than a system designed to control win rates, seeing as a controlled system might be more predictable.

 

I did a report a while back on flipping a quarter to simulate randomness and streaks that occur naturally. If you look at the results, you'll see that the daily heads/tails actually mimic an average player's potential win/loss tabulations. Being that the quarter test only shows one factor it is far from a comparison to this game, but it does show some interesting results (namely the few days with wild streaks).

 

Quarter Test: http://forum.worldoftanks.com/index.php?/topic/555032-rigged-matches-the-quarter-test/

 

In the end, the better a player is, the more wins they will likely receive. However, even the best players can have bad losing streaks. Two weeks ago I had a 10-match losing streak, but the next day I had both a 12-match and 15-match winning streak, almost back to back. Crazy happens in small samples. Trust your 1,000 match averages to indicate a better picture of your progression.

 

x.



NeatoMan #9 Posted Sep 15 2018 - 12:27

    Major

  • Players
  • 25918 battles
  • 17,823
  • Member since:
    06-28-2011

View Postgpsjake, on Sep 14 2018 - 22:48, said:

My question is....Am I looking at this wrong? Since it works out overall does that mean its purely due to the small sample sizes of nightly sessions? or does this seem a little wonky from a math/statistical viewpoint? ( Seems wonky to me)

Since you haven't done any kind of statistical analysis you can't really say how wonky it is.

 

Sample size will be one cause of a wide variation of win rates.  Theoretically if everything else is the same from session to session, you would see more 5-5 results than 9-1 result's, but your level of play probably isn't exactly the same from session to session either, nor are the tanks you play, the missions going on, etc.



Hauptmann_Ludwig #10 Posted Sep 15 2018 - 12:39

    First lieutenant

  • Players
  • 63178 battles
  • 890
  • Member since:
    11-26-2011

View Postgpsjake, on Sep 14 2018 - 22:48, said:

 

Looking for some understanding of session win rates.

So, overall I am right at a 50% win rate and right at 52% for my last 1000 games or so.  I have no problem with overall win rate, I feel like overall it reflects how much I contribute to wins or losses.

When I hang out in tiers six or seven my win rate responds positively compared to higher tiers, so I get the connection to overall win rates. My performance is better in the lower tiers so I win more.

What I do not understand is why my session win rates are so very random. It is constantly an all or nothing outing as far as wins or losses. My WN8 stays fairly level but I get 1-9 nights followed by 9-1 nights regularly. It is rare when I end a session around 50%. I typically try to play 10-15 games in a session if I can. 

If I am fairly consistent shouldn't my average sessions be around 50% more often than not.  I feel like I understand statistics and I understand small sample sizes, but it still seems like my mean ought to be centered more often than not. 

 

So jumping from math to the conspiracy stuff....

I have read on these forums that War Gaming does have some kind of patent on an MM that raised and lowered quality of teams/ matches based on recent win rates or performance of players. I am not saying this is happening, but is there any credence to this?

 

My question is....Am I looking at this wrong? Since it works out overall does that mean its purely due to the small sample sizes of nightly sessions? or does this seem a little wonky from a math/statistical viewpoint? ( Seems wonky to me)

 

Also since I know how some of these discussions go... I am aware that my overall stats may say that I should have a lower win rate than I do. This is due to my first 15,000 or so games being on satellite internet with 600+ping. I have substantially raised my stats/gameplay in the last 8,000 or so games. Judge me on more recent efforts than on my early stuff.

 

Thanks for your input ahead of time.

 

 

1)  I concur with those who say one should ignore the conspiracy mavins.  There is no 'Pay to Win' or such thing going on.

Having said that...…

2) Anyone thinking there's no problem are deluding themselves.  When players consistently end up in the lower tiers in a battle (90-95% of the time) regardless of what tier they run (of course tiers IX and X exempted), there's something wrong.

Let's face facts, one's 'skill' plays a role but a very minor one.  The game, by it's nature is pure chance and luck.

You get to choose what vehicle in your garage you're going to run.  After that, it's all random chance.

Random chance what team you end up on, with what mix of vehicles and on what map. etc.  Then the random chance of RNG in game.  Do you see your target, does he see you?  Do you hit him?  If so do you do any damage and if so how much?

So, the game is a series of random dice rolls after dice rolls with skill having little to no affect.

And saying that, statistics should follow that wins and losses, heads and tails even out over time.

But it doesn't then explain that many have experienced consistently lopsided losses with only a handful of wins.  Not 9-1 followed by 1-9's.  But 9 and 1, followed by 5 then 1, then 12 and 1.

Conspiracy?  No.  But either people have an incredibly long (months) run of bad luck or a logarithm is messed up somewhere.  Either way, I'm not sure WG can fix it.  Though they say they're going to......again.



__Hawkeye_ #11 Posted Sep 15 2018 - 13:01

    Sergeant

  • -Players-
  • 26137 battles
  • 227
  • [SIDES] SIDES
  • Member since:
    05-07-2014

I checked a site that said I have approx. 6 wins in a row and my WR will go up. Well after that, for the past 2 days. I literally could not purchase 2 wins in a row. It is pure insanity. I play light tanks mostly. 80% of my games I am literally low tank. Which means usually out camo, out spotted, out gunned and also every enemy will literally not even care about any other tank on the field. I can be chased all the way from the front line to the rear, run around 3 alllies heavies and td's and still the enemy will destroy me and sometimes even get all the way back to their great team coverage. 

It's not a conspiracy, this is just the true insanity of random battles. 

To have 2-13 end result so much to where you just walk away for a while and look to see what other games may be played. 

There is something going on where a team gets more than half the players can play against a team of maybe 2 that try to win. There is zero communication, except for the constant rants, or the map pinging that is purly a shame that wargaming will not stop you from pinging like that, they do it for the hot keys, you can't just constantly click the hot keys over and over again.

But to be back on track, yes for such a long time, I have been on the losing sides, which are not even a game. 

So, I turned off the team chat, I can't receive messages from other "players" and now it's just to get through the missions. 

This is not a team game, once I learned that, it's easier to just do my own damage, or passive spotting. 

 



PuddleSplasher #12 Posted Sep 15 2018 - 13:02

    Major

  • Players
  • 36599 battles
  • 2,730
  • Member since:
    07-11-2011

View Postgpsjake, on Sep 15 2018 - 03:48, said:

 

Looking for some understanding of session win rates.  

 

 

 

The best advice I can offer is to forget win rates and just have some FUN.

Playing to enjoy your game and forgetting about the stats will make the game more enjoyable and that leads to a better and more relaxed game. A relaxed attitude will offer more potential to influence the game outcome and will more than likely lead to the increase in your stats without even thinking about them.



KingofDragons #13 Posted Sep 15 2018 - 13:50

    First lieutenant

  • -Players-
  • 21116 battles
  • 581
  • Member since:
    05-14-2011

Do yourself a favour ( this will also gather stats to help everyone )... For a week just play your top WR% tanks that have 100+ games.. Starting with your G.W Panther as the Top than  MT-25 / AMX12t / Type 64 / Ram ll / M22 Locus /  is3 / T29 / T25 AT / is2 (both)  ... These are all 55+ WR% Tanks.. In theory if you only play these Tanks your WR should go up ..

 

Stats say that you do a lot better with these Tanks than any other.. 100 games sets the tone on your ability to play those individual Tanks .. For whatever reason you help yourself and your team more often with these..

 

Save your stats at the beginning and compare.. 

 



FrozenKemp #14 Posted Sep 15 2018 - 14:20

    Major

  • Players
  • 47917 battles
  • 7,677
  • [SWU] SWU
  • Member since:
    04-24-2011

Random numbers have random variation, and in smaller samples, the variation is greater - when you average things out more, values tend towards expected values.

 

Here's a little test you can try about randomness.

 

Go find a board game and get a die out of it.  Or maybe you play D&D etc and have a bunch.  Let's just assume you have a normal six-sided die.

 

Roll it in sets of 10 and write down the numbers in a line.  (Use a text editor if you don't have paper)

Do that 10 times.

On each line, count the number of 4s, 5s, and 6s, and write that number down in a separate column. 

In each line, how many 4s, 5s, and 6s are there.  Is it consistently 5, or does it vary?0

If you add up all of those totals, how close to 50 do you get?

 

Ok here I go:

2 5 4 3 4 2 3 3 5 6

1 6 4 6 4 4 5 4 1 2

3 6 2 4 2 4 5 1 4 6

6 4 2 3 6 3 5 2 4 2

3 6 2 1 4 6 6 4 2 1

4 1 1 3 4 3 3 4 3 5

1 4 4 4 6 1 1 2 6 1

2 3 3 6 4 2 4 2 1 4

3 3 3 5 3 3 1 4 3 4

6 2 1 5 2 3 4 5 2 3

 

My totals (4+) per row are: 5, 7, 5, 5, 5, 4, 5, 4, 3, 4

My overall total: 47

 

So my total number of rolls 4+ is just a little below average.  Many of the sub-samples are average, but we still have a 7 and a 3. 



GeorgePreddy #15 Posted Sep 15 2018 - 15:17

    Major

  • Players
  • 14345 battles
  • 9,797
  • [L_LEG] L_LEG
  • Member since:
    04-11-2013

View Postgpsjake, on Sep 15 2018 - 00:48, said:

Looking for some understanding of session win rates.  

So, overall I am right at a 50% win rate and right at 52% for my last 1000 games or so.

 

When I hang out in tiers six or seven my win rate responds positively compared to higher tiers, so I get the connection to overall win rates. My performance is better in the lower tiers so I win more.  You understand more than the avg. pubby.

 

What I do not understand is why my session win rates are so very random.   You used the right word here.

 

It is constantly an all or nothing outing as far as wins or losses. My WN8 stays fairly level but I get 1-9 nights followed by 9-1 nights regularly. It is rare when I end a session around 50%. I typically try to play 10-15 games in a session if I can. 

If I am fairly consistent shouldn't my average sessions be around 50% more often than not.  I feel like I understand statistics and I understand small sample sizes, but it still seems like my mean ought to be centered more often than not.  You're on the right track with these thoughts, but you are not giving nearly enough weight to the fact that this game puts you on a 15 man team for every battle. So that's 15 "effecters" of each battle outcome and YOU are only ONE of them, that is the main reason the game is all about The Law of Large Numbers, meaning large sample sizes are everything in this game. You have an effect that is positive, that is plain due to your 52% recent winrate... and you ARE consistent, over ~ 1k battles you consistently win 52%.

 

So jumping from math to the conspiracy stuff....

I have read on these forums that War Gaming does have some kind of patent on an MM that raised and lowered quality of teams/ matches based on recent win rates or performance of players. I am not saying this is happening, but is there any credence to this?  The bold part of this statement is wrong, WG does have a paragraph in their MM patent that mentions POSSIBLE raising or lowering TIER PLACING ( not "quality of teams/ matches" ) of players who are winning/losing very often.

I tested this over 2,000 consecutive battles back before 3/5/7 and found conclusively that it was not being used. I had a 60+% w/r over those 2,000 battles and yet my tier placement remained normal.  Now that we have 3/5/7 MM, it is even more obvious to anyone that they are not placing people in tiers by winrates.

 

My question is....Am I looking at this wrong? Since it works out overall does that mean its purely due to the small sample sizes of nightly sessions?  YES !

 

 

 



Rompin #16 Posted Sep 15 2018 - 15:25

    Corporal

  • Players
  • 26528 battles
  • 11
  • Member since:
    08-02-2013

Battle engagements in WoT are not random events! That is they don't even remotely resemble any justifiably random cohort regardless of sample size.

If something is not random, then statistical significance is oxymoronic.

 

(Disclaimer - I am not claiming WoT is rigged (it makes me sad to think I even need to say this...))

 

WR, particularly individual WR are not domains where statistical methods would provide useful insight.



FrozenKemp #17 Posted Sep 15 2018 - 15:29

    Major

  • Players
  • 47917 battles
  • 7,677
  • [SWU] SWU
  • Member since:
    04-24-2011

View PostRompin, on Sep 15 2018 - 09:25, said:

Battle engagements in WoT are not random events! That is they don't even remotely resemble any justifiably random cohort regardless of sample size.

If something is not random, then statistical significance is oxymoronic.

 

(Disclaimer - I am not claiming WoT is rigged (it makes me sad to think I even need to say this...))

 

WR, particularly individual WR are not domains where statistical methods would provide useful insight.

 

They're events with unknown outcomes.  Your and the enemy team composition are effectively random. 

Rompin #18 Posted Sep 15 2018 - 15:42

    Corporal

  • Players
  • 26528 battles
  • 11
  • Member since:
    08-02-2013

View PostFrozenKemp, on Sep 15 2018 - 14:29, said:

 

They're events with unknown outcomes.  Your and the enemy team composition are effectively random. 

 

Respectfully, and without malice - you are wrong.

 

Having said this, whatever you or I say doesn't matter where a fact is concerned.

Choose well and be at peace.



DEADTIME #19 Posted Sep 15 2018 - 15:56

    Sergeant

  • Players
  • 9321 battles
  • 238
  • [1IWH] 1IWH
  • Member since:
    04-03-2011

View PostPuddleSplasher, on Sep 15 2018 - 04:02, said:

 

The best advice I can offer is to forget win rates and just have some FUN.

Playing to enjoy your game and forgetting about the stats will make the game more enjoyable and that leads to a better and more relaxed game. A relaxed attitude will offer more potential to influence the game outcome and will more than likely lead to the increase in your stats without even thinking about them.

Yup, I just quit caring about being a purple long ago, and would rather just enjoy the game and try to get better while being happy. It helps to have a good attitude and just enjoy the game then the losses don't seem to matter to much.



Themothius #20 Posted Sep 15 2018 - 16:52

    Corporal

  • -Players-
  • 5241 battles
  • 20
  • [BRIG] BRIG
  • Member since:
    07-03-2015

View PostUS_Cavalryman, on Sep 15 2018 - 05:39, said:

 

1)  I concur with those who say one should ignore the conspiracy mavins.  There is no 'Pay to Win' or such thing going on.

Having said that...…

2) Anyone thinking there's no problem are deluding themselves.  When players consistently end up in the lower tiers in a battle (90-95% of the time) regardless of what tier they run (of course tiers IX and X exempted), there's something wrong.

Let's face facts, one's 'skill' plays a role but a very minor one.  The game, by it's nature is pure chance and luck.

You get to choose what vehicle in your garage you're going to run.  After that, it's all random chance.

Random chance what team you end up on, with what mix of vehicles and on what map. etc.  Then the random chance of RNG in game.  Do you see your target, does he see you?  Do you hit him?  If so do you do any damage and if so how much?

So, the game is a series of random dice rolls after dice rolls with skill having little to no affect.

And saying that, statistics should follow that wins and losses, heads and tails even out over time.

But it doesn't then explain that many have experienced consistently lopsided losses with only a handful of wins.  Not 9-1 followed by 1-9's.  But 9 and 1, followed by 5 then 1, then 12 and 1.

Conspiracy?  No.  But either people have an incredibly long (months) run of bad luck or a logarithm is messed up somewhere.  Either way, I'm not sure WG can fix it.  Though they say they're going to......again.

 

ok I just have to let you know you are bottom tier 41% of the time. In 10 games maybe not but over 100s and 1000s of games its 41% bottom tier something like 29% top tier and 30% mid tier.  Just for the record that math is on the forums somewhere and the FULL DATA SETS where posted to.  Yes it's easy to say think it 90% but it's not please make down ur games if your top/mid/bottom tier in EVERY game then after two weeks do the math you will be shocked.

 

Also to the Charle Epps here(numbers get it) shouldn't it be 15 or 30 coins being flipped and if you get say 8 or 16 heads then it counts as heads, cause it's not a one person game.  Might be wrong about that but just seems more accurate to this game to me.  Also if I'm not mistaken you need to flip a coin 100s of time to get even close to 50% on it, so simple size is VERY VERY important.

 

Lastly since lots of % are being thrown around always remember, 74% of the time people quote statistics they make it up on the spot and only 47% of the time are they right.






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users