Jump to content


Blitz Economics - Crate Mechanics, Odds and Pricing - T26E5 Event (average £80, Gold in crates)


  • Please log in to reply
33 replies to this topic

stubbo66 #1 Posted 19 August 2019 - 03:41 PM

    First Sergeant

  • Player
  • 70012 battles
  • 1,387
  • [49DIV]
  • Member since:
    06-29-2014

This will be another ongoing series of posts where I will put up a table of the actual odds of getting the tank and the costs based on the event offer to help you understand a bit more about the costs involved.

 

Go to the later pages to find the latest event info.


Edited by stubbo66, Yesterday, 12:38 PM.


stubbo66 #2 Posted 19 August 2019 - 03:46 PM

    First Sergeant

  • Player
  • 70012 battles
  • 1,387
  • [49DIV]
  • Member since:
    06-29-2014

So for the T-22 event, there are three sales options.

 

1+1 free crates for £9.99, single crates for £9.99 and 5+1 crates for £48.99. I'll stick with British pounds, I may try and evolve this later for euros etc.

 

For the sake of this article, I will also assume that everyone buys bulk and not single crates

 

I've created a statistical model that starts with a pool of 10,000 people all willing to keep buying until they get the tank. Assuming they buy the 1+1 crates (both of them) and then continue with the 5+1 then the chances of you getting one and the costs. This doesn't count players that might get 2 tanks in their sets, there are impacts here I will explain later.

 

Now into my model I've also built in a droppoff rate, 5% if you don't get the tank from the cheaper crates, and 40% if you don't get them from the larger ones....it's a work in progress but this is where I am so far.

 

Players Gambling Crates Bought Players Winning Tank Players giving up Cumulative Cost for winners %age of gamblers winning tank (total) WG money made
10,000 1+1 (2) 1000 450 £9.99 10% £99,900
8,550 1+1 (4) 855 3078 £19.98 18.55% £185,314
4,617 5+1 (10) 1385 1292 £68.97 32.4% £411,501
1940 5+1 (16) 582 543 £117.96 38.22% £506,541
815 5+1 (22) 244 228 £166.95 40.66% £546.468
343 5+1 (28) 102 96 £215.94 41.68% £563.272
145 5+1 (34) 43 40 £264.93 42.11% £570.375
62 5+1 (40) 18 17 £313.92 42.29% £573.413

 

So you should get the idea. The percentages are of the original 10K customer that have got the tank for that amount of money shown or less.

 

So there is a 18.55% chance it would cost you £19.98 or less to get the tank, going up to a 40.66% chance it will cost you under £167. But that still means there is a 60% chance you haven't won the tank after spending £167.

 

Also, a significant note here, is the smaller the pool of players there are opening tanks, the higher the probability that one player gets more than one tank in his batch of crates, worsening the odds for the other players trying to get it.

 

So take a look at the table above, determine what you might be willing to spend and look at the chance that you might get that tank for that money....that's the real stake here...ignore the 5% as it means nothing on an individual basis.

 

The above also explains why you hear about a lot of players opening only a crate or two and getting the tank, as initially with the pool of purchasers high, the chances are almost 20% of getting the tank for under £20.

 

I will add further to this in time but thought it was about time to provide a lot more detail than the 5% chance and notional average costs that we all have worked to until now.

 

The drop off curve based on amount spent needs refining above, but it's good enough for an indicative model right now.


Edited by stubbo66, 19 August 2019 - 05:07 PM.


gallabru #3 Posted 19 August 2019 - 04:53 PM

    Sergeant

  • Player
  • 17512 battles
  • 414
  • [LAGER]
  • Member since:
    09-01-2016
Does that mean that you are you assuming that the odds of getting a tank in each crate are not independent events ?

My team stats in real time  : http://forum.wotblit...ats-repository/

Carlsberg clan family is recruiting. Contact us.


O_o__O_o__O_o #4 Posted 19 August 2019 - 10:03 PM

    Lance-corporal

  • Player
  • 36611 battles
  • 74
  • [AFK]
  • Member since:
    12-16-2015
hang on is these 5%drop rate calculated by the number of crates bought ?

i was thinking its fair for everyone and you get 5% every crate doesnt matter how many bought it

anyway congrats to all stupid people who buy crates

Jukkis74 #5 Posted 20 August 2019 - 04:47 AM

    Senior Sergeant

  • Player
  • 32334 battles
  • 709
  • [AFK]
  • Member since:
    06-29-2014
In this case, I think stubbo66 is overcomplicating things. I would be really surprised if WG would have modeled the event prizes as anything else than independent events. And if they have, then they (WG) are incorrect in their marketing statements (and not complying to Apple guidance either).

MarsIsBrightTonight #6 Posted 20 August 2019 - 06:07 AM

    Senior Sergeant

  • Player
  • 28769 battles
  • 539
  • [OBB]
  • Member since:
    11-03-2016

As usual the dedication you put into this is commendable.

I just don't understand where this part comes from:

Quote

Also, a significant note here, is the smaller the pool of players there are opening tanks, the higher the probability that one player gets more than one tank in his batch of crates, worsening the odds for the other players trying to get it.

To me it is still the chance of rolling a natural 20, the fact that there are other people trying does not factor into it. But I'm mediocre at maths on my better days, so.


"Gambling: The sure way of getting nothing for something." (Wilson Mizner, US playwriter, 1876-1933)


Mjr_Eazy #7 Posted 20 August 2019 - 07:55 AM

    First Sergeant

  • Player
  • 27184 battles
  • 5,756
  • [BT_SV]
  • Member since:
    04-26-2017

Stubbo, how did you arrive at this theory?  I’m curious as for them to market it as a 5% drop chance each crate must have a 5% chance and therefore be independent events...

I really hope you’re wrong - no offence intended m8 you do brilliant work and I am one of many here who find it indispensable! - because if you are right then WG are even more despicable liars than I already believed them to be and the crates really are a rip-off and a con!


547146227.png

 


Replicant_2 #8 Posted 20 August 2019 - 08:11 AM

    Sergeant

  • Player
  • 5469 battles
  • 443
  • [HRG]
  • Member since:
    10-08-2018

View PostMjr_Eazy, on 20 August 2019 - 07:55 AM, said:

Stubbo, how did you arrive at this theory?  I’m curious as for them to market it as a 5% drop chance each crate must have a 5% chance and therefore be independent events...

I really hope you’re wrong - no offence intended m8 you do brilliant work and I am one of many here who find it indispensable! - because if you are right then WG are even more despicable liars than I already believed them to be and the crates really are a rip-off and a con!

 

I’d think you may be confusing single repetition event and probability which is theoretical over v. large/ infinite number of repetitions, Mjr.

 


All those moments will be lost in time, like tears in rain.

Time to die.


stubbo66 #9 Posted 20 August 2019 - 10:10 AM

    First Sergeant

  • Player
  • 70012 battles
  • 1,387
  • [49DIV]
  • Member since:
    06-29-2014

I've worked with a lot of RNG generators in my time as a software developer, and you can very easily model and demonstrate what I'm calculating here.

 

What I'm saying is that RNG does not guarantee a 5% drop rate, the RNG generator may award a tank on a roll of 1-5 out of 100, but there is no guarantee the generator will evenly generate 1-5, or it may over generate them, especially on a small sample size. But on larger sample sizes (10,000 in my example) then the results tend to normalize.

 

But don't think that anyone players odds improves from one crate opening to another, they don't, but they are influenced by other players opening crates. Each crate being opened is subject to the single RNG distribution, so if 10,000 people open a crate at the same time, 5% will on average get the tank.

 

RNG generators also require a seed to guarantee a base and even distribution, often the systems will use the clock timing in milliseconds of the time the RNG generator is corrected. You can't create a new RNG generator on every request because that will not provide you the average distribution and can skew results either for or against your required distribution criteria.

 

So there should only be a single RNG generator for the event (or for the server) to provide fair results. And that generator is shared across everyone opening the crate. Now whether the contents are allocated at the point of purchase or point of opening doesn't really matter (it's probably the latter).

 

So to the point about every crate having a 5% chance...it does, but that is not based on your openings alone and other people opening crates will skew your chances either positively or negatively....it's how RNG works and how the drop rate is maintained.

 

So when WG says there is a 5% chance of getting the tank, that is not for you personally, that is for everyone combined opening the crates.

 

I will go and write some programs to model the above and run some tests to validate the results above, but I'm confident that it would be not far off.



stubbo66 #10 Posted 20 August 2019 - 10:11 AM

    First Sergeant

  • Player
  • 70012 battles
  • 1,387
  • [49DIV]
  • Member since:
    06-29-2014

View PostMjr_Eazy, on 20 August 2019 - 07:55 AM, said:

Stubbo, how did you arrive at this theory?  I’m curious as for them to market it as a 5% drop chance each crate must have a 5% chance and therefore be independent events...

I really hope you’re wrong - no offence intended m8 you do brilliant work and I am one of many here who find it indispensable! - because if you are right then WG are even more despicable liars than I already believed them to be and the crates really are a rip-off and a con!

 

Hi Mjr_Eazy, No, they aren't lying, it's about how RNG works and the fact that 5% is across all crate openings, not your personal openings. Every crate has a 5% chance, but odds are influenced by other openings.



stubbo66 #11 Posted 20 August 2019 - 10:23 AM

    First Sergeant

  • Player
  • 70012 battles
  • 1,387
  • [49DIV]
  • Member since:
    06-29-2014

One important note from the above as well, players giving up and dropping out (not wanting to spend any more money). They obviously add to the number of players not getting the tank. So if everyone kept going then firstly WG would be very happy indeed, and secondly, the number of players getting the tank would trend to 100% eventually.

 

And on my point about a smaller sample size resulting in a higher probability of getting the tanks multiple times, this again comes back to the number of times the RNG generator is queried and the distribution model. Again, I should be able to simulate that.

 

I'm writing a program as I type this to prove (or disprove) all of the above.



stubbo66 #12 Posted 20 August 2019 - 12:39 PM

    First Sergeant

  • Player
  • 70012 battles
  • 1,387
  • [49DIV]
  • Member since:
    06-29-2014

OK, I wrote a simulator for the event in Java, with 10,000 independent players, every crate opening with a 5% drop rate.

 

I assumed 5% of players give up after the first 1+1 crates, then 40% give up after the second 1+1, 40% after the 5+1 and so on.

 

I ran the program 10 times, then averaged the results...and they are.....

 

Players gambling Crates Bought total Players who only bought these crates Total WG income Individual Player Cost Players Winning Tank Multi Tank Winners Players Giving Up %age of players with the tank
10000 2 1428 £14269.72 £9.99 974 22 455 9.74%
8572 4 3931 £78537.38 £19.98 839 20 3092 18.13%
4641 10 2586 £178370.21 £68.97 1213 144 1373 30.26%
2055 16 1150 £135606.82 £117.96 538 66 612 35.64%
905 22 511 £85227.98 £166.95 241 30 270 38.04%
395 28 220 £47485.21 £215.94 101 13 119 39.06%
175 34 99 £26307.55 £264.93 49 5 50 39.55%
75 40 39 £12305.66 £313.92 19 3 20 39.55%
36 46 20 £7185.62 £362.91 11 1 9 39.85%
16 52 16 £6755.16 £411.90 4 0 5 39.89%

 

Now, this model has everyone opening their crates at once, which isn't obviously what happens in the real world, but it's a start. In reality, a smaller number of people open their crates at the same time, and in reality that will likely improve their odds....that's for v2 of my program!



stubbo66 #13 Posted 20 August 2019 - 01:02 PM

    First Sergeant

  • Player
  • 70012 battles
  • 1,387
  • [49DIV]
  • Member since:
    06-29-2014

So the above model was with 10,000 players all opening crates at the same time, so I created another version of the model where every person opens their own crates until they win the tank or give up spending to see if that would help improve your chances of getting the tank or not....here's the results of that.

 

Players gambling Crates Bought total Players who only bought these crates Total WG income Individual Player Cost Players Winning Tank Multi Tank Winners Players Giving Up %age of players with the tank
10000 2 1445 £14430.56 £9.99 983 26 461 9.83%
8556 4 3927 £78453.47 £19.98 839 22 3087 18.22%
4629 10 2587 £178446.08 £68.97 1239 144 1349 30.61%
2042 16 1136 £133955.38 £117.96 541 68 594 36.03%
906 22 500 £83541.78 £166.95 235 28 266 38.37%
406 28 231 £49860.55 £215.94 106 11 125 39.44%
175 34 99 £26254.56 £264.93 46 7 53 39.50%
76 40 44 £13686.91 £313.92 21 1 23 40.10%
32 46 18 £6459.80 £362.91 8 1 10 40.18%
14 52 14 £5848.98 £411.90 4 1 4 40.23%

 

So as you can see, even with a completely different model of deciding when to open the crates, it hardly changes the outcome in terms of what percentage get tanks and how much they would have to spend to get them.

 

It's a sobering set of results, but one that's extremely important to show you.

 

Some explanation of the last column as well, this represents the percentage of players that decide to gamble on at least one crate that will get it, but recognizing that players will stop spending at different points. This doesn't mean that if you keep on spending you won't get the tank, the model shows that your odds do improve the longer you go on, but there are no guarantees.

 

What is interesting from the model is that instead of getting a 30% chance of getting the tank from 6 create bundles, it's nearer 25%.

 

So open to criticism and suggestions here, obviously things like drop out rates are subjective, but this gives you the idea anyway.

 



nazan_ir #14 Posted 20 August 2019 - 02:11 PM

    Lance-corporal

  • Player
  • 5377 battles
  • 90
  • [PN1]
  • Member since:
    12-17-2015

View Poststubbo66, on 20 August 2019 - 10:23 AM, said:

And on my point about a smaller sample size resulting in a higher probability of getting the tanks multiple times, this again comes back to the number of times the RNG generator is queried and the distribution model. Again, I should be able to simulate that..

 

It makes sense that the chances are affected by the openings of other players. I don't want to draw a correlation between my chest openings on different times of the day since the sample sizes are very small and I'm in no position to send a few thousand dollars to confirm this hypothesis.

 

However, for the giggles I twice bought the 1+1 crates and opened them around 14:00 +1 GMT in the weekend. The best crate out of the four was 250 gold and some boosters.

 

Yesterday, I read this thread and started thinking about the "others affecting your chances" statement. I had recently received my salary from work and thought to myself let's see if this could make any sense. I bought the 5+1 crates between 12-1am +1 GMT and I got the T-22 medium on the first crate I opened from this crates pack. The rest of the crates were I think 2x 500 gold and 3x 750 gold with several Epic experience boosters.

 

Now of course drawing any conclusion out of this is probably not accurate and biased. On top the sample size is very small. We probably need at least a few hundred pack openings to properly draw a conclusion based on the statistics. What I experienced could be just my small luck.

 

Interesting post anyways.


98% of all the battles I have played are in Soviet tanks

a.k.a. OCCupants [Asia]


Warren5 #15 Posted 20 August 2019 - 02:59 PM

    First Sergeant

  • Player
  • 36950 battles
  • 1,115
  • [VG]
  • Member since:
    12-06-2014
WG = WarGambling ...... again. 


 


MarsIsBrightTonight #16 Posted 21 August 2019 - 10:40 AM

    Senior Sergeant

  • Player
  • 28769 battles
  • 539
  • [OBB]
  • Member since:
    11-03-2016

Clearly you've thought this in depth and it's my problem that I can't wrap my head around it.

Question: does the "trending" mean that a player willing to mortgage the family house still has less than a chance in two of getting the goddamn thing?

(Yeah, I know, functions, limits ad infinitum... but I peaked maths in my fifth year and that was still in 19-canteen)


"Gambling: The sure way of getting nothing for something." (Wilson Mizner, US playwriter, 1876-1933)


stubbo66 #17 Posted 21 August 2019 - 11:40 AM

    First Sergeant

  • Player
  • 70012 battles
  • 1,387
  • [49DIV]
  • Member since:
    06-29-2014

View PostMarsIsBrightTonight, on 21 August 2019 - 10:40 AM, said:

Clearly you've thought this in depth and it's my problem that I can't wrap my head around it.

Question: does the "trending" mean that a player willing to mortgage the family house still has less than a chance in two of getting the goddamn thing?

(Yeah, I know, functions, limits ad infinitum... but I peaked maths in my fifth year and that was still in 19-canteen)

 

No, there is a slightly under 10% chance of getting the tank in the first 2 crates (£10), and just under 19% if you buy both of the 1+1 bundles (£20). After that you're in the lap of the gods, buying 6 crates at a time does give you a 25% chance of getting the tank in that set of 6, or any subsequent set.

 

What was interesting was that even if you were the only person buying crares, your odds were no better than if you were in a huge pool of people doing the same thing.

 

Probability dictates you will get it the more you buy, but there are no guarantees!

 

Overall, you have around about a 40% chance of getting the tank for £70



ponedelj #18 Posted 21 August 2019 - 08:28 PM

    Senior Sergeant

  • Player
  • 14393 battles
  • 557
  • [LV-S]
  • Member since:
    11-05-2013
So how understand it (tnx bw) to get tank fm crate (i strongly advise against it) U buy 1+1 crate and than 2x 5+1 change to get tank are 30% if u are rich guy u can go for one more 5+1 after that with each opening chance to get tank increase very slow. In end depenind of your luck u will spend 9.99£/ 107.97£ or 156.96£
Btw in € its 10.99 and 54.99 so it will be 10.99/120.97/175.96 €
Its mobile game... I dont know...u can buy something great for that money.  


Fugit555 #19 Posted 23 August 2019 - 05:07 AM

    First Sergeant

  • Player
  • 26769 battles
  • 1,309
  • [49DIV]
  • Member since:
    09-30-2014

Just to add some fuel to this fire:

 

During the 24hr live stream I did with 3very good name is taken last weekend I decided to give this a bash and test something. Whilst I already have the tank on my main account, got it first time round in the New Years Tree event, I was interested to see if it would drop quickly - after all 15k gold would be compensation. Alas I didn't get it on my main account after using the special 2 for 1 crates, of which there are 2 thus making 4 crates and also I stupidly bought 2 x 5 plus 1 crate making 6. So after 16 crates i didn't get the tank! 

 

I then went onto my re-roll account, with less battles (a whole lot less) and bang! 2nd crate! I then went onto the NA account, again less that 100 battles and bang! 2nd crate again! What are the odds? Now I am not saying that those with lower battle counts have more chance of getting the tank than those with high battle counts, but with my 28k main account I couldn't get it after 16 crates, but I got it almost immediately on my 2 accounts that have less than 100 battles - coincidence or just very very very lucky?


Ours is not to reason why, ours is just to kill, kill kill - War is hell, maybe :justwait:

VISIT MY YT CHANNELhttps://www.youtube....2L0jDqH-GsulArQ


Buckleyteer2 #20 Posted 23 August 2019 - 07:39 AM

    Junior Sergeant

  • Player
  • 40020 battles
  • 162
  • [7GD]
  • Member since:
    10-16-2015

Im with you FUGIT I notice that nearly  all the you tubers got them dropped in the 2 crate deal also which told me its clever marketing and its pushed again to the poor new players while us old ones suffer the realms of the ever changing WG machine... Its a shame as Tier x was really a prestige club where you have gained your place...

 Now its just......






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users