Brit Global PP

[Guest]

Welcome, Guest. Please login or register.
Did you miss your activation email?

1 Hour 1 Day 1 Week 1 Month Forever

Login with username, password and session length

[XIIcorps]

But creeping is balanced as its a specialisation where as Fire barrage is a T4 and come with other benefits

[Hicks58]

At the moment the better players are flitting between Allies and Axis, with a preference to Allies due to shiny new Airborne. Axis has a dedicated section of players, who are low skill level. That kinda leads to the Axis getting pushed back hard on PP's with the warmap.

[nikomas]

Nevermind the fact that incendiary barrage is a T3 and that only ChemFire is a T4...

[Mysthalin]

Plus, allies is boring as shit. Nobody plays them. Anybody that choses to take boring allies get to pick and chose games. Say, AveragePlayer1 and AveragePlayer2 go on allies and get a choice between playing vs Skiffi and ProGren or fighting StoneStander and IcanCountToPotato they'll probably take on the easier deal... And win.

One game wouldn't change one, but this can very easily turn systematic, and it has.

[UndeathWrath]

For an overwhelming victory, 6 doctrine points, x 30, is around 180 points (given fractions) - which, as you say is ~2PP.

That number reflects the number of CW PLAYERS there are, which there isnt many.

Mathematically, it all works out.  The more CW players play, the closer to 10 it ends up.  It just so happens that no-one has been playing CW, so the ratio is well out of whack.

The more pressing issue I am seeing, is why the F**k are the Axis sides sucking so badly?

Ok let me kindly explain why I think this system is rubbish from what I understand of it.

You ASSUME that because there are less CW players, less games will be played by CW players.

So let's make an example to demonstrate:

Let's pretend there are 15 american players, 5 CW players, 15 wehr players, and 15 PE players in total.

Let's say 1 night 3 of those CW player play against 3 wehr players 4 times in a row.

2 times the allies get overwhelming victory and 2 times the axis get overwhelming victory. All players dump their points into doctrine, resulting in each side dumping 42 doctrine points total. (14 for each player (6x2 for the 2 overwhelming victories and 1x2 for the 2 losses)

Now according to what you're saying, because there are only 5 CW players compared to the 15 ameri players, their PP gains will be 42x30= 1260 which divided by 100 = +12.6PP for the CW faction.

In contrast, the Wehr players, because are in the same numbers as the PE players, will receive:
42x10=420 which divided by 100 = +4.2PP for the Wehr faction.

In the exact same amount of games and wins/losses and doctrine points spent, the CW gets triple the PP's. So how is this fair?

To me this is just a prime example of an overly complicated fix to a simple issue, just like the WEAPONS CACHE system was compared to simply using pool values.

Why not just have both races of each faction share the same PP pool and call it a day?

This is based on what I understand of it so don't crucify me if I got it wrong.

[nikomas]

Pretty sure he means that it's base on games played, for example...

US/WHER/PE = 60 games = x1
CW             = 20 games = x3

Now here is were it goes south a little as it currently stands, at least earlier in the war the players playing brit were often of higher skill on average than the other factions players meaning that the few games the brits play they are likely to score more overwhelming victories, leading to what we see now.

[Mysthalin]

The system would be fine if it used the right units, and the multiplier effects were not done the way they are (right now the system double-changes gains to invert the original gains). Let me show you why:

Right now it's using companies total, whereas for it to work as intended it should calculate effective companies fielded. I.E. Active companies being used each game.

Let's say we have 100 3v3s played in a time frame of 12 hours (I know, lulz, this is to make the maths easy). That is 300 companies fielded from each side (a company can be used multiple times, but that doesn't change the maths). Let's say these included 225 US, 75 CW, 150 WM, 150 PE companies, and both sides have had an equal amount of overwhelming wins as well as overwhelming defeats. Let us also assume that there are, in actual fact, an equal amount of companies made for each faction by the playerbase, but this has been the hours in the early morning and then day for Europe, whereas it was prime gaming time in the US, which made US a more commonly played faction than CW.

Under Luci's current system what would happen with these assumptions (under the current total companies created system.. which is also easy to "cheat" by creating a bunch of one faction's companies to buff the other) is that these doctrine point gains would be seen:

US: 225*0.5*7*10=7875
CW: 75*0.5*7*10=2625
WM: 150*0.5*7*10=5250
PE: 150*0.5*7*10=5250

Under an effective companies fielded system without accounting for double modification:

US 225*0.5*7*10/3=2625
CW: 75*0.5*7*10*3=7875
WM: 150*0.5*7*10=5250
PE: 150*0.5*7*10=5250

(US and CW simply switch places in how many points were obtained)

Basically what we need is a normalization modifier that would make US point gain 2/3rds of it's original size, and CW points gain - twice it's original size. In this arbitrary scenario these modifiers are clear, but I'll try and solve for a general case that the system could actually use based on real used company numbers.

I know the system collects data on companies that are used in battle, so it should be fine in terms of coding - and the modifier could be based on all games ever played, or reset to 1:1 every day and be re-constructed based on daily data (though I don't see why that would be necessary - over time both should end up being very similar, with less deviations in the full-time modifier case).


Or, we could simply reduce point gain by half across the board and have US+CW as well as WM+PE share a PP pool. Same speed in achieving doctrines, less difficulties arising from mathematics.


TL:DR : Current system uses wrong units, on top of essentially inverting point gain, rather than normalizing it. US should have CW PPs, and WM should have PE PPs.

[Heartmann]

Hahaha oh myst i love the fact that after you did the super log explenation you added the tldr at he end :-P

[Mysthalin]

Hahaha oh myst i love the fact that after you did the super log explenation you added the tldr at he end :-P

You owe me documentz.

[TheWindCriesMary]

Tbh something really needs to be done, because right now I am steamrolling the global PP in favour of brits and it's a train that's only picking up speed.

[EIRRMod]

The system would be fine if it used the right units, and the multiplier effects were not done the way they are (right now the system double-changes gains to invert the original gains). Let me show you why:

Yeah, I came to the same conclusion last night.

Thanks for the maths for everyone else though

[EIRRMod]

Ok let me kindly explain why I think this system is rubbish from what I understand of it.

You failed at the 'kindly' part.  Read Mysts response further in this thread for a civil example.

You ASSUME that because there are less CW players, less games will be played by CW players.

There are no assumptions.  The calculation is a 1 liner, with the ratio vs 10.

So let's make an example to demonstrate:

Let's pretend there are 15 american players, 5 CW players, 15 wehr players, and 15 PE players in total.

Let's say 1 night 3 of those CW player play against 3 wehr players 4 times in a row.

2 times the allies get overwhelming victory and 2 times the axis get overwhelming victory. All players dump their points into doctrine, resulting in each side dumping 42 doctrine points total. (14 for each player (6x2 for the 2 overwhelming victories and 1x2 for the 2 losses)

Now according to what you're saying, because there are only 5 CW players compared to the 15 ameri players, their PP gains will be 42x30= 1260 which divided by 100 = +12.6PP for the CW faction.

In contrast, the Wehr players, because are in the same numbers as the PE players, will receive:
42x10=420 which divided by 100 = +4.2PP for the Wehr faction.

In the exact same amount of games and wins/losses and doctrine points spent, the CW gets triple the PP's. So how is this fair?

Your assumption is incorrect.  Each game players modifier is at the point of calculation.
So, for the first CW and US player, the ratio is still 1:1.
If 3 more US players play, then they gain 10, 9, 8 (roughly) and so on as the ratio becomes 0.9:1.1 and so on.  This is being 'double' calculated though, which Myst points out in his post.

To me this is just a prime example of an overly complicated fix to a simple issue, just like the WEAPONS CACHE system was compared to simply using pool values.

Why not just have both races of each faction share the same PP pool and call it a day?

Because the idea of fighting for your side, not just faction had a nice appeal.
If you see 1 line of calculation for PP (To stop US being waaaaay ahead of CW etc) is too complicated, well, sorry I couldnt find a suitable solution.  Im not perfect, but you dont need to be a douchebag.

This is based on what I understand of it so don't crucify me if I got it wrong.

Some parts were wrong, but the crucification would be because you knock a process of development that is multifaceted, complicated, and most of all - done for free in our free time.  Myst has solved the issue, and it was something I was going to look at. 

But quite frankly - I didnt expect US to be 4x as popular as the CW, which exasperated the results.

[EIRRMod]

The system would be fine if it used the right units, and the multiplier effects were not done the way they are (right now the system double-changes gains to invert the original gains). Let me show you why:

Right now it's using companies total, whereas for it to work as intended it should calculate effective companies fielded. I.E. Active companies being used each game.

Im not sure where I stated it is using companies total.  It is not.
It has a counter for each side, which gets a +1 to either CW, US, WM, PE for each player (of that game) who played.

So heres an example:
Say the totals are at CW: 100, US: 100, WM: 100, PE: 100 (Coincedently, thats what I start the war numbers at, so the ratio doesnt go all funny straight away).
A 3v3 with 1xCW, 2xUS, 3xWM makes the totals 101,102,103,100.  Altered each game.
 

Let's say we have 100 3v3s played in a time frame of 12 hours (I know, lulz, this is to make the maths easy). That is 300 companies fielded from each side (a company can be used multiple times, but that doesn't change the maths). Let's say these included 225 US, 75 CW, 150 WM, 150 PE companies, and both sides have had an equal amount of overwhelming wins as well as overwhelming defeats. Let us also assume that there are, in actual fact, an equal amount of companies made for each faction by the playerbase, but this has been the hours in the early morning and then day for Europe, whereas it was prime gaming time in the US, which made US a more commonly played faction than CW.

Under Luci's current system what would happen with these assumptions (under the current total companies created system.. which is also easy to "cheat" by creating a bunch of one faction's companies to buff the other) is that these doctrine point gains would be seen:

US: 225*0.5*7*10=7875
CW: 75*0.5*7*10=2625
WM: 150*0.5*7*10=5250
PE: 150*0.5*7*10=5250

Under an effective companies fielded system without accounting for double modification:

US 225*0.5*7*10/3=2625
CW: 75*0.5*7*10*3=7875
WM: 150*0.5*7*10=5250
PE: 150*0.5*7*10=5250

(US and CW simply switch places in how many points were obtained)

Sorry myst, the initial assumption is incorrect, so the math is not proven.
See my next comment to your post.

Basically what we need is a normalization modifier that would make US point gain 2/3rds of it's original size, and CW points gain - twice it's original size. In this arbitrary scenario these modifiers are clear, but I'll try and solve for a general case that the system could actually use based on real used company numbers.

I know the system collects data on companies that are used in battle, so it should be fine in terms of coding - and the modifier could be based on all games ever played, or reset to 1:1 every day and be re-constructed based on daily data (though I don't see why that would be necessary - over time both should end up being very similar, with less deviations in the full-time modifier case).

Or, we could simply reduce point gain by half across the board and have US+CW as well as WM+PE share a PP pool. Same speed in achieving doctrines, less difficulties arising from mathematics.

This is what we have at the moment.  However if you look at the modifier as it stands:
CW: 213, US: 507.
Now, - 100 per side, means 113 CW players have played a game while 407 have played as US.
This is a HUGE discrepancy, and similar seen with WM/PE.

The problem, if you can help solve it Myst.  Is that because the modifier does the CW at 10 * (507/213) per pt in doctrine, AND the US is 10 * (213/507) - when the numbers have such a wide gap, the numbers go crazy!  In effect, with this calculation, we are accounting for the descrepany TWICE.

So, I was thinking, what if the higher side was ALWAYS 10, and only affect the lower played side as a sort of 'Catch-up'.  The problem is, when I ran tests, it was never enough.

TL:DR : Current system uses wrong units, on top of essentially inverting point gain, rather than normalizing it. US should have CW PPs, and WM should have PE PPs.

Unfortunately based on incorrect assumptions Myst =(
The same may turn out true with the correct input, but it cant be calculated with a 1 pass math test, it needs a 100+ strong pool of random 'battle' numbers.

Im kinda stumped.  Im open to suggestions.  I dont want to modify the PP pools to Allied and Axis, which would work, but we'd lose a bit of flavour.  (There are warmap cards coming out that boost PP etc)

[Mysthalin]

Ah, great if it's using effective companies fielded. Makes it a lot better - but the maths remains transitive for the normalisation modifier...

Yeah, that's what I mean about double calculation. I tried solving it in two different ways as well and it gave me the wrong (current) normalisation modifier both times...


I've got one that, more or less - works:

sqrt(x/(1-x))

Where x is the proportion of games played with the company in question. In the case of US that would currently be 507/720, and in the case of CW - 213/720.

However, it does end up giving a slightly lower number of total PPs gained overall than a nice 50-50 split :S.

I'll try and think of a formula that works more precisely. Question, though - do you want the system to over-produce points for less prevalent companies slightly, or do you want it to just more or less equalise it?

[UndeathWrath]

Hey I didn't mean to sound rude. For the record my hat is ever off to you and tank and the original creators for making and keeping this mod going (not to mention evolving. And future EIR2!)

It's been a really rough week...not even over yet.

Still imo there should be equal incentive to play whatever of the 4 races the player desires, and the gap in pp's only works against that.

[CommieKiller]

the amount of allied pp advantage is getting ridiculous

[tank130]

the amount of allied pp advantage is getting ridiculous

We know..... we are trying to calculate a solution, but it is turning out to be very complex. Please read Mysthalin and Eirrmods previous post to get a better idea of just how complex it is.

[EIRRMod]

I'll try and think of a formula that works more precisely. Question, though - do you want the system to over-produce points for less prevalent companies slightly, or do you want it to just more or less equalise it?

More or less equalise it.

The PP number will be higher or lower depending on the success of that side (Due to the pure number of Doctrine points for a win) so will reflect that Id hope.

The speed at which PP increases can be adjusted up or down by the base 10 number, I just need a formula that doesnt go haywire with a high ratio like this.

[EIRRMod]

Hey I didn't mean to sound rude. For the record my hat is ever off to you and tank and the original creators for making and keeping this mod going (not to mention evolving. And future EIR2!)

It's been a really rough week...not even over yet.

Still imo there should be equal incentive to play whatever of the 4 races the player desires, and the gap in pp's only works against that.

Well yes I can understand that.  The gap was what this calculation was meant to avoid, rather than aggravate.

I guess I read your tone wrong ;P  Sorry mate, and thanks for the input!

[Mysthalin]

Tbh the simplest formula is simply (Faction 1 Points + Faction 2 points)/2 and boom, perfect equalisation without any sort of complicated maths behind it that gets all messed up with silly ratios. But you don't want that since that is bland.


Thought about it - you could code it in to work based on the distance between the currently top faction and the faction in question. I'd go for a sqrt(Leading PP point faction PPs/Played faction PPs) modifier to PP gain.

So, if say, CW are currently top of the pack, they'll just be gaining normal PP gains. If, however, you chose PE who are basically at a quarter of that level, then they will get a sqrt(4)=2 modifier to their PP gains.

Probably the easiest way to do it while still retaining flavour between factions, and giving a leg up to those clearly lagging behind.