PDA

View Full Version : Warhammer Championship Series



Feefait
25-09-2011, 14:18
Ok so this may have already have been done, or is being done but I have a thought to 'figure out' tiers. That would basically be to create a Warseer BCS formula. For those not familiar with college football it determines it's championship and ranking based on a super complicated formula of wins, margin of victory, strength of schedule, etc. It's as far as I know just about impossible for mortal men to comprehend or figure out. :)

I was thinking we could do something similar, and less complicated for warhammer. Here is my formula thoughts:
Initial rankings are created by a month long tally of wins v losses for each army,and then ranking top to bottom based on winning percentage (wps). After month 1 we would recalculate the next months wps, but add in their percintile ranking from the previous month.

Example ( This is of course rounding all decimals up to the nearest 10th. ):

Beastmen Month 1 win 45 of 100 games for a .45 wps. Dwarves win 26 of 75 for a .35 wps. Dark elves win 56 of 89 games for a .63 wps. As dark elves are closer to 1 (100%) for month 1 they would get top ranking, followed by Beastmen then Dwarves.

Next month beastmen win 67 of 125 for a .54, Dwarves have a good month winning 45 of 50 games for a wps of .90 while Dark elves go 60 of 70 for a .86 wps.

Month 2 WPS rankings:
Dwarves
Dark Elves
Beastmen

But we have to figure in their previous showings, so we add their percentage of previous ranking:
Dark elves were 1/3 for .66 points. (1/3-100)
Beastmen were 2/3 for .33 points
Dwarves 3/3 for 0 points.

Our new totals for month 2 are:

Dark elves .66+.86= 1.52
Dwarves 0+.90= .9
Beastmen .33+.54= .87

So by our completely made up numbers through 2 months Dark elves are the better army. :)

This system obviously has some flaws. We need mass data to average out any anomalies. We also need honest and frequent participation. This won't tell us everything, but I think after a while we can see trends, army usage numbers, etc. I think we would need to say that an army needs a certain amount of games before they could be counted. If Brettonia only gets 2 reports a month, and both are victories that obviously invalidates the results. I would say minimum 10 games reported to rank, otherwise they get 0. Or something like that.



I would also say that this should be limited ot 'average' game size - meaning no grand armies or smaller armies. 750 pt or 3000 pt armies change the basic game. Because of that I would limit participating games to 1500-2500 point values.

Some armies excel at higher points values, and others at lower values. Skaven, generally I would say have an advantage over ogres at 1500 just by sheer numbers and model points cost. However, using a wide spread of games would balance this out hopefully.

I am also not concerned over opponents, special characters, scenarios or 'power builds'. This is just a sampling of anything and everything that can be accomplished by an army against their peers, and then ranking those results mathematically.

As I said, there is probably a better way to do this. I am hardly the smartest person out there. Heck, I'm not the smartest person in this room and it's just me and the dog here. If someone has a better formula I'm willing to listen.


Ok - so what do we think? Is this possible? Will people participate if I am willing to do the legwork and run the numbers?

Thanks a lot,
Fee.

TheMadMarquis
25-09-2011, 16:43
Seems unnecessary when RankingsHQ already shows army rankings based on tournaments, which seem roughly what you'd expect (taking into account Tomb Kings and Ogres being dragged down by the results from their old books):

UK:


Vampire Counts
Dark Elves
Daemons Of Chaos
Lizardmen
Warriors Of Chaos
Skaven
Empire
Bretonnians
Orcs & Goblins
High Elves
Beastmen
Wood Elves
Dwarfs
Ogre Kingdoms
Tomb Kings

Note that there's quite a big gap between Skaven and Empire which relates pretty well to what we intuitively think about "tiers".

and USA:



Warriors Of Chaos
Lizardmen
Dark Elves
Skaven
Daemons Of Chaos
Dwarfs
High Elves
Orcs & Goblins
Empire
Vampire Counts
Ogre Kingdoms
Wood Elves
Tomb Kings
The Daemons-Dwarf gap in the States also shows a "tier" effect, with the top five armies being, very clearly, Daemons, Warriors, Lizardmen, Skaven and Dark Elves.

A couple of things that do suggest a bit of caution using these figures are the incredible discrepency in how Vampire Counts perform, and the fact that apparently NOBODY in the US ever uses Bretonnians.

Actually, the Vampire Counts divide might explain why there's quite a bit of controversy in their update thread on this forum about whether Ghouls and Grave Guard need a nerf. What on earth are UK VC players doing that doesn't translate across the atlantic?