Jump to content
Gryphon_

WN8 Expected Values Update

Recommended Posts

Personally, I think it should work as a curve that applies for all players (well, at least for players who played most of their battles while the metagame was largely similar to the way it is now - basically after the MM tier spread reduction). I don't see any purpose in purposefully gearing it towards "most" players in the middle of the sample, and it's just a little selfish to do it only for the top 1%, not caring about sub-yellow players (although technically, if it's accurate for the top 1%, it's probably accurate for them too).

 

Moreover, throwing away 1/2 of the accuracy for the top 1% just so the bottom 99% is 1/10 more accurate is stupid to me.

Share this post


Link to post
Share on other sites

Why do you need to do extra work to "fix" it for obvious clubbers? They get purple stats regardless so it is still down to not being a moron and looking further than the WN rating. How do you fix players like XDeathLordX with a 50% win rate and blue WN8 in tier 10? His recent is sub 48% and dark green WN8....actively harming his teams chances by hitting the battle button - still manages dark green WN8. Doesn't say maybe it's a little broken there?

 

Believe it or not, but when you work on a rating like Eff or WN you have a certain amount of responsibility. Dont you remember what happend after a while when Eff was introduced or widespread? Suddenly "80%" of the players where sitting half the game in that stupid cap circle. That didnt promote skill and was really bad for the gameplay. With WN4-7 we managed at least to a certain degree get rid of that nonsense and the gameplay improved. Some of us spend dozens of hours talking sense into people who were convinced that 4+ cap/battle was the epitome of pixel tanking.

WN4-7 had its own issues, but these werent that great compared to the cap nonsense and the gameplay overall improved, when it comes to that topic of cap.

With WN8 we made an other big step forward and even critics like Garbad, who seemed of the opinion he wasnt purple enough on some tanks - my perception, acknowledged that its an improvement. Preator and the rest of the gang here even managed to get a handle on sealclubbers to a certain degree, LTs and SPGs were rated fair, though not perfect, for the first time since ratings exist. Wannabe ProScouts still bitch about it because we didnt serve purple on a silver tablet, and real unicums can do some serious padding with certain light tanks, but overall it works even for lights.

 

So why should we throw away parts of the progress we made?

 

 

WN8's credibility is built around it being a great metric for comparing damage padders. I doubt fixing issues will really change that perception.

 

There is no doubt a certain amount of empty damage farming is going on with WN8. On the other hand a great deal of players still havnt understood how WN8 really works. I m running the german WN8 thread in the Eu-Forum, were I had a player "complaining" just a few days ago, that when comparing 2 games with the same tank, the higher damage game didnt yield a higher Wn8. He was suprised that Kills/Def/Spots make a difference.

I think over time that knowledge will spread and if not it will not hurt the real purples, because they arnt only farming empty damage.

 

 

Realistically you'd probably need to split the metrics into one for those padding it and one for the rest.... but good luck with that.

 

 

Impossible - and get rid of your fixation on padding. Usually things like that arnt healthy

 

                                                                  

 

Edit:

 

What I dont want see happening is that suddenly blue/purples drop down 2-4 avg tier points. That would even be worse for the game overall than the cap madness. I dont give a shit if [EFE] had to rename their slogan from "Everything for Eff" to "sealclubbing ahoi", but culling the WoT newbies wont do us any good in the long run especially on servers where the population isnt that great to begin with.

Share this post


Link to post
Share on other sites

I'm mostly ignorant here, so I apologize in advance if my comments are stupid. Reading through the thread it appears that updating the expected values is causing some consternation because the expected valued are going up for higher tier tanks and down for lower tier tanks. Is that correct?

I also saw a comment that the initial set of values used was biased upwards for low tiers and downwards for high tiers, which makes sense as part of the reason for moving to WN8 in the first place was to highlight those folks "padding" WN7 in lower tier vehicles.

It seems to me that this is kind of a natural byproduct of chasing stats. People saw they could get "better stats" by playing tier 10 tanks (which is something I've seen commented on many times), so they played tier 10 tanks. Since more above average players were playing these tanks, the tank's average stats increased. Now if you re-baseline these tanks people will have their stats effected.

The opposite happened at low tier, the "good" players stopped playing low tier tanks so more poor players dragged the numbers down, and the expected values for at least some of those low tier tanks were inflated to begin with.

Share this post


Link to post
Share on other sites

I'm mostly ignorant here, so I apologize in advance if my comments are stupid. Reading through the thread it appears that updating the expected values is causing some consternation because the expected valued are going up for higher tier tanks and down for lower tier tanks. Is that correct?

I also saw a comment that the initial set of values used was biased upwards for low tiers and downwards for high tiers, which makes sense as part of the reason for moving to WN8 in the first place was to highlight those folks "padding" WN7 in lower tier vehicles.

It seems to me that this is kind of a natural byproduct of chasing stats. People saw they could get "better stats" by playing tier 10 tanks (which is something I've seen commented on many times), so they played tier 10 tanks. Since more above average players were playing these tanks, the tank's average stats increased. Now if you re-baseline these tanks people will have their stats effected.

The opposite happened at low tier, the "good" players stopped playing low tier tanks so more poor players dragged the numbers down, and the expected values for at least some of those low tier tanks were inflated to begin with.

 

I cant comment on the NA server, but the thing is that WN8 didnt really reduce the amount sealclubbing on the EU server afaik. Though it made it harder to get a good rating doing it and that was intentional.

Share this post


Link to post
Share on other sites

 

Impossible - and get rid of your fixation on padding. Usually things like that arnt healthy

Indeed - there are already deep purple WN8 players getting about with high 50s win rates. I guess this is what happens when people shift their priorities to 'less healthy' things - assuming that's what you meant by healthy.

 

But basically Treeburst looked to have hit the nail on the head. Those keen to pad WN8 (call it what you like) favoured the easiest method of doing so. Moving the goal posts will likely change the habits of those most concerned with the metric. There's usually a week or two lag between a new metric formula being released and people figuring how best to game it.

Share this post


Link to post
Share on other sites

I dont know what s going on that strange servers of yours, but I m confident in my claim that switching from solo random to platoon at a certain skill level, will give 10%+ WR on Eu server and wont decrease WN. Some people it seems already lost the sense of reality when it comes to winrates.

 

 

unbenannt1yjk5.jpg

Share this post


Link to post
Share on other sites

I dont know what s going on that strange servers of yours, but I m confident in my claim that switching from solo random to platoon at a certain skill level, will give 10%+ WR on Eu server and wont decrease WN. Some people it seems already lost the sense of reality when it comes to winrates.

 

It kind of effects WN too I am finding out going 100% solo now.

 

You just have so many battles where 2 more decent players on your team would have allowed you to at least do decent damage in a lost battle.

Share this post


Link to post
Share on other sites

It kind of effects WN too I am finding out going 100% solo now.

 

You just have so many battles where 2 more decent players on your team would have allowed you to at least to decent damage in a lost battle.

 

I was cautios on purpose here.

Share this post


Link to post
Share on other sites

I dont know what s going on that strange servers of yours, but I m confident in my claim that switching from solo random to platoon at a certain skill level, will give 10%+ WR on Eu server and wont decrease WN. Some people it seems already lost the sense of reality when it comes to winrates.

Absolutely. My current theorem is that you can express ones positive or negative influence on a battle based on a standard deviation from a mean. So putting 2 x +1SD players in platoon will get the win rate of roughly a +2SD solo player. I don't know the actual spread of WR but for illustrative purposes this would be roughly 2x 52%ers would win at about 58% when platooned. Which would be roughly a high range solo player. 3 at maybe 65% - or amongst the very best solo players going around. My guess is this would hold for any of the servers - but perhaps have a stronger influence on the servers where the average ability is thought to be lower (i personally doubt there is that much variation - but others love to go on about it).

 

As to the effect on WN... hard to say. Survivability is clearly going to help WN8 and therefore be helped by platooning. Having to share damage may not... but overall i've seen marked improvements in my WN8 when i platoon.

 

But the point about padding WN8 relates more to playstyle than platooning. Whether it's healthy really comes down to whether a player feels winning is more important than WN8 i guess.

Share this post


Link to post
Share on other sites

Who decides whats right?

 

Clearly the results show that people with an above 6 average tier gets less WN8 rating with this expStat Table.

And those who enjoy sealclubbing at 3-5 receives a massive bonus to their WN8 score.

 

 

Find out what calculations Praetor, or try to replicate it, otherwise it would just be like a total new rating, and there are enough of them.

 

 

I already did find out the calculations that Praetor used to derive expected values. I worked with him for weeks in Jan-Feb to help him to release an update. The numbers he (didnt release) then are pretty much the numbers I have released now - high tiers are set too low, low tiers are set too high. I will produce a side by side comparison of the expected values for the 'current' table, the 'Feb' table I did for him using his method,and the 'May' values I just put out. I will include a huge graph that shows that for damage, both potential updates agree that the current values were low in high tiers and high in low tiers. I will do this tonight.

 

As to who decides what is right, I think the answer is the website owners like you,

 

The way ahead here is for me to give everyone visibility of the numbers that came out of that previous work, and also to put up some scatterplot graphs showing the damage done by high tier tanks - to illustrate that players are hitting those numbers. Owners of other websites are most welcome to throw out data and graphs.too. 

Share this post


Link to post
Share on other sites

Pointless. WN8 has to work for the vast majority of players, not well-known outliers, and most purples (and terribads) fall into that category.

 

This has to to be right for the 95%, not the 5%. I know that's not going to sit well with all the purples in here, but most of you are smart enough to see that. 

 

I'm not overly concerned with what is going to sit well with the purples here.  I am, on the other hand, concerned with how this metric will affect the game.  

 

 

Believe it or not, but when you work on a rating like Eff or WN you have a certain amount of responsibility. 

 

 

This exactly.

 

 

 

Gather round, kiddies.  Let ol' uncle Deus tell you stories from the history of the world... of tanks.  

 

In the beginning, SerB created the tanks and the maps.  And he placeth the tanks upon the maps, and madeth half appear hostile to the other half.  And he sayeth unto them, "Thou shalt kill."  And so they rolleth about the maps, shooting at their red foes until death.

 

In these days, the world was void and without form or physics.  Surely fortune alone determined victory, for the light of wisdom had not been made known to the tanks.  Until tanks drove deep into the caverns of Pagorki, and therein found The Statistician.  And The Statistician sayeth unto them, "Vieweth ye the Service Record"  And the tanks replied "Knoweth ye not the words of SerB?  That whosoever vieweth the Service Record shall surely die?"  And The Statistician said "Ye pubbie shitlourde, ye surely shall not die.  For SerB knoweth that when you vieweth the Service Record, ye shall become as gods, knowing wins from losses."

 

And so they viewed the Service Record, and their eyes were opened, and they knew they were bad@tenks.  They left the cavern, and fashioned repair kits and fire extinguishers to cover their badness, and began to seek wins.  And that place was thereafter known as "Mines", for the knowledge that was mined from the depths.  And young tanks still seek the central cavern, hoping to once again gain wisdom from The Statistician.

 

Thus ended the age of ignorance, and began the age of winning.  Wins were sought above all else.  Wins were known as the mark of greatness.  But there existed among them fools, who thought themselves wise.  They sought to betray the eyes of their comrades, by amassing wins through idol worship and sorcery low tier clubbing and tank companies.  They cried "Behold, I am unicum.  My 60% win rate makes it so."  And the people said "You're terrible in anything higher than a tier 2"

 

Into this darkness a new light of wisdom shone.  The light was called Efficiency.  Efficiency measured the individual observable components that contributed to wins, as they were understood in that time.  Efficiency rewarded damage, and kills, and spotting, and defense, and capping.  The win manipulators tore their clothing and gnashed their teeth.  In sackcloth and ashes, they mourned.

 

But their mourning did not persist, for they learned the terrible flaw of Efficiency.  They came to know that spots and flag points were as valuable as kills.  And thus, the brightness of Efficiency began to fade.  White circles throughout the world were sought out, and parked in.  Efficiency scores rose, no longer representing the skill of the players.

 

Wars raged between those faithful to wins, and those faithful to efficiency.  The wise sought alternate means by which to judge a player.  Many potential faiths rose and fell in these troubled times.  The darkness was finally pierced by the WNx system.  Rather than requiring faith, it attempted to use reason.  Conclusions were based on observations.  WN evolved, and improved.  Each improvement earned new followers.  Efficiency faded to insignificance, but the twin terrors of clubbing and TCs once again threatened to obscure the light of reason.

 

As WN evolved, it tried to limit the effect of running a single OP tank to pad stats.  Players reacted by diversifying.  WN tried averaging tiers.  Padders balanced low and high tier play.  Finally, the 8th version of WN judged each tank on its individual merits.  Still, padders sought out tanks with low expected values, and began farming damage.

 

Short story, for you apathetic ADHD teens:

 

Each popular skill metric has caused a change in game play.  Win focus caused clubbing and TC abuse.  Efficiency caused spot/cap abuse.  WN has caused a few abuses, depending on which version you look at.  Some caused certain vehicles in a tier to be overplayed.  Others caused clubbing.  Others caused damage farming, which is probably the least offensive abuse.

 

If the proposed changes are added in as they stand, it absolutely will cause a shift to lower tiers.  People want their purple numbers.  Purple numbers are an admirable goal, but we must remember that purple is supposed to indicate the top ~1%.  Providing an easy way to pad your way to purple completely defeats the purpose of having a purple category.  

 

I think there is, however, some merit in the numbers.  I was surprised at how low the expected values for some high tier tanks were.  They probably should rise.  Xvalues for the popular farming tanks should probably rise as well.  But we need to keep in mind the fact that purple level play in low tiers FAR exceeds the results that same level of play earns in top tiers.  Hitting 60% and 2 kills per battle in an MS-1, where most of your opponents have 50% crews and no equipment, is a much lesser accomplishment than hitting 60%/2kpb in a tier 10.  

Share this post


Link to post
Share on other sites

4a18e881da.png

 

 

Already posted some EU burbles, so here are some NA #goldenshowers, and #sealclubbers. * ( Pinkman_Jesse and Dyaebl are EU. )

 

Gibe me names and I'll check em up.

 

Also, here be table "neatly" readable by the human eye.

expected_values_9_0.html

Can you do mine?  I'm curious how a guy who mostly plays tier 7/8/9 does..

Share this post


Link to post
Share on other sites

I think there is, however, some merit in the numbers.  I was surprised at how low the expected values for some high tier tanks were.  They probably should rise.  Xvalues for the popular farming tanks should probably rise as well.  But we need to keep in mind the fact that purple level play in low tiers FAR exceeds the results that same level of play earns in top tiers.  Hitting 60% and 2 kills per battle in an MS-1, where most of your opponents have 50% crews and no equipment, is a much lesser accomplishment than hitting 60%/2kpb in a tier 10.  

 

This is exactly how I feel.  It just looks to me like Gryphon is putting blinders on with regards to be bolded part, which is worrisome.

Share this post


Link to post
Share on other sites

I'm not overly concerned with what is going to sit well with the purples here.  I am, on the other hand, concerned with how this metric will affect the game.  

 

 

 

This exactly.

 

 

 

Gather round, kiddies.  Let ol' uncle Deus tell you stories from the history of the world... of tanks.  

 

In the beginning, SerB created the tanks and the maps.  And he placeth the tanks upon the maps, and madeth half appear hostile to the other half.  And he sayeth unto them, "Thou shalt kill."  And so they rolleth about the maps, shooting at their red foes until death.

 

In these days, the world was void and without form or physics.  Surely fortune alone determined victory, for the light of wisdom had not been made known to the tanks.  Until tanks drove deep into the caverns of Pagorki, and therein found The Statistician.  And The Statistician sayeth unto them, "Vieweth ye the Service Record"  And the tanks replied "Knoweth ye not the words of SerB?  That whosoever vieweth the Service Record shall surely die?"  And The Statistician said "Ye pubbie shitlourde, ye surely shall not die.  For SerB knoweth that when you vieweth the Service Record, ye shall become as gods, knowing wins from losses."

 

And so they viewed the Service Record, and their eyes were opened, and they knew they were bad@tenks.  They left the cavern, and fashioned repair kits and fire extinguishers to cover their badness, and began to seek wins.  And that place was thereafter known as "Mines", for the knowledge that was mined from the depths.  And young tanks still seek the central cavern, hoping to once again gain wisdom from The Statistician.

 

Thus ended the age of ignorance, and began the age of winning.  Wins were sought above all else.  Wins were known as the mark of greatness.  But there existed among them fools, who thought themselves wise.  They sought to betray the eyes of their comrades, by amassing wins through idol worship and sorcery low tier clubbing and tank companies.  They cried "Behold, I am unicum.  My 60% win rate makes it so."  And the people said "You're terrible in anything higher than a tier 2"

 

Into this darkness a new light of wisdom shone.  The light was called Efficiency.  Efficiency measured the individual observable components that contributed to wins, as they were understood in that time.  Efficiency rewarded damage, and kills, and spotting, and defense, and capping.  The win manipulators tore their clothing and gnashed their teeth.  In sackcloth and ashes, they mourned.

 

But their mourning did not persist, for they learned the terrible flaw of Efficiency.  They came to know that spots and flag points were as valuable as kills.  And thus, the brightness of Efficiency began to fade.  White circles throughout the world were sought out, and parked in.  Efficiency scores rose, no longer representing the skill of the players.

 

Wars raged between those faithful to wins, and those faithful to efficiency.  The wise sought alternate means by which to judge a player.  Many potential faiths rose and fell in these troubled times.  The darkness was finally pierced by the WNx system.  Rather than requiring faith, it attempted to use reason.  Conclusions were based on observations.  WN evolved, and improved.  Each improvement earned new followers.  Efficiency faded to insignificance, but the twin terrors of clubbing and TCs once again threatened to obscure the light of reason.

 

As WN evolved, it tried to limit the effect of running a single OP tank to pad stats.  Players reacted by diversifying.  WN tried averaging tiers.  Padders balanced low and high tier play.  Finally, the 8th version of WN judged each tank on its individual merits.  Still, padders sought out tanks with low expected values, and began farming damage.

 

Short story, for you apathetic ADHD teens:

 

Each popular skill metric has caused a change in game play.  Win focus caused clubbing and TC abuse.  Efficiency caused spot/cap abuse.  WN has caused a few abuses, depending on which version you look at.  Some caused certain vehicles in a tier to be overplayed.  Others caused clubbing.  Others caused damage farming, which is probably the least offensive abuse.

 

If the proposed changes are added in as they stand, it absolutely will cause a shift to lower tiers.  People want their purple numbers.  Purple numbers are an admirable goal, but we must remember that purple is supposed to indicate the top ~1%.  Providing an easy way to pad your way to purple completely defeats the purpose of having a purple category.  

 

I think there is, however, some merit in the numbers.  I was surprised at how low the expected values for some high tier tanks were.  They probably should rise.  Xvalues for the popular farming tanks should probably rise as well.  But we need to keep in mind the fact that purple level play in low tiers FAR exceeds the results that same level of play earns in top tiers.  Hitting 60% and 2 kills per battle in an MS-1, where most of your opponents have 50% crews and no equipment, is a much lesser accomplishment than hitting 60%/2kpb in a tier 10.  

So stop trying to define purple play as 2x good, and start defining purple play as top 1% of players in that tank/tier/etc.

Share this post


Link to post
Share on other sites

Well it was worth all the work just to read what Deusmortis just posted.....!

I hear you about the potential meta changes but don't you think that raising the bar on high tiers will keep the bads out of tier ten? I think this would also reduce the number of players in tier ten TDs , which would open up the game a lot more at that level, and result in more aggressive and less campy play.

However, I stress that I will not support any update that adjusts the values based on meta; the values must reflect the true performance of the players.

Share this post


Link to post
Share on other sites

... It just looks to me like Gryphon is putting blinders on with regards to be bolded part, which is worrisome.

 

 

Calm down pls

 

Gryphon isnt doing anything on purpose here, he is just presenting the numbers. Attacking the numbers is ok, and one of the main reasons I think Praetor tried to come up with a 2 point system were these numbers, but attacking Gryphon who is doing the number magic on his free time for us wont be received well here.

Share this post


Link to post
Share on other sites

Who said I wasn't calm?  I'm not trying to attack Gyphon himself, it's called having a discussion.  My point is that a valid concern hasn't been addressed even though it's been brought up by several people throughout the thread.  Instead he kept pushing the linear scale model, which is inherently flawed for the reasons above.

Share this post


Link to post
Share on other sites

So stop trying to define purple play as 2x good, and start defining purple play as top 1% of players in that tank/tier/etc.

 

Strongly agree.

 

I haven't been keeping up with the development, but wasn't this one of the goals for WN9?  One data point to determine average, expected values, and a second data point to define unicum level play?

 

Well it was worth all the work just to read what Deusmortis just posted.....!

I hear you about the potential meta changes but don't you think that raising the bar on high tiers will keep the bads out of tier ten? I think this would also reduce the number of players in tier ten TDs , which would open up the game a lot more at that level, and result in more aggressive and less campy play.

However, I stress that I will not support any update that adjusts the values based on meta; the values must reflect the true performance of the players.

 

No, bads are generally unaware or deniers.  They lose because they get bad teams.  They have bad stats because they do great things that aren't captured by stats.  They play tier 10 because they're convince they're great, and will always be great as the top tank, and anyone who beats them is cheating/hacking/firing gold/etc.  

 

I'm not sure if it will keep the TD population down either.  Bads play TDs because they're easy.  Goods play them because they're effective.  I honestly can't predict how a bar raising would affect that paradigm.  

 

Once again, I have no issue with raising the bar for high tiers.  Even if that drops me back into the blue range (I consider myself a borderline uni at best.)  I

just don't want to see a situation where players can easily farm 5-6k rated games in tiers 1-2, and start thinking themselves great.  Seriously, my M5 Stuart has over 6k WN8 at 77 games with the proposed numbers.  Those average numbers will never shift much, since there are a hundred new bads for every low tier clubber.  

Share this post


Link to post
Share on other sites

I'm not pushing anything new, the WN8 system is what it is and I'm not trying to change it; on the contrary, Im trying to sustain it

Regarding the ease of achieving purple scores in a tier one, if you really want to inflict the misery of low tier play on yourself then go for it, but by virtue of the design of the WN8 system you will have to make almost double the expected values to get purple. WN8 is way better than WN7 in that regard. Also, the average tier stat will mark the csealclubbrs for us all to see.

Perhaps we should get some visibility on sealclubbing to see if its worth worrying about. If one of the site owners could link me a csv file with userid, global winrate, WN8, and average tier I'll see what graphs I can produce.

Share this post


Link to post
Share on other sites

I had PM'd BJ with this, but I guess I can ask here, and if one of the mods feel like it should be split from this, by all means delete and I can make a new topic:

 


 

I want to see how a metric (or 4 metrics to be precise, still kinda brainstorming) that was tier 8-10 class-based would look.

 

I'm guessing I have to figure out how to shave the number of steps in the equation to make it workable, but in the meanwhile i was thinking of just averaging existing wn8 scores in each group. 

 

Lastly, was thinking of a way to take these values and assign a variable based on games played per-class,IE someone has an aggregate wn8 of 2000 in 8-10 med/light, but has significantly more arty games in 8-10, so end result would be 2000A.

 

If you can help point me in the right direction, maybe tell me which of these isn't feasible, I'd like to see what I can do to help assist in metric creation. 

 

 

 

I'm not looking to foist my project on someone else, so if anyone could offer tips on where to start, and what's pants on head retarded from a modelling standpoint, I'd appreciate it. 

Share this post


Link to post
Share on other sites

I don't think there's a massive seal clubbing problem currently because of how WN8 is implemented.  The bigger worry is that by lowering the expected values in low tier tanks it will create a seal clubbing problem.  Like Deus said above, people will chase the purple stats no matter what drives the metric.  If tier 1s are the easiest way to achieve purple, people will flock to tier 1.

Share this post


Link to post
Share on other sites

I judge rating systems by the accuracy they offer as a tool to evaluate my own play or that of others. WN8 is so far clearly the best we ever had. But I really don't like what I saw in those new proposed number sheets.

 

I believe we can all agree that every rating has some effects on the behaviour of those players that are aware of stats: increasing the bar for high tiers while lowering it for low tiers is contra-productive. Of all ratings WN8 had the least harmful side-effect. I wouldn't receive an overhaul well that encouraged seal-clubbing.

 

If you use the same bar for high and low tiers you get a flawed rating since it is considerably more difficult to dominate high tier games than to club the crap out of some helplesss newbies. Praetor gathered his low tier expected values from accounts with a high battle number for a reason.

 

And although it came initially as a shocker for me, I believe the original WN8 creators were correct in adjusting some expected values by hand (Bison is one example if I remember correctly). In your new set of data one pair of tanks that caught my eye were the E-100 and the VK 72. The expected values for the latter are considerably higher in your new version. This isn't a suprise since ownership is limited to a group of players that has an above average skill-level. But in reality they probably should be lower since very few would seriously argue that the VK72 is the superior tank.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...