Risk Aversion Rewared

[nrshida]

Today I ran into a player who has PMed me in the past to smack-talk me using relative rank as the core of his 'argument'. Today he was flying a Zero quite high and only getting brave when I was already 2 on 1 etc. Otherwise he'd just climb neutrally. Finally I found him alone and co-e and in two turns only shot him down. Howwibly. Like a dirty rat. As some kind of yellow coolant began leaking from his aircraft. I would rate his ACM understanding at below intermediate level although I've been told he's played for years. My Rank is something over 1100. His: 48.

It got me thinking why is AH scoring apparently focussed on promoting risk aversive gameplay? Is this by design? Could a more sophisticated scoring system promote more combat?

1 - Why is one pattern of behaviour / skill set anticipated for all players, when evidentially there is much variety?

2 - Why is the time of day / density of players not taken into consideration?

3 - Why is there absolutely no circumstance of the fight accounted for?

4 - Why is there no sort of handicap system in place? Now we have an achievement system, can the type and nature of training you do not be incorporated into the scoring system?

5 - When a player drags you into ack why does he automatically get a kill?

6 - If certain aircraft are perked, why aren't less-capable-than-average ones more rewarded in score (especially the specialists players in those)?

7 - Why aren't certain styles of ACM moderated versus others when it comes to scoring?

8 - When a squad is flying in cohesive groups why are they scored equally against those who fly without this advantage?

9 - As additional players join a ganging, shouldn't all those involved get an assist instead of a kill?

10 - Why does Rank, for instance, require you to carry out a spectrum of AH activities when some only choose to fly fighters or bombers or drive vehicles?

11 - In addition to a kill/death ratio why isn't a you/opponents average ratio also not recorded?

12 - Could a subjective-qualitive feedback system balance those purely playing for score and add (or re-add even) a qualitive later to gameplay?

[Vulcan]

Rank is fairly meaningless in the MA. People can milk run stuff and crank it up. Do 1 or 2 bombing runs, attack runs, and vehicle runs and you''ll be ranked in the top 150 easy. Just play and have fun is what I say.

[Crash Orange]

6 - If certain aircraft are perked, why aren't less-capable-than-average ones more rewarded in score (especially the specialists players in those)?

Aren't they reflected in the kill and damage points? You don't get the same score per town building destroyed in a Lanc as you get destroying the same building in a G5M.

5 - When a player drags you into ack why does he automatically get a kill?

Because you'd get a kill if you vulched him wheels down on the runway.

10 - Why does Rank, for instance, require you to carry out a spectrum of AH activities when some only choose to fly fighters or bombers or drive vehicles?

Because it's overall rank. If you're a specialist your rank in that individual category will benefit.

1 - Why is one pattern of behaviour / skill set anticipated for all players, when evidentially there is much variety?

3 - Why is there absolutely no circumstance of the fight accounted for?

7 - Why aren't certain styles of ACM moderated versus others when it comes to scoring?

8 - When a squad is flying in cohesive groups why are they scored equally against those who fly without this advantage?

I don't see how the game could possibly know most of these things to take them into account. How could the program possibly tell the difference between a squad flying cohesively and a bunch of people who are near each other but not cooperating at all? How would the game evaluate what style of ACM you're performing? You'd need about 50 or 100 employees at HTC doing nothing but watching films all day and manually adjusting skill based on play style. Not to mention that it would be absolutely impossible to do this objectively or to set standards that wouldn't have half the players screaming bloody murder because behavior they see as dweebish or gamey is being rewarded.

[nrshida]

Just play and have fun is what I say.

Was just thinking how influential playing for score was on the enjoyment of others. And what would happen if the scoring system became more capable.

Aren't they reflected in the kill and damage points? You don't get the same score per town building destroyed in a Lanc as you get destroying the same building in a G5M.

I don't know to be honest. I heard the Germans recorded two kills for each Mosquito at one point, and the Japanese had a kill-share system.

Because you'd get a kill if you vulched him wheels down on the runway.

I'm not talking about people rtbing, I mean those who deliberately use the ack to solve their ACM problems for them. Vulching should also score less I think.

Because it's overall rank. If you're a specialist your rank in that individual category will benefit.

Why is overall rank listed on the roster and not your speciality / best one then?

I don't see how the game could possibly know most of these things to take them into account. How could the program possibly tell the difference between a squad flying cohesively and a bunch of people who are near each other but not cooperating at all? How would the game evaluate what style of ACM you're performing?

I think most of those are quite straightforward to implement automatically.


You think everyting's fine with the scoring system exacty the way it is then?

[Randy1]

The rank has no meaning.   I don't know anyone that pays any attention to ranking since it can easily be manipulated.

I play a lot so my rank is meaningless compared to someone who can not play as often.

Then there are the players who use low risk attacks like vulching.  These people rarely engage in air fights without backup but they often abandon the players that save their butt.

I would bet if htc stopped posting rank, very few would even notice.

[crazyivan]

nrshisa was in AH2 ..so this doenst app.. was seeing if Ah II stil lworked as you killed the 109 ,,with 5 players on..

[crazyivan]

My old pc unable to get a 4 hour erection waa.. If i spend a grand ona new cupter which makes no sense ..titanfall? If dumbarse fugutive guy who you base your cleintel on show's up to nerf me..well Hitech nuff said y brotha.

170 player s in FSO last week!

[Lusche]

10 - Why does Rank, for instance, require you to carry out a spectrum of AH activities when some only choose to fly fighters or bombers or drive vehicles?

It doesn't.

What you refer to is, obviously, the overall rank. Which, by the very definition of overall, includes all of AH (except manned guns). But you still have also the various subranks. When someone really insists using rank to show off his 1337 fighter skilzz (which is wrong anyway), you gotta look at his fighter rank. If anything, overall rank just shows how much of a 'complete AH player' one is.

And when it comes to fighter rank, AH doesn't propagate total risk aversive gameplay, just partially .
For there is also k/h and k/s, kill points, hit%. In AH, it's very common to talk about 'fighter score' when you actually just mean K/D. But the K/D aIn't your score, and the K/D is only a small part of the fighter rank calculation. K/D is usually totally overrated by players when it comes to score/rank (not to speak of game enjoyment). If someone does only constantly run away from any kind of risk, he may 'protect' his pretty k/d, but not his score at all.

But in the end, score just tells you how good you are fitting into the AH scoring system. See it as a different sub-game. If someone else is using his rank to boast about his fighting skills, just ignore it. (However, the 'anybody can become #1' is very wrong either)

On a personal note: Ten times I was featured on the AH front page as the top fighter pilot. At no point I ever did believe I actually was. There always had been plenty of fighter pilots in the MA which thoroughly reminded me by mopping the floor with my plane that I am not. 

[Lusche]

I don't know to be honest. I heard the Germans recorded two kills for each Mosquito at one point,

Never. One kill is one kill, there was no multiplying nor any sharing.
What existed was, for some time, a point system for for medals, where, for example a fighter got you one point, a heavy bomber four.

[nrshida]

What you refer to is, obviously, the overall rank.

I feel that is inapplicable to someone who never drives a tank of flies a bomber for instance. Perhaps if a player has 0 in tanks, bombers etc, that shouldn't drag his average down?

And when it comes to fighter rank, AH doesn't propagate total risk aversive gameplay, just partially .
For there is also k/h and k/s, kill points, hit%. In AH, it's very common to talk about 'fighter score' when you actually just mean K/D. But the K/D ain't your score, and the K/D is only a small part of the fighter rank calculation. K/D is usually totally overrated by players when it comes to score/rank (not to speak of game enjoyment). If someone does only constantly run away from any kind of risk, he may 'protect' his pretty k/d, but not his score at all.

I was thinking that a more intelligent combination of those factors would be more useful. What would you design as the 'ideal' scoring system Lusche?

But in the end, score just tells you how good you are fitting into the AH scoring system.

Yes I agree. I didn't accept his 'argument' by the way. He just wanted to shut me up and thought that would do it. Lucky for him logical argument isn't scored.

I've always paid very little attention to the score. I realize now it's completely misaligned with my style of gameplay, for example. Especially with my more intelligent (system) ideas, I was thinking that vector information could also be combined with ordinary metrics for a more useful scoring system. I was also thinking to delimit reality from the game more clearly, like the squad example. In WWII everyone was in a communicating squad, essentially. Also no stats webpage in WWII.

Alternatively offering an opt out of scoring entirely would be interesting to me.


Is CrazyIvan telling me I'm old hat? 

[caldera]

Add in a few new scoring categories:


H/D, or Hit per death, as in how many players it takes to kill you.  This also may help indicate how much danger you fly into.
If you are an easy kill, your H/D will be close to 1.0, meaning it takes only 1 player to kill you.
If you typically own 1v1 engagements, it may take assists by many players to kill you.  The higher the number, the better.

Plane choice and luck of the draw on opponents can effect this, but over the course of a month, it would be another barometer of "skill" to incorporate with the others used to determine score. 

Hit per sortie (H/S) would be the same idea, only not requiring a death to count for your score. 



ENY/K, or ENY per kill.  This would account for what you use to make your kills in. 
An ENY/K of 5.0 would indicate someone who, all else being equal, takes the easy road. 
A player with an ENY/K of 29.8 would indicate a player of higher skill, than the above player.  Again, provided all else as being equal.

As score stands now, it is far easier to have a better score with the best planes.  Adding this in would require score oriented players to get some kills in a middling or lower end ride, or else get passed in rank by someone with similar numbers, but a better ENY/K.





Player A using current score model:          Player B using current score model

K/D  5.0                                                            K/D 4.5
K/S  3.0                                                             K/S 2.5
K/H  7.0                                                            K/H 6.5
Hit%  9.0                                                          Hit% 8.5
Kill Points  25,000                                              Kill Points 24,500

Player A obviously has a better score, but what if we made it tougher to have a high score?

Now adding in new score categories, with all others unchanged:

Player A                                                    Player B

H/D 1.5                                                              H/D 3.5
ENY/K  5.0                                                          ENY/K  29.5


Player B didn't score as well in the other categories, but did so in inferior planes and fared much better in dangerous situations against multiple enemies.   Adding in these score categories would give a better overall picture of skill than the current method.

[The Fugitive]

I think the time factor also plays into it. Those that get to play more have a number of categories bumped up, not to mention all that practice time.

Nrshida's fighter ranking at about 8.5 hours playing time,

Score      Rank 
Kills per Death + 1    0.59    637
Kills per Sortie    0.36    690
Kills per Hour of Flight    1.87    798
Kills Hit Percentage    1.16    1013
Kill Points    886.86    583

His nemesis's fighter ranking same day at just under 75 hours playing time.

Kills per Death + 1    0.78    387
Kills per Sortie    0.49    399
Kills per Hour of Flight    2.92    441
Kills Hit Percentage    4.74    340
Kill Points    38067.60    2

The kill points is a huge difference, and you just need time to build that one.

K/D and K/S is really two factors. Skill and style. The better fighter will have better numbers here, but if you fly in furball to the death all the time it will bring those numbers down. Those that fly careful..... AKA as nrshidas nemesis in this instance can inflate those numbers a bit.

Kills per hour of flight, this is another number that can be inflated due to time playing. Lets say you find a single great fight where you can get a bunch of kills every ten hours. Nrshida gets one such fight were his nemesis gets 7 of them. In these opportunities your kills for that hour can be very high over your average and so bump it up.

Hit %, is just aim right? Well you can bump this one as well just by hitting buffs. They take a lot more rounds to drop, and as a bigger target are much easier to hit. Find a few dive bombing wonders and rake the crap out of them as they lawndart and your hit% will climb.

Luckily rank is only important in the game when it comes to CV groups, which is too bad but there it is. Just because someone can tweak there rank, or is willing to spend the time to do it certainly doesnt make them a better fleet admiral.

There are a bunch of loop holes in the scoring system. There will always be those that will look for and exploit those loop holes making any type of ranking/comparison a waste of time. I don't think there is any way to make it "fair" as there are just too many factors involved due to time, styles, arena numbers and even the maps that can help a player tweak his/her rank.

I look at it this way, if a beat a player consistently, no matter their rank, I fly this cartoon game better than them. Unfortunately, there are not very many I can beat consistently     

[Lusche]

I feel that is inapplicable to someone who never drives a tank of flies a bomber for instance. Perhaps if a player has 0 in tanks, bombers etc, that shouldn't drag his average down?

Why should it not? After all, it's overall meaning how well you do fare in all things combined. If you never tank, your overall rank should 'suffer' accordingly, as you are not that good overall.
If I want to compare myself with others only in certain categories, I'd use the specific sub ranks.

By the way, I actually look at lot at other people's scores (apart from the obvious reasons of having to do so in my position as the unofficial AH statistician), but I only 'judge' people by combination of score and plane statistics. I entirely ignore the ranks. I don't care about who's number one in this and that, and I'd bet neither do the huge majority of AH players.

What would you design as the 'ideal' scoring system Lusche?

I don't know anything like that. While I do would have two rather minuscule changes to it (with one being entirely for statistical reasons), I don't see a way to make it that much more fair. And I'm afraid, many have a very different, distinguished conception of a 'better scoring system' based entirely upon their own playing style

[Lusche]

Add in a few new scoring categories:


H/D, or Hit per death, as in how many players it takes to kill you.  This also may help indicate how much danger you fly into.

Which may be fun, but in my book that could also be a sign for an incompetent fighter pilot with a huge lack in situational awareness.

As score stands now, it is far easier to have a better score with the best planes.

That's why the only major modification I'd propose would be replacing the fighter points category with perks earned. this simple change would factor in your ride, the enemy ride, and your side's perk modifier. Right now I massively benefit from flying late war perk monsters and shooting up enemy bombers. (Not only 3 quick kills, but also heaps of score points)

But even that would be gameable, so to speak. The hot shots of AH would congregate onto a very few "sweet spot" ENY 15-20 rides, and the ENY value debates would get quite heated.

[caldera]

Which may be fun, but in my book that could also be a sign for an incompetent fighter pilot with a huge lack in situational awareness.

An incompetent fighter pilot could never do well enough in K/D, K/S or Hit % to ever have a high score.


Oh, and a modification to your idea would be to change from "perks earned" to "perks earned per sortie".  Having a cumulative total favors players with more hours, while an average per sortie is fair for all.