Aces High Bulletin Board
General Forums => Aces High General Discussion => Topic started by: dtango on September 22, 2006, 03:19:55 AM
-
The debates about gameplay have raged even back in 2001 and I suspect even longer than that. It's been really tiring to see the same ol' stuff be argued ad-naseum. If you don't believe me just do a few searches in the BBS and see. Mostly I get a chuckle out of how zealous folks can get on the topic regardless of which religion they profess.
There is an aspect of the gameplay debate though that has never failed to disturb me and that is how little recognition there is of just how terribly fallible we are as human beings to our own biases and filters and how easily we miss our own blinders.
"...humans are poor at evaluating probability. We tend to see patterns where there aren't, overestimate the probability of low frequency events and, most importantly, have a tendency towards comfirmation bias -- looking for evidence that confirm our preexisting beliefs."
With that in mind I thought I would try to demonstrate ways of challenging ourselves to be more objective and intellectually honest about the reality of things. We tend to extrapolate our individual experiences as a given that the majority of folks experience the same things. This is a terribly inaccurate way of sizing up what is happening as a whole.
So let's take a look at how much or how has gameplay changed in the Aces High MA of old etc? Are there some ways to tangibly measure the above available to us? A general proxy could be to look at stats available to the public AH community through the monthly kill stats database. With that as a premise I went and pulled stats from 2001 to 2006. To make it a somewhat reasonable exercise for me timewise I just did snapshots for the months of July and Dec from 2001 to 2006 (July 2001 was the earliest I could find monthly kill stats recorded for AH).
HOW HAS THE MA PLAYER BASE FLUCTUATED AS A WHOLE?
===========================================
If we assume that total number of kills (which also equals total deaths) in the MA as a proxy for the total MA player base population what does that tell us about the population?
(http://thetongsweb.net/412th/ahmatotalkills.gif)
So how has the MA population changed? From Jul 2001 to Jul 2002 AH experienced a whopping 250% rate of growth. From Jul 2002 onwards until now the rate of growth has essentially leveled off with roughly a +/- 25% change. One of the possible conclusions is that players coming roughly equals the players leaving therefore balancing each other out which is resulting in a plateau in growth for AH over the last 4 years.
Folks who believe that a controversial change will be the death knell of HTC should take note of this because in those 4 years there have been some big changes with the AH2 and the ENY limiter system happening in that span yet there is no evidence that those changes adversely impacted their business if these stats are representative of total HTC player base subscription.
Let's take a look at some other stats. The following are stats based on essentially the top 30 models in AH in terms of kills % and deaths %. I arbitrarily cut it off at the top 30 to keep the graphs and tables somewhat readable. It's also important to note that the top 30 make up about 80-85% of the total kills and deaths in the old MA. There are 86 model types available in the planeset and vehicle set in AH. Someone has already pointed this out recently (sorry I can't remember who it was that posted the stats!) but it's worth repeating. That means 1/3 of the planeset and vehicle set in AH account for 80-85% of the kills/deaths in the MA while the other 2/3 account for the last 15-20% (which is a noteworthy stat).
HOW HAS THE PLANESET AND VEHICHLE SET USAGE CHANGED IN THE MA?
===========================================
One way we might extrapolate this is to look at distribution of % of total kills or deaths among planes and vehicles. The following charts are the top 30 models sorted in order of average % of total kills or deaths.
(http://thetongsweb.net/412th/ahmakillpct.gif)
(http://thetongsweb.net/412th/ahmadeathpct.gif)
This is a pretty revealing set of tables and graphs that reveal some fundamental shifts in model usage in AH which are bit surprising. Several of the significant global trends stick out.
(1) By far the greatest increase in usage is the Panzer IV compared to all other models of planes and aircraft.
(2) The Spitfire XVI quickly replaced the usage of Spitfire V’s and Spitfire IX’s.
This would indicate that of all the gameplay changes in AH, the one that has been significant is that of the increase in the GV battles.
2ndly, the data appears to indicate that overall the other aspects of gameplay in terms of plane and vehicle usage hasn’t dramatically shifted and suggests that the idea that a significant increase or decrease in furballing or base capturing just isn’t reality over the course the AH history since 2001. If they did, I would expect to see things like a drop in the use of N1K’s, a rise in the use of B-17’s, etc. etc.
All that being said these charts might not tell us the whole story because it is hard to gauge the volatility in kills and deaths of models given the scale being examined above. The numbers represent % of the total so it’s harder to understand what a .5 % change for a model is for instance.
Another way to get a better sense of the volatility and trends in planeset and vehicle-set usage is to look at specific model and to understand the % above or below the median the kill and deaths fluctuate. In other words if I take a specific plane and look at the median % of total kills or deaths in the MA, the amount of flunctuation above or below the median for that plane will tell me how much change that plane has experienced relative to it’s own usage.
Here are the tables and graphs:
(http://thetongsweb.net/412th/ahmakillspctvol.gif)
(http://thetongsweb.net/412th/ahmadeathpctvol.gif)
Note that the median min and max volatility % for the top 30 models are:
- for kills: -50% and +33% respectively
- for deaths: -39% and +33% respectively
These tables give us a better idea of how much volatility each of the top 30 killers or death models experienced from 2001-2006. The obvious ones from the previous charts stickout such as the Panzer IV, Spit IX, and Spit V changes.
Some other ones that we now see that have experienced noticeable changes are (fluctuations above and below the overall top 30 min/max medians).
(1) drop in P-51D usage
(2) drop in P-38L usage
(3) rise in bF110G-2 usage
(4) rise in Hurri II C usage
(5) drop in bF-109K4 usage
(6) drop in P-47D-40 usage
(7) drop in F6F-5 usage
(8) drop in B-17 usage
(9) crop in c-47 usage
Everything else with a few exceptions seems to have remained fairly steady over the course of time in AH.
Anyway, these are but some stats that might bring some rationality to the gameplay discussion. Overall what I'm struck with is..
(1) how significant the ground war is growing in AH
(2) the slow reduction in diversity of aircraft usage in AH
What I don't see evidence of is that "furballing" is diminishing - but rather the diversity of the types of planes engaged in a furball are, nor do I see the base capture / strat gameplay increasing.
It'll be interesting to revisit this topic over time with the new arena settings.
Tango, XO
412th FS Braunco Mustangs
-
Heck-of-a-post,dtango! :aok
I'll need some time to digest this data.
Cool!
Thanks
Sun
-
Can drop in B-17 use be attrbuted to appearance of B-24?
Likewise, the P47-D40 is replaced by P-47N quite often now.
-
Originally posted by dtango
It'll be interesting to revisit this topic over time with the new arena settings.
Tango, XO
412th FS Braunco Mustangs [/B]
I expect Hitech will have an interest also.
Phew - a mighty tome dtango, but good to see some science applied to a topic normally buried in anecdote and personal bias.
Thanks for that:aok
-
Originally posted by Booz
Can drop in B-17 use be attrbuted to appearance of B-24?
As a bomber jock, that would be my guess. Also, I fear my lack of skill brought down the 109K :cry lol. Just to stick it to the man and be non-conformist, I swear not to touch a K for 6 months. Instead, I will fly the G-14! Get that pretty little bird on the list.
-
Great job. But I need to go sit down. My brain hurts.
-
WOW! Good post dtango! :eek:
-
Aside from my pesonal feelings on this subject let me add a thought i have had for a number of years....
First a question:
WHO REALLY PLAYS AH? especially vs other games on the internet...
I am 37 yo and developed my love for WW2 air war because i spent a year in a British boarding school...absolutely fell in love with the battle of britain.
when i look around the members of my squad and the other peeps i know, i would say that by far the average subscriber is 30-60 yo. i am guessing that we are the generation who remember ww2 or had relatives in ww2 and therefore have an interest in the gameplay that gets us to do a trial period.
My fear has akways been that this game would eventually dry up as our "generation" got older and that there would not be enuf teenagers/20 yo to replace our ranks. most of the 20ish guys i know are active or were active military....(persoanlly i would advertise on armed forces radio and other military locales all day long) i think they have to be the future of this game.
the second question is of course: What creates a long time customer?
lol..gotta run...will post more later today:aok
-
great thread btw...hope we can take the discussion in this direction:aok
-
Originally posted by SuperDud
WOW! Good post dtango! :eek:
Agreed. Thank you.
- oldman
-
Originally posted by FALCONWING
the second question is of course: What creates a long time customer?
Community.
-
Originally posted by FALCONWING
Aside from my pesonal feelings on this subject let me add a thought i have had for a number of years....
First a question:
WHO REALLY PLAYS AH? especially vs other games on the internet...
I am 37 yo and developed my love for WW2 air war because i spent a year in a British boarding school...absolutely fell in love with the battle of britain.
when i look around the members of my squad and the other peeps i know, i would say that by far the average subscriber is 30-60 yo. i am guessing that we are the generation who remember ww2 or had relatives in ww2 and therefore have an interest in the gameplay that gets us to do a trial period.
My fear has akways been that this game would eventually dry up as our "generation" got older and that there would not be enuf teenagers/20 yo to replace our ranks. most of the 20ish guys i know are active or were active military....(persoanlly i would advertise on armed forces radio and other military locales all day long) i think they have to be the future of this game.
the second question is of course: What creates a long time customer?
lol..gotta run...will post more later today:aok
<---- 14 years old. Hate most games, because of lack of realism. Fell in love with world war two when I was about 3. My grandpa was a test pilot. He decided not to go into the glider corps and test instead. The rest of his unit that joined the gliders (They came from the same fighter unit, which was disbanned before the war) died in normandy. When I was in 3rd grade, my friends and I would draw out imaginary battles in World War Two, spending hours on each aircraft. We even did several large ship murals, including every single stairwell (took me 3 weeks). I bought CFS soon as I saw it, as well as pearl harbor zero hour and everything else. I feel we are drifting away from the large numbers, but closer twoard realism which is fine with me. Most of my squad is under the age of 18, and yet we still do the long, boring, realistic runs. So there are more teenagers, and although they dont live up to yours and my expectations as far as responsibility and maturity, trust me when I say I have my best people working on that ;)
-
Fantastic analysis dtango. Just amazing and thorough.
You know what else strikes me? I think it also gives an insight as to why HT made some of the changes he did. Think about it....
Variety of aircraft used was shrinking
ground war growing
what was the first arena setup HT switched to?
EW1
MW1
LW1 (Ground Vehicles)
LW2
That would triple the variety of aircraft used in theory, as well as give the ground pounders thier own haven for war.
I know that the decision probably goes much deeper, but it kind of makes sense, doesn't it?
-
I would love to see something other than the top 30%
I think that would be revealing too.
The early war planes need to be on that list. lets see how they are being used over the years with the old MA and how they are being used this month.
HTC created a large toolbox... some who claim that they are WWII aficianados and battle of britan fans... would never get into a spit one in the old MA... they simply did not. they weren't fans of WWII they were fans of the best planes in WWII with the biggest guns.
Those who really flew the early and slow planes against such "aficianados" knew how unfun it was. Only something like the Hurri 2 with 4 monster cannons was much used.
All the charts and graphs are pretty but... what they show is....
How people liked, or were forced, to play in the old MA.
It will be fun to see how many planes make the "top 30" hit parade in a year or so and what new planes we will see..
I think everyone was about sick of the hit parade of the old MA.
I also think that the late war only and whorde arena also helped to get people out of fighters and into GV's..... they did it out of disgust for what was becomeing a very stale air war.
HT's new changes are a breath of fresh air to what the charts prove was a very stale MA.
lazs
Public Relations Officer for the BK's
-
Oh... a list of what new planes were developed over the last year or two and the ones developed over the next year might be revealing too...
why put all the effort into modeling some of the most important planes in the war when the "aficianados" would just turn their nose up (and vultch em) in the old MA?
So what did we get? What will we get now?
lazs
Public Relations Officer for the BK's
-
Originally posted by lazs2
HTC created a large toolbox... some who claim that they are WWII aficianados and battle of britan fans... would never get into a spit one in the old MA... they simply did not. they weren't fans of WWII they were fans of the best planes in WWII with the biggest guns.
Those who really flew the early and slow planes against such "aficianados" knew how unfun it was. Only something like the Hurri 2 with 4 monster cannons was much used.
I agree. In my whole time in the MA I have flown the Emil quite a bit, just because its a great plane, and only seen one other guy do the same, and he was excellent in it. I have seen Mako in a Murri I I beleive, or maybe it was a Spit I, but it was one of the two early war planes.
-
Great post and nice work DT :aok
-
I still think moving fields closer, creating more interesting terrain to fight in, and removing barriers to player fights such as heavy flak, will keep players more interested. The arena should foster player interaction, period. The fields are so far apart now, it takes a good 20 min to find a good fight. There are few neat terrain features to draw players together to fight. The flak often drives players away from areas where they would otherwise find people to fight. All of this degrades gameplay because it discourages player vs. player interaction.
It makes the arena boring.
If I want to fly around enjoying the view, not seeing other players for 20 min at a time, and having AI push me away from places where I might encounter other players, I'll fire up second life. I play AH to have player vs. player encounters, and anything that discourages that will keep me out of the game. It's been a problem for a couple of years (IMHO) and it's even more obvious now that the arena is split into the three war periods. There is no reason why 32 people in a mid-war arena shouldn't be fun, but when I logged in, it was a 20 minute flight to another empty field over flat featureless terrain, only to be beaten back by heavy flak as I approached the other field before I ever saw an enemy aircraft. Oh joy.
What happened to the good old days when you could take off, set auto-climb at 5 deg nose high, and arrive at 10k over the enemy field and find 10 willing opponents battling in a deep canyon, on the side of a mountain, or in a volcano? That was awesome. Never find that nowadays, since a 5 deg climb will get you to the enemy field at 20k+ assuming the flak doesn't shoot you down en-route, over nearly featureless terrain.
Bring back the 4-cornered symmetrical WB 1.11 arena... That was fun. I never wandered around the arena looking for a fight back then, and I never worried about some AI shooting me down during a dogfight. It was me and my friends vs. the other player and his friends, in an interesting terrain environment. That rocked. Why don't we still have an arena for players who just want to play at air combat?
-
quote:
--------------------------------------------------------------------------------
Originally posted by FALCONWING
the second question is of course: What creates a long time customer?
--------------------------------------------------------------------------------
Community.
__________________
NoBaddy "NB"
"Ego is an anesthetic that dulls the pain of stupidity."
" COMMUNITY " ! .. :aok
Bob/CHECKERS
-
Panzer kills come from camping. It's incredibly easy to rack up large numbers of kills in short periods of time, without having to do much more than click a button. In some setups, the spawns put you in range of other spawns, so there's no transit time, just click a button and shoot.
While that may technically be the ground war, it's really just spawn and vulch, rinse, repeat. I don't know if you can call that a tremendous growth in the ground war, or simply people being attracted to the easiest method of getting kills that the game currently offers.
By no means am I saying all kills, but I would wager that it is the source of a majority of kills, probably by a disturbing margin, but that last part is speculation only.
-
Originally posted by hubsonfire
Panzer kills come from camping. It's incredibly easy to rack up large numbers of kills in short periods of time, without having to do much more than click a button. In some setups, the spawns put you in range of other spawns, so there's no transit time, just click a button and shoot.
While that may technically be the ground war, it's really just spawn and vulch, rinse, repeat. I don't know if you can call that a tremendous growth in the ground war, or simply people being attracted to the easiest method of getting kills that the game currently offers.
You are overlooking the very few large maps (now gone) that had either a string of vbases or at least two next to each other. Led to good GV battles.
Also when the large forests were introduced to provide some cover making things a lot more interesting/surviveable (much easier to sneak up on a Tigger now).
No doubt a lot come from camping, but you have to add the above factors in also.
-
Originally posted by FALCONWING
the second question is of course: What creates a long time customer?
Woot! 2 different pilots have alredy posted what i was thinking of!!!!
1. Community
Absolutely! we obviously have our own ideas of what community is...but suffice to say community is very important!
2. Game play
Again i will refrain from going into what type is better or not but will instead try to stay with generalizations.
a: how easy is it to pickup and learn
b: how great a computer is needed to run it
c: how expensive will it be after i buy extra equipment and pay monthly fee
d: how quickly can i get into action
e: how sustainable is the action
f: how long does it take for me to get bored (or what keeps me from getting bored)
-
I didn't fly the Spit Mk I in the MA very much, but it was often fun when I did.
I remember getting in a fight with a Bf109 in the middle of a furball and getting close enough to see that it was a Bf109E-4. In the middle of a brawl of 1944/45 monsters a Spit I and 109E fought as though in a one v one. I shot the 109E down with no interference from the others.
In another case I took off in a Spit I as the sole defender of a base with a large attack force coming in. I shot down a Spit IX, La-7 and P-47D-40 before they got me.
-
FALCONWING: Congrats, you have just discoverd the 2 fundementals of online computer games.
How to implenent both conecepts is what pyro and I have doing since we began 12 years ago.
-
They have a saying: there's lies, there's damn lies, and then there's statistics...
OK - first let me say that I appreciate all the data, and I find it interesting, but there are some assumptions that I can't agree are valid without a lot more information.
First, that the number of kills in the MA has a direct correlation to the number of players since there are a number of factors that can (and probably have) thrown that data into a cocked hat. The first thing that comes to my mind are the maps - when were big maps introduced? When did we see the first FT and TT maps? There are a fixed number of game hours per month - a map that increases or decreases the average time between kills simply by the close proximity of enemy bases or requiring a longer flying time to reach a furball would affect your data, as would a map that invites vehicle spawn camping or easily vulchable fields. The introduction of such a map between two points on your graph would throw a lot of error into your data.
Another factor that might throw that kill data off was the introduction of the two weeks free promo. When did that start? How many two week wonders appeared, got shot down a lot, and then disappeared? How about changes in flight models, cannon lethality, number of rounds carried (for us old Spit V drivers), introduction of new aircraft models (and the learning curve associated with them). There's a lot of factors that you can see might affect those montly kill numbers besides just the raw membership.
Something big must've happened in 2004, because between Dec 2003 and Jul 2004 there was almost a 1/3 drop in kills, which has taken nearly 3 years to fix now that the July 2006 kill number is finally back up to where July 2003 was.
The real game growth data is at HTC - the number of paid memberships. That is the real indicator, and that data we don't have.
Again, I'm not trying to flame ya or anything, and I do appeciate seeing that data. I'm just not sure that the kills data equates to membership as directly as you assume.
EagleDNY
$.02
-
The aircraft data charts are also interesting. Regardless of other factors, it looks like the La-7, Niki, P-51, and whatever hi-end Spitfire is available have pretty much ruled the roost since the game began.
I assume the 109K-4 data has been merged with the old 109G-10 data. I'm not sure that was a good idea. I was a PT G-10 driver, but I don't tend to take up the K-4 since I can't get gondolas (or a single 20mm) any longer, and it just doesn't seem like the G-14 can compete against other late war rides. I don't know how many other 109 drivers switched over to something else when the K-4 came out, but I do remember there was a lot of people complaining that the 109s had been neutered.
There are plenty of factors that affect this data as well - changes in flight model, cannon lethality changes, etc. I am still of the opinion that the Hispano Mk.II 20mm cannons are a bit more lethal and the German Mk.151s are a bit less lethal than they ought to be, especially given that the Germans were designing their cannons to blast bombers from the sky. We did ROF testing that seemed to show that the Hispanos were shooting faster and the Mk.151s were shooting slower than was documented, but nothing ever came of it (other than a few accusations of luft-whining).
I think the aircraft usage data shows that the arena splits were really a good idea. Although we are still going to have some tweaking issues for a while, I think that it will all be for the best in the long run.
EagleDNY
$.02
-
Phweeeww
More questions than answers!!!
Thanks for the head ache:huh
Seriously though.. THANKS! Very nice post and so far rational discussion has developed.
PS. Ross Perot errr dtango for president. :D
-
Originally posted by EagleDNY
The aircraft data charts are also interesting. Regardless of other factors, it looks like the La-7, Niki, P-51, and whatever hi-end Spitfire is available have pretty much ruled the roost since the game began.
Almost - last year the trend started to become to choose the Vc over the IX.
Once people realised that MA fight alts (ie down low) were more suited to the low alt Merlinned Vc (with it's uber boost) than the high alt Merlinned IX.
It was a given once the low alt Merlin XVI was introduced it would become the free Spit of choice, even though both it and the VIII share the same motor.
Human nature says XVI is higher than VIII so it must be a far better Spit.
-
Great post, dtango.
One question:
With the kill stats that you collected, did you include air-to-ground kill information (for bombers and attack)? Not that it's crucial but if not, it might be more revealing as to bomber usage instead of just a2a stats..
Awesome article.
-
EagleDNY- There's another saying - statistics don't lie, people do. And that is the problem. The way that usually happens is through selection bias - when people look for data that confirms a predetermined belief while conveniently ignoring or dismissing the others.
A few points:
(1) I think 2 week free trials have been around since 2001. (I want to say I may have even joined in 01 on a free 2-weeker).
(2) I accept the argument that the total kills/deaths may not indicate the total player base. It seems like a reasonable proxy however. If not player base it at least represents the amount of activity in the MA. I'd be curious if HT would weigh in on the correlation between total kills/death as a proxy of playerbase population.
(3) The kill/deaths % stats are what they are and include the effects of map size changes, lethality change, flight model changes, ENY limitations etc.
(4) The % of total kill/death stats are independent of the amount of time it takes to get kills. They show the relative change or lack there of between the distribution of kills among the different vehicles and planes no matter how fast or drawn out it takes to get kills. It would be interesting to plot on the timeline when different maps were introduced and might give us a start at analyzing any correlation.
But let's just talk about big vs. small map adjustments just a bit. For instance If it takes longer to find the fights for larger maps, how come the distribution of kills/deaths for the short-legged aircraft like the La-7 or the Spitfire hasn't changed? Instead we see a drop in some of the longer-legged aircraft instead. That would seem to indicate that map sizes haven't changed the game play dramatically.
(5) 03-04 drop was ~25% so let's be clear about that. 1/3 is exagerating the drop. That was the biggest 6 month drop registered, but overall the drop was still in the +/- 25% range from Jul 02 on. If I had to guess, I would theorize that might be due to the switch from AH1 to AH2 (it seems to correspond with the AH2 introduction Dec 03-Jan04).
If you believe there are holes in the data then by all means fill it in :). It's available. I just didn't have the time to collect it for every month from Jul 2001 - 2006 (and I'm not sure HTC would be happy with me doing that since it's a drain on their SQL servers for all those queries!).
Tango, XO
412th FS Braunco Mustangs
-
It looks like the "new" Spit XVI and VIII have pretty much replaced the "old" IX and V in the kills list. I'm still pretty attached to the Seafire, but I'll have to take an VIII up and play with it for a while as well, just to see how much like the old V it really is.
I've tried out the XVI, and it is an easy ride to get kills in. I can see why newbies especially would want to take it up - it gives you a lot of options and will get you out of a lot of jams. My only problem with Spits in general is the "seems like a unicycle" landing gear... getting back down without putting a wingtip in the dirt ;)
For gameplay purposes, newbies do need a ride that gives them a chance against the more "seasoned" members of the community. I'm not one of those who thinks that anyone up in a Spit or La is some kind of "dweeb" that needs to be mocked on channel 200. We were all newbies once.
If a newbie is up in an La-7 or Spit XVI, just think of it as an opportunity for greater perkies when you rip them up in a lesser ride.
EagleDNY
$.02
-
Originally posted by EagleDNY
They have a saying: there's lies, there's damn lies, and then there's statistics...
OK - first let me say that I appreciate all the data, and I find it interesting, but there are some assumptions that I can't agree are valid without a lot more information.
Ya beat me to it, one of the first lines is "if we assume", well you know what they say about THAT word?
This analysis is all fine and good, and I DEFINITELY appreciate the work that went into it and it's presentation, as well as the intellectual discussion sans dogmatic accusations. But I beleive we, the players, simply do not have access to the correct dataum to make such assumptions. What you really need is the number of downloads/trials/subscribtions/cancellations broken down by demographics to take this where you intended it to go. If I were running HTC that information would not be made availble to anyone outside of the executive offices.
Bottom line is this:
The game has changed much since it was called "Confirmed Kill", so has the player-base, so has the player mentality. I suspect that change is the only constant and only those willing to adapt to change will ever be long-term subscriber. I do not think there is anyone on the planet with more experience and expertise at making MMOL air-combat sims than either Dale or Doug, and HTC gets them both. I for one intend to sit back, trust their judgement, and enjoy the ride for less than the cost of a coke per day.
-
scottydawg - that's possible but I didn't do it. Someone else is welcome to do it. One proxy could be to look at B-24 vs. all models, B-17's vs. all models etc. in the killstats. That's a lot of work though to run each of those reports per bomber model per number of months you would want to collect.
Tango, XO
412th FS Braunco Mustangs
-
Originally posted by dtango
scottydawg - that's possible but I didn't do it. Someone else is welcome to do it. One proxy could be to look at B-24 vs. all models, B-17's vs. all models etc. in the killstats. That's a lot of work though to run each of those reports per bomber model per number of months you would want to collect.
Tango, XO
412th FS Braunco Mustangs
Absolutely, and you've done a lot of work to pull that data together. Thanks for that, I'm sure a lot of people appreciate it. The rest will say "Holy crap, Einstein, I'm not reading all that".
-
edbert wrote:
Bottom line is this:
The game has changed much since it was called "Confirmed Kill", so has the player-base, so has the player mentality. I suspect that change is the only constant and only those willing to adapt to change will ever be long-term subscriber. I do not think there is anyone on the planet with more experience and expertise at making MMOL air-combat sims than either Dale or Doug, and HTC gets them both. I for one intend to sit back, trust their judgement, and enjoy the ride for less than the cost of a coke per day.
I accept the idea that the game has changed but objectively looking at some of the data available to us it hasn't changed in ways that some folks have been talking about.
2ndly, let's not confuse the issue of the total player population statistics with the aircraft and vehicle usage statistics.
Tango, XO
412th FS Braunco Mustangs
-
Originally posted by dtango
EagleDNY- There's another saying - statistics don't lie, people do. And that is the problem. The way that usually happens is through selection bias - when people look for data that confirms a predetermined belief while conveniently ignoring or dismissing the others.
(2) I accept the argument that the total kills/deaths may not indicate the total player base. It seems like a reasonable proxy however. If not player base it at least represents the amount of activity in the MA.
But let's just talk about big vs. small map adjustments just a bit. For instance If it takes longer to find the fights for larger maps, how come the distribution of kills/deaths for the short-legged aircraft like the La-7 or the Spitfire hasn't changed? Instead we see a drop in some of the longer-legged aircraft instead. That would seem to indicate that map sizes haven't changed the game play dramatically.
Agree with your argument on selection bias - we all do it every day. Every time we pick one ride over another and take off, we are expressing our bias and creating another piece of data for your graph.
Agree that the total kills/deaths does represent the level of activity in the MA, and even must be representative of the total player base in some fashion.
In the Big Maps vs Small Maps argument, I would submit that the distribution of kills/deaths isn't really changed because all you have to do is up with enough gas to get you to the nearest enemy dar bar and get into a fight to provide data for your statistics (either a kill, death, or both).
The average time it takes to get you to that fight is probably more to do with the average distance between bases on big vs small maps, which would tell us how much deviation in the data a big map might cause over a small map.
Is there data available on LANDED kills vs just kills & deaths? That might provide some interesting info as well. I'd rather like to know what the average landed per plane is, and the percentage of players that actually get home with their kills as opposed to getting a kill and gettting subsequently killed themselves.
Again - appreciate the research, I agree that it does show that AH can take quite a few changes and keep on going just fine.
EagleDNY
$.02
-
EagleDNY - no way that I know of to get landed kills stats easily & publicly.
I've thought a bit about the whole kills/sortie, landed-kills issue a bit. On the surface they may seem useful but without getting into the gory details of it all the more I thought about them the more I'm not sure I could find a meaningful use for them at least for some of the analysis I was doing.
Tango, XO
412th FS Braunco Mustangs
-
Originally posted by Edbert
Ya beat me to it, one of the first lines is "if we assume", well you know what they say about THAT word?
This analysis is all fine and good, and I DEFINITELY appreciate the work that went into it and it's presentation, as well as the intellectual discussion sans dogmatic accusations. But I beleive we, the players, simply do not have access to the correct dataum to make such assumptions. What you really need is the number of downloads/trials/subscribtions/cancellations broken down by demographics to take this where you intended it to go. If I were running HTC that information would not be made availble to anyone outside of the executive offices.
Bottom line is this:
The game has changed much since it was called "Confirmed Kill", so has the player-base, so has the player mentality. I suspect that change is the only constant and only those willing to adapt to change will ever be long-term subscriber. I do not think there is anyone on the planet with more experience and expertise at making MMOL air-combat sims than either Dale or Doug, and HTC gets them both. I for one intend to sit back, trust their judgement, and enjoy the ride for less than the cost of a coke per day.
Yup, me to!:aok
-
If we assume that total number of kills (which also equals total deaths) in the MA as a proxy for the total MA player base population what does that tell us about the population?
I would have assumed so also, until I looked at your chart and compared it to our subscriber base, but it dosn't.
HiTech
-
Reading between the lines of Dale's post...the number of kills, while generally climing over time, has declined over time when viewed as a per-subscriber figure.
So even in a more crowded arena there's less killin per-capita (pure conjecture on my part, and of course tainted by my own personal bias, I realize).
-
Originally posted by hitech
FALCONWING: Congrats, you have just discoverd the 2 fundementals of online computer games.
How to implenent both conecepts is what pyro and I have doing since we began 12 years ago.
and doing a great job of it btw imho anyways!
-
Originally posted by hitech
I would have assumed so also, until I looked at your chart and compared it to our subscriber base, but it dosn't.
HiTech
I didn't think there would be an exact correlation (too many other factors involved). And I know better than to ask if the subscriber base data is available.
HT: might be interesting to track some additional data (or post some if you are already tracking it). Sorties upped, kills landed, etc. Might make some more interesting reading.
EagleDNY
$.02
-
Sir Tango most excellent post sir!...in my post "integrity" It was suppose to be about the integrity of the decsion ...ie show us some data....Unfortunatly some took this as a direct attack used the propriatary data sheild to not devulgle the "skunkworks best and deepest secrets. an not other less sensitive data.
Hopefully their can be more thoughtfull information data and discusions..sometimes the answers to problems come from two bad ideas that spark the thought of a good idea in someone.
I even wonder if health of an areana can be measured buy the total number of player hours played a month...And think it would be interesting to see if the hours played "satisfaction" has increased or not since the new changes.
hoping for the best!
999000
-
Originally posted by Edbert
The game has changed much since it was called "Confirmed Kill", so has the player-base, so has the player mentality. I suspect that change is the only constant and only those willing to adapt to change will ever be long-term subscriber. I do not think there is anyone on the planet with more experience and expertise at making MMOL air-combat sims than either Dale or Doug, and HTC gets them both. I for one intend to sit back, trust their judgement, and enjoy the ride for less than the cost of a coke per day.
Many people might be surprised how little has changed since Dale first laid fingers to keyboard to produce "Confirmed Kill". Most everything I see today, in game and on this BBS, is pretty much the same stuff I was seeing 16 years ago in AW...just a heck of a lot more of it. The most positive change I have seen is the one made most recently. It appears to be having a very positive impact on the overall community of the game. Time will tell, but, the first steps appear to be in the right direction.
BTW, just so you know...Doug is the REAL brains behind the operation. Dale is just The PUTZ. :D
-
EagleDNY: While I agree on the more interesting reading. It realy is not in HTC's best interst to release data.
HiTech
-
Holy metrics Batman!:eek:
-
it is an interesting post, but i would warn against using only numbers to try to quantify behavior. there is a fine line between reading numbers correctly and misinterpreting data.
there are many, many variables in the MA, and as we have seen, even numbers that were thought to represent the population of the MA have turned out to be inaccurate.
here is my example: say i have marbles, red, blue, green. on every marble there is a not that says each marble will give you magic powers, red makes you strong, blue makes you fast, and green makes you smart. every day i drop these marbles and count how many are picked up. over time i found that blue was more popular, and that more marbles are being picked up. from these results i can conclude that...
1. people want to be fast more than anything else
2. more and more people like my marbles.
what someone else might say is that really there are only 10 guys, 7 of which are spending all day picking up blue marbles and beating up everyone who goes near the marbles. more information shows that only 3 of those 7 are spending the majority of time picking up marbles.
so i then determine that...
1. there is a problem
2. if i make more marbles then those 10 guys will have a harder time beating up the others chasing the marbles...more ground to cover. making more room for others who enjoy picking up marbles.
the numbers are interesting, but without all of them, i think it is very difficult equate online gaming quality by numbers alone....especially when you don't have all the numbers. just my idea.
-
Originally posted by 999000
Sir Tango most excellent post sir!...in my post "integrity" It was suppose to be about the integrity of the decsion ...ie show us some data....Unfortunatly some took this as a direct attack used the propriatary data sheild to not devulgle the "skunkworks best and deepest secrets. an not other less sensitive data.
Hopefully their can be more thoughtfull information data and discusions..sometimes the answers to problems come from two bad ideas that spark the thought of a good idea in someone.
I even wonder if health of an areana can be measured buy the total number of player hours played a month...And think it would be interesting to see if the hours played "satisfaction" has increased or not since the new changes.
hoping for the best!
999000
maybe HTC can just worry about that.
honestly how can you view this as not a direct attack when you accuse them of using "propriatary data shield" and then go on to assume that it's not important to HT via "skunkworks." give it a rest, we are having a discussion about gameplay, not to evaluate HTC.
-
lol pluck, my theory is you've lost your marbles.
Just kidding, couldn't resist.
-
lol, i thought someone would:) the sad thing is you may be right:aok
-
HTC, HiTech, I just wanna say its so beautiful to see the creators of this game getting so involved, and reminding us that they care, and actually read our posts ;)
-
Originally posted by pluck
lol, i thought someone would:) the sad thing is you may be right:aok
you actually had marbles? ;)
couldn't resist that one myself
-
Originally posted by hitech
FALCONWING: Congrats, you have just discoverd the 2 fundementals of online computer gaming
Outstanding....took me 5 minutes over lunch!:D
In all seriousness though...you have created a good game...and the opportunity for community..kudos:aok
-
Originally posted by hitech
I would have assumed so also, until I looked at your chart and compared it to our subscriber base, but it dosn't.
HiTech
Interesting. Lots of different things I'm sure you guys can digest from this. Thanks for weighing in about the lack of correlation.
Hopefully folks won't ignore the other analysis because of my self-inflicted red herring I introduced by tying the total kills stats to total MA player base. I believe it's still a good measure of MA activity. It's also independent of the kill and death %.
Tango, XO
412th FS Braunco Mustangs
-
Originally posted by dtango
Interesting. Lots of different things I'm sure you guys can digest from this. Thanks for weighing in about the lack of correlation.
Hopefully folks won't ignore the other analysis because of my self-inflicted red herring I introduced by tying the total kills stats to total MA player base. I believe it's still a good measure of MA activity. It's also independent of the kill and death %.
Tango, XO
412th FS Braunco Mustangs
Excelent work dtango
Hitech: The data may not reflect player base, but it does reflect MA activity or is this a statistical Illusion..... ?
-
Originally posted by dtango
Interesting. Lots of different things I'm sure you guys can digest from this. Thanks for weighing in about the lack of correlation.
Hopefully folks won't ignore the other analysis because of my self-inflicted red herring I introduced by tying the total kills stats to total MA player base. I believe it's still a good measure of MA activity. It's also independent of the kill and death %.
Tango, XO
412th FS Braunco Mustangs
tango, i think it is a very interesting display, and i'm sure quite a bit of work went into it; and is informative. i like just like to play devil's advocate for the sake of discussion.
-
With respect to all concerned, I think these metrics are more of a scoreboard, like one kept by a scorekeeper, and not nearly the metrics the coach and players use to affect the score.
It's like looking at a P&L. Many people can tell you where all the numbers come from and derive some meaning (real or imagined) to them, but that doesn't mean they are equipped to understand or manage the changes needed to improve subsequent P&L's.
No one would expect HTC to give a data dump of all they've learned and all they know to mold a self-sustaining, independent company with a sole proprietary game. I'm no game expert, but the game experts do say that it's almost impossible to do.
Anyway, it's curious to look at. I think the only thing that should be assumed from it is that nothing should be assumed from it.
-
Dt, nice job!
You use S-Plus to make the graphs? I've seen formatting like that before, but I can't remember what I was using.
I think that the research you've done is interesting, I enjoy looking at that sort of thing.
Keep it up :).
-
Thanks Urchin :). Actually I'm a little surprised at the number of folks that have been interested in analysis. I didn't do anything fancy for the graphs etc. Just plain MS Excel, cut and paste into PS and saved as a .gif file.
Tango, XO
412th FS Braunco Mustangs
-
Originally posted by dtango
Interesting. Lots of different things I'm sure you guys can digest from this. Thanks for weighing in about the lack of correlation.
Hopefully folks won't ignore the other analysis because of my self-inflicted red herring I introduced by tying the total kills stats to total MA player base. I believe it's still a good measure of MA activity. It's also independent of the kill and death %.
Exactly. I for one want to say thanks for all the hard work you did.
-
Originally posted by hitech
EagleDNY: While I agree on the more interesting reading. It realy is not in HTC's best interst to release data.
HiTech
Don't be so sure. I know it wouldn't be prudent to release data like the number of paid subscribers per month or number of free 2 week trials started (thats clearly proprietary), but releasing data on the number of kills landed per ride wouldn't be a bad thing. You might get some folks interested in trying some new rides out, or it might tell you something about how some of your rides are really being used (suicide rate indicator for example) so that you can do some more tweaking.
EagleDNY
$.02
-
EagleDNY: Agreed on data like plane stats.
HiTech
-
Be interesting to know how the ratio of kills/hours flown has changed over time and particularly what effect the new arena setup has had.
-
Poppycock!
The sky IS falling and HTC must cater to my demands!
:D
-
Originally posted by EagleDNY
Don't be so sure. I know it wouldn't be prudent to release data like the number of paid subscribers per month or number of free 2 week trials started (thats clearly proprietary), but releasing data on the number of kills landed per ride wouldn't be a bad thing. You might get some folks interested in trying some new rides out, or it might tell you something about how some of your rides are really being used (suicide rate indicator for example) so that you can do some more tweaking.
EagleDNY
$.02
How could you distinguish between a suicide and something esle?
Eg -
Every pork run I go on I intend to live.
Doesn't always work out that way.
a) 1 ping pilot kill from ack/
b) Lose a wing or something to ack.
c) Hit by the gazillions of cons.
This could happen
a) Just after releasing eggs.
b) On climb out.
Could be a newb in a P-38/190/etc for the first time and lawndarts accidently.
So how do you tell the difference?
I think the true AH suicider is a rare animal most oftenly confused with genuine attempts or newbs.
Another ex -
Sinking a CV.
We'll up 6-7 Hellcats to sink it.
You know you'll lose a few to the 5" on the way in, is this suicide? Or in fact more closer to reality.
-
Reading between the lines, although the player base has enjoyed modest growth, the average kill per player has gone down...........
This means that the average fun per average player has gone down..............
on top of this the kill spread has shifted to favour a smaller % of experten....
If community spirit is poor or medi-ocre then what incentive is there for non experten to stay?
New subscribers most definately come under the heading of non experten
The above would cause a high turnover of new subscribers............ retain the new subscribers for a longer period and modest growth becomes more significant growth.
HT would know the effect on his business model if he could just retain each new subscriber an average of # months longer than he does now.
If there was something he could do to brush up the welcome mat and create more fun for new subscribers he would do it.
All very interesting..........
except I cannot see a firm correlation between the above and how it is served by EW, MW and LW (MA) arena's.
What I do see is that these arenas effectively enlarge the usable plane set. They make ac that are not viable (in the old MA) into very viable and fun rides within their own arenas.
This effectively gives us more choice.....hurrah :)
1)I think there is too much choice for off peak periods.....low numbers effect game play
2)I think the ENY side balance incentive is presently poorly implemented for lower numbers.........
3)I think limiting the arena size of the LW is an error. (although I can see the incentive I do not think you can force folk into other arenas and sustain it)
4)I think there are some errors re the ac allowed in the EW and MW arenas.
However all of the above can be adjusted as HTC fine tune what they have done....and 4) may be just my opinion.
This still leaves the original challenge however.......how do we make it more fun for newbies?