Keep in mind that modern graphics cards (even entry-level) can easily support upwards of a million polys onscreen, with the high-end ones handling over thirty million at 60fps. Most of the stress in modern games is from the enormous shader load associated with everything onscreen, for literally every single lighting calculation and special effect (both textures and geometry) that exists in the game.
I'd worry not so much about the unit's polycount (1-7k is perfectly fine), but rather the types of textures you put on it and the shaders that are applied to that texture, and other effects the unit has.
I learned (the hard way) that Battle.net does not count all of your map's hours played toward the overall total that counts for your map's place on the Top Played list.
Instead, battle.net only counts the TOP-MOST VARIANT (NOT necessarily the default, but literally the one on the top of the list) toward the map's popularity. The other variants do not raise the map's popularity at all.
I discovered this bug when I added an update to Mafia introducing the Ranked variant. On the list, Ranked happened to be on the top and Custom (the old, default mode) on the bottom. Literally overnight following the update, Mafia's popularity divebombed from #6 to #61. On the EU server, the popularity dropped from #7 to #74 in the same amount of time. This problem eluded me for a few days before I realized there was literally no explanation for the drop in popularity; people still played at about half the rate that they used to. By hours played, that would still put Mafia on the first two pages, not on page 6. So I thought for a bit and conjectured that Mafia was only counting the hours from Ranked, since not many people play Ranked compared to Custom. At the rate that people play Ranked, Mafia would indeed be around #60 if that were the only mode. I thought of a random possibility: Ranked was at the top of the list, so maybe changing the order would cause it to go back to counting Custom's hours. Turns out I was right; in a matter of hours, Mafia rocketed up to #20 and continues to climb.
For obvious reasons, this can be a huge issue for maps that have multiple variants. BattleCraft, for example, once had four variants. The topmost one was 1v1, hardly the most popular. As a result, the game's popularity was very low even though before 1.5 it was always on the first two pages. More recently, the 1v1 mode was dropped and now 2v2 is the highest, and easily the most popular. Since then, the map reached page 2.
You'd be hard-pressed to prove that the majority of players regularly play Squad TD. I remember, before the update, counting up the hours played of the top few pages. Squad TD would be around 25k, while the rest of the maps combined would be around 70k.
My point is that "mass appeal" is arbitrary. You won't get the majority of players but you can get enough that you're still a dominant figure. There's no secret to making a game that appeals to the masses; it's just a matter of carefully adjusting your plans as you make/update your game to keep gaining support.
Squad TD is very, very good at appealing to the Tower Defense target market, which is already pretty large. That's all it is.
That definition has literally nothing to do with my argument. I didn't say ANYTHING about the game needing perfection; I was strictly talking about its ability to generate mass appeal.
if a == b, then that means you're dividing by a - a. Dividing by zero. You can't do that in this kind of function.
You didn't take into account the fact that people can buy StarCraft 2 for the purposes of playing custom maps. People have done that to play Mafia, after all. Also, you completely ignored "got offended by some random thing about your game, dropped it from the box art or title, won't buy it because of rating, etc." There are people out there who only ever like MOBA games like SotIS and will never be convinced to take seriously anything else. Same goes for TD players. Same goes for observer games like 1v1 Obs. You can't mix every genre to accomplish your goal, either, because it will alienate some of the players you're trying to grab just by having a mixed-genre title, plus such a convoluted mix is really hard to make into a good game in the first place.
1 != 0. It's invalid to say 1 == 0 if 1-1 == 0. If you take away 1 from 1, it's no longer 1. Therefore, that stupid train of thought can't be used to say that 1 can equal 0.
The effect you mention is called "mob mentality" sometimes, but has many terms associated with it. It helps with getting more people to play the game but doesn't stand up on its own. And it's not quite as effective as you make it out to be.
Also, he said he wanted to appeal to all types of people. He didn't say what the types are, so I used the fact that everyone is their own type of person. Appealing to the majority of people in the world means that you need to appeal to 3,500,000,000 people. But, then you have to narrow down the age range. Your game probably won't be played by many 5 year olds nor will it be played by 70 year olds; cut it down to around 2,000,000,000 people. Now, cut it down to all the people who actually own a machine capable of running your game: 200,000,000 people, roughly. Already we're not even close to a majority of people in the world. Now cut it for people who don't play games, don't buy any new games, got offended by some random thing about your game, dropped it from the box art or title, can't buy it because of legal or international reasons, won't buy it because of rating, etc. This is before genre comes into play, and no matter what you do, a lot of people will not play your game on the sole reasoning that it is a genre they don't like, regardless of how good your game is or how appealing it is to everyone.
The quote is plain bullshit. 1 does not equal 0 even if you set your mind to it, unless you're just delusional or insane.
Anyway, as for your goal, it's impossible to make a game appeal to everyone unless you kill a large chunk of the world's population. It's already extremely difficult to make a game appeal to the majority of people.
There is a concept called the "target market" that nearly all designers and salesmen use. Your game or product has a targeted demographic (the target market) that it will appeal the most to. You must design the game and its marketing around appealing to the target market to maximize sales within that demographic. All other attempts to appeal to other markets are secondary and, often, can actually hurt sales by alienating both demographics.
Polls like this can only ever give you the answer to the question: which demographic is larger? If you know which type of game is most-liked, you can cater to the largest demographic. It's a boring, but ultimately effective way of maximizing your game's chances at making money.
I refuse to consider a map whose name is so horribly and deliberately misspelled to be a better-quality game than mine. Call me biased. It's like a guy called Captin Poopiepants beating you at chess.
Also, the mean score is around 4.2 and the lowest is about 3.4. Plus, individual scores and rankings fluctuate every day. There is also little correlation between ratings and objective quality of the map. How can Special Forces be ranked 1? The map is simplistic and almost sloppy.
I use them as an example of why a 5-star system will never work. Scores will be inflated to an artificial high or dragged to an artificial low no matter what you do. Mafia, being a deep-rooted mod, will see very little disadvantage from having lower ratings. People will play it anyway.
Thus, the top two reviews for Mafia (the only two you can see without scrolling) are 1/5. Four of the top seven reviews are 1/5.
Too lazy to update the whole list but now Mafia is ranked 30/36 with a score of 3.94. Now, even Mineralz Evolution is allegedly better than Mafia.
By definition, Mafia is worse than a large majority of all regularly played maps on Battle.net. Mafia also has the distinction of having the worst quality-to-popularity ratio of any StarCraft 2 map.
Logically, this means that Battle.net is filled with amazing maps of such a radiance that they utterly overshadow even a well-made game like Mafia
Yeah, what my data proves is that there is no way to salvage the 5-star system. Even with statistical analysis I can't get accurate ratings out of the system. Besides, it wasn't like I had much data to go on. Helpfuls, review ratings, the map rating and the number of ratings are literally all that is available to us.
I wish Blizzard had a YouTube style like/dislike bar. At least it would be a more accurate portrayal of a map's quality.
Fun fact: Mafia had the 2nd highest number of helpfuls on its reviews, next to Squadron TD. Mafia also has the distinction of having the most polarized reviews of any map on battle.net. Seriously, go look at the reviews. In the top 9 reviews alone, 5 of them are 5-star and 4 of them are 1-star. Of course, the people who wrote those particular 1-star reviews are mostly guys who were banlisted.
I looked at a bunch of the top played maps on the NA Battle.net, then recorded their average rating and the number of ratings. Then I went into the reviews section for each map and counted up an average of each review score, weighted by the number of helpfuls per review. I only counted reviews with 5 or more "helpfuls" and then only counted maps with around 200 helpfuls or more. I noticed that each of these maps had around 300 ratings or more, so I set the minimums at 200 helpfuls and 300 ratings.
I then used a Bayesian algorithm to determine a more accurate score for each map, based on the minimum number of helpfuls/ratings needed to qualify and the average rating a person is likely to give (based on the data of the 36 maps I analyzed). I did this both for Ratings and Reviews, then averaged the two scores together.
My results:
Rank
Map
Score
1
Special Forces
4.78
2
TD Tycoon
4.70
3
1v1 Obs
4.69
4
Squadron TD
4.68
5
Night of the Dead
4.67
6
Undead Assault 3
4.66
7
Photon Discs
4.65
8
HOTS Custom
4.64
9
MageCraft
4.63
10
Bx Monobattle
4.62
11
Civilization Sapphire
4.60
12
Star Battle
4.59
13
Marine Arena
4.58
14
Nexus Wars
4.55
15
Eras Zombie Invasion
4.55
16
StarJeweled
4.54
17
Desert Strike
4.52
18
A Song of Ice and Fire
4.52
19
The Troll Tribes
4.49
20
Galaxy Vampirism
4.47
21
The Star Strikers
4.41
22
Left 2 Die
4.40
23
Cortex Roleplay
4.39
24
Pobes vs Zeelot
4.34
25
New Established Order
4.30
26
Mafia
4.24
27
Aeon of Storms
4.14
28
Hero Attack
4.14
29
Mineralz Evolution
4.11
30
Auir Chef
3.98
31
Runling Run
3.91
32
Build and Fight
3.79
33
God's Land
3.71
34
Cruiser Command
3.69
35
Heaven Besieged
3.67
36
Survival Man
3.38
As you can see, although the ratings are a bit better than the completely nonsensical crap we have on battle.net right now, it still doesn't make much sense. Special Forces ranked 1? Mafia ranked lower than Pobes vs Zeelot?
0
Keep in mind that modern graphics cards (even entry-level) can easily support upwards of a million polys onscreen, with the high-end ones handling over thirty million at 60fps. Most of the stress in modern games is from the enormous shader load associated with everything onscreen, for literally every single lighting calculation and special effect (both textures and geometry) that exists in the game.
I'd worry not so much about the unit's polycount (1-7k is perfectly fine), but rather the types of textures you put on it and the shaders that are applied to that texture, and other effects the unit has.
0
In starcraft, it takes up 32 bits just like any other variable.
In languages like C, it takes up 8 bits.
0
They read, albeit indirectly, this site too.
Also. I'm ahead of you: http://us.battle.net/sc2/en/forum/topic/6413262517#1
0
I learned (the hard way) that Battle.net does not count all of your map's hours played toward the overall total that counts for your map's place on the Top Played list.
Instead, battle.net only counts the TOP-MOST VARIANT (NOT necessarily the default, but literally the one on the top of the list) toward the map's popularity. The other variants do not raise the map's popularity at all.
I discovered this bug when I added an update to Mafia introducing the Ranked variant. On the list, Ranked happened to be on the top and Custom (the old, default mode) on the bottom. Literally overnight following the update, Mafia's popularity divebombed from #6 to #61. On the EU server, the popularity dropped from #7 to #74 in the same amount of time. This problem eluded me for a few days before I realized there was literally no explanation for the drop in popularity; people still played at about half the rate that they used to. By hours played, that would still put Mafia on the first two pages, not on page 6. So I thought for a bit and conjectured that Mafia was only counting the hours from Ranked, since not many people play Ranked compared to Custom. At the rate that people play Ranked, Mafia would indeed be around #60 if that were the only mode. I thought of a random possibility: Ranked was at the top of the list, so maybe changing the order would cause it to go back to counting Custom's hours. Turns out I was right; in a matter of hours, Mafia rocketed up to #20 and continues to climb.
For obvious reasons, this can be a huge issue for maps that have multiple variants. BattleCraft, for example, once had four variants. The topmost one was 1v1, hardly the most popular. As a result, the game's popularity was very low even though before 1.5 it was always on the first two pages. More recently, the 1v1 mode was dropped and now 2v2 is the highest, and easily the most popular. Since then, the map reached page 2.
0
@Taintedwisp: Go
You'd be hard-pressed to prove that the majority of players regularly play Squad TD. I remember, before the update, counting up the hours played of the top few pages. Squad TD would be around 25k, while the rest of the maps combined would be around 70k.
My point is that "mass appeal" is arbitrary. You won't get the majority of players but you can get enough that you're still a dominant figure. There's no secret to making a game that appeals to the masses; it's just a matter of carefully adjusting your plans as you make/update your game to keep gaining support.
Squad TD is very, very good at appealing to the Tower Defense target market, which is already pretty large. That's all it is.
0
@Taintedwisp: Go
That definition has literally nothing to do with my argument. I didn't say ANYTHING about the game needing perfection; I was strictly talking about its ability to generate mass appeal.
0
@Taintedwisp: Go
Divide each side by "a-b", you say?
if a == b, then that means you're dividing by a - a. Dividing by zero. You can't do that in this kind of function.
You didn't take into account the fact that people can buy StarCraft 2 for the purposes of playing custom maps. People have done that to play Mafia, after all. Also, you completely ignored "got offended by some random thing about your game, dropped it from the box art or title, won't buy it because of rating, etc." There are people out there who only ever like MOBA games like SotIS and will never be convinced to take seriously anything else. Same goes for TD players. Same goes for observer games like 1v1 Obs. You can't mix every genre to accomplish your goal, either, because it will alienate some of the players you're trying to grab just by having a mixed-genre title, plus such a convoluted mix is really hard to make into a good game in the first place.
0
1 != 0. It's invalid to say 1 == 0 if 1-1 == 0. If you take away 1 from 1, it's no longer 1. Therefore, that stupid train of thought can't be used to say that 1 can equal 0.
The effect you mention is called "mob mentality" sometimes, but has many terms associated with it. It helps with getting more people to play the game but doesn't stand up on its own. And it's not quite as effective as you make it out to be.
Also, he said he wanted to appeal to all types of people. He didn't say what the types are, so I used the fact that everyone is their own type of person. Appealing to the majority of people in the world means that you need to appeal to 3,500,000,000 people. But, then you have to narrow down the age range. Your game probably won't be played by many 5 year olds nor will it be played by 70 year olds; cut it down to around 2,000,000,000 people. Now, cut it down to all the people who actually own a machine capable of running your game: 200,000,000 people, roughly. Already we're not even close to a majority of people in the world. Now cut it for people who don't play games, don't buy any new games, got offended by some random thing about your game, dropped it from the box art or title, can't buy it because of legal or international reasons, won't buy it because of rating, etc. This is before genre comes into play, and no matter what you do, a lot of people will not play your game on the sole reasoning that it is a genre they don't like, regardless of how good your game is or how appealing it is to everyone.
0
The quote is plain bullshit. 1 does not equal 0 even if you set your mind to it, unless you're just delusional or insane.
Anyway, as for your goal, it's impossible to make a game appeal to everyone unless you kill a large chunk of the world's population. It's already extremely difficult to make a game appeal to the majority of people.
There is a concept called the "target market" that nearly all designers and salesmen use. Your game or product has a targeted demographic (the target market) that it will appeal the most to. You must design the game and its marketing around appealing to the target market to maximize sales within that demographic. All other attempts to appeal to other markets are secondary and, often, can actually hurt sales by alienating both demographics.
Polls like this can only ever give you the answer to the question: which demographic is larger? If you know which type of game is most-liked, you can cater to the largest demographic. It's a boring, but ultimately effective way of maximizing your game's chances at making money.
0
Two hours is too much time on my hands?
0
I refuse to consider a map whose name is so horribly and deliberately misspelled to be a better-quality game than mine. Call me biased. It's like a guy called Captin Poopiepants beating you at chess.
Also, the mean score is around 4.2 and the lowest is about 3.4. Plus, individual scores and rankings fluctuate every day. There is also little correlation between ratings and objective quality of the map. How can Special Forces be ranked 1? The map is simplistic and almost sloppy.
0
Heh, I never cared about the reviews.
I use them as an example of why a 5-star system will never work. Scores will be inflated to an artificial high or dragged to an artificial low no matter what you do. Mafia, being a deep-rooted mod, will see very little disadvantage from having lower ratings. People will play it anyway.
0
Oh, Blizzard deleted the top review for Mafia.
Thus, the top two reviews for Mafia (the only two you can see without scrolling) are 1/5. Four of the top seven reviews are 1/5.
Too lazy to update the whole list but now Mafia is ranked 30/36 with a score of 3.94. Now, even Mineralz Evolution is allegedly better than Mafia.
By definition, Mafia is worse than a large majority of all regularly played maps on Battle.net. Mafia also has the distinction of having the worst quality-to-popularity ratio of any StarCraft 2 map.
Logically, this means that Battle.net is filled with amazing maps of such a radiance that they utterly overshadow even a well-made game like Mafia
0
Yeah, what my data proves is that there is no way to salvage the 5-star system. Even with statistical analysis I can't get accurate ratings out of the system. Besides, it wasn't like I had much data to go on. Helpfuls, review ratings, the map rating and the number of ratings are literally all that is available to us.
I wish Blizzard had a YouTube style like/dislike bar. At least it would be a more accurate portrayal of a map's quality.
Fun fact: Mafia had the 2nd highest number of helpfuls on its reviews, next to Squadron TD. Mafia also has the distinction of having the most polarized reviews of any map on battle.net. Seriously, go look at the reviews. In the top 9 reviews alone, 5 of them are 5-star and 4 of them are 1-star. Of course, the people who wrote those particular 1-star reviews are mostly guys who were banlisted.
0
Edit: Title changed to appeal to the dumb masses.
I looked at a bunch of the top played maps on the NA Battle.net, then recorded their average rating and the number of ratings. Then I went into the reviews section for each map and counted up an average of each review score, weighted by the number of helpfuls per review. I only counted reviews with 5 or more "helpfuls" and then only counted maps with around 200 helpfuls or more. I noticed that each of these maps had around 300 ratings or more, so I set the minimums at 200 helpfuls and 300 ratings.
I then used a Bayesian algorithm to determine a more accurate score for each map, based on the minimum number of helpfuls/ratings needed to qualify and the average rating a person is likely to give (based on the data of the 36 maps I analyzed). I did this both for Ratings and Reviews, then averaged the two scores together.
My results:
As you can see, although the ratings are a bit better than the completely nonsensical crap we have on battle.net right now, it still doesn't make much sense. Special Forces ranked 1? Mafia ranked lower than Pobes vs Zeelot?