Getting over 90% ?

[jamie.santavy]

So I've had the game for a little less than a week. With a little practice and the help of the guide pinned at the top of the forum here I've gotten the hang of getting my company started and working well.

I seem to have hit a plateau around 88-89% game quality though, I can never quite break 90. I know that part of gamescore is random, but I played a game all the way through the year 2004 without getting over that hump and I'm not sure what I'm missing.

In that game I got the graphics and music department going in the early 90s and was using all the available features / improvements available from both, as well as all the improvements from the QA dept but it never seemed to make a difference.

I have also used as many as 16 employees working on the game.

In my current game (1989) Im experiencing the same problem and don't know what to do.

Using game reports I'm sure I have the right sliders and allocations, but I still haven't nailed down the right target audiences. I always use an engine that is made for the genre of the game I'm making and I always keep up to date on features.

Anything else I'm missing to get games up over 90% ?

[yutterh]

This guide has everything, it is my preference guide. I get 90's in the mid 80's

http://steamcommunity.com/sharedfiles/filedetails/?id=616265616

Also have you made any game reports? They tell you he target audience and let you know if any of your sliders are off. I did that one time and never realized my slider was off by one point.

[Varnhagen]

I'm currently in the process of writing a guide for the review process - so I can tell you a bit about how reviews work.
You have pretty much two distinct parts to the gamedevelopment process:

1. Point accumulation
   
2. Settings (Sliders, Options, earned Stars. Trends)

Points are earned at random during the development of the features that games are comprised of. You can influence the amount of points with your team composition.
Let's say you are developing the first 2-feature game and your dude has 10s in all 4 game-dev-skills. You don't use a specialized engine or any genre- or feature-specs to keep the argument simple.
Your dude (or dudette) will roll ten times on all 4 skills, to see if points are produced. For every skill that fires, a random number between the bounds (1) and (skill /10) is chosen.
To keep it simple again, in every roll all four skills fire:
Bounds 1 - 10/10 = 1 --> mean: 1;
--> You will develop a game that has 10 points in all categories.
Let's do the process with 2 dudes - the one from the former and a new hire who has 40 in all his skills. They have equal workwill so neither of them will work faster.
10 rolls / 2 dudes = 5 rolls per dude.

First Dev: 5 * [1;1] = (5 + 5) /2 = 5 points for all categories.
Sec Dev: 5* [1;4] = (5 + 20) /2 = 12,5 points for all categories (on average).
By adding a marginally skilled dev to your team, the output has improved by 75%.

Let's say some time has passed and for old times sake both decide to do one of their old games anew. They are both now maxed out in all skills.
First Dev: 5 * [1;10] = (5+50)/2 = 27,5 points on average
Sec Dev: the same.
By maxing out all stats, your game has improved by 450%.

Above should illustrate, that it doesn't matter how many people are developing a game, but how good they are. Ideally you'd only want the best. In a perfect world, new hires should never by allowed anywhere near a development room.
Depending on the type of game you are developing you should bias your dudes towards some skills.

During the first stage of the review process, the points are divided by a factor that itself is a quotient of game-year and difficulty level. Afterwards these values are multiplied by their importance, added together and possibly topped off.

An example:
You started in 1980. It is now year 1990 = year 10.
You are developing a skill game. 10% GFX, 10% SFX, 30% Tech, 50% gameplay.
Your game has 1000 points in every of the four tallys.
Your difficulty is normal ( difficulty factor 0.55)

1000 / (10 / 0.55) = 1000 / 18,2 = 55
55*0,1 + 55*0,1 + 55*0,3 + 55*0,5 = 5,5 + 5,5 + 16,5 + 27,5 = 55 total review score.

Let's do another one: RPG (0,25; 0,15; 0,2; 0,4), year 15, diff: legendary, points (2700, 1300, 1000, 2400)
difficulty- and time-dependant factor:
15 / 0,7 = 21,4

GFX = (2700 / 21,4) * 0,25 = 31,5
SFX = (1300 / 21,4) * 0,15 = 9,1
CON = (1000/ 21,4) * 0,2 = 9,3
GP = (2400 / 21,4) * 0,4 = 44,9 TOTAL = 94,8. --> Topped off at 88.

What should be evident is, that different genres require a different approach to team selection:
E.g.: Do you really need the best Sound and GFX specialists for skill games? If one earned point equates to a tenth of review importance shouldn't you better focus on tech and gameplay?

That's the point where your gameplan kicks into action. How do you want to build your team to create the best possible games? What kind of skill do your employees need?
That's for you the player to determine and the reason why we play the game.

You can test any game you develop with the above approach. If you score below 88% then your team needs improvement.

The next stage in the review process is where all the options come into play: right settings give bonus percentages wrong settings deduct from your score.
Is your game a sequel? How many experience-stars do you have? Is the genre/theme in the trend? Depending on the year, percentages will be subtracted if you lack improvements, you get a bonus in the first few years, and quite some shenanigans more.
But those aren't that interesting, if your game report shows, that you have done everything right. The gameplay lies in reaching the first 88% threshold.

I hope that helps...

Beste Grüße

Addendum:
One last thing - the sliders governing the percentages don't come into play here. If you develop an RPG with 0% - 0% - 50% - 50% sliders your game will still be reviewed with the importance values listed in the game report.
What these sliders actually govern is the probability of the skills generating points during development. Thus you'd want to fiddle around with them to make up for deficiencies of your team. You don't have to set them to "THE PERFECT SETTINGS", as is so often told on these forums...

[jamie.santavy]

Thanks to both of you, Varnhagen based on what you're saying it's safe to assume my problem is point allocation, and it may actually be caused by me replicating the priority sliders from the game reports. My main development team is farily well trained and balanced across the board. While they dont have all skills high I have graphics and music people to balance out the design / tech people.

But my last game published in 1990 was an RPG and it still only got an 87% rating after over a year of training for the dev team. The sliders were right, audience was right, genre and topic were popular.

My points were 3500 gameplay. 1500 graphics. 1500 sound, but tech was only 981. However in the reviews section graphics sound and tech were my highly reviewed areas. Gameplay was still only at 75%

I'm playing on normal difficulty my priority sliders were the ones recommended by game reports. Right now my dev team consists of 8 and I only plan to add a quality person very occasionally. Is my team just not good enough in gameplay?

[lordi]

Originally posted by Varnhagen:

I'm currently in the process of writing a guide for the review process - so I can tell you a bit about how reviews work

sounds great. where will we able to read this guide? here at the steam forums?
i'm still having questions about the review-score, and especially about the genres and topics and their relationships. so i'm waiting for your guide. looks like you know a lot about the mechanics.

best regards

grüße

ps: wird der guide auf deutsch sein? ich vermute mal nicht...? ^^

[Varnhagen]

Let's do go through CalcReview() step by step.

RPG(0.25;0.15;0.2;0.4) Points(1500;1500;1000;3500) Year(10) Normal(0.55)

ScoreGameplay = 0;
ScoreGFX = 0;
ScoreSound = 0;
ScoreControl = 0;
ScoreTotal;

year = 10;

ScoreGameplay = 3500 / 18,2 = 192
ScoreGFX = 1500 / 18,2 = 82,4
ScoreSFX = 1500 / 18,2 = 82,4
ScoreControl = 1000 / 18,2 = 54,9 // That looks bad

ScoreTotal = 192 * 0,4 + 82,4 * 0,25 + 82,4 * 0,2 + 54,9 * 0,2 = 120,9 //That looks fine

Top the scores off:

ScoreGameplay = 192 --> 80 //Huge waste
ScoreGFX = 82,4 --> 82,4 //No change
ScoreSFX = 82,4 --> 82,4 //No change
ScoreControl = 54,9 --> 54,9 // No change
ScoreTotal = 120,9 --> 88 // The gameplay waste is cut out here too

I'm assuming you have no bugs or else their would be some deductions now.

Since year == 10, you need:
- all gameplay improvements or else there is a 0.85 deduction for each missing.
- first five sound improvements
- all, but the fourth, gfx improvements

Do you have a good genre combination? Would add +5 to total and gameplay
Add your genre and theme stars to gameplay and total
Your gameplay <> gfx sliders are perfect? Otherwise each point off gives 1 deduction
Do you have all languages? Every missing language translates to 0.2 deduction
Do genre and them match? Otherwise you'll get a -2.55 deduction to ScoreTotal and ScoreGameplay

ScoreGameplay = 80+5 + 3 + 3 = 91;
ScoreTotal = 88 +5 +3 +3 = 99;

New Topoff Round

ScoreGFX = 82,4 --> 82,4
ScoreSFX = 82,4 --> 82,4
ScoreControl = 54,9 --> 54,9
ScoreGameplay = 91 --> 90 +- 4
ScoreTotal = 99;
//No changes

ScoreSpread Deductions:
if (ScoreTotal > 80) //True; 99
80 < ScoreSFX < 90 --> ScoreTotal -1; ScoreGameplay -2;
80 < ScoreGFX < 90 --> ScoreTotal -1; ScoreGameplay -2;
50 < ScoreControl < 60 --> ScoreTotal -10;

ScoreGFX = 82,4
ScoreSFX = 82,4
ScoreControl = 54,9
ScoreGameplay = 86 +- 4
ScoreTotal = 87;


In summation;

Your control value su*ks and you waste too much effort on gameplay.

1. Train programming. Either with courses or by using your staff in rooms that require programming. Don't train gamedesign
   
2. Set Gameplay to 10% or even lower in the game.design. Your current output is more than double the necessary points.
   You'll get a lot of GPpoints from your QA anyway. Crank your control up to 50%.
   
3. Program games that don't need no programming.
   
4. Have fun!

[Varnhagen]

Originally posted by lordi:

ps: wird der guide auf deutsch sein? ich vermute mal nicht...? ^^

Nope, aber Du darfst ihn gerne übersetzen. Ich hab kein Ahnung wann ich fertig bin.Kann noch eine ganze Weile dauern, immerhin will ich ja auch Screenshots einbinden und ein paar Grafiken, die den doch sehr trockenen Stoff verständlich machen.

No eta on a guide. But not anytime soon.

Beste Grüße

[jamie.santavy]

Thank you very much Varnhagen. It would be great to see your guide when you're finished and learn the formulas and how they work. The game is a little counterintuitive about the scoring in a way. When my game gets reviewed my lowest % is gameplay so it led me to think that's where I needed the extra points

[jamie.santavy]

btw got 92% on my next game after shifting sliders

[Varnhagen]

Yes, it is pretty counterintuitive.
The distinction between less-deduction-affected GFX and SFX on the one hand, and more-deduction-affected Gameplay scores (provided optimal settings) leads to a lesser representation of Gameplay scores.

Two rules of thumb:

1. If GFX and SFX get good scores (90+) use their points tally as your baseline. Since all scores are calculated the same way at the start (points divided by factor), you can gauge where the necessary maximum for your time and difficulty resides.
   
   
2. Try to reach 90+ in all subscores, so you don't get these massive deductions that brought your score down from 99 to 87.
   If you have a capacity to generate more points than the current necessary maximum, you can try to funnel them into the review-prioritised areas.

If you get a game that scores at least 96% the last part of the review is some RNG-nonsense to randomize your score and give you a shot at the elusive 100%.
But to get there, the margin for deductions is extremely low.

I'm glad that I could help. Hope you'll have a lot of fun figuring out some other of the basics, and enjoy some good legendary/slow games, where every decision counts.
Have fun.

[jamie.santavy]

It's funny after picking this trick up the game started feeling a little too easy so after making a few billion in my last game I decided to start up a hard game and see the difference. I have to give this game credit for providing challeng beyond a lot of other games in the genre.

But I'm getting the impression from what you've already said that as features / improvements become unlocked over time that games that don't have those features / improvements will suffer a penalty in their review score.

What makes normal so easy is the ability ot hire 3 employees and create an 88% game immediately, and you'll have enough money to develop an engine and its easy enough to keep up with researching the early features.

I tried this strategy on hard (borrowed 50k to do it) and what I found was this strategy isnt sustainable. My first game was 55% with genre in trend and all languages and my second game was 56% both made some modest profit, but not enough to keep up with salaries.

So I'm assuming I need to work the contracts early on and then its either publish games as cheaply as possible to unlock dev kits, or build the engine and try and build games with features + languages that can hopefully eek out small profits.

My concern is that time is your biggest enemy in the early game. Once your company is at a point that it can keep up with all the latest research and improvements the game is simple, but it definitely seems like a big task on hard forget legendary lol.

What also stands out to me is there's no training room until you upgrade your building, so outside of features and hiring staff there's no way to improve game quality. I can forsee a lot of frustrating games because of this. It's not like you can just sit there for 5 years and build up a nest egg and try and hire a few people, train them, and develop a game that will set you up with good cashflow.

[Varnhagen]

Features are just containers for work-units. The more modern a feature is, the more work-units it contains, the more opportunities for your employees to produce points, thus raising your points output.

Normal is easy because you can create good games with bad teams. On harder difficulty you can still produce good games by utilising all advantages, especially, engine and genre specialisations of your employees.

You might not be able to improve your peeps by training, but you can improve their output.

Recall points generation:
Random(1, Skill / 10 + genre.spec + engine.spec + feature.spec)

If you start with a dev that has 20 across the board, you'd normally expect them to generate either 1 or 2 points in every workunit. If that dev has a genre="skill"-spec and an feature="ascii"-spec and you develop a "without engine" skill game, you'll notice that your dev produces points in the range [1;7].
That makes 35 points per category on average. Exactly what you need to produce a 100% game in the first year on hard. On top you'll notice that features give you an added bonus to one or more of the point totals. The randomness of the process will prevent you from ever producing the perfect game, but you can do better than 80%.

It doesn't matter how many people are developing a game, it only matters how good they are and how you are utilising them.

There's no better time to produce games than the first 3 years. You'll get bonus points during the review process to make up for the volatility of early game. (Actually it is a multiplier which decreases every year. In the 4th year it is just barely larger than 1, while in the 5th year the multiplier vanishes)

With the introduction of slow-mode I can't attest to any concerns that time is the limiting factor - never felt like that.

[Varnhagen]

To add a bit of context to that,

Originally posted by jamie.santavy:

But I'm getting the impression from what you've already said that as features / improvements become unlocked over time that games that don't have those features / improvements will suffer a penalty in their review score.

If improvements aren't part of a game once they are unlocked there will be penalties in the review, that's right. This lends to tedious repetition where you have to select all your special departments, click the 'all improvements' button. There is pretty much no decision to be taken in that.
But at least you get an opportunity to employ all those employees with specialisations you don't need in your main dev groups. For me it isn't unusual to have small core development departments with up to ten employees and large supporting gfx, sfx staff groups.

Feature selection is a bit different. You can create somewhat competitive games in late game, if you disavow the notion that the name of a feature has any bearing on the quality of the game. If you produce a B+ 10-feature game in 2010 eg, you can still earn all necessary points if you get bonus points from the improvement departments and use the most modern features, i.e. the features with the highest amount of workunits.

[yutterh]

Originally posted by jamie.santavy:

It's funny after picking this trick up the game started feeling a little too easy so after making a few billion in my last game I decided to start up a hard game and see the difference. I have to give this game credit for providing challeng beyond a lot of other games in the genre.

But I'm getting the impression from what you've already said that as features / improvements become unlocked over time that games that don't have those features / improvements will suffer a penalty in their review score.

What makes normal so easy is the ability ot hire 3 employees and create an 88% game immediately, and you'll have enough money to develop an engine and its easy enough to keep up with researching the early features.

I tried this strategy on hard (borrowed 50k to do it) and what I found was this strategy isnt sustainable. My first game was 55% with genre in trend and all languages and my second game was 56% both made some modest profit, but not enough to keep up with salaries.

So I'm assuming I need to work the contracts early on and then its either publish games as cheaply as possible to unlock dev kits, or build the engine and try and build games with features + languages that can hopefully eek out small profits.

My concern is that time is your biggest enemy in the early game. Once your company is at a point that it can keep up with all the latest research and improvements the game is simple, but it definitely seems like a big task on hard forget legendary lol.

What also stands out to me is there's no training room until you upgrade your building, so outside of features and hiring staff there's no way to improve game quality. I can forsee a lot of frustrating games because of this. It's not like you can just sit there for 5 years and build up a nest egg and try and hire a few people, train them, and develop a game that will set you up with good cashflow.

My strategy is this and i play on legendary. I make a 5x4 development room in the top left corner. in the top right to bottom right corner should have enough room for a 3x9 research room. then i put in 4 of the wooden desks in, reasonibg is they are smaller and i can get mor epeople into a small area. across the way from this i make another 5x4 development room. Now i usually have this second room focus on contracts only. but lately i have been using them as another game publisher too. but i do this later when i have money to build it. exact amount is fine but i like having extra just incase.

I do borrow from the bank if i need the money. just remember to pay as much as you can back by week four and never borrow on week four. you wont even get to spend the money before you have to pay interest on it.

Now i get one Game designer, programmer, graphics, and music focused guy and put them in the development room. everyone else goes to research. When i make my first game I usuaully use a guide, if i cannot remember from memory. but i got arcade down pretty well now lol. I make my first game with only the native language in. native language is free to add. everything else is like 5,000 which adds up in the first part of the game. so if the games gets higher then a 70% i will add others in a update. Most money can come from contracts or your own engine. I have my engines sold for $250,000 and 20%. I sale quite a bit and try and keep them updated. I only make the genre's that i am using cause they are expensive to make for each genre.

I keep that up tell i get 500,000 and then i buy the next building which is $400,000. i keep rooms to 5x4 and give only four desks in each. I also make a production room and a storage room. Making your own games is expensive to start off as but it extremely rewarding. when i do produce my own games i make them $30 and put one of each item on with a $1. you get a contents bonus to your game and the rating goes up. once that starts to slide off i wait tell i am almost out of stock or intintionally go out of stock and i lower the game price to $20 and remove everything except the manual. I get a few more game sales that way and once that starts to dip i drop it to $10. i do this for everygame until i can support a successful market hype increase to 100 hype then i put everything into the game box and charge $49 instead. once that starts to dip i go out of stock lower it to $30 and take off anything over a $1. then i continue that process.

[Doctor Bog]

Originally posted by yutterh:

This guide has everything, it is my preference guide. I get 90's in the mid 80's

http://steamcommunity.com/sharedfiles/filedetails/?id=616265616

Also have you made any game reports? They tell you he target audience and let you know if any of your sliders are off. I did that one time and never realized my slider was off by one point.

Really dumb question here -- but how do you generate a game report??

[lordi]

Originally posted by Doctor Bog:

Really dumb question here -- but how do you generate a game report??

it's not a dumb question... i was wonderung myself about reports at the beginning... :D

u can create game reports in the quality assurance room.

copied from the wiki:
"Quality AssuranceEdit
The quality assurance is locked at the beginning, but will be unlocked when you develop a game with more than 50 bugs. In this room you can remove bugs during development or add some improvements that will raise the games ratings."