Blood Bowl 3

[Smiling Tom]

yeah, quite sure that s what will happen. But is as true that the 2016 brought in some new blood and brought the game back to the spotlight

[Juriel]

There might be a single book collecting all the randomly released Bloodbowl rules in 2020, but I doubt there'd be a new tabletop version, since all the changes that GW did in 2016 were... basically nerf Piling On.

 

Which, while I think there should be something to curb the herpderp Chaos kill teams, was an absolutely boneheaded maneuver. It still leaves Chaos with ClawMB on everyone, while every other team with natural S access (most impacted, Norse, who absolutely need to remove people fast) gets a hugely important tool taken away. And now any elf with a single double is as good at hitting as any S access player can ever hope to be. And any players with only S access (Yhetee, Ogres, Underworld Trolls) just got one skill removed, with nothing taking its place.

 

...so I guess, the less GW tries to modify the game, the better. The game's balance is a crap-shoot, and there's no reason to play anything but chaos or elf teams in a persistent league, but anything GW does will just make it worse.

Edited May 16, 2019 by Juriel

[Juriel]

That said, if GW were serious about Bloodbowl as a tabletop experience, the current format (and match durations!) was never going to work to widen the player base beyond old grognards. It needs something more than what Blitz Bowl does (though that has some good ideas).

[Juriel]

Oh yeah, BB3 just makes sense. It's been years since BB2 came out, and they've once again released almost all the teams.

 

Which would be fine, if BB2 hadn't taken so many steps back... basically no lobby chat, no spectator chat, lacking league management like not being able to set stadium upgrades off, half-hearted attempt at micro-transactions, no customization of players, same player appearance across all levels which actually hinders gameplay because you cannot tell them apart, etc.

Edited May 16, 2019 by Juriel

[C2MC]

11 minutes ago, Juriel said:

same player appearance across all levels which actually hinders gameplay because you cannot tell them apart, etc.

 

I think this was my biggest disappointment in BB2. The ability to view skills above players' heads did sort of make up for it on a gameplay-level. But aesthetically, it was a giant step back.

[Verbal_HM]

The rumours I heard said that BB3 was coming, but Cyanide had been kicked out of the party as GW wanted a more competent developer.

 

Cos, you can believe everything you read on the internets...

 

We'll upgrade in time, as, as far I can can comprehend. Our little community did actually convince GW of the monetary returns of making direct ports of their TT games digital.

 

That said, looking at how they're trying to make new things, only to have 3rd party smalls completely out pace them on releasing actual models to go along with the rule changes. I can understand why they're pissed, that said, it has to be viewed as their own fault for being an old, and slow, international cooperation these days?

[Rymdkejsaren]

I do wish that's true. BB3 by a competent and responsive developer would get me back in the game, big time.

[Verbal_HM]

1 minute ago, Rymdkejsaren said:

I do wish that's true. BB3 by a competent and responsive developer would get me back in the game, big time.

 

You say that, but the sad fact is that if it wasn't for Cyanide we wouldn't exist as a community?

 

As bad as they are at making games, they were the company that made OCC a thing.

[Verbal_HM]

Also, just the rant on, can you imagine a fresh team trying to take on this problem?

 

Cyanide have made a fuck tonne of mistakes, and they have mostly learnt from them.

 

Back to a new team is kinda grim, we'll go back to the days of Tackle still being active on a prone player, crap like that.

 

It's a complex game, and lazy coding by people who don't play it hurts. It hurts a lot.

[GaryW]

Which I think any update will really rather be BB2.5 than 3. I doubt cynadie will write the game from the ground up again. It wouldn't make much sense as the current engine would be good enough for a few more years yet. Can only hope they don't start again with 8 teams and more options for the admins.

[Verbal_HM]

You have to appreciate that current changes are being driven by GW, and not Cyanide.

 

We, us, we pissed a fuck tonne of money into the pot, thus GW are now paying attention to us.

 

They didn't necro a dead game by random accident. We, us, we made this happen.

 

It's a business decision, cos it makes them money?

[Smiling Tom]

What I fear is that a new coder going up from clean slate might for some reason skyrocket the requisites for playing something that ain't that far from chess. 

[Juriel]

The main complex thing about BB to code would be AI. The rest are just a flowchart of checking for exceptions, which is more down to understanding the game... which is in no way a certainty if GW were to pick a 3rd party randomly. 

 

'why would anyone want to NOT use their Dodge when hit?' 

[Verbal_HM]

@Juriel just plain eat a dick on that comment.

 

AI, it's truly flipping irrelevant?

 

This game is best played against other fully aware humans, and you know it.

 

 

Concept is to get the code in order, then start working on AI.

 

Your puzzle is all mixed up.

[Smiling Tom]

6 minutes ago, Verbal_HM said:

@Juriel just plain eat a dick on that comment.

 

AI, it's truly flipping irrelevant?

 

This game is best played against other fully aware humans, and you know it.

 

 

Concept is to get the code in order, then start working on AI.

 

Your puzzle is all mixed up.

Dude, that's exactly his point.

But you both seem to ignore that what hogs resources is the interface/graphics/animated toons.

[Verbal_HM]

@Smiling Tom, honestly, I'm seriously confused.

 

Truthfully, that may be down to my blood alcohol content. 

[GaryW]

Basically the easy part to program is the logic involved going from A to B to C etc. I.e you roll a dice and check the result on a logic table, then the processor goes on to next thing to check. I'm no programmer so it maybe I a little more involved but have dabbled with PIC's doing some basic stuff like a traffic light sequence . But what will hog your processing power is the graphics and AI.

 

As you pointed out the, programming an AI would require a lot of work and is almost pointless to get the best out of the game. Surprisingly there are quite few on the Steam forums that do want a competent single player game...

Edited May 17, 2019 by GaryW

[Verbal_HM]

@GaryW, I still claim stupid, thus I need you to reinforce that fact.

[GaryW]

24 minutes ago, Verbal_HM said:

@GaryW, I still claim stupid, thus I need you to reinforce that fact.

 

I'll try :) Using my example of a traffic light sequence. Just a quick explanation a PIC is a microcontroller, a chip that is a processor that has ROM and RAM all built in, a mini computer if you will. Used in loads of devices from washing machines to alarms. There are many variants available. Basically you start the program by waiting (polling) for a switch to be pressed, or in a game your mouse button. Once pressed it will execute the next line of code. In a traffic light sequence, if I remember it correctly, the light will start at green, using a LED, you hit the button, your next line of code is to put a delay of few seconds, the next line is to switch to from a green LED to a Yellow LED, then next line of code is a delay again before switching to the final Red LED. Some more lines of code are needed to follow the sequence back until you end up back to the beginning again with a green LED and the processor then waiting for you press the button again. 

 

In a game such as BB, you have set sequence you follow to implement the action you wish to take, which you all do naturally when you play table top. That sequence is pretty easy to program, you can write that out in flow diagram if you wish before you code that and is supposedly good programming practice to do so. 

 

With modern computers, that sort of programming is pretty straight forward, if you take chess which again has easily defined rule set, these where programmable on the earliest computers with ease. AI programming is whole different kettle of fish though. I have no experience of how you do that, but if you take something like the film I Robot you could do it on a probability. I.e in that movie the robot chose to save Will Smith's character rather than his son, as the probability was higher. That will be very hard to program a computer to play a good game of BB. Heck it took super computers to play a good game of chess to beat a grand master!

 

I will say much of the graphics work is not really handled by the actual processor nowadays as we have dedicated processors for that now, but the main cpu still has to tell what the graphic chip to draw etc. But modern CPU's are so quick it might not be that big strain on it.

 

Hope that make sense :)

[Verbal_HM]

@GaryW, it kinda made sense.

 

But what you're saying is...

 

 

You're stuck in a problem of mathematics vs design?

 

1 vs 0, it's a binary sum. I don't own the words to explain that fact to you people.

 

It's back at the time in the universe where everything was one, we all started at Hydrogen?

 

A massive ball of 'one', a single atom, a single entity.

 

A single ball of energy, a closed experiment which only had a single master?

 

That cunt, that single fucking cunt. The bastard who made us.

 

 

 

 

 

I do believe he wore red socks on a Tuesday, and if you don't believe that you're dead to me.

 

 

 

 

 

I also believe I've just typed this message in the wrong thread, this should be 'Verbal Rants'.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Meh, fuck Labour, just fuck Jerry. He doesn't speak for the working class.

[GaryW]

Kinda off, but the rule set for blood bowl is 1 or 0, or pass or fail if you will. To get the AI to play a game would surely have to based on a probability.

 

As for the universe, you could perhaps boil that down to something and nothing, or 1 or 0. But that is philosophical debate belonging else where. And the less said the better of current political situation ?

[danton]

19 hours ago, Juriel said:

The main complex thing about BB to code would be AI. The rest are just a flowchart of checking for exceptions, which is more down to understanding the game... which is in no way a certainty if GW were to pick a 3rd party randomly. 

 

'why would anyone want to NOT use their Dodge when hit?' 

 

The only way to get a decent AI into the game would be to leverage something like https://njustesen.files.wordpress.com/2018/06/justesen2018blood1.pdf

 

I just can’t see any Game Development team creating a decent AI in conjunction with doing a good job with the UI, graphics, admin features, etc.

 

Creating a competitive AI is a huge challenge on its own for something like BB, but definitely doable given the recent advances with games like Go and League of Legends. 

 

Making skills like Dodge optional has some rare applications (avoiding surfs, chain pushes, etc), but that is more of a UI challenge. 

Edited May 18, 2019 by danton

[Juriel]

1 hour ago, danton said:

I just can’t see any Game Development team creating a decent AI in conjunction with doing a good job with the UI, graphics, admin features, etc.

 

Yes, playing BB is on the stupid side of simple but complex, that is why I listed it as an example of something that is actually hard.

 

1 hour ago, danton said:

Making skills like Dodge optional has some rare applications (avoiding surfs, chain pushes, etc), but that is more of a UI challenge. 

 

Point wasn't to look at this all-encompassing example, but to put forth a corner case that nevertheless is something that WILL come up and is easy to account for using the same systems you use for other things... which is not only an UI challenge, when the existence of such options is best realized during early design (so they can during coding make one broad skill toggle, rather than hardcoding the one exception they could think of).

 

Anyway, point being, people making the game have to know how to play the game to realize such options even exist, is the hardest part. If you have that, then the rest of it (barring AI) is simple, from the coding side of things.

 

Graphics etc will of course take up most of the actual effort, yet the jump from BB1 to BB2 gave us worse art direction in many cases, and no player customization.

[GaryW]

2 hours ago, danton said:

 

The only way to get a decent AI into the game would be to leverage something like https://njustesen.files.wordpress.com/2018/06/justesen2018blood1.pdf

 

I just can’t see any Game Development team creating a decent AI in conjunction with doing a good job with the UI, graphics, admin features, etc.

 

Creating a competitive AI is a huge challenge on its own for something like BB, but definitely doable given the recent advances with games like Go and League of Legends. 

 

Making skills like Dodge optional has some rare applications (avoiding surfs, chain pushes, etc), but that is more of a UI challenge. 

 

An interesting read, doesn't explain much of what they are actually doing but must be some PhD graduates who are fans of the game. Must be pretty cool to get some funding to do this kinda of research.

 

I wouldn't say never on a good AI on any game. I have seen some recent trends in indie games that are using random generators for world generation. I guess there is alogrithim that is now taught at university or is well known in the programming community that a lot of indie developers use.

 

Once some clever people have worked out a set of code for a smarter AI these will be used everywhere. As there is no point in reinventing the wheel once it is widely known. I do remember seeing a news report years ago where a university taught a computer to play football, it started of not been able to play at all and leaving the computer alone it figured out how to play. It was just dots on a screen but still pretty neat.

[Borke]

On 5/18/2019 at 6:06 PM, GaryW said:

Once some clever people have worked out a set of code for a smarter AI these will be used everywhere. As there is no point in reinventing the wheel once it is widely known. I do remember seeing a news report years ago where a university taught a computer to play football, it started of not been able to play at all and leaving the computer alone it figured out how to play. It was just dots on a screen but still pretty neat. 

 

It's my impression that this is indeed the latest trend in AI development. You teach a program the rules of a game, you then put it into a "room" with another copy of itself, then you lock the door and tell the program to find a way to beat itself. This is how Go was mastered, that's how the OpenAI for Dota 2 was developed (with the cool added concept that the AI is playing as 5 seperate entities, who have learned to cooperate with each other), and I imagine it's the same for League of Legends. The fact that this approach was applicable to Go makes me think that it's applicable to Blood Bowl as well. It just costs a LOT to do so, still. The efforts for Go and Dota 2 were large, sponsored investments. I'm not sure this is ready yet to the point that any game development company can just say: "We want a strong AI for my new game, so we'll just bring in company XYZ to develop one for us". It would drive up the costs too much.