Spectating AI matches and some worrying AI findings

[Langeveldt]

I appreciate the thanks, I love my cricket.

I am using the on disk teams, although I have edited the names to real and reduced the batting stats of all tailenders to one dot.

I cannot upload the videos although I wish I could. My internet isn't great

Just played an ODI
Australia 312 (Smith 68, Rodgers 102)
Zimbabwe 120 (Masakadza 28)

And now halfway through a test
Australia 110 (Rodgers 68, Panesar 4 for 24)
England 95 for 2 (Cook 55)

 

[whiteninness]

I appreciate the thanks, I love my cricket.

I am using the on disk teams, although I have edited the names to real and reduced the batting stats of all tailenders to one dot.

I cannot upload the videos although I wish I could. My internet isn't great

Just played an ODI
Australia 312 (Smith 68, Rodgers 102)
Zimbabwe 120 (Masakadza 28)

And now halfway through a test
Australia 110 (Rodgers 68, Panesar 4 for 24)
England 95 for 2 (Cook 55)

Do you think perhaps it's best not to copy the attributes over when linking teams with CA but just the look and name. It seems like BA tested the game with the default teams and, as expected, produces better results with them. Although Australia being bowled out for 110 is a bit unrealistic (well, not totally )

Do you think there is room to distinguish between different tailenders though? Jimmy Anderson deserves to be a tad higher than Monty Panesar, for example.

 

[BagoZA]

Do you think perhaps it's best not to copy the attributes over when linking teams with CA but just the look and name. It seems like BA tested the game with the default teams and, as expected, produces better results with them. Although Australia being bowled out for 110 is a bit unrealistic (well, not totally )

Do you think there is room to distinguish between different tailenders though? Jimmy Anderson deserves to be a tad higher than Monty Panesar, for example.

As long as the skill set influences the AI in the way that its supposed to, there should be no problem with downloading stats for created players. What you are suggesting, IMO, would be trying to work around a flaw. Besides as I've said, I've played with user created teams, and they perform as expected.

 

[Langeveldt]

I am using on disc teams and have only changed the names and tailder stats. I have guys like Tredwell slightly higher, let's see how he goes.

My argument was that the skill set is poorly calibrated. And by that you have guys with 25 percent skills whacking fifties all the
time. We need to be aware of that when downloading and editing teams.

 

[Langeveldt]

Anderson 35 not out with one dot of skill.

I have no explanation for this lower order bug. I wish I knew something from big ant about how these attributes are applied to how AI batsmen behave

 

[whiteninness]

Anderson 35 not out with one dot of skill.

I have no explanation for this lower order bug. I wish I knew something from big ant about how these attributes are applied to how AI batsmen behave

Is it not so much a lower order bug but a last wicket partnership bug?

 

[Langeveldt]

Is it not so much a lower order bug but a last wicket partnership bug?

Quite likely, but what causes it? After watching a lot of cricket on dbc I'm none the wiser. I assume it also comes into play when the player is bowling, which is a bit disconcerting

 

[zimrahil]

Don't forget it's still early days - more testing is probably needed really but I have found this thread really interesting

 

[Langeveldt]

Don't forget it's still early days - more testing is probably needed really but I have found this thread really interesting

If anyone wants to spectate games it requires no human input and the scoresheet stays up after every innings and session, so just leave it running while doing something else and post the results

Remember to turn replays off for faster results

 

[isrararrafi]

Yeah, the attributes thing really isn't a DaVinci Code to break, it'll be a long term tweaking thing as we get to grips with how they're implemented in game. The academy created teams just need to be rebalanced by the creators, not a big job...

says you.

 

[twobob]

Have you considered becoming a tester?

The pay isn't great and the hours are awful but I would guarantee that a test team somewhere would be delighted with the effort and thought behind this investigation.

As a side note, imagine the resource it would take for BigAnt to test these things thoroughly (and i'm sure they do as much as they can). You've tested part of this scenario at such a small scale as to what would actually need to be tested to verify this properly such as the different match conditions, length, type, players and teams. It would take multiple resources weeks, then it would be tweaked & tested again and so on.

I just think i's worth putting into perspective when people complain about bugs or issues that they may encounter when they play any game.

The best thing that somebody can do in this situation is exactly what you have done, I don't doubt that Ross or the Big Ant team will be looking at this information and going back to see if they can improve on this.

cheers

 

[Langeveldt]

If tailenders are slapping runs with near zero skill, I don't see how we can ever make a genuine Chris Martin, until a patch at least.

 

[BRAB]

The great thing is this is BigAnt so we know they'll support the game post release, at the very bare minimum with mod support so PC can do all sorts of wacky things to it

 

[Langeveldt]

Have you considered becoming a tester?

The pay isn't great and the hours are awful but I would guarantee that a test team somewhere would be delighted with the effort and thought behind this investigation.

As a side note, imagine the resource it would take for BigAnt to test these things thoroughly (and i'm sure they do as much as they can). You've tested part of this scenario at such a small scale as to what would actually need to be tested to verify this properly such as the different match conditions, length, type, players and teams. It would take multiple resources weeks, then it would be tweaked & tested again and so on.

I just think i's worth putting into perspective when people complain about bugs or issues that they may encounter when they play any game.

The best thing that somebody can do in this situation is exactly what you have done, I don't doubt that Ross or the Big Ant team will be looking at this information and going back to see if they can improve on this.

cheers

I've logged every innings since picking up my copy of dbc14. I will post averages and things in due course.

I always use legendary difficulty when spectating a game. The conditions are set to random.

I'm going to spectate an entire ashes series, ill post batting and bowling averages soon.

 

[BRAB]

Langeveldt you're an absolute star!

Yell out if you need any help with some simulated games or anything, I can set the game up exactly how you want it and you could test more than one thing at once