What level is giving you the best experience

[Fluid Druid]

I started playing easy aim Vs Hard A.I. but got fed up with the A.I. making almost impossible shots and winning frames from them, a better balance seems to be Pro Vs medium A.I.
I can make century and frame winning breaks in pro but can easily mess one up too. The A.i. on medium still seem to make amazing shots but they seem better balanced in regards to clinching the frame from them. This i find is giving me the best experience right now.

[jinx20001]

ive played pro from the start, never tried anything else yet. knew straight away online play was using pro aiming aids so that was the one to get good with.

[Paulie]

Originally posted by jinx20001:

ive played pro from the start, never tried anything else yet. knew straight away online play was using pro aiming aids so that was the one to get good with.

Same as you Jinx. Overall I like Pro, though (as suggested elsewhere in this forum) requires tweaking to make long potting more difficult.

Adding optional toggles to each of the aiming aids would enable players to fine-tune / find their ideal level of "difficulty".

[Fluid Druid]

I wanted the A.I. to be tough yet fair and with easy aim it could have provided a good challenge with both yourself and the A.I. but it just does not appear to work well that way, so we will see what pro and medium A.I. are like..so far its a better balance

[Paulie]

Originally posted by DoGgs:

I wanted the A.I. to be tough yet fair and with easy aim it could have provided a good challenge with both yourself and the A.I. but it just does not appear to work well that way, so we will see what pro and medium A.I. are like..so far its a better balance

Let's see what the Devs do with the A.I. updates based on player feedback moving forwards. It's a fine balancing act and would have been nigh on impossible to nail it first attempt. Sports games like FIFA have constant balancing tweaks through the year, so expect a similar process with Snooker 19.

Having watched some of the shot making and break building at the World Championships so far, I think the Hard AI is pretty accurate!

Having random elements such as kicks to interrupt a break would be great, though not sure how you'd accurately simulate that

[Fluid Druid]

The Hard A.I. are accurate in terms of winning frames in 1 or 2 visits but the constant rolling in of very difficult long shots is rendering back to baulk safety play pointless because you only have to leave a shot on and they pot it 9/10, this is far from accurate.

[jinx20001]

long pots is a problem, i was discussing this with somebody yesterday and we come up with this:

pro players dont have an issue lining up long shots, they are typically accurate but the issue is long potting requires excellent cueing so how could we implement that into the game? well... we thought as a start and for online play we should not reduce the yellow line as this reduces how accurate the shot is but instead we should perhaps make the ''white zone'' of the power bar much smaller for long pots which would simulate poor cueing if you cant hit the zone and thus throw the shot off target as this is more accurate to real life on a pro level, long shots that hit the centre of the pocket are typically cue'd excellently.

so... how to translate this to AI play? well first of all we need to understand if the same rules apply to the AI, do they have a ''chance'' to cue the ball poorly if so this chance needs to increase with long potting, if they are indeed scripted as we suspect they might be then yes the accuracy simply needs to be reduced for long range pots as it seems it is as you say and 99 percent long pots are successful.

[Jud]

Definitely medium, as the AI miss a realistic percentage of difficult pots.

[Fluid Druid]

Just played 2 matches vs the A.I. on pro/medium i managed 2 centuries and a few frame winning breaks, gave the A.I. chances to pinch frames but on medium they failed to take any really. These were however against players outside the top 20. Currently playing the number 16 player luca brecel and i'm 3-1 up, so he's a bit better but not challenging enough.

[Jud]

Originally posted by jinx20001:

so... how to translate this to AI play? well first of all we need to understand if the same rules apply to the AI, do they have a ''chance'' to cue the ball poorly if so this chance needs to increase with long potting, if they are indeed scripted as we suspect they might be then yes the accuracy simply needs to be reduced for long range pots as it seems it is as you say and 99 percent long pots are successful.

I'm not sure any of us here can claim to understand how AI is coded (unless of course they happen to do that kind of thing for a living) but human behaviour and its thought process is clearly a very very difficult thing to simulate accurately. Okay, it's particularly poor in this game, but all games that use AI are lacking - even AAA games.

But just think about what that coding has to do in a snooker game. It has to calculate a pot, along with how best to get the white into a position for the next ball. If there's no pot on it has to 'work out' a safety shot. And on top of all that, some kind of 'miss' percentage has to be calculated.

No wonder, then, that we're never going to see complex shots, such as the 'shot to nothing', where a speculative pot is taken on, but very much second to the safety aspect of the shot.

It does need improving, but how that's done is very much beyond most of here, I think.